England and Wales High Court (Administrative Court) Decisions

IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
ADMINISTRATIVE COURT

		CO/313/2016, CO/314/2016, CO/315/2016 Royal Courts of Justice Strand, London, WC2A 2LL
		28/07/2017

Introduction

1. This is the judgment of the Court to which both members of the Court have contributed.
2. This is an appeal by the Government of Rwanda ["GoR"] against the decision of the then Deputy Senior District Judge Arbuthnot, declining to permit the extradition of these five men to Rwanda. The requested Parties ["RPs"] cross-appeal. The decision of Senior District Judge Arbuthnot [hereafter "the SDJ"], as she now is, was given on 22 December 2015, following a hearing which occupied the Court for 66 days. She declined to order extradition. Her decision in respect of the first four named Respondents was founded on her conclusion that there was a real risk they might suffer a flagrant breach of their rights to a fair trial if extradited, such that there would be a breach of Article 6 of the European Convention of Human Rights, and of Section 87 of the Extradition Act 2003 ["the 2003 Act"].
3. The decision in respect of the Respondent Mutabaruka, was also based on Section 80 of the 2003 Act. SDJ Arbuthnot found that his extradition was barred by the Rule against double-jeopardy. However, she indicated that she would also have discharged the fourth Respondent pursuant to Section 87 of the 2003 Act.
4. The GoR seeks to try the five RPs on various charges arising from the killings in Rwanda in 1994. It is not suggested that the charges faced by these RPs are connected, or that they acted together.
5. In the course of her judgment now appealed before us, SDJ Arbuthnot gave a helpful summary of the allegations against these five RPs, which reads as follows:

"44. Vincent Brown, it is said (…): 1. That VB was a close associate of President Habyarimana, a member of the Akazu (the President's inner circle), he participated in MRND party meetings prior to April 1994 and was a member of MRND until 1993 when CDR was founded. In 1993 he attended a meeting in Myamirambo [sic: read Nyamirambo] Stadium in Kigali where Hutus were called to disassociate themselves from the Tutsi who were said to be the enemy, at this meeting VB was said to be in charge of protocol. Finally in 1 [sic]. He attended the swearing in ceremony for the interim government on 4th July 1994 at Kibehehank where he collected financial contributions. 2. He established and supervised the manning of roadblocks in Rugenge (Kigali) and near to Kibihekane School in the North-West of Rwanda where killings took place. He is said to have participated in the killings of 3. Dominique, 4. Leandre in Rugenge, 5. Charlotte Kamugaja and baby in Rugenge.

45. In the case of Charles Munyaneza, it is alleged that as the bourgmestre of Kinyamakara in Gikongoro province he encouraged others to kill Tutsis, chaired meetings and organised roadblocks. CM took part in the looting of a property and he punished those who looted without killing the owners first; he also led a number of attacks over some days on Ruhashya over the Mwogo River which killed thousands of Tutsis.

46. Emmanuel Nteziryayo was the bourgmestre in Mudasomwa commune and therefore was another bourgmestre of the Gikongoro province (see Munyaneza above), it is alleged against him that as a bourgmestre he held meetings with conseillers and gave them guns, he told them to set up roadblocks and to kill Tutsis, he told people that Tutsi bodies should be hidden by being buried, he was present at the same meetings that CM was at on 13^th April and on 26^th April 1994 when the bourgmestres reported on the numbers of Tutsis that had been killed in their areas. In May 1994 he did nothing to prevent some people being beaten at a roadblock. He later fled to the Congo and was a leader in a refugee camp.

47. The allegations against Celestin Ugirashebuja is that he was the bourgmestre of Kigoma commune and a long standing member of the MRND. CU would pass by the road blocks and find out how many had been killed. He held a meeting of the conseillers and responsables at the commune office and instructed them to set up roadblocks and bring any "inyenzi" (cockroaches) to him. The commune policemen who were under his direct control played an important part in the killings. CU urged people attending meetings to kill Tutsis. He gave instructions that certain Tutsis were to be tricked to come out of hiding so they could be killed, he ordered that Tutsi bodies should be moved so they would not be seen by foreigners. He is also alleged to have distributed guns. In May 1994 he addressed about 300 people and urged them to destroy property belonging to Tutsis.

48. The evidence against Celestin Mutabaruka are that he took over the running of a forest management company in Musebeya, Gikongoro called Crete Zaire Nil ("CZN") in January 1992. When he took over he brought in ethnic segregation and by June 1992 it is alleged he was persecuting and discriminating against Tutsis. In November or December 1993 there was an attempt to remove him from his post when he was to be replaced by a Tutsi. CMU refused to go and claimed that his removal was a political issue between two parties, the MRND and PSD. There is documentary evidence which confirms his writing to the President of Rwanda saying he had set up a political party and asking the President to intervene in his removal from CZN.

49. The evidence shows he set up a political party on 20th October 1993 and called it "UNISODEC". It had 50 members … He describes himself as the President of UNISODEC in a letter dated 4^th January 1994. Mr Ngoga describes the party as being an ally of MRND formed as a satellite party in order to strengthen opposition to the RPF ... CMU was said to enjoy the support of the top leadership of the MRND.

50. Importantly as far as the prima facie case is concerned, it is alleged that during the genocide he played an important role along with his close friend and colleague Jonas Kanyarutoki in the killings which took place at Gatare when people who had taken shelter in the Presbyterian Church were persuaded to leave and then killed on 17^th April 1994. He was an organiser and is mentioned by a number of those involved. Once the killing had finished he checked on where the bodies had been dumped.

51. In mid May 1994, CMU led a gang of killers that murdered many people on Muyira hill in Bisesero. Using CZN vehicles, he led Interahamwe and fired into the crowd killing one person through the eye. He returned the following day and finished off any survivors. He was said to be one of the leaders who would stand with the Interahamwe and give them briefings. About 40,000 Tutsis are said to have died in those attacks.

52. CMU fled to Congo with his wife in July 1994. One witness who replaced him as director of CZN found several of his documents in his former office. Various letters are exhibits including his asking the President, Prime Minister and Commanding Officer of the Gikongoro gendarmerie for a firearm, another a letter from a CZN employee complaining that he had sacked a number of Tutsis unfairly and a letter from Oswald Rugema complaining of unfair dismissal and of being arrested at the behest of the CMU on what he says is a trumped up charge. In one of the letters which is undated CMU denies the accusations of ethnic segregation at CZN (…). These letters continue into 1994."

6. It will be noted that throughout this part of the SDJ's judgment, and indeed throughout the whole of the judgment, the Respondents were, for convenience, referred to by their initials. Thus the First Respondent, Emmanuel Nteziryayo, was 'EN', the Second Respondent, Vincent Brown/Bajinya, was 'VB', the Third Respondent, Charles Munyaneza, was 'CM', the Fourth Respondent, Celestin Mutabaruka, was 'CMU', the Fifth Respondent, Celestin Ugirashebuja, was 'CU'. We propose to refer to each individual by his surname throughout this judgment.
7. In relation to four of the RPs, the attempt at extradition is not new. Messrs Brown/Bajinya, Munyaneza, Nteziryayo and Ugirashebuja were the subject of extradition proceedings in 2007/2008. An extradition hearing took place in the Westminster Magistrates' Court over a series of dates between September 2007 and May 2008. DJ Evans heard some 18 days of oral evidence. He decided to send the cases to the Secretary of State, who signed extradition orders in respect of the four on 1 August 2008.
8. The four RPs appealed successfully to the Divisional Court. Their appeal occupied ten days in December 2008.
9. There were, and are, no general treaty arrangements covering extradition between the United Kingdom and Rwanda. In 2007/9, and now, extradition was and is sought pursuant to Memoranda of Understanding ["MoU"] between the British and Rwandan governments. The terms of the MoUs may be summarised shortly. In each case the "Judicial Authority" of the Requested State (the United Kingdom) must consider the request for extradition. Extradition may be refused if "it appears to the Judicial Authority that extradition would be incompatible with [the Requested Person's] human rights": MoU paragraph 4(d).
10. In such circumstances, the application for extradition falls under Section 194 of the 2003 Act. We address the law below. In these earlier proceedings, the four RPs raised a series of objections to extradition, but the major common position, then as now, was a claim that they would not receive fair trials in Rwanda. The Divisional Court, consisting of Laws and Sullivan LJJ, agreed, and in paragraph 121 of their judgment concluded that the four would be at real risk of a flagrant denial of justice if extradited to Rwanda. The four RPs were discharged, see R (Brown and Others) v Government of Rwanda and Another [2009] EWHC 770 (Admin).
11. The GoR made its second extradition request on 2 April 2013, naming all five RPs. They were arrested on 29 May 2013, and the proceedings followed. Interpolated in the proceedings at Westminster were hearings in the Divisional Court (VB and Others v Rwanda [2014] EWHC 889 (Admin)) and the Supreme Court (VB and Others v Westminster Magistrates' Court [2015] AC 1195), on the issue as to whether the RPs could call evidence in a Closed Material Procedure. It was ruled that they could not.
12. The outcome of the proceedings below were as we have indicated.
13. The GoR raises the following Grounds of Appeal:

i) The Senior District Judge wrongly concluded that the Respondents' extradition was incompatible with Article 6 of the ECHR [and thus Section 81(b) of the 2003 Act].

ii) The Senior District Judge erred in [her] conclusions on Section 80 of the 2003 Act (double jeopardy) in relation to Mutabaruka and ought not to have discharged him.

14. The RPs' position is that the Senior District Judge's conclusions were "simply justified", and they advance various arguments to support her conclusions. They disagree with some aspects of the GoR's interpretation of her findings. The RPs submit that the Senior District Judge must be understood to have found there was a risk of interference by the GoR with the judges who would try the RPs. If she did not so find, then the RPs say she should have done so, and they cross-appeal on that Ground. In a similar vein, the RPs submit that the Senior District Judge found that there remained a real risk (as found by the Divisional Court in 2008/9) that witnesses would be too intimidated to come forward. If SDJ Arbuthnot did not so find, she should have done, and the RPs cross-appeal on that point.
15. Brown cross-appeals on various grounds, submitting that SDJ Arbuthnot should have discharged him (1) under Section 79(1)(b) and Section 1 of the 2003 Act, that for various specific reasons to do with his own history, he would not be given a fair trial; (2) under Section 82 of the 2003 Act, that such time has passed since the alleged extradition offence that it would be unjust and oppressive to extradite him; (3) that there was no sufficient case to answer against him, and he should have been discharged under Section 84(5) of the 2003 Act; (4) that his extradition would be "incompatible with his Convention Rights" on wider grounds than those found below, and he should be discharged under Section 87(2) of the 2003 Act; and (5) that these proceedings represent an abuse of process, both because these matters could have been tried in the United Kingdom following the amendment of the International Criminal Court Act 2001 ["the ICC Act"] effected by the Coroners and Justice Act 2009, but the GoR refused to co-operate in such a process, and because evidence against Dr Brown has been fabricated.
16. Ugirashebuja cross-appeals on the basis that it would be oppressive and an abuse of process to extradite him, given that he had been previously both acquitted and convicted by local Rwandan "gacaca" courts.
17. In addition to the points of cross-appeal already touched on, Mutabaruka seeks to cross-appeal on the ground that he was and is an active politician in opposition to the GoR and will for that reason not be accorded a fair trial. The Senior District Judge should have found that extradition was barred on that basis, pursuant to Section 81(b) of the 2003 Act. In addition, Mutabaruka cross-appeals, submitting that SDJ Arbuthnot failed to give sufficient weight to the evidence of executive influence on the judiciary, particularly encapsulated by the evidence of Professor Longman. There should have been a finding that a fair trial was impossible on those grounds.
18. Nteziryayo cross-appeals on the basis of double jeopardy and/or abuse of power, under Section 80 of the 2003 Act, on the basis of extraneous matters, namely the existence of contradictory and exculpatory statements by prosecution witnesses under Section 81 of the Act; on the basis there is no adequate prima facie case against him under Section 84 and finally under ECHR Article 8.
19. It follows that the central point common to all, arising from the appeal and cross-appeals alike, is whether the RPs are at risk of a flagrant denial of justice if returned to Rwanda.

Our Task

20. We have reminded ourselves that this is an appeal, not a first instance hearing. That remains so, even though some important additional fresh material is available to us beyond that before SDJ Arbuthnot. It is limited, particularly compared with the enormous quantity of evidence below. This case was a truly formidable undertaking, clearly so at first instance and we have found it so on appeal. We pay tribute to the care taken by the judge below. The volume and complexity of the evidence, the disparate jurisdictions under consideration, the multiple levels of Courts through which many of the cases have passed; the fact that some of the evidence is incomplete and a great deal of it in writing only, all these factors must have added to her difficulties, and certainly have to ours.
21. The approach of this Court to its task of deciding whether the SDJ should have decided the case differently, particularly in the context of its evaluation of the evidence before a District Judge, was set out in Wiejak v Olsztyn Circuit Court of Poland [2007] EWHC 2123 (Admin) at paragraph 23 and has been acted upon regularly since then (see, e.g., AM v Examining Magistrate's Court No. 4 Murcia, Spain [2014] EWHC 1113 (Admin) at [38]). In Wiejak Sedley LJ said this:

"23. The effect of sections 27(2) and (3) of the Extradition Act 2003 is that an appeal may be allowed only if, in this court's judgment, the District Judge ought to have decided a question before her differently. This places the original issues very nearly at large before us, but with the obvious restrictions, first, that this court must consider the District Judge's reasons with great care in order to decide whether it differs from her and, secondly, that her fact-findings, at least where she has heard evidence, should ordinarily be respected in their entirety."

22. We direct ourselves that our task is not to decide the case afresh, but to decide if it has been shown the Senior District Judge was wrong.

The Legislation

23. The critical provisions of the 2003 Act are sections 79, 80, 81, 82, 84, 93 and 95. For convenience, the relevant parts of those sections are reproduced and appended to this judgment as Annex 1.

The History of the Rwandan Genocide

24. The history of the genocide was summarised with clarity by the Divisional Court in 2009. We do not intend to repeat that exercise. For convenience, paragraphs 7 to 14 of the 2009 judgment, setting out the summary history, are reproduced as Annex 2 to this judgment.

The gacaca Courts

25. The terrible events in Rwanda overwhelmed the existing institutions in the country. A great number of judges and lawyers died in the massacres. At the same time, once (relatively) stable government returned, there was an enormous requirement for some form of justice, beyond the capacity of the formal court system. Therefore the system of gacaca courts was established. The term means "lawn", because the hearings often took place outside on the grass or on an open space. The gacaca court was typically formed by elders or responsible members of the community. They heard witnesses, reached verdicts and passed sentences, on genocide cases, in many instances, sentences of very great length. There was an appeal system. Three of the Respondents had gacaca hearings. The system and its operation will be examined in more detail when we consider double jeopardy and abuse of process.
26. However, it is important to note from the outset that, despite their informality of process and the accepted fact that the procedures of these Courts are agreed to have fallen very far short of the requirements of fair trial, they were the established Court system for addressing genocidal offending. They were established in Rwandan law, established in practice, and their sentences were carried into execution, with very long terms of imprisonment being served as a consequence of gacaca court verdicts.

The ICTR and MICT

27. Following the genocide, the Rwandan Patriotic Front ["RPF"], which had been formed by displaced Tutsis in 1990, took power in 1994, and formed a government of National Unity, which lasted until 2003. In November 1994, the UN Security Council established the International Criminal Tribunal for Rwanda ["the ICTR"]. It was intended to bring to trial those responsible for the genocide and other serious violations of law perpetrated in Rwanda, or by Rwandan citizens in neighbouring countries. It operated from 1995 in Arusha, Tanzania. A Security Council Resolution of July 2008 called for the completion of the work of the ICTR by 2010.
28. The operation of the ICTR and the International Residual Mechanism for Criminal Tribunals, ["MICT" or "the Mechanism"]and the relationships with the Rwandan Courts are of importance in these appeals. It is helpful to touch on this before considering the judgment under appeal.
29. The work of the ICTR was not completed in 2010. Further impetus to the conclusion of its work was, however, given in 2010. This is recorded in a number of the judgments to which we have been referred. In 2010, at the request of the Security Council of the United Nations, the tribunal started to gradually reduce their activities. The remaining duties of the tribunal were handed over to the MICT, which started its work on 1 July 2012.
30. We understand from other material put before us (for example, Mr Arguin's report: see below) that the ICTR closed formally on 31 December 2015. However, of some significance in this case is the 5-year period leading to this closure. The UN Security Council passed Resolution 1966 on 22 December 2010. It is a lengthy resolution, but in short it noted the delays in the completion of the work of the ICTR (and the International Criminal Tribunal for the Former Yugoslavia set up in 1993 to deal with issues arising from the massacres in that region) and established the International Residual Mechanism for Criminal Tribunals ["the Mechanism"] to finish the remaining tasks of the ICTR and the ICTY. The branch of the Mechanism dealing with Rwanda was to begin work on 1 July 2012.
31. The ICTR was requested "to take all possible measures to expeditiously complete all their remaining work … no later than 31 December 2014, to prepare their closure and to ensure a smooth transition to the Mechanism, including through advance teams in each of the Tribunals". The Resolution urged the Tribunals and the Mechanism:

"… to actively undertake every effort to refer those cases which do not involve the most senior leaders suspected of being most responsible for crimes to competent national jurisdictions in accordance with their respective Statutes and Rules of Procedure and Evidence".

32. The distinction is made in the Resolution between "the most senior leaders suspected of being most responsible for the crimes" who may be the subject of an indictment before the ICTR and those "who are not among the most senior leaders". The Mechanism retained the unconditional power to prosecute those in the former category (Article 1(2)), but its power to prosecute those in the second category was subject to the proviso that "it has exhausted all reasonable efforts to refer the case" to the State authorities (Article 1(3)).
33. The President and the Prosecutor of the Mechanism were requested "to submit six-monthly reports to the Security Council on the progress of the work of the Mechanism." The work of the ICTR was not completed in time for a formal closure on 31 December 2014, that closure being delayed by a further year.
34. Any case referred to the jurisdiction of the Rwandan Courts, either by the ICTR or the MICT, is generally called a "referred case" or a "transferred case". What has happened in Rwanda in relation to a number of referred or transferred cases is of importance to the principal issue in these proceedings. Transfer takes place to the Specialized Chamber of the High Court in Kigali, which will be the Court of trial for the RPs if extradited. The cases of Jean Uwinkindi, Charles Bandora, Bernard Munyagishari, Léon Mugesera and Emmanuel Mbarushimana will be of interest. They have all taken place since 2009.

The 2009 Decision

35. In 2009, the Divisional Court were reaching their decision following a period during which the ICTR had declined to transfer cases to Rwanda. As the Divisional Court recited, the ICTR trial chamber in the case of Kanyarukiga had praised the "notable progress" in improving the Rwandan judicial system. The Human Rights Watch report of July 2008 had opened by stating:

"The Rwandan authorities have improved the delivery of justice in the last five years, a noteworthy achievement given the problems they have faced."

The Danish Institute for Human Rights had made positive observations about the "significant achievements in the justice sector of the past twelve years" in a submission that was put before the District Judge in 2008: see Brown, paragraph 69. The Court was therefore fully aware of the continuing efforts at that stage to improve justice in Rwanda.

36. They also took as their starting point the political context in which the Rwandan Courts operated:

"68. Moreover, the question whether a court is independent and impartial cannot be answered without considering the qualities of the political frame in which it is located. If the political regime is autocratic, betrays an intolerance of dissent, and entertains scant regard for the rule of law, the judicial arm of the State may be infected by the same vices; and even if it is not, it may be subject to political pressures at the hands of those who are, so that at the least the courts may find it difficult to deliver objective justice with even-handed procedures for every litigant whatever the nature of his background or the colour of his opinions. We must take care, of course, to avoid crude assumptions as to the quality of a State's judiciary based on the quality of the State's politics. There are, thankfully, many instances of independent judges delivering robust and balanced justice in a harsh and inimical environment; but it takes courage and steadfastness of a high order."

37. Between paragraphs 67 and 118 of the Brown decision, they recited the evidence placed before them. We do not intend to repeat any of that, noting as we do that there is a degree of overlap in some of the material relied on at that stage and the material now relied on by the Respondents.
38. It was on that background and material that the Court in Brown reached their conclusions as follows:

"119. As will be apparent from this judgment, we accord great respect to the ICTR's decisions. However, the Appeals Chamber's finding that no reasonable Trial Chamber would have concluded that there was sufficient risk of government interference with the Rwandan judiciary to warrant denying the prosecution's transfer request was based only on the record before it, and in particular on the failure to mention any specific incidents of judicial interference (paragraph 78 above). We have had the advantage of being able to consider not only the HRW Report of July 2008, including its treatment of the Bizimungu case (paragraph 104 above), but also the evidence of Professor Reyntjens, Professor Sands and Professor Schabas, and in particular the acceptance by Professor Schabas in cross-examination on 21 April 2008 that there probably was executive interference in the Bizimungu case (see paragraph 101 above). Thus we have the evidence of a specific incident of judicial interference that the Appeals Chamber lacked.

120. More generally, we have not forgotten the scale of the dreadful tribulations suffered in Rwanda in 1994. Nor have we ignored the real and substantial measures taken to establish a judicial system capable of delivering criminal justice to acceptable standards. But our duty is to apply an objective test – real risk of flagrant denial of justice. We certainly cannot sanction extradition as a means of encouraging the Rwandan authorities to redouble their efforts to achieve a justice system that guarantees due process. That might serve a political aspiration, but would amount to denial of legal principle.

121. We stated earlier (paragraph 68) that the question whether a court is independent and impartial cannot be answered without considering the qualities of the political frame in which it is located. We have had no day-by-day details from the GoR of the conduct of the Rwandan High Court's business. No details of trials; of defences run, successfully or unsuccessfully; no details of any of the myriad events that show a court is working justly. We have reached a firm conclusion as to the gravity of the problems that would face these appellants as regards witnesses if they were returned for trial in Rwanda. Those very problems do not promise well for the judiciary's impartiality and independence. The general evidence as to the nature of the Rwandan polity offers no better promise… "

39. The Divisional Court then set these findings against the other aspects affecting the justice system when reaching their overall conclusion to decline extradition.
40. The Divisional Court in Brown went on to criticise the first instance decision in 2007 in the following terms:

"34. …The test is correctly stated in the opening sentence of paragraph 536. Notwithstanding that, the judge appears to have directed himself that the appellants carried the burden of proving on the balance of probabilities that there would be a flagrant denial of justice if they were extradited. But "real risk" does not mean proof on the balance of probabilities. It means a risk which is substantial and not merely fanciful; and it may be established by something less than proof of a 51% probability. The approach is the same as that taken in refugee cases, where the asylum seeker has to show a real risk that if he is returned to his home State he will be persecuted on any of the grounds set out in the 1951 United Nations Refugee Convention …. We think that despite his citation of the correct test the judge fell into error here. He may have been distracted by the second part of the test - "flagrant denial": so much is suggested by his repeated references to the "high" or "very high" test."

As will already be clear, consistent with the concession then made, the Divisional Court concluded that there was a lack of independence on the part of the Rwandan judiciary sufficient to represent a real risk of a flagrant denial of a fair trial.

The Basis of the Fresh Application for Extradition

41. The basis of the fresh application for extradition was summarised in the deposition or affidavit of Martin Ngoga, then Prosecutor General of Rwanda, sworn on 2 April 2013, in support of the application. The deposition begins from the refusal in Brown, and the core case can be summarised as follows: there have been material changes in the provision of video conferencing facilities in Rwanda; changes in witness protection facilities in Rwanda and changes in Rwandan law allowing international judges to try cases of genocidaires transferred or extradited to Rwanda. There were, in addition, some changed practices as to representation. The affidavit went on to recite, as "developments in relation to extradition requests by Rwanda for genocidaires", a number of decisions by the ICTR Referral and Appeals Chambers and other non-Rwandan Courts, permitting transfer or extradition to Rwanda.
42. The essence of the claim is that these changes in Rwandan law are sufficient to make extradition safe, at least so as to pass the threshold for Article 6, as witness the decisions of the ICTR and other Courts. It was to that end that the Court below looked in detail at the changes advanced, and how far they had changed things in practice, and also looked at the decisions of other Courts, and the evidence and reasoning behind those decisions.
43. Mr Ngoga's deposition also outlined the social and political history of Rwanda and of the genocide, and the allegations against the five Respondents. This account of the killings varies in detail, and to some degree in tone, from that outlined by the Court in Brown, but the differences are not sufficient to warrant an account here.

The Decision of the Senior District Judge

44. As we have indicated, the hearing was very extensive. There were over 23,000 pages of evidence. Given the length of hearing and the breadth of the evidence, SDJ Arbuthnot's judgment was economical and to the point, but even so covers 128 pages. We do not intend to rehearse all of her findings.
45. It is inevitable in proceedings which extend over this length of time, and in which a Court is looking forward, essentially assessing a future risk, that evidence will emerge as matters proceed. That happened below, as SDJ Arbuthnot noted. Broadly, she permitted material to be considered as it arose, considering some of the trials of transferred defendants and developing practice in Rwanda. We have taken the same course. There was a significant body of live evidence below, and to that extent SDJ Arbuthnot was better positioned to judge at least some of the evidence than we are able to. However onerous the process has been, we do not consider we are in any difficulty as an appellate Court, given the material we have in writing, and the clarity of SDJ Arbuthnot's findings.
46. Early in her judgment (paragraphs 11 to 32), SDJ Arbuthnot recorded the submission of the GoR that the system of justice in Rwanda had undergone a "sea-change" since 2009. This underpins many of the arguments made below and to us. Much of this bears on the central issue of fair trial, but also on other grounds advanced by the RPs. It was said then by the GoR that no Court had refused to extradite, transfer or deport to Rwanda since 2009, in relation to those accused of genocide charges, although a late exception arose in November 2015, when a Dutch Court in The Hague refused to extradite a Rwandan citizen, Mr Iyamuremye. This was of particular interest, since a key witness in that case was Mr Martin Witteveen, who played a very significant part in these proceedings. However, since this matter arose so very late, and since so little detail had surfaced for her to consider, SDJ Arbuthnot declined to rely on this decision. We touch on the case later in this judgment.
47. The submission on behalf of Vincent Brown is that the decision of the Divisional Court in 2009 was "binding" on SDJ Arbuthnot. We reject this submission, if it is intended to mean that there can be no variation from that Court's conclusions reflecting developments since that hearing concluded. Even in respect of the factual conclusions reached by that Court, there is no strict rule binding any Court in these proceedings to follow them. However, the latter point does not seem to us to arise. There is no evidence which would cause doubt to be thrown on the conclusion in 2009 based on the evidence then available. On the contrary, we accept the accuracy and validity of that judgment, which inevitably forms the starting point for this case.
48. The key points advanced in favour of the suggested "sea change" were outlined in Ngoga's affidavit as we have indicated. His points included reliance on a new provision under the 2009 Transfer Amendment Law providing for a quorum of three or more judges to try cases; immunity from prosecution for witnesses in respect of anything said or done in the course of a trial and the capacity to take evidence by deposition or video-link, in Rwanda or abroad, provided there is good reason. There is a right to legal aid, and an obligation placed on all members of the Bar in Rwanda to provide pro bono legal services. The Rwandan Bar Association ["RBA"], later renamed the Kigali Bar Association ["KBA"], has an obligation to provide for legal services to the indigent. The Ministry of Justice pays a monthly stipend for such legal services and was said to have specific funds available for transferred and extradited defendants. Mr Ngoga described arrangements for facilitating defence investigations, suggesting that the judicial police conduct investigations "for and against" the accused. There were said to be proper provisions for witness protection. There were two bodies responsible for that programme. The first was the Witness and Victim Support Unit ["WVSU"] housed within the public prosecutor's office. The second body was the Witness Protection Unit ["WPU"] set up more recently under the direction of the judiciary, in response to concerns that "some defence witnesses might be reluctant to use the WVSU", given its association with the prosecutors.
49. Mr Ngoga deposed that the Rwandan judiciary were now impartial and independent. In 2008, their tenure had been changed, from office for life to a determinate term of office, renewable by the High Council of the Judiciary. Hearings are in public and reasoned decisions given. The acquittal rate in High Court criminal cases was about 30%, although SDJ Arbuthnot noted that Mr Ngoga quoted no separate acquittal rate for genocide cases. Mr Ngoga noted the new provision in a law of November 2011 permitting the President of the Supreme Court to request judges from abroad to come and sit with Rwandan judges on cases such as these.
50. In the light of those changes, Mr Ngoga pointed to the significance of the recent cases transferred to Rwanda from the ICTR and elsewhere.
51. SDJ Arbuthnot noted two further changes following the close of evidence before her. In August 2015 there was a change in Rwandan law permitting applications to be made to the Registrar of the Specialized Chamber of the High Court to grant funding in relation to international and cross-border crimes for defence investigations. It had also emerged that the Transfer Law had been amended to provide for a measure of witness immunity.
52. Many of these points formed an important backdrop to the fair trial consideration by SDJ Arbuthnot and will do so for us. Very many of those points are contested, at least as existing only in theory, but belied by the practice.
53. SDJ Arbuthnot accepted the plea of double jeopardy in relation to Mutabaruka (paragraphs 58-70) and rejected it in the case of Nteziryayo (paragraphs 71-77). In relation to the alleged bar of "extraneous considerations", based on judicial bias and hostility, made by Brown/Bajinya, Munyaneza and Nteziryayo, she ruled there was no such bar: see paragraphs 80-96. She rejected arguments on behalf of Brown/Bajinya and Nteziryayo based on oppression or injustice because of delay: see paragraphs 97-104. She rejected arguments that there existed no prima facie case against them from Brown/Bajinya (paragraphs 110-111, and Appendix 1, paragraphs 1-133), from Nteziryayo (paragraphs 123-125 and Appendix 1, paragraphs 134-203), Munyaneza (paragraphs 112-122), Ugirashebuja (paragraphs 126-131) and Mutabaruka (paragraph 133). She rejected arguments from Brown/Bajinya, Munyaneza, Ugirashebuja and Nteziryayo based on Article 3 (paragraphs 135-146). She rejected submissions based on Article 8 from Brown/Bajinya and Nteziryayo (paragraphs 632-643) and an argument based on forms from Brown/Bajinya (paragraphs 644 to 648) and residual arguments from Brown/Bajinya, Nteziryayo and Ugirashebuja on abuse of process (paragraphs 649 to 664). She rejected all other arguments from Respondents on extraneous considerations.
54. However, the greater part of the judgment is devoted to the issue which for brevity we simply refer to as "fair trial". This occupied paragraphs 147 to 631 of the judgment.

Summary of SDJ Arbuthnot's Conclusions on "Fair Trial"

55. SDJ Arbuthnot reached a rather nuanced conclusion on the impartiality of the judiciary in genocide cases as follows:

"545. To conclude, I am drawing a distinction between the way the Rwandan High Court tries cases with a political flavour and the way they try genocide allegations. This is based on the clear evidence I have seen about the approach taken by the Specialised Chamber towards the five transferred genocide cases. Having considered all the evidence, I cannot exclude a risk of interference but judging from the transferred defendants the highest risk is from the pressure exerted by GoR ministers' comments in public and in the press. I consider any such risk would be reduced by a robust, able and experienced defence team with an ability to investigate the defence case and international monitoring of some sort. I consider that without both of these the RPs would be at a greater risk of judges behaving partially and being influenced by factors outside the evidence."

56. It will be evident that these conclusions throw great weight on the quality of representation in the trial, on the capacity of the defence representatives to investigate, prepare and present the defence case, and on effective international monitoring. In its essence, SDJ Arbuthnot's conclusion was, in effect, that vulnerability to pressure on the Bench and other weaknesses in the system might be sufficiently offset by effective representation and international monitoring. She then proceeded to consider the defence capacity and effectiveness of representation.
57. In relation to the availability of defence witnesses, SDJ Arbuthnot's conclusions are set out in paragraphs 593 to 599 as follows:

"593. I accept that a number of witnesses in this case have told investigators that they are too frightened to give evidence in Rwanda. I find that they are frightened and are expressing their genuinely held views. Many of the witnesses are from rural backgrounds and are of relatively low educational attainment. The reputation of the GoR at home and abroad as I have found in Paragraphs 221-223 cannot be of assistance either but Witteveen's evidence of his experience, which I accept, is that the witnesses are more frightened of local repercussions rather than national ones although both are feared. The witnesses' fears vary from a concern that they will be killed or imprisoned and tortured to a fear of losing their benefits as genocide survivors. Some fear they may be charged with genocide minimization type offences or prosecuted for offences arising out of the genocide.

594. At the same time 14 defence witnesses have been heard in the transfer case of Uwinkindi and nine in the case of Bandora. Defence witnesses were heard in the Gacaca proceedings against the acquitted RPs. There is the example of the nine defence witnesses who gave evidence in the case against EN in Tare II when he was acquitted. There is also the nine defence witnesses in the case of Celestin Mutabaruka who are willing to give evidence in the Specialised Chamber of the High Court were he to be extradited.

595. It was said in the Uwinkindi Referral Chamber in 2011 that between 2005 and 2010 at the ICTR, 357 witnesses had testified for the defence and 424 for the prosecution and there was no evidence that witnesses who returned to Rwanda subsequently raised security concerns. The same Chamber considered the question of defence witnesses in the High Court in Kigali and even then without the protection of Transfer Laws, the ICTR noted that in most of the then 36 genocide cases tried in the High Court, defence witnesses were available. I noted that it was not disputed that thousands of witnesses have given evidence for the defence in Gacaca proceedings.

596. I noted that the ICTR in Munyagishari in June 2012 found that legal protection for defence witnesses was adequate on the basis of the Transfer Law and this was despite the court receiving statements from 16 witnesses who wanted to be remain anonymous and said they did not want to testify in Rwanda.

597. I find the following provisions would be provided by the judicial authorities in Rwanda for frightened witnesses giving evidence.

A guarantee that defence witnesses could not be prosecuted for anything said or done in the course of a trial

A facility to enable them to give evidence anonymously

A facility to enable them to give evidence by video-link or by deposition in Rwanda or abroad. Video-link includes voice distortion and other methods of disguising a witness' identity

The use of the WPU which is now fully functional which can provide a safe house, assistance with travel and other protective measures. It is independent of the GoR and is run by the Registrars of the High Court and Supreme Court.

598. There are many reports of witnesses being willing to give evidence in genocide trials in the High Court in Kigali. I consider that although there can never be any guarantee of safety at least some of the frightened defence witnesses are likely to give evidence for the defence were the defendants to be returned for trial. I do not accept that the defence in this case will be unable to marshall a sufficient number of defence witnesses if the defendants are properly represented with adequate assistance and the new provisions for video-link, anonymity, the WPU etc are explained to them which will enable them to give evidence in a protected environment.

599. I find that there is no real risk of a flagrant denial of fair trial in relation to the availability of defence witnesses as long as the RPs, if returned, are represented by able and effective representatives who are able to investigate and put together the case for the defence."

58. It follows once again that the Senior District Judge's conclusions on the adequate preparation and deployment of defence witnesses were dependent on the quality and effectiveness of representation. And it was on this latter point, after a careful review of the evidence, that SDJ Arbuthnot decided against extradition. Her conclusion on this aspect of the case, critical as she found it to be - in a sense the keystone of the structure - was as follows. It is clear from the way her conclusions are expressed, that the evidence of the expert witness Martin Witteveen was crucial to the outcome:

"630. From all the evidence I have read and heard I concur with Witteveen's Final Conclusions in his Additional Report, he is certain that the facts in genocide cases can be established but "only under the condition of high quality and professional investigations, applying internationally accepted standards. Part of this professionalism and these standards is the necessity to have defence attorneys who possess the knowledge, experience and the resources to conduct investigations for the defence, including the capabilities to conduct investigations abroad". He has profound doubts that the defence lawyers assigned to the transfer cases can do that. His solution is still to extradite but to provide a lawyer with the right experience to work alongside Rwandan lawyers who would be provided with appropriate funds to conduct investigations. In his view (and in my view) "it ensures the necessary adequate defence capabilities for the defendant that meets the required standard and guarantees not only procedurally fair trial but also a fairness to the trial". Unfortunately this court does not have the power to order extradition on conditions and I am applying a different test.

631. I find that if extradited, as things presently stand, the defendants would be denied the effective representation of counsel in cases which so obviously call for effective and skilled representation by suitably experienced and resourced defence lawyers. It is too early to say that sufficient funding for defence investigations in relation to witnesses abroad will be provided. These defendants are legally aided in this country and will be indigent in Rwanda. I have seen in this case what the effective representation by counsel can achieve. Without such representation and funding, the High Court in Rwanda would be presented with the prosecution case and the RPs would find it impossible to present their side of what happened. I find the RPs would be exposed to a real risk of a flagrant denial of justice and a breach of Article 6."

59. Drawing the threads together, SDJ Arbuthnot focussed on four key issues when deciding whether there would arise a "real risk of a flagrant denial of justice": the independence of the judiciary and pressure from the Executive; the availability of, and conditions for, defence witnesses; adequate legal representation for those facing genocide trials; and more briefly, on international supervision. Those matters interlock. SDJ Arbuthnot recognised concerns about the independence of the judiciary and about defence witnesses. Reciting the correct legal test, that is to say not full procedural fairness as would be required for conformity with ECHR Article 6 in a State which is a member of the Council of Europe, but the lower standard of "a real risk of a flagrant denial of justice", SDJ Arbuthnot concluded that the difficulties she identified with the independence of the judiciary and the position of defence witnesses might be, but in the event were not sufficiently offset, by robust and effective representation.
60. We now proceed to focus firstly on the legal test to be applied, and then to consider these critical areas affecting fair trial. As indicated, we are prepared to consider not only the material before the Court below, but some material which has emerged since the conclusion of the hearing below. In general terms, subject to any particular point arising, we accept that the additional material advanced by either side passes the test laid down in Hungary v Fenyvesi [2009] EWHC 2313 (Admin). For reasons explained below, we do not follow this approach when considering the issue of double jeopardy or abuse of process for Mutabaruka.

The Legal Test: a real risk of a flagrant denial of justice

61. We have reviewed the relevant law with some care, given the central importance of this issue in the case.
62. In R (Ullah) v Special Adjudicator [2004] 2 AC 323, the House of Lords had to consider the circumstances in which it would be unlawful for a state which is party to the European Convention of Human Rights to remove an alien to a country where he or she was foreseeably at real risk of ill treatment infringing Articles of the Convention other than Article 3. The overall conclusion, enshrined in the headnote was that:

"Reliance on such Articles required presentation of an exceptionally strong case, such that the actual or threatened treatment would amount to a flagrant denial or gross violation of the relevant right."

63. The facts in Ullah were very different to those arising here. The case concerned the practice of religion. In the course of his leading speech, Lord Bingham made reference to the well-known earlier authority of Soering v The United Kingdom (1989) 11 EHRR 439 where the ECtHR, while dismissing the applicant's complaint under Article 6 on the facts, did not reject it as ill-founded in principle. The ECtHR observed (paragraph 113):

"113.  The right to a fair trial in criminal proceedings, as embodied in Article 6, holds a prominent place in a democratic society. The Court does not exclude that an issue might exceptionally be raised under Article 6 by an extradition decision in circumstances where the fugitive has suffered or risks suffering a flagrant denial of a fair trial in the requesting country."

64. Lord Bingham went on to review the post-Soering Strasbourg authority on the potential applicability of Article 6, although observing (paragraph 17) that the authority "remains somewhat tentative". Following his review, Lord Bingham recorded his conclusions as follows:

"24. While the Strasbourg jurisprudence does not preclude reliance on articles other than article 3 as a ground for resisting extradition or expulsion, it makes it quite clear that successful reliance demands presentation of a very strong case. In relation to article 3, it is necessary to show strong grounds for believing that the person, if returned, faces a real risk of being subjected to torture or to inhuman or degrading treatment or punishment: Soering, paragraph 91; Cruz Varas, paragraph 69; Vilvarajah, paragraph 103. In Dehwari, paragraph 61 (see paragraph 13 above) the Commission doubted whether a real risk was enough to resist removal under article 2, suggesting that the loss of life must be shown to be a "near-certainty". Where reliance is placed on article 6 it must be shown that a person has suffered or risks suffering a flagrant denial of a fair trial in the receiving state: Soering, paragraph 113 (see paragraph 10 above); Drodz, paragraph 110; Einhorn, paragraph 32; Razaghi v Sweden; Tomic v United Kingdom. Successful reliance on article 5 would have to meet no less exacting a test. The lack of success of applicants relying on articles 2, 5 and 6 before the Strasbourg court highlights the difficulty of meeting the stringent test which that court imposes. This difficulty will not be less where reliance is placed on articles such as 8 or 9, which provide for the striking of a balance between the right of the individual and the wider interests of the community even in a case where a serious interference is shown. This is not a balance which the Strasbourg court ought ordinarily to strike in the first instance, nor is it a balance which that court is well placed to assess in the absence of representations by the receiving state whose laws, institutions or practices are the subject of criticism. On the other hand, the removing state will always have what will usually be strong grounds for justifying its own conduct: the great importance of operating firm and orderly immigration control in an expulsion case; the great desirability of honouring extradition treaties made with other states. The correct approach in cases involving qualified rights such as those under articles 8 and 9 is in my opinion that indicated by the Immigration Appeal Tribunal (Mr C M G Ockelton, deputy president, Mr Allen and Mr Moulden) in Devaseelan v Secretary of State for the Home Department [2002] IAT 702, [2003] Imm AR 1, paragraph 111:

"The reason why flagrant denial or gross violation is to be taken into account is that it is only in such a case - where the right will be completely denied or nullified in the destination country - that it can be said that removal will breach the treaty obligations of the signatory state however those obligations might be interpreted or whatever might be said by or on behalf of the destination state"."

65. It is worth considering carefully the balancing considerations touched on by Lord Bingham, including as they did "the great desirability of honouring extradition treaties made with other states". As we have said, there is in the instant case no extradition treaty with Rwanda. The MoU might be thought to involve more specific and more nuanced mutual obligations than the more general arrangements enshrined in an extradition treaty.
66. Lord Steyn also reviewed the Strasbourg jurisprudence and concluded that:

"It will be necessary to establish at least a real risk of a flagrant violation of the very essence of the right before other articles [than Article 3] could become engaged." (paragraph 50)

Lord Walker and Baroness Hale agreed with Lord Bingham. Lord Carswell followed a similar line, although he emphasised that:

"I do regard it as important, however, that member states should not attempt to impose Convention standards on other countries by decisions which have the effect of requiring adherence to those standards in those countries." (paragraph 63)

67. Specifically, in relation to Article 6, Lord Carswell said the following:

"69. The adjective "flagrant" has been repeated in many statements where the Court has kept open the possibility of engagement of articles of the Convention other than article 3, a number of which are enumerated in paragraph 24 of the opinion of Lord Bingham of Cornhill in the present appeal. The concept of a flagrant breach or violation may not always be easy for domestic courts to apply - one is put in mind of the difficulties which they have had in applying that of gross negligence - but it seems to me that it was well expressed by the Immigration Appeal Tribunal in Devaseelan v Secretary of State for the Home Department [2003] Imm AR 1 at p 34, para 111, when it applied the criterion that the right in question would be completely denied or nullified in the destination country. This would harmonise with the concept of a fundamental breach, with which courts in this jurisdiction are familiar."

68. In EM (Lebanon) v Home Secretary [2009] 1 AC 1198 the House of Lords was concerned with a case arising from the possible engagement and infringement of Article 8 of the Convention. Although Article 6 was not directly in point, the proper approach to infringement of Articles other than Article 3 was relevant. The House of Lords reviewed earlier authority including Ullah. The leading speech was given by Lord Hope who considered Ullah, but also more recent European authority, and recorded his conclusions in two key paragraphs as follows:

"3. I take the wording of the test to be applied to determine whether there would be a flagrant denial of this right from what Judges Bratza, Bonello and Hedigan said in their joint partly dissenting opinion in Mamatkulov and Askarov v Turkey (2005) 41 EHRR 25, 537-539. That was a case where political dissidents claimed that they would not receive a fair trial if they were extradited to Uzbekistan because, among other things, torture was routinely used to secure guilty verdicts and because suspects were frequently denied access to a lawyer. Their case was that they ran a real risk of a flagrant denial of justice. In para O-III14 the judges said:

"In our view, what the word 'flagrant' is intended to convey is a breach of the principles of fair trial guaranteed by article 6 which is so fundamental as to amount to a nullification, or destruction of the very essence, of the right guaranteed by that article."

In paras O-III17 and O-III19 they used the expression "a real risk" to describe the standard which the evidence has to achieve in order to show that the expulsion or extradition of the individual would, if carried out, violate the article.

4. I have gone directly to what those judges said about the test in Mamatkulov rather than to what was said in R (Ullah) v Special Adjudicator, Do v Immigration Appeal Tribunal [2004] 2 AC 323 and R (Razgar) v Secretary of State for the Home Department [2004] 2 AC 368 for several reasons. First, their description of it is the most up to date guidance that is available from Strasbourg. Second, it combines in a simple formula the approach described in Devaseelan v Secretary of State for the Home Department [2003] Imm AR 1, para 111 referred to with approval by Lord Bingham of Cornhill and Lord Carswell in paras 24 and 69 of Ullah with Lord Steyn's use of the expression "the very essence of the right" in para 50 of Ullah. And, third, it shows that Carnwath LJ in the Court of Appeal [2007] UKHRR 1, paras 37-38 was, with great respect, wrong to regard words such as "complete denial" or "nullification" on the one hand and "flagrant breach" or "gross invasion" on the other as indicating different tests. Attempts to explain or analyse the formula should be resisted, in the absence of further guidance from Strasbourg. There is only one test, although I think that how it is to be applied in an article 8 read with article 14 case needs some explanation. The use by the partly dissenting judges of the expression "a real risk" is also significant. It shows that what was said about the standard of proof in the context of article 3 in Soering v United Kingdom (1989) 11 EHRR 439, para 91, applies to cases such as this where the rights in issue are among the qualified rights to be found elsewhere in the Convention."

69. Further, in paragraph 13 of his speech, Lord Hope made a distinction between the rights enshrined in Articles 2, 3 and 6, and other Convention rights, in the following terms:

"13. Running through these three recent cases is a recognition by the Strasbourg court that, while the Contracting States are obliged to protect those from other jurisdictions who can show that for whatever reason they will suffer persecution or are at real risk of death or serious ill-treatment or will face arbitrary detention or a flagrant denial of a fair trial in the receiving country, limits must be set on the extent to which they can be held responsible outside the areas that are prescribed by articles 2 and 3 and by the fundamental right under article 6 to a fair trial. Those limits must be seen against the background of the general principle of international law that states have the right to control the entry, residence and expulsion of aliens."

70. The second speech in EM (Lebanon) was given by Lord Bingham. He reviewed previous authority, and in particular reprised the decision of the House in Ullah (paragraphs 32 and 33). In his view the threshold test laid down in Ullah could not be said to misrepresent or understate the effect of Strasbourg authority (paragraph 34). His conclusions as to the slightly differing formulation, he set down as follows:

"35. In adopting and endorsing the test formulated by the AIT in Devaseelan I did not in para 24 of my opinion in Ullah [2004] 2 AC 323 understand that tribunal to be distinguishing a "flagrant denial or gross violation" of a right from a complete denial or nullification of it but rather to be assimilating those expressions. This was how the point had been put to the House by the Attorney General for the Secretary of State, as is evidenced from the report of his argument (p 337D):

"If other articles can be engaged the threshold test will require a flagrant breach of the relevant right, such as will completely deny or nullify the right in the destination country: see Devaseelan v Secretary of State for the Home Department [2003] Imm AR 1. A serious or discriminatory interference with the right protected would be insufficient."

It is difficult, with respect, to see how the point could be put more clearly, and any attempt at paraphrase runs the risk of causing confusion."

71. In her speech, Baroness Hale agreed with Lord Bingham (paragraph 45), as did Lord Carswell (paragraph 51). Lord Brown indicated agreement with both Lord Bingham and Lord Hope (paragraph 60).
72. The case of Othman v SSHD [2008] EWCA Civ 290 did directly concern the risk of breach of Article 6. The decision in the Court of Appeal preceded the decision of the House of Lords in EM (Lebanon). One key question in Othman was whether evidence would be used in a Jordanian trial which was the product of torture. The case was an appeal from the decision of SIAC, which concluded that there would not be a flagrant denial of the right to a fair trial, because of other safeguards which they found present in the Jordanian trial system. That conclusion was disapproved by the Court of Appeal, essentially on the ground that SIAC had considered such evidence merely as one element affecting the issue of fair trial, rather than considering the potential breach of the prohibition on torture itself (paragraphs 53 and 67). We are not convinced that the decision in Othman at this level adds anything to the analysis which must be applied here.
73. The Othman case reached the House of Lords as RB (Algeria) v Secretary of State [2010] 2 AC 110. In his leading opinion, Lord Phillips addressed the question of breach of Article 6, again recapitulating earlier authority, but adding emphasis to the consequences of a breach of Article 6. In paragraph 133 he said:

"133. I have described earlier the origin of the phrase "flagrant breach" in relation to foreign Convention cases and expressed the view that, where article 5 is engaged, the potential consequence of a breach of that article must be severe before the breach can properly be described as flagrant. This approach to the meaning of "flagrant breach" should pose little difficulty where the right engaged is substantive. It is not so easy where the right arises under article 6, for that right is not substantive but procedural. As there is no reported foreign case where article 6 has successfully been invoked, there is a lack of authoritative guidance as to what will amount to a "flagrant breach" of that article."

74. In the ensuing parts of his speech, Lord Phillips went on to consider the minority opinion of the judges of the Grand Chamber of the ECtHR in Mamatkulov referred to above, the speech of Lord Bingham in EM, and then went on to consider the application of the test by SIAC:

"135. SIAC adopted this test when considering Mr Othman's case. They held that to succeed under article 6 he needed to establish "a real risk of a total denial of the right to a fair trial" para 451. The Court of Appeal also held at paragraphs 15 to 19 of the judgment in Mr Othman's appeal that the test was whether there had been a "complete denial or nullification of the Convention right."

136. This is neither an easy nor an adequate test of whether article 6 should bar the deportation of an alien. In the first place it is not easy to postulate what amounts to "a complete denial or nullification of the right to a fair trial" That phrase cannot require that every aspect of the trial process should be unfair. A trial that is fair in part may be no more acceptable than the curate's egg. What is required is that the deficiency or deficiencies in the trial process should be such as fundamentally to destroy the fairness of the prospective trial.

137. In the second place, the fact that the deportee may find himself subject in the receiving country to a legal process that is blatantly unfair cannot, of itself, justify placing an embargo on his deportation. The focus must be not simply on the unfairness of the trial process but on its potential consequences. An unfair trial is likely to lead to the violation of substantive human rights and the extent of that prospective violation must plainly be an important factor in deciding whether deportation is precluded.

138. A conviction that results from a flagrantly unfair trial cannot be relied upon under article 5(1)(a) as justifying detention.

"It is the Convention organs' case law that the requirement of Art.5(1)(a) that a person be lawfully detained after "conviction by a competent court" does not imply that the Court has to subject the proceedings leading to that conviction to a comprehensive scrutiny and verify whether they have fully complied with all the requirements of Art.6 of the Convention. However, the Court has also held that if a "conviction" is the result of proceedings which were a "flagrant denial of justice", that is were "manifestly contrary to the provisions of Article 6 or the principles embodied therein", the resulting deprivation of liberty would not be justified under Art.5(1)(a)."– Stoichkov v Bulgaria (2005) 44 EHRR 276."

Nor can such a conviction justify the imposition and execution of the death penalty. In either case the breach of the procedural rights guaranteed by article 6 will result in a breach of a substantive right. If an alien is to avoid deportation because he faces unfair legal process in the receiving state he must show that there are substantial grounds for believing that there is a real risk not merely that he will suffer a flagrant breach of his article 6 rights, but that the consequence will be a serious violation of a substantive right or rights. Quite how serious that violation must be has yet to be made clear by the Strasbourg jurisprudence. Plainly a sentence of death will be sufficient but the ECtHR has expressed doubts as to whether the risk of violation of article 5 can suffice to prevent the expulsion of an alien. In Tomic v United Kingdom (application no. 17837/03) an admissibility decision of 14 October 2003, the court commented:

"The Court does not exclude that an issue might exceptionally be raised under Article 6 by an expulsion decision in circumstances where the person being expelled has suffered or risks suffering a flagrant denial of a fair trial in the receiving country, particularly where there is the risk of execution (see, mutatis mutandis, Soering v United Kingdom, [(1989) 11 EHRR 439], § 113; Ocalan v Turkey, [(2005) 41 EHRR 985], §§ 199-213). Whether an issue could be raised by the prospect of arbitrary detention contrary to Article 5 is even less clear."

A similar comment was made in Z and T v United Kingdom (application no. 27034/05) decision of 28 February 2006.

139. In Mamatkulov the Grand Chamber was prepared to contemplate, and the minority to find, a violation of article 6 in circumstances where the extradition of the applicants had resulted in lengthy prison sentences. A different approach will, however, be appropriate in an extradition case. There it is the prospective trial that is relied on to justify the deportation. If there is a real risk that the trial will be flagrantly unfair, that is likely to be enough of itself to prevent extradition regardless of the likely consequences of the unfair trial. 140. In Bader v Sweden (2005) 46 EHRR 1497 the applicant successfully resisted deportation on the ground that if sent back to Syria he would be at risk of being executed pursuant to a trial that had been held in his absence. He relied on articles 2 and 3, but did not expressly aver a breach of article 6. Such a breach was, however, a necessary element in his case in relation to articles 2 and 3. At para 42 the ECtHR observed that to implement a death sentence following an unfair trial would violate article 2 and continued:

"Moreover, to impose a death sentence on a person after an unfair trial would generate, in circumstances where there exists a real possibility that the sentence will be enforced, a significant degree of human anguish and fear, bringing the treatment within the scope of Art.3 of the Convention.

In this connection it should also be noted that the Court has acknowledged that an issue might exceptionally be raised under Art.6 of the Convention by an extradition decision in circumstances where the fugitive has suffered or risks suffering a flagrant denial of a fair trial in the requesting country.

It follows that an issue may arise under Arts 2 and 3 of the Convention if a contracting state deports an alien who has suffered or risks suffering a flagrant denial of a fair trial in the receiving State, the outcome of which was or is likely to be the death penalty."

This decision exemplifies the approach of considering in combination the risk of a violation of Article 6 and articles guaranteeing substantive rights, the articles in question being 2 and 3. Although there is no authority that establishes this, I think that it is likely that the Strasbourg Court would hold article 6 and article 5 to be violated if an applicant were to be deported in circumstances where there were substantial grounds for believing that he would face a real risk of a flagrantly unfair trial and that the defects in the trial would lead to conviction and a sentence of many years imprisonment.

141. In summary, the Strasbourg jurisprudence, tentative though it is, has led me to these conclusions. Before the deportation of an alien will be capable of violating article 6 there must be substantial grounds for believing that there is a real risk (i) that there will be a fundamental breach of the principles of a fair trial guaranteed by article 6 and (ii) that this failure will lead to a miscarriage of justice that itself constitutes a flagrant violation of the victim's fundamental rights [emphasis added]. I turn to consider, having regard to this test, whether SIAC erred in law in concluding that Article 6 posed no bar to Mr Othman's deportation.

142. The potential consequences to Mr Othman of conviction of the offences for which he will be tried have already been established by the trials that took place in his absence: life imprisonment in the first instance and fifteen years imprisonment in the other. These consequences are, I believe, sufficiently severe to satisfy the second limb of the test. The vital issue that SIAC had to address was whether there were substantial grounds for believing that Mr Othman faced a real risk of a fundamental breach of the principles of a fair trial as recognised by the Strasbourg court."

75. In paragraphs 144 to 146, Lord Phillips went on to consider the composition of the Jordanian Military Court, consisting of three judges, of whom two at least would be senior military officers. The state prosecutors would be military officers. The judges would have no security of tenure and would be:

"…subject to being replaced by executive decision. They would be subject to the influence of the executive. While not independent there was no reason to suspect them of partiality." (paragraph 144)

There had been in the past acquittals and successful appeals. Lord Phillips observed that such a court would not satisfy the Article 6 requirement of an independent and impartial tribunal (paragraph 145) and to conclude that:

"146. While in a domestic case the composition of the SSCt would violate article 6, it does not follow that this would, of itself, constitute a flagrant breach of article 6 sufficient to prevent deportation in a foreign case. The Court of Appeal considered this question at paragraphs 33 to 42 of its judgment and, in agreement with SIAC, concluded that it would not. I have reached the same conclusion and would endorse the reasoning on this point of the Court of Appeal."

76. The decision in relation to these Claimants in the 2009 judgment was given on 8 April 2009 but the hearing was concluded on 19 December 2008, after argument but before the decision in RB (Algeria), and thus that decision was not cited to the Divisional Court.
77. In Brown and Others [2009], the Court was faced with an argument on behalf of Ugirashebuja that by reason of the terms of the MoU "a different, and lower, test than that of flagrant denial of a fair trial" applied (paragraph 26). The Court rejected that, concluding that:

"… it is not to be supposed that the Secretary of State intended that the MoU would require a different test or standard for the fair trial guarantee than, advisedly, the Appellants would enjoy under ECHR Article 6 mediated in this context by Section 87 of the 2003 Act. Had he done so, he would have introduced an appropriate modification pursuant to Section 194(4)(b)."

78. For the purposes of the hearing before the Divisional Court in 2009, the Secretary of State had made a concession, recorded in the judgment as follows:

"31. In any event it is of the first importance to notice the concession made - plainly rightly - by Mr Lewis QC for the GoR to the effect that if the appellants were brought to trial before a tribunal that was not impartial and independent, that would indeed constitute a flagrant breach of their rights under Article 6; and this is a large dimension in the case the appellants seek to make. In our judgment nothing turns on the epithet "flagrant" in these appeals' particular context if the appellants' whole case on fair trial, or the want of it, is substantially established; for if it is, a flagrant violation will be made out."

79. It is to be noted that the concession made then is now withdrawn. That appears not to have been clear in advance of the hearing before us. It is now clear and, as we shall see, Mr Knowles QC cites authority which he submits means that any such concession would now be at variance with the law. We address this aspect of the case below.
80. The Appellant places particular reliance on the case of Samantha O (Orobator) v Governor of HMP Holloway and Another [2010] EWHC 58 (Admin). It is helpful to consider the facts and legal context of the case. The Claimant was a British citizen serving a term of life imprisonment in the United Kingdom following repatriation from Laos. She had been tried and convicted in Laos for a drugs offence. It was her case that she had been convicted and sentenced in circumstances amounting to a flagrant denial of justice and a flagrant breach of Article 6 of the Convention. As a consequence, she claimed that her continued detention in the United Kingdom was arbitrary for the purpose of Article 5 of the ECHR and was therefore unlawful. The Court consisted of Dyson LJ and Tugendhat J. Mr Fitzgerald QC for the Claimant in Orobator accepted that:

"The bar of 'flagrant denial of justice' is set very high … but he says that there is one feature which stands out and which alone suffices. This is the fact that the court which convicted and sentenced the Claimant was neither independent nor impartial. Mr Fitzgerald submits that this fact alone necessarily means that the Claimant suffered a flagrant denial of justice such that the conviction was not by a 'competent court' within the meaning of Article 5 of the ECHR."

81. The judgment of the Court analysed the facts in terms which make it clear that there was a real lack of independence of the judiciary in Laos. The Court then reviewed authority including RB (Algeria), EM (Lebanon), Othman and Brown and Others.
82. In relation to the decision in Othman, the Court in Orobator observed:

"88. We would make two points about this decision. First, it is not clear whether SIAC would have reached the same conclusion if the lack of independence issue had stood alone. Secondly, even if the lack of independence issue had stood alone, it is not clear how critical was the fact that not only were the judges institutionally lacking in independence and impartiality, but there was evidence that they were in fact influenced, decisively so, by what Colonel Quadafi said both in public and in private."

83. Further, in relation to the decision in Brown, the Court in Orobator concluded:

"92. We observe that the court's conclusion that there was a real risk of a flagrant denial of justice was not based on the simple proposition that lack of judicial independence and impartiality would of itself involve a flagrant denial of justice. The court adopted a more calibrated approach than that. In assessing the seriousness of the lack of independence and impartiality, they took into account the fact that there was evidence of actual governmental interference in particular cases. Furthermore and in any event, it is not at all certain that the lack of independence and impartiality of the judiciary in Rwanda was a decisive factor in the court's overall conclusion stated at para 121. It is clear that the lack of independence and impartiality was not the only feature of the Rwandan justice system which led the court to its overall conclusion.

93. In our judgment, Mr Fitzgerald seeks to place a weight on Brown which it cannot bear. It does not purport to state any general principle. It is no more than an example of a case where the high threshold of "flagrant denial of justice" was satisfied on the particular facts of the case having regard to the situation likely to face persons in Rwanda accused of genocide."

84. In Orobator the Court went on to make observations upon which Mr Knowles QC places considerable emphasis:

"99. It is a striking fact that there is no case of which we are aware in which this test has been successfully invoked in any context in relation to article 6 on the grounds of lack of independence and impartiality of a court. We recognise that judicial independence and impartiality are cornerstones of a democratic society and that their absence will without more involve a breach of article 6. But we cannot accept that lack of judicial independence and impartiality will necessarily involve a flagrant denial of justice or the "nullification or destruction of the very essence of the right guaranteed" by article 6. Whether the lack of independence and impartiality has that effect must depend on the particular facts of the case, examined critically as a whole. Regrettably, there are many states throughout the world where judges are less independent and less impartial than they are in the UK and other democratic societies which are fully committed to the rule of law. But even where the judiciary are not fully independent and impartial, it is possible for a trial to take place which does not involve the complete nullification or destruction of the very essence of the right guaranteed by article 6."

85. In the course of Orobator, Mr Fitzgerald argued there were other features which, taken together, demonstrated a flagrant denial of justice. At critical points of the procedure, the Claimant had no legal representation. She was "frequently threatened and intimidated into making statements", was denied "a proper opportunity to prepare her case" and "had incompetent legal representation by a lawyer who was not independent of the Executive". On the facts in the case itself, the Court rejected that there had been a flagrant denial of justice. However, in the rather different context with which they were dealing, the Court was assessing the procedural failures and, unsurprisingly, doing so on the specific facts. Ms Orobator admitted being in possession of the drugs in question. Her defence was necessity, or duress. The case was not complicated. She did have sufficient access to her legal representative for him to appreciate the nature of her defence and to present mitigation. The legal representative:

"was able to present a written submission and to supplement his written material with oral submissions so as to deal with the issues in the case. In particular, he was able to advance the claimant's defence of necessity or duress."

On that basis, the case came:

"nowhere near amounting to a nullification or obstruction of the very essence of the right guaranteed by Article 6."

86. In respect of admissions made by the Claimant, the Court was not informed as to what admissions the Claimant had made in statements signed allegedly under duress. Further, there was nothing in the judgment of the Court to suggest that its decision was based on the content of her statements, beyond the "uncontroversial admission" that she was in possession of the drugs. The Court found there was no basis for a claim of a really significant failure of presentation of her case.
87. In our judgment, there is force in the submission from Mr Fitzgerald to us, when considering Orobator, that a retrospective examination of whether there was in fact a "flagrant denial of justice", in a completed case where the facts are known, may be significantly different from considering whether there is a real risk of a flagrant denial of justice in the future. The one is history, the other an assessment of future risk.
88. Mr Fitzgerald also submits that the observation by Dyson LJ in paragraph 99 of the judgment quoted above, that there was no case that the Court was aware in which the test had been successfully invoked in relation to Article 6, arose simply because the Court had not been fully informed. Mr Fitzgerald was able to cite to us a considerable number of decisions where this was the outcome. A number of these were first instance decisions, thus Termerko (SDJ Workman, December 2005), DD and AS v SSHD (SIAC, April 2007), Izmaylov and Another v Russia (SDJ Workman, December 2008), Kononko (SDJ Riddle, 2015), Alpha Bah v Gambia (DJ Purdy 2012), Yasar Afsar v United Arab Emirates (DJ Zani, 2013).
89. In June 2012, the ECtHR handed down their judgment in Ahorugeze v Sweden (Application No. 37075/09) which was a decision relating to extradition to Rwanda. The Court reviewed the Strasbourg case law and set out the relevant principles of law very shortly as follows:

"113. According to the Court's case-law, an issue might exceptionally arise under Article 6 by an extradition decision in circumstances where the individual would risk suffering a flagrant denial of a fair trial in the requesting country. The principle was first set out in Soering v. the United Kingdom (cited above, § 113) and has been subsequently confirmed by the Court in a number of cases (see, for instance, Mamatkulov and Askarov, cited above, paragraphs 90-91).

114. The term "flagrant denial of justice" has been considered synonymous with a trial which is manifestly contrary to the provisions of Article 6 or the principles embodied therein (see, among other authorities, Sejdovic v. Italy [GC], no. 56581/00, paragraph 84, ECHR 2006-II).

115. It should be noted that, in the twenty-two years since the Soering judgment, the Court has never found that an extradition or expulsion would be in violation of Article 6. This indicates that the "flagrant denial of justice" test is a stringent one. A flagrant denial of justice goes beyond mere irregularities or lack of safeguards in the trial procedures such as might result in a breach of Article 6 if occurring within the Contracting State itself. What is required is a breach of the principles of fair trial guaranteed by Article 6 which is so fundamental as to amount to a nullification, or destruction of the very essence, of the right guaranteed by that Article.

116. In executing this test, the Court considers that the same standard and burden of proof should apply as in the examination of extraditions and expulsions under Article 3. Accordingly, it is for the applicant to adduce evidence capable of proving that there are substantial grounds for believing that, if removed from a Contracting State, he would be exposed to a real risk of being subjected to a flagrant denial of justice. Where such evidence is adduced, it is for the Government to dispel any doubts about it (see, mutatis mutandis, Saadi v. Italy [GC], no. 37201/06, paragraph 129)."

90. The final authority said to be of significance is the decision of Kapri v Her Majesty's Advocate (for the Republic of Albania) [2014] HCJAC 33, a decision of the Appeal Court, High Court of Justiciary, of Scotland. That case concerned the alleged corruption and lack of independence prevalent in the Albanian judiciary. The judgment of the Court was given by Lord Carloway, Lord Justice Clerk. The Court reviewed the European and English authorities, including most of those considered in this judgment. The Court also touched on the decisions of Dzhaksybergenov v Ukraine and Dzhurayev v Russia (2013) 57 EHRR 22, and the reasoning of the Court continued:

"122. …Although the general situation is relevant, it is necessary to look at the particular circumstances of the suspect and not just at the general situation (see also Sharipov v Russia (no 18414/10), 11 October 2011, at paras 33-37; Yefimova v Russia (no 39786/09), 19 February 2013). Thus, an allegation that any suspect in Kazakhstan or Tajikistan runs the risk of ill-treatment is too general. It is only in "the most extreme cases", where the general situation is such that any removal to the particular country would necessarily violate a Convention right, that the court should proceed purely on the basis of generality (Dzhurayev v Russia (supra at para 153).

123. It was this type of reasoning which influenced the court in deciding that the evidence about the general situation in Albania was irrelevant, in that the material produced contained nothing to suggest a concern in the appellant's particular case. It is this area which the court now requires to review. The court assumes that the United Kingdom Supreme Court was not disagreeing with the European Court in Dzhaksybergenov v Ukraine (supra) when it stated (para 32) that "systemic corruption in a judicial system affects everyone who is subject to it" and that thus "No tribunal that operates within it can be relied upon to be independent and impartial". Rather, it must be asking this court to consider whether the general circumstances in Albania are so extreme as would inevitably lead to a violation in the appellant's and any other person's case. If that is correct, it would theoretically be sufficient for the appellant to demonstrate that the general situation in Albania, is in the "most extreme" category, thus avoiding the need for any reference to the particular circumstances of the appellant's own case. This would result in a "total ban" (Dzhaksybergenov (supra) para 37; Shapirov (supra) at para 35) on extradition to Albania. However, conversely, presumably if it is positively demonstrated that a particular person will in fact receive a fair trial in Albania, the general situation in Albania cannot fall into the "most extreme" category, since it would have been demonstrated that the extradition of the particular individual would not result in a violation. It must therefore follow, if that were demonstrated, that extraditions to Albania as a generality will not necessarily violate a Convention right.

124. The court is, accordingly, acutely conscious that its remit may be a curious one. It must, first, determine whether there is "systemic" corruption in the Albanian judicial system on the hypothesis (which this court must accept) that, if there is, it must affect everyone involved in that system (including the appellant) on the basis that no "tribunal that operates within it can be relied upon to be independent and impartial". Secondly, it must determine whether there are substantial grounds for believing that there is a real risk of a flagrant denial of justice in the appellant's case if he were returned to Albania. That does not require an appellant to prove the existence of a risk on the balance of probabilities. He merely has to show "substantial grounds" and leave it to the respondent to dispel any doubts. It would seem, however, that the evidence demonstrating the grounds has to be "compelling" in the form of something approaching an international consensus."

Our Conclusions on the Law

91. As a number of the authorities have made clear, the general question is not controversial: is there a real risk of a given Respondent suffering a flagrant denial of a fair trial in Rwanda? This is a much narrower test than simple conformity with the standards required by Article 6. The Convention is essentially territorial in jurisdiction. Lord Carswell in Ullah reminds us that Member States cannot export the standards and obligations to other States not signatories. On the contrary, Member States have international obligations, in this context most often derived from extradition treaties, in this instance from the MoUs.
92. The burden of raising such an objection to extradition lies on the Requested Person, although if an objection of substance is raised, it is for the Requesting State to counter the objection. The standard of proof is probability. It is a projection of risk: the burden on the RP is to show, to the standard of probability, that a real risk of the specified kind arises. As the criticism of DJ Evans in Brown (paragraph 34) makes clear, the height of the bar is not derived from the standard of proof, or the establishment of "real risk", as opposed to straight probability: the "high test" lies in the degree of denial of fair trial if the risk eventuates.
93. The rarity of cases where extradition has been rendered unlawful by reference to Article 6 does not, of course, amount to any kind of presumption against such a bar, much less a rule of law. It does serve to emphasise how significant must be the denial of justice in question, before an effective bar is raised.
94. The speech of Lord Phillips in RB (Algeria) (paragraph 133) is a useful guide. The consideration of the risk of denial of justice must go beyond the procedural. Article 6 is a procedural right. The Court should be giving consideration to the outcome of a breach, if it eventuates. Here, a significant adverse outcome, in terms of the nature of conviction and sentence, cannot be in doubt.
95. The GoR having withdrawn their earlier concession, we must approach the application of the legal test without reference to that concession, and to that extent we are differently positioned from the Divisional Court in 2008/2009. We have recited above a number of the judicial dicta relied on suggesting that a lack of independence or impartiality on the part of the Court or Tribunal before whom an extraditee will be tried will not, on its own, establish a flagrant "denial of justice". Indeed, Mr Knowles goes so far as to say that the Divisional Court in Brown fell into error on this point. We reject Mr Knowles' proposition.
96. The first point is that we do not understand the Divisional Court in Brown to have reached such a conclusion. We agree with the Divisional Court in Orabator: the Court in Brown had in mind a number of factors affecting trial in Rwanda, much in the way we have understood SDJ Arbuthnot to have done in her decision.
97. The second point is that we cannot accept that prejudice or bias in a tribunal or judiciary, no doubt almost always arising from political or other pressure, can never amount to a flagrant denial of justice. That would seem to us an unwise proposition, bound eventually to be confounded by a bad case. What can be said, reflecting the thinking of Lord Phillips in RB (Algeria), is that where proper procedure, arrangements for witnesses, and representation are all available, it may often be that the effects of a lack of independence on the part of the tribunal will be sufficiently mitigated by such other adequate features of trial, so that incursion on fair trial process will fall short of a "flagrant denial" of justice. For the great part, all these aspects of trial will have to be weighed together and an overview reached.
98. The necessary risk will not be established by merely formal badges of lack of independence (military tribunals, judges without security of tenure, a judiciary without some form of inspectorate, a judiciary with a weak or mixed reputation with the public): the risk required must comprise a risk of real substance, a risk of a truly serious denial of justice. Anything short of that would represent an unwarranted export of European Convention standards to States not subject to the Convention.
99. We consider that the Divisional Court in Brown directed themselves correctly, in accordance with the principles properly derived from authority. We too apply an objective test of a real risk of a flagrant denial of justice, and with a potentially very serious outcome. In applying that test we will hold in mind all significant elements of the Rwandan criminal justice system.

The Experts and Legally Qualified Witnesses in the Case

100. Before plunging deeper into the detail of the case, there are one or two matters which should be addressed in order to make clearer what follows. The first is some account of the experts and/or professionally qualified witnesses whose evidence is in the case, or who play important roles. We include those who gave evidence in the proceedings in 2008/2009, since their evidence and the Courts' conclusions have been considered in the later litigation and because in the cases of two witnesses (Nerad and Reyntjens), their evidence was introduced directly before SDJ Arbuthnot.
101. James Arguin: Arguin's views were never before SDJ Arbuthnot (or, of course, the Court in Brown). Although never advanced as a witness as such, and never in Court, Arguin's account, particularly as set down in his report "Rwanda's Capacity to Ensure an Effective Defense in Transferred Cases" dated 22 May 2016 was admitted before us. In addition, his exchanges with Witteveen are relied on. Arguin is a US-trained lawyer who served as a federal and state prosecutor, working with the US Department of Justice and the Massachusetts Attorney General's office on "a broad range of policy matters". That took place over a period of ten years or so. From 1992 to 2000 he was an equity partner and associate with K and L Gates in Boston. He says that he has "qualified for judicial appointment at both the state and national levels in the United States."
102. In 2010 he was appointed as Chief of the Appeals and Legal Advisory Division of the ICTR Office of the Prosecutor in Arusha. During the five-year period ending in 2015 he "directed all litigation relating to [the] referral proceedings" in eight genocide cases including Uwinkindi and Munyagishari. As indicated above, in 2010 the United Nations passed a resolution designed to hasten the closure of the ICTR and to bring to an end its role in prosecuting or over-seeing the prosecution of Rwandan genocide suspects. Mr Arguin's service in the role we have indicated appears to have ended in December 2015 (when the ICTR was closed), since when he has served as Deputy Chief of the Rule of Law Unit with the UN Assistance Mission in Afghanistan. He has never been questioned orally in any case of which we are aware. If he were to be questioned, we consider he would be liable to be pressed whether part of his responsibility was to try to achieve the objectives of the UN Security Council expressed in that Resolution: in other words to transfer the caseload to Rwanda. Securing the transfer of suspects and cases that fell within Article 1(3) to Rwanda within the time frame set was a prime objective. He was part of the prosecution team resisting the revocation of the transfer of the Uwinkindi case to Rwanda.
103. Johnston Busingye: was in 2014 the Minister of Justice and Attorney General of Rwanda. He provided a witness statement supporting the extradition of Brown/Bajinya. He previously attested to political influence over judges.
104. Dr Phil Clark: Reader in Comparative and International Politics, School of Oriental and African Studies [SOAS], and expert in Rwandan politics and genocide-related law. He has spent more than ten years of study, including considerable fieldwork in Rwanda. He was instructed in the current litigation on behalf of GoR, providing a main report dated 19 February 2014, a copy of a report "Testifying to Genocide: Victims and Witness Protection in Rwanda" published in 2012, of which he was co-author, and some of his field notes.
105. Florida Kabasinga: this witness, called on behalf of the GoR, was a Senior Legal Advisor within the International Crimes Unit of the NPPA. She is a dual qualified attorney in the United States and Rwanda, with experience as a Prosecutor with the ICTR. In March 2016 she left the NPPA and set up in private practice in Kigali. Her principal evidence concerns trial in Rwanda, but she gave more recent evidence concerning the failure of disclosure of the Witteveen memorandum. She was present in court for much of the hearing before SDJ Arbuthnot and was acting, in effect, as part of the CPS team. She was not presented as an expert but treated by the Court as a "vehicle" by which documentary evidence might be introduced. She is important in relation to the production of documents and evidence relating to Witteveen and Arguin, and the issue of representation.
106. Professor Timothy Longman: is the Director of the African Studies Centre at Boston University, and a specialist in Rwanda, Burundi and the Democratic Republic of Congo. He has spent considerable periods in Rwanda and in the region. He has focussed on state-society relations, especially ethnic conflict, human rights and, inter alia, served from 2001 to 2006 as Director of Rwanda Research for the Human Rights Center at the University of California, Berkeley. He last visited Rwanda in 2006, but has kept abreast of developments since. In common with Professor Reyntjens, he supports the view that the international community has" preferred to see the RPF government's decent technocratic governance while ignoring its deeply flawed political governance (Longman report 9 December 2013, paragraph 17).
107. Scarlet Nerad: is an American Licensed Investigator, instructed for the defence in the earlier litigation and again in the recent phase of the case. She gave evidence as to the difficulty of obtaining witness evidence, the fears of witnesses and the difficulty of deploying evidence in Rwanda. Her evidence was relied on by the Divisional Court in Brown (see paragraphs 50 et seq).
108. Martin Ngoga: for a number of years Prosecutor General of the Republic of Rwanda. Mr Ngoga is not an expert witness, but is obviously legally qualified and experienced. His deposition was the principal support for the extradition request. He is no longer Prosecutor General. His evidence sets out the case, on behalf of the GoR, that the changes in the Rwandan judiciary and legal system have transformed matters since 2008/2009, suggesting that extradition can now safely be ordered.
109. Professor Filip Reyntjens: Professor of Law and Politics at the University of Antwerp, was instructed in 2008/2009 by Brown/Bajinya. He gave evidence on the nature of the Rwandan state and the operation of the judicial system; he was found to be reliable by the Divisional Court in Brown, and is relied on still by the Appellants. He is persona non grata in Rwanda, and has not been able to return there since 1994. His account is that that situation arises because he has been critical of the RPF and their successive governments. The GoR is heavily critical of him as an alleged associate of the Habyarimana government.
110. Professor Philippe Sands QC: then of University College, London, Professor of Public International Law, gave evidence in 2009. His report was based on publicly available source material as well as interviews he conducted. He was not involved in the later litigation.
111. Professor William Schabas: with a chair at the National University of Ireland, Galway, called on behalf of the GoR, was the only expert to give evidence to the earlier Courts that there could be safe extradition to Rwanda. He was extensively cross-examined. It was he who advanced the view that the decision of the ICTR as to whether to transfer cases to the Rwandan Courts should be regarded as a "litmus test". He was heavily challenged by Ms Ellis QC, inter alia, as to whether he had fully disclosed material (including his own notes) relevant to the proceedings, and attacked as to whether he had given partial or misleading citations. The Divisional Court found there was substance in these suggestions (see Brown, paragraphs 109/118). Although some reliance has been placed by the Appellant on some of his observations, he was not called in the later proceedings. The earlier proceedings in relation to this witness emphasise the need for full disclosure by an expert.
112. John Bosco Siboyintore: a legal adviser in the NPPA, and colleague of Florida Kabasinga. He was at material times head of the Genocide Fugitive Tracking Unit ("GFTU")
113. Martin Witteveen: is a Dutch lawyer and investigative judge, with long experience of international criminal work. In June 2014 he became an Advisor on International Crimes to the National Public Prosecution Authority ["the NPPA"]. He was instructed on behalf of the GoR. The sequence and thrust of his reports is of significance. His first report of 19 September 2014 (three months' post-appointment) was succeeded by an Addendum report of 19 May 2015. He then drafted an Additional Expert Report of 3 June 2015. Crucially, he then sent a memo dated 1 June 2016 recording further views, unfavourable to the case of the GoR, to James Arguin (q.v.). This memorandum was never disclosed to SDJ Arbuthnot. Despite that, his views as to the inadequacies of the system of trials, and in particular the arrangements for defence representation, were crucial in the decision of SDJ Arbuthnot to decline extradition. In his final Memorandum, Witteveen records that he has been warned by three separate individuals not to return to Rwanda, and that "some Rwandan authorities have called me a traitor" (memo 1 June 2016, paragraph 16).

The "Fair Trial" Issue:

(A) The political system and the courts

114. In paragraphs 156 and 157, SDJ Arbuthnot accepted the conclusions of the Divisional Court in 2009 as to the connected nature of the political system and the Courts, quoting part of paragraph 68 of the Divisional Court judgment:

"68. Moreover the question whether a court is independent and impartial cannot be answered without considering the qualities of the political frame in which it is located."

She went on to consider some aspects of the political system and turned first to the "table of violations" prepared by counsel, which was available in an earlier form to the Divisional Court in 2009 for the purpose of the Magistrates' Court hearing to July 2015. We have been able to consider the updated version of these tables.

115. It is worth emphasising how this table of evidence was compiled. This was explained to us by Ms Ellis QC. A draft of the table was provided to the GoR who were "invited to provide for inclusion in these bundles any further documentation which falls within this category". The Appellant provided no such documentation, nor has there been any evidential basis advanced for questioning the material included. We regard that as significant.
116. This table was available for the expert witnesses called on either side, who could comment if desired. In addition, SDJ Arbuthnot was able to hear oral evidence from a number of the witnesses involved in the events tabulated.
117. SDJ Arbuthnot found this material helpful in setting the context for her examination of the Rwandan criminal process in general, and the independence of the judiciary in particular. She found that the table reflected accurately what is to be found in the different source documents (paragraph 166). She entered a limited degree of qualification for the material, accepting that the source of some of the information, although coming from respected bodies such as the US Department of State, Human Rights Watch, the UN Committee Against Torture and Amnesty International, might be secondary or based on hearsay. She accepted that there was a certain amount of cross-reference, in that some matters of complaint could be found under more than one sub-heading in the table. Despite these qualifications, she concluded that:

"…overall the tables paint a clear and compelling picture."

118. She correlated the table of violations with the views of the expert witnesses as to the up-to-date position in relation to the Rwandan political position. Dr Clark, called by the GoR, accepted the substance of the picture painted, and accepted that: "…the regime was authoritarian" as he said "in the broadest possible sense – you could argue over some of the specifics". He also accepted that there have been "extra-judicial killings and disappearances" and that there was a "shrinking political space in Rwanda", including "a clamp- down on opposition parties, on critical media and local human rights groups and on the local population expressing critical views". It is important to emphasise these observations date from the recent hearing, not 2008/2009.
119. Dr Clark was positive about some aspects of the developments in Rwanda, in particular the living standards of the population, and political stability. SDJ Arbuthnot found, based on his evidence, that "that the cost of this was a reduction in a number of basic freedoms".
120. Martin Witteveen on behalf of the GoR, whilst disclaiming that the focus of his evidence was directly concerning the political situation in Rwanda, accepted that Rwanda was a repressive and autocratic regime.
121. For the defence, two experts gave evidence on the point. Professor Longman had not been to Rwanda since 2006 but made it clear he had kept up to date through multiple contacts inside the country. He had acted as a prosecution witness in cases against Rwandan nationals. SDJ Arbuthnot found some of his conclusions important, and we think it helpful to include the following paragraphs from her judgment dealing with Longman's evidence and that of Professor Reyntjens:

"210. Professor Longman in his first report dated 9th December 2013 said that Rwanda had gone through an ostensible transition to democracy with the reality that power had become more centralised and the government more authoritarian (Page 4, Paragraph 13). Power had increasingly concentrated in the hands of the RPF, Tutsi and Anglophones (Paragraph 14) with President Kagame increasing his personal power.

211. Professor Longman in his report at Page 12, made the telling observation that "The public and official policies of the government often appear moderate and consistent with standards of democracy and human rights, but the realities in practice are quite different, revealing a regime that maintains strict control and is deeply authoritarian. For example, de jure power may be shared among parties and ethnic groups, but de facto power is firmly in the hands of Tutsi from the RPF".

212. I noted that Bertelsmann Stiftung in their 2014 report (referred to above) at 'Objective' Bundle 9, Pages 54 onwards, at Page 61 summarised the system as "a skillfully designed institutional facade that conceals the real distribution of power. All major political and power-related matters are decided by the president, together with his key advisers".

213. Professor Longman recognised that the GoR had achieved much, driven by their "Vision 2020" policy that Rwanda will transform from an agricultural to a knowledge based economy by 2020. It will do this by "promoting good governance and efficient administration, training the population for work in sciences and information technology, developing the infrastructure and encouraging private business" (Paragraph 15). The government is fighting corruption and Rwanda has moved from 158th to 58th in terms of the ease of doing business.

214. Professor Longman recognised the many positives of the Rwandan regime whilst pointing out that in his view the international community is being manipulated and conned into seeing the positives of good technocratic governance whilst ignoring the authoritarian RPF's grip on politics. Professor Longman explained that a key element of the RPF's strategy to maintain strong support from the international community is by justifying many of its policies by referring to the genocide and its legacy. It has suppressed discussion about its own violence against civilians as it took power and in its invasions of Congo. The RPF justifies its behaviour by saying that it is preventing another genocide. The restrictions on free speech and the laws banning "divisionism" and "genocide ideology" are said to be because of a necessity to prevent extremism from gaining strength in Rwanda (Paragraph 21).

215. In this expert's view since 1999 the RPF had given up the extensive violence it used to control the population after the genocide to using more subtle means to silence it and keep it fearful (Page 6, Paragraph 19). Laws had been passed to regulate speech, judicial processes are used to silence critics, opponents are arrested whether they are journalists or politicians. Professor Longman's evidence was that all the human rights organisations have either been taken over, coerced or closed. The press is similarly controlled. His evidence is that the population from the "most local communities to the highest levels of political and civil society, feel threatened and fearful". [Emphasis of SDJ Arbuthnot]

216. Professor Reyntjens (VB File 1, Page 1 onwards) was another expert who had not been able to go back to Rwanda since 1994 when he was critical of human rights' violations carried out by the new RPF Government. For his information he relies on a network of researchers in Rwanda as well as academics based there. He did not name them at their request and explained at Page 5 "Both they and I hold the view that to identify them would likely have an adverse impact on their personal security and their ability to undertake the work they do".

217. Reyntjens said the regime exercises strong control within the country. Opponents are arrested, they disappear or they flee the country. He described the government as highly autocratic and authoritarian (Page 10, Paragraph 20) and said matters had deteriorated since his first statement for the 2007 proceedings. He held the strong view that the political landscape had closed even further since 2010. He points out that in the Democracy Index 2012 of the Economist Intelligence Unit, Rwanda now occupies the 132^nd place out of 148."

122. The table of violations is divided into differing sections. Table A records 15 examples of extra-judicial killings and/or inadequate investigations into such deaths, between 2005 and 2007. The updated schedule records 21 episodes from 2008 through to 2015. In each case the episode may refer to a single or to multiple deaths. SDJ Arbuthnot singled out one or two episodes by way of illustration in her judgment, but looking at the underlying material it is clear she could have made other selections of equal potency. There is a clear pattern to be discerned from the episodes digested. A preponderance of those who have died or disappeared have been political opponents of the existing government, journalists who have been critical of the government, or former associates of the President and the government who have fallen out of favour.
123. In paragraphs 170 and 171 of her judgment, SDJ Arbuthnot picks out the case of Patrick Karegeya, the former Rwandan Intelligence Chief found murdered in South Africa in early 2014. He had fallen out with President Kagame and been granted political asylum in South Africa in 2009. He became an opposition party supporter. As SDJ Arbuthnot indicates, there were hostile statements from a range of Rwandan government ministers following his killing, culminating in a speech on 12 January 2014 when President Kagame said:

"Whoever betrays the country will pay the price, I assure you. Letting down a country, wishing harm on people, you end up suffering the negative consequences."

The President denied responsibility for the death but said that "you should be doing it", which in effect meant that others should be killing traitors on behalf of the country. In an interview with the Wall Street Journal he again denied Rwanda had killed Karegeya but added "I actually wished Rwanda did it".

124. On 17 January 2014, the US State Department spokeswoman stated that President Kagame's comments in relation to the death of Karegeya were viewed with "deep concern" by the United States, who were "troubled by the succession of what appeared to be politically motivated murders of prominent Rwandan exiles".
125. From Table B, SDJ Arbuthnot chose the example of Emanuel Mughisa, alternatively known as Emile Gafirita, a former Rwandan soldier who was due to give evidence to a French judge in December 2014. His anticipated evidence was that President Kagame was responsible for the shooting down of the aeroplane of President Habyarimana in 1994. He never gave evidence. In November 2014 Gafirita/Mughisa was kidnapped in Nairobi. The Kenyan authorities say he had not been arrested by them. He has not been seen since.
126. Another example from Table B is that of Theogene Turatsinze, the former managing director of the Rwandan Development Board. He was found tied up, dead, floating in a lake in Mozambique, in October 2012. The table commentary includes the following:

"Domestic political observers commented that Turatsinze had access to politically sensitive financial information related to certain Rwandan government insiders."

127. President Kagame's former personal physician, Dr Emmanuel Gasakure was killed in a police station in Kigali, Rwanda in February 2015.
128. Not all of the episodes reported involve prominent individuals. For example, Human Rights Watch in May 2014 reported that an increasing number of people have been forcibly disappeared or reported missing since March 2014, including 14 people forcibly disappeared in the Rubavu district, with further cases elsewhere. According to Human Rights Watch, in at least eight of the Rubavu cases there were indications that state agents were involved. In June 2014, a US Government press statement expressed concern about the arrest and disappearance of "dozens of people". Sixteen people who had been reported missing appeared before a court in Rubavu in the same month accused of endangering state security.
129. Table C of the schedule addresses episodes of torture or inhuman or degrading treatment. The updated table beginning at May 2011 and including episodes up to June 2014 sets out a number of episodes of such abuse. They include a report by Amnesty International, documented and accepted by the United States Department of State, relating to the torture or other inhumane treatment of detainees by military intelligence and other state personnel in 2010 and 2011. These episodes were said to have involved a considerable range of torture techniques. Another example drawn from the schedule is the arrest in September 2012 of two university students attempting to deliver a petition to the office of the Prime Minister. The episode was again reported by the State Department. The students were, it is said, detained in solitary confinement without food or water for two days by police in Kigali and beaten with metal rods. Although torture was reported to a judge, no investigation of their allegations ensued.
130. Human Rights Watch reported in January 2015 on the trial in Kigali before a Military Court of 16 defendants. They included a defendant named Mutabazi, who had been abducted from Uganda. Mutabazi and several co-defendants stated in Court they had been tortured and forced to sign statements.
131. SDJ Arbuthnot heard evidence via video link from Mutabazi's wife and brother. Mutabazi had worked for President Kagame as Commander in Charge of Security at the Presidential Palace in Rwanda. He had been arrested in 2010, detained and tortured. Taken to hospital he was seen by his wife, who confirmed his position. After a period he was released to home detention. The couple left for Uganda and moved to a safe house provided by the United Nations. It was from that house he was abducted on 25 October 2013 and taken back to Rwanda. Despite his claim of torture before the Military Court, he was convicted of various offences and in October 2014, sentenced to life imprisonment.
132. SDJ Arbuthnot, having the benefit of direct evidence from Mutabazi's wife and brother, found their account of the consequences of torture "compelling".
133. Table D in the schedule sets out a number of examples running up to 2015 of illegal detention and of detention in military camps with prisoners being held incommunicado. Conditions were often said to be very poor, with frequent accounts of mistreatment by police or other guards. The United Kingdom Foreign and Commonwealth Office, as part of the Human Rights and Democracy Report of 2014, noted that "dozens of people" have been arrested and held incommunicado in the northwest of Rwanda: "the UK regrets that legal due process was not followed". Evidence before the United States House Committee on Foreign Affairs and an account from the Washington Director of Human Rights Watch dating from 2015 both confirm a considerable range of arrest and detentions involving "scores of cases" of unlawful or unofficial detention, on occasion for several weeks or months.
134. SDJ Arbuthnot gives specific examples (paragraphs 184 and 185), including that of Peter Erlinder, a US lawyer arrested in May 2010 as he entered Rwanda in order to act as a defence lawyer at the trial of the former presidential candidate Victoire Ingabire. He was arrested on accusations of "genocide denial". When the trial took place in the spring of 2013, others were arrested outside the courtroom while demonstrating in her support. They were charged with public order offences and "contempt of public officials".
135. Table F of the schedule addresses the ways in which press freedom has been restricted. In paragraphs 188 to 193 of her judgment, SDJ Arbuthnot considered this area. One of the most striking examples is set out in paragraph 190. SDJ Arbuthnot there addressed the events of July 2013 concerning a group known as LIPRODHOR, one of the last remaining human rights advocacy organisations in Rwanda. The group was taken over, ousting the existing leadership. The ousted president of LIPRODHOR challenged the takeover legally. He then received threats of violence from the Rwandan Governance Board, a state body responsible for relations with civil organisations. In the middle of the dispute in June 2014, the New Times newspapers, a supporter of the Kagame government, stated that Human Rights Watch was an organisation that supported the Democratic Forces for the Liberation of Rwanda ["FDLR"]. FDLR is an organisation with a military wing opposing the GoR and operating in the Eastern Congo. SDJ Arbuthnot dealt with this episode at some length, her concerns being firstly that the human rights organisation was subverted and secondly that there was a pro-government attempt to undermine or remove authority from Human Rights Watch.
136. SDJ Arbuthnot also quoted a critical passage from the Human Rights Watch World Report of 2015 which reads:

"…the Government continues to impose severe restrictions on freedom of expression and association and does not tolerate dissent. Political space is extremely limited and independent and civil society remains weak. Real or suspected opponents inside and outside the country continue to be targeted. The ruling Rwandan Patriotic Front dominates all aspects of political and public life. Opposition parties cannot operate in a meaningful way."

137. We find the material reviewed very highly concerning as to the context in turn in which judges must operate in Rwanda.

(B) Osman Warnings to Rwandans in the United Kingdom

138. Between paragraphs 194 and 201 of her judgment, SDJ Arbuthnot addressed the evidence she heard of two Rwandan exiles living in this country who had been given "Osman" warnings by British police. The first, Rene Mugenzi, made critical remarks in a BBC radio phone-in programme in 2011, in which President Kagame was a participant. The President was "clearly furious" at this intervention, which accused the GoR of human rights abuses and of crushing the opposition. A few weeks later Mugenzi received a visit from two members of Special Branch, telling him, as the judgment puts it, "they had credible and reliable information from the Security Service that the GoR was trying to assassinate him".
139. The second witness was Mr Marara. He was a former close protection officer of President Kagame. After an error organising cars during a visit by President Kagame, he was sent to prison for 15 days. He fled Rwanda and came to the United Kingdom. In 2009 or 2010 he was given an Osman warning by Special Branch. Following the death in January 2014 of Karegeya, Marara was approached again by Special Branch and advised that he should change his name, which he has done. As he came to the Westminster Magistrates' Court to give evidence, he stated that a person from the Rwandan embassy had photographed him.
140. In June 2015, President Kagame gave a speech when he mentioned both Mugenzi and Marara. In the context of discussing the arrest of another Rwandan (Lieutenant General Karake) in London on a European Arrest Warrant, President Kagame complained that the arrest of Karake was intended to show contempt for Rwanda and to destabilise Rwanda. President Kagame blamed "criminals" from the United Kingdom who gave advice to those who arrested General Karake. President Kagame named Mugenzi "and another thug from our army who ran away, a criminal called Marara".
141. The observations by SDJ Arbuthnot in paragraph 200 were stark:

"I find that the UK authorities had received information that led them to believe that the GoR was threatening to kill UK citizens on UK soil. I was not in a position to see the evidence relied upon by the police before they gave those warnings but I had to accept there was sufficient concern for the police to take what is a very unusual step in each of these three situations. If the police were correct in their concerns, it is not a satisfactory state of affairs that a foreign government thought it appropriate to plan to kill those taking refuge here at a time they were seeking to take advantage of a memorandum of understanding brought about by diplomatic ties between the two countries, which they hoped would lead to extradition."

142. In our view, the remarks by SDJ Arbuthnot could have been even more trenchant.

Fair Trial:

(C) Rwanda and the Rule of Law - Gerald Gahima and Theogene Rudasingwa

143. Part of the evidence bearing on the independence of the judiciary came from two brothers, both formerly senior members of the RPF party and of the Rwandan State. Each gave significant witness statements, and each appeared before SDJ Arbuthnot. Dr Gerald Gahima and Dr Theogene Rudasingwa are brothers born into a prominent Tutsi family. Their father was killed in Rwanda at the end of the 1950s and the family left for Uganda. Dr Rudasingwa trained as a medical doctor. Dr Gahima trained as a lawyer, in Uganda and then at the London School of Economics, before entering private legal practice in Nairobi.
144. In 1990, both brothers gave up their professional engagement to work full time for the Rwandese Alliance for National Unity, the organisation which became the Rwandese Patriotic Front, the party now led by President Kagame. Each of them was associated with President Kagame from that era. Dr Rudasingwa rose to become the Secretary General of the RPF and subsequently the Rwandan Ambassador to the United States. In his witness statement, Dr Rudasingwa, amongst other things, outlines the degree of personal control exerted by President Kagame over the institutions of Rwanda, even before he became President. On his return from the United States, Dr Rudasingwa became chief of staff to Mr Kagame. He describes the position of Mr Kagame at that point in the following terms:

"By mid-1999 Paul Kagame was the ultimate boss of all security institutions, the army, the police, intelligence and was chairman of the RPF. The speaker of parliament, Joseph Sebarenzi, had been forced to resign; Seth Sendashonga had been forced into exile and he too was assassinated just after it had become know that he proposed testifying for the defence at the ICTR in Arusha; Theoneste Lizinde, also forced into exile was then assassinated. Paul Kagame held meetings with United States, United Kingdom and German ambassadors and claimed that [former President] Pasteur Bizimungu was corrupt. He had to resign in 2000 and Paul Kagame became president in his place. Any pretence of power sharing had gone. Almost the entire workforce at the presidency was Tutsi. The Hutu were marginalised and held no positions of power and influence."

145. Dr Rudasingwa fell from grace with President Kagame between 2004 and 2005, and left the country in April 2005 to become a refugee in the United States. He maintains, he says, close links with Rwanda and remains fully informed about developments within the country. His account of the country is that:

"I watched Rwanda develop a political system which maintains a false appearance of multi-party democracy whereas in reality no opposing political views or criticism are tolerated. Whilst in Rwanda I was close to the centre of power for many years and this has allowed me to gain considerable insight and understanding of the way in which the country is governed. The Rwandan Patriotic Front, under President Kagame, has engineered changes which have led to Rwanda becoming a one party secretive police state with a façade of democracy and in which there is no space for political participation and where political opposition is not tolerated. The Rwandan Patriotic Front ensures it maintains its monopoly of power by use of draconian measures to restrict the exercise by Rwandan citizens of their fundamental human rights. The media, civil society and opposition political parties are not permitted to operate openly and freely."

146. SDJ Arbuthnot recited Dr Rudasingwa's evidence, and the GoR's response in relation to fair trial for those who held the post of bourgmestre, as follows:

"Dr Rudasingwa … stated it was not possible for a former bourgmestre to have a fair trial (EN Volume 4, Tab 22 Page 235 and in cross examination on 16^th June 2014). Dr Clark the GoR's expert did not deny the point."

147. Dr Gahima's career paralleled that of his brother, but as a senior lawyer it brought him closer to the legal and judicial systems in Rwanda. At the time of the invasion of Rwanda by the RPF, he was a civilian member of the executive committee of the RPF. He became successively Deputy Minister of the Ministries of Public Service and of Justice, the Prosecutor General and the Vice President of the Supreme Court of Rwanda. He describes himself as "the principal policy maker on matters relating to accountability for genocide and justice sector reforms between 1994 and 2004". From being appointed Deputy Minister of Justice in 1996, he led the reconstruction of the justice sector. In his witness statement, he dealt in some detail with the system of gacaca courts, and we will return to his evidence on that topic later in this judgment. He, too, has had extreme difficulty with President Kagame. He left the country in 2003. He was subsequently tried in absentia by a Military Court in Kigali in 2011. He was accused alongside the three other authors of a document entitled "Rwanda Briefing", which constituted an analysis of political governance in Rwanda and an assessment of the leadership of President Kagame. He summarises its conclusion in very similar terms to his brother, Dr Rudasingwa, who was one of the co-authors. The conclusion was that Rwanda has become "a hard-line, one-party, secretive police state with a façade of democracy".
148. Any Court will be cautious in assessing the evidence of those who have become public critics and political opponents in exile from a situation such as this. It is not easy to tease out all of the relevant political twists and turns, so as to make a detailed assessment of the witness and whether the evidence advanced is fair. However, in this instance there is a helpful judgement on the balance and fairness of Dr Gahima from the Appellant's expert witness Dr Clark. As SDJ Arbuthnot recorded in her judgment at paragraph 467:

"I noted that Dr Clark considered that Dr Gahima was a respected academic and when he lived in Rwanda had been dedicated to cross ethnic dialogue."

Dr Clark also accepted that whilst Dr Gahima was in Rwanda until 2003/4 he had an intimate knowledge of the judicial system (evidence of Dr Clark, 17 June 2015) and in her judgment, SDJ Arbuthnot accepted that Dr Gahima had kept in close contact with a number of people still in Rwanda.

149. SDJ Arbuthnot also accepted the written submissions made by Ms Ellis QC, as containing an accurate summary of the evidence of Dr Gahima before her. It is helpful to include here the critical passages from that summary:

"i. In Rwanda … the RPF is itself controlled by the military and security services and it exercises absolute control over all branches of government including the judiciary;

ii. The judiciary, prosecution service and law enforcement agencies are dominated by RPF members and more especially Tutsi, particularly those who came from Uganda to Rwanda in 1994;

iii. Organs which appoint judges, prosecutors and police officers responsible for criminal investigations merely ratify decisions which have been made internally by the party leaderships, upon the advice of state intelligence services;

iv. A large number of judges, including the majority of judges of the most senior courts, are active RPF members;

v. Judges and prosecutors I know from personal experience are compelled to pay financial contributions to the RPF party and to participate in its activities;

vi. The primary obligation of the judges is to the party, not the state or the people of Rwanda. The party has the power to remove them from office at will;

vii. When I was Prosecutor General, I became aware of many attempts by the Executive to influence prosecutorial and judicial decisions;

viii. Local government, security and military officials […] interfered with the independence of the judiciary and the prosecution service;

ix. Even the highest levels of the leadership of the judiciary and prosecution service were not immune from attempts to interfere with their independent functioning;

x. On many occasions, President Kagame and security services personnel have overruled the decisions of the judiciary and have removed judges from office for failing to toe the line;

xi. The Courts have been used as a mechanism to silence opposition leaders and government critics such as the former President Pasteur Bizimungu, Charles Ntakirutinka, Victoire Ingabire, Bernard Ntaganda, Deo Mushayidi, Dr Niyitegeka, Nkusi Uwimana Agens and Mukakibibi Saidata;

xii. The judiciary is vulnerable to political manipulation … in cases where the executive has an interest; and

xiii. While I believe that many in Rwanda's judiciary are persons of integrity, the judiciary is not independent. The executive and particularly the security services, wish to influence the outcome of a case then they can and will do so because members of the RPF have been placed in all strategic positions of leadership and are often called upon to assign cases to judges willing to do what the government wants done.

(b) Dr Gahima also gave evidence of two specific attempts which were made to interfere with his independent functioning as Prosecutor General. The first incident related to Pasteur Bizimungu as set out above and was made on behalf of President Kagame himself. The second attempt at interference was made by James Musoni (President Kagame's personal assistant on RPF matters) and Colonel Jack Nziza who asked Dr Gahima to arrest, detain and institute criminal proceedings against several opposition politicians in order to prevent them from participating in forthcoming elections. Although Dr Gahima declined this request, the police nonetheless opened fake criminal investigations against most of the politicians in question, arrested them and instructed the politicians to report for questioning on many days during the course of the campaign period. Both of these individuals still hold high office in Rwanda."

Fair Trial:

(D) Political or merely Prominent Cases?

150. We consider that the material we have digested above, taken together, justifies a bleak view of the political context within which the Rwandan judiciary are obliged to operate. Certainly, were these Respondents to be thought of as political opponents of the Rwandan government, a very high level of anxiety would be justified as to their prospects of a fair trial. As SDJ Arbuthnot made clear (see paragraphs 522-526 and 545 of her judgment), it is an important consideration whether the Respondents fall into such a category, or whether they will be regarded as "merely" alleged genocidaires, and of less interest to the authorities.
151. Mr Jones for Brown/Bajinya in particular, argues against this distinction, a distinction supported by the Appellant. Mr Jones submits that the absence of direct evidence of interference does not mean that interference is absent: interference is clandestine. Mr Jones cites some of the conclusions of the judge herself. In paragraph 479 she noted the evidence of Dr Clark (for the GoR) who was told by Minister of Justice Busingye that "executive interference has continued" in trials (judgment paragraphs 479-480). The previous Minister of Justice Karugarama confirmed that "some politicians might ring judges but they just ignored the calls …" (paragraph 481). President Kagame gave a speech on 1 March 2015 claiming that "corruption even in the judiciary continues as a way of business…"
152. Dr Clark interviewed the ombudsman Tito Rutaremara, who stated there were "far fewer complaints of political interference in trials" (judgment paragraph 479), but the notes of evidence of Dr Clark confirm that he was not asked his opinion about the range of cases prone to interference, or whether such interference extended to genocide trials as opposed to those with direct political implications [Bundle 14, p6930].
153. The Respondents submit that genocide trials in Rwanda are "inherently political" and that the description of them as having merely a "political flavour" is inadequate. Professor Reyntjens' view was that genocide trials are indeed political, and we see no logical basis for rejecting his view out of hand, coming from an expert whose evidence was heavily relied on by the Divisional Court in Brown [see judgment in Brown paragraphs 71, 94 and 99].
154. Mr Jones cites the remark of the Appeals Chamber of the ICTR in Prosecutor v Munyakazi, speaking of "politically sensitive cases, such as genocide cases" [Munyakazi judgment, paragraph 26], particularly where the cases have "an international profile and where they involve allegations against people who held political office under the previous regime and/or who are alleged to have been part of the former 'inner circle'. The SDJ agreed with the evidence of Professors Longman and Reyntjens that the cases of these five Respondents would be "high profile", see judgment paragraph 527. Someone alleged to have been part of the "Akazu" elite would draw attention as would the position of a former Bourgmestre. However, the Respondents go on to criticize the distinction made by SDJ Arbuthnot between these trials and political cases, and her observation that:

"The GoR will use the return of the Respondents as examples to show that the state's justice system is recognised internationally as a system which can try defendants fairly [judgment para. 527]."

155. SDJ Arbuthnot was clear that she could make the necessary distinction between political cases and genocide trials by looking at the five transferred cases and others tried in Rwanda in the years since 2009. She had read the extensive monitors' reports on these trials, read the decisions and considered the expert evidence in regard to the trials, particularly that from Witteveen. Before returning to that aspect of the reasoning of SDJ Arbuthnot, it will be helpful to consider the evidence in relation to the transferred cases, as it was before her, and as it is now, and to consider the other defence evidence and the expert evidence, as presented in the hearing.

Fair Trial:

(E) The Transferred Cases

156. We have identified above the cases which have been transferred to Rwanda for trial since 2009. The principal cases are as follows. Uwinkindi was transferred by the ICTR Referral Chamber on 28 June 2011, a decision that was confirmed by the Appeal Chamber on 16 December 2011. In his case there were subsequent revocation proceedings that are of significance. Bandora was extradited from Norway by virtue of a decision of the Oslo District Court on 11 July 2011, as confirmed by the Borgarting Appeal Court on 19 September 2011. Munyagishari was transferred by the ICTR Referral Chamber on 6 June 2012, a decision that was confirmed by the Appeal Chamber on 3 May 2013. Mugesera was made the subject of a deportation order by the Canadian government with which, on 23 January 2012, the Federal Court declined to interfere by way of judicial review or to prevent Mugesera's deportation. A further application, instigated by the UN Committee against Torture, was dismissed by the District Court of Montreal on 23 January 2012. Mbarushimana was made the subject of an extradition order by the Danish High Court on 22 March 2013, a decision confirmed by the Danish Supreme Court on 6 November 2013. In addition, SDJ Arbuthnot considered the trials of Ntaganda and Ingabire.
157. These cases are significant because they represent the track record of the Rwandan Court, and they were the subject of comment by the experts, in particular by Witteveen and Arguin. Later in this judgment we analyse that debate closely, and those passages should be read alongside the broader survey which follows here.
158. In paragraphs 221 to 227 SDJ Arbuthnot deals with the trial in 2010/2011 of Bernard Ntaganda. He was the leader of an opposition party, on the point of registering to stand as a Presidential candidate for election in 2010. He was arrested, kept in custody, and then tried for genocide ideology, divisionism, creating a criminal organisation and threatening national security. It should be noted that "genocide ideology" in Rwanda need not involve a denial of genocide by Hutus on Tutsis, but can be established by the assertion that killings of Hutus by Tutsis were racially motivated.
159. There is no need to repeat all the details of this trial, since they are summarised in the judgment and analysed more fully in a Note by counsel for Brown/Bajinya. It gives sufficient flavour to say that the Defendant was prevented from cross-examining prosecution witnesses and prevented from calling defence witnesses. Two key prosecution witnesses claimed that their evidence was false, having been obtained from them by threats and pressure. Despite their retractions, the Rwandan Supreme Court declined to admit their revised evidence or consider it on the appeal.
160. None of this is contradicted by the GoR. In his response to the written submissions of the Respondents, Mr Knowles QC made the matter plain in his summary of the judge's findings: "most if not all of the examples of unfair trials in Rwanda are in connection to political opponents of the GoR" (GoR's reply, paragraph 28(b)). His principal submission is that these Respondents "are not political opponents of the regime that the regime fears" (Skeleton Argument on the Cross Appeal, paragraph 128).
161. SDJ Arbuthnot next considered the trial of another Presidential hopeful, Victoire Ingabire, who gave oral evidence to her by video link from prison in Kigali. Here the judge considered the most reliable account of events came from Amnesty International. Ms Ingabire was tried in 2011 to 2012. One of the key charges against Ms Ingabire was based on a speech she delivered on 16 January 2010. This was said to amount to "genocide ideology". The critical passage reads (in translation) as follows:

"It is clear that the path to reconciliation is long. It has a long way to go because if you look at the number of people who were killed in this country, it is not something that one can get over quickly. But still, if one looks around, one realises that there is no strong political policy designed to help the Rwandan people achieve genuine reconciliation. For example, if we look at this memorial, it only refers to the people who died during the genocide against the Tutsis. There is another untold story with regard to the crimes against humanity committed against the Hutus. The Hutus who lost their loved ones are also suffering; they think about the loved ones who perished and are wondering "When will our dead ones also be remembered?"

For us to reach reconciliation, we need to empathize with everyone's suffering. It is necessary that for the Tutsis who were killed, those Hutus who killed them understand that they have to be punished for it. It is also necessary that for the Hutus who were killed, those people who killed them understand that they have to be punished for it too. Furthermore, it is important that all of us, the Rwandan people, from our different ethnic groups, understand that we need to unite, respect each other, and build together our country in peace."

162. The case was characterised by outspoken political comments hostile to the defendant before and during her trial, including from President Kagame himself, such that it was said to be "impossible" for the judges to acquit. In addition to Amnesty International, Human Rights Watch and the European Parliament found the trial to be unfair. The judge recited the account from Amnesty International that the trial judges showed "signs of hostility and anger" towards the defendant, interrupting her regularly. The defendant and her lawyers withdrew from the trial, late in the process, after an incident where a defence witness's cell was searched and documents removed, after he had given evidence. As with Ntaganda, key prosecution evidence came from witnesses who were credibly believed to have come under pressure to testify. On appeal to the Rwanda Supreme Court, the Court quashed one conviction, but convicted Ingabire of treason with intent to undermine the existing government, and imposed sentences of nearly double the length (15 years instead of 8 years) passed below.
163. We have referred above to the treatment of the American lawyer Peter Erlinder, as he travelled to Rwanda to assist Ms Ingabire.
164. SDJ Arbuthnot's conclusions relating to Ingabire were as follows:

"266. The conclusions I have drawn from the evidence of the conduct of the High Court trial is that the Court was not fair to Ms Ingabire and that political considerations may have been at play. It is clear that the President and the GoR were very concerned about the prosecution against Ingabire but at the same time there is no evidence that the judges were directly interfered with by the executive.

267. I cannot exclude the fact that the High Court judges may have been reacting to Kagame and his ministers' repeated comments, the press reporting and other factors that are unknown, when they convicted her. These comments undermined the presumption of innocence in her case and I give more weight to the influence that ministers' comments may have in Rwanda than I would give to such comments in a genuinely democratic country. I noted, against the above, that the High Court did acquit her of four out of the six charges she faced. The Supreme Court although doubling her sentence and convicting her of a new charge, did not uphold all aspects of her conviction. The Supreme Court judgment is detailed and I have no evidence that that Court was unfair in its conclusions other than it is upholding (as indeed the Court must) the law of minimizing genocide, which interferes with free speech and did so, in Amnesty's view, on this occasion. "

165. SDJ Arbuthnot spent a considerable time focussing on the trial of Jean Uwinkindi. This case occupies paragraphs 282 to 370 of the judgment, in addition to later comments from SDJ Arbuthnot about the quality of Uwinkindi's representation. The referral decision from the ICTR to the Rwandan Courts was itself hotly contested. One of the principal planks advanced by Uwinkindi against transfer was that defence witnesses were afraid of harassment, imprisonment or even death for them or their relatives if they gave evidence in defence. This was countered by the prosecution, relying on the fact that very many witnesses from Rwanda had testified for the defence at the Rwandan tribunal and no witness had subsequently raised security concerns with the ICTR witnesses and victims support section. The ICTR judges were concerned "not with whether the fears of potential witnesses were legitimate, but with whether the accused would be able to secure the appearance of defence witnesses and thus obtain a fair trial" (judgment, paragraph 287). As to the impartiality and independence of the Rwandan judiciary, the ICTR found the Rwandan higher judiciary to be "suitably qualified and experienced and have the necessary skills" (judgment, paragraph 291). The prosecution relied on the acquittal rates in Rwanda.
166. The ICTR Chamber also clearly distinguished between political cases and genocide charges such as those faced by Uwinkindi. An appeal to the Appeals Chamber of ICTR failed. Uwinkindi arrived in Rwanda in April 2012. Thereafter, the progress of Rwandan litigation could be examined at least to some degree through the reports of the ICTR monitors, as well as Rebuttal Material produced by Ms Kabasinga and the additional report from Mr Witteveen (judgment, paragraph 295). The proceedings commenced in June 2012 but the trial itself did not begin until 14 May 2014.
167. The Senior District Judge considered that the monitors' reports give a clear and accurate picture of the criminal proceedings in Rwanda, but she qualified this by saying:

"The monitors rarely make comments on what they see, so are not reliable sources when the quality of work is being considered." (Judgment, paragraph 297)

168. The nub of Uwinkindi's complaints is that the majority of his witnesses for the defence are and would remain outside Rwanda. In order to achieve a fair trial he required legal facilities sufficient to find the witnesses he required, proof them and organise their evidence to be delivered within Rwanda. Had his case remained with the ICTR, such resources would have been available. This picture was confirmed to SDJ Arbuthnot by one of the ICTR monitors (Ms Buff) whose monitoring reports addressed the matter and established that 74 per cent of the defence witnesses called in the 57 cases tried at the ICTR lived outside Rwanda when they testified. Ms Buff also commented to the Senior District Judge that she had reviewed the submissions made by the GoR and the KBA and concluded that:

"…there is no evidence that these institutions have anticipated the need to pursue investigations outside Rwandan territory." (Judgment, paragraph 301)

Nor could the GoR point to any defence witnesses coming from abroad called in Rwandan cases, other than before the gacaca courts. Thus the question arose whether the resources were available that would permit investigations abroad and the obtaining of evidence abroad.

169. There was a further issue about the Rwandan system for obtaining evidence within the country. The system is that the judicial police, at least in theory, conduct investigations both for the prosecution and for the defence, performing the functions that an investigative judge would perform in the civil law system. When Ms Buff (the ICTR monitor) raised this with Uwinkindi, he told her "that he would rather die than provide the names of his witnesses to the prosecution" (judgment, paragraph 302).
170. In this context, SDJ Arbuthnot went on to deal with the two systems of witness protection or witness contact. We have touched on that already in the course of this judgment. The system run by the State (the WVSU) was fully functioning but not trusted by defence witnesses. At the time of the Uwinkindi proceedings, Mr Buff noted that the second witness protection unit, initiated by the judiciary (the WPU) "was in a fledgling state and had yet to be set up" (judgment, paragraph 304).
171. SDJ Arbuthnot went on in her judgment to outline the extended twists and turns of the dispute between Uwinkindi, the Minister of Justice of Rwanda, Uwinkindi's lawyers and the Courts. It is not sensible to try and reproduce that here. The nub of it is that Uwinkindi's first lawyers wished to be instructed on terms comparable to those which apply in the ICTR, which would allow resources for obtaining evidence from abroad. An hourly work rate was agreed initially, but the Rwandan authorities insisted on a change to those arrangements because they were too costly. The financial arrangements were reduced in November 2014, and in December 2014 the Rwandan Ministry of Justice adopted a new legal aid policy, with the effect of a further reduction in the funding for defence representation. Moreover, the new contract included a controversial Article 6, which provided for a unilateral cancellation of the representation contract in the event that "counsel make any statements aimed at discrediting the Government or the Ministry in the course of their work, either to the press or during the trial". Uwinkindi's first team of counsel declined to accept the agreement and the Ministry of Justice terminated their contract with a notice period of three months, during which the lawyers were expected to go on assisting the defendant but would be paid at the reduced rate (judgment paragraph 331).
172. To that point no funds had been released for gathering the testimony of witnesses abroad. The lawyers objected to Article 6, and the lawyers withdrew, leaving Uwinkindi unrepresented. In January 2015, Uwinkindi asked for an adjournment, but the High Court ruled that the proceedings should continue. His existing team of lawyers were present for a period but after an adjournment did not return to the courtroom. The High Court found the lawyers had done this in order to delay the trial and they were promptly fined (judgment, paragraph 336).
173. In January 2015, new lawyers were appointed. It is not clear if the contested Article 6 was placed in their agreement. However, Uwinkindi refused to cooperate with them since they were not of his choosing. During March 2015, these new representatives remained in court but did not participate in the hearing, during a period in which fourteen prosecution witnesses and nine defence witnesses gave evidence without being questioned, either by Uwinkindi or his new defence lawyers. As SDJ Arbuthnot observed in paragraph 340, this meant that all the prosecution witnesses had been heard without being questioned on his behalf. After further months without decisive action, Uwinkindi asked the MICT to revoke the decision to refer his case. Before that was concluded his trial continued. The High Court decided his new counsel should continue to represent him but that the witnesses who had been heard in March should be re-heard and granted an adjournment for preparation. In October 2015, the application to revoke the transfer failed.
174. The application to the ICTR concerned the progress of the case. It was said by the Rwandan government that by November 2014 Uwinkindi's defence team had received 83 per cent of all the budget available for all transfer cases (judgment, paragraph 350). The trial chamber found that:

"equality of arms did not require material equality between the parties. Rwanda had introduced legal aid programmes and had provided a budget of 100 million RwFrs for all such transfer cases. The 2014 15 million RwFrs (about £12,500) allocated to counsel for the whole case did not include fees for additional defence investigations" (judgment, paragraph 351).

Such investigations had to be addressed through a new practice direction. SDJ Arbuthnot requested a copy of that and it was supplied. The practice direction requires a number of detailed steps to justify any expenditure abroad. There is no evidence that it has been successfully operated.

175. SDJ Arbuthnot notes (judgment, paragraph 357) the evidence given to the English Court by Witteveen as to the capacity of defence lawyers in Rwanda to conduct such cases and the judge continued in paragraph 358 as follows:

"The defence make the valid point that 'equality of arms' connotes a value judgment as to the adequacy of defence assistance available to the defendant and MICT has not grappled with that problem. Another concern raised by the defence is that MICT notes that the judicial police will be responsible for gathering evidence for the prosecution and defence. I agree with the defence in this respect that I have heard no evidence about how this police force functions and I do not consider the defence witnesses with the sort of fears I have heard about will be able to be marshalled by the judicial police." (Paragraph 358)

176. We consider below the evidence of Martin Witteveen as to the defence capacity in Rwandan trials.
177. SDJ Arbuthnot went on to consider the evidence as to whether the Court in Uwinkindi's case displayed any bias or prejudice. She did so after entering a notable (and thoughtful) caveat:

"The defence would argue that pressure happens behind closed doors; it can be consciously applied or applied through comments made in the press by politicians and others. It is very rare that a situation occurs, such as in the case of Bizimungu, when a judge confesses that he was pressurised into giving a particular verdict. The defence is of course right. Inappropriate pressure is almost impossible to expose and the only possible way to look at whether it is being applied in the Uwinkindi case is [to] consider the way the case is being conducted by the parties in particular by the judges. I saw no sign of anything of out of the ordinary in the hearings that were described by the monitors."

Having noted the reports of the monitors, SDJ Arbuthnot recorded that she had read the summaries of the judgments in the High Court and Supreme Court on Uwinkindi and there was nothing in those to cause her any concern.

178. In considering this conclusion, Mr Moloney QC for Munyaneza, who focussed on behalf of the Respondents on the Uwinkindi trial, points to the mandate to the monitors given by the President of the MICT. In dealing with the decision on the Registrar's Submission in the Uwinkindi and Munyagishari cases, 15 November 2013, the President of the MICT stated:

"I accordingly consider that the monitors in the Uwinkindi and Munyagishari cases should limit themselves to providing objective information relevant to any possible violations or impediments to the fair trial rights of Mr Uwinkindi and Mr Munyagishari in their reports, and refrain from including in their reports any opinion, assessment, or conclusions regarding such violations or impediments unless otherwise directed."

179. Mr Moloney also submits that SDJ Arbuthnot fell into error when assuming that the MICT Trial Chamber did not have Mr Witteveen's report. In fact he submits that it is clear they did, as paragraph 40 of the MICT Appeal's Chamber decision of 4 October 2016 reads:

"40. As to the Witteveen report, the Appeals Chamber observes that the Trial Chamber did not expressly refer to it in the impugned decision even though it was included in written submissions and was raised in Uwinkindi's arguments before the Trial Chamber…. There may be an indication of disregard when evidence, which is clearly relevant to the findings, is not addressed in the Trial Chamber's reasoning … The Appeals Chamber finds that the Trial Chamber erred in failing to address the Witteveen report in the impugned decision." (Paragraph 40, Prosecutor v Jean Uwinkindi MICT-12-25- R14.1) (cf paragraph 357 Arbuthnot)

180. As we set out below, the Uwinkindi case was important in the judge's reasoning that defence representation was inadequate. Before turning to the next case in the sequence of those under consideration, we record the GoR's criticisms of her approach to Uwinkindi.
181. The GoR suggest that a review of the papers from the MICT Chamber would demonstrate that "defence counsel's approach to defence funding is a tactical approach in an attempt to frustrate the trials. The SDJ failed to consider this crucial aspect when considering the issue of defence representation" (Appellant's Skeleton, paragraph 46).
182. The GoR rely upon the findings of the MICT Chamber: (a) the right of an accused to be represented by a counsel of his own choosing is not absolute, Article 14(3)(d) of the ICCPR; (b) Uwinkindi had not shown that it was unreasonable for the High Court to appoint new counsel where counsel of his choosing "failed to attend at two consecutive hearings"; (c) Uwinkindi failed to show that newly-appointed counsel had insufficient experience or that others were unavailable; (d) Uwinkindi was not justified in failing to cooperate with his newly-appointed counsel; (e) the High Court did reconsider its decision to proceed with the trial in March 2015 and decided to recall witnesses who had been heard whilst Uwinkindi was not represented; (f) "the Chamber recalled that an accused does not have the right to claim a breakdown in communication through unilateral actions in the hope that it will result in the withdrawal of counsel or in Uwinkindi's case revocation of the referral order".
183. The message from the MICT judgment is that Uwinkindi has been "obstructive in his own proceedings".
184. In relation to defence investigations, the GoR submits that SDJ Arbuthnot was in error in concluding funds were not available for such investigation. The Practice Direction of August 2015 provides for such investigations, subject to detailed application and authorisation. The flat rate for legal representation does not include defence investigation, which either is to be separately funded under the Practice Direction or is to be performed by the Judicial Police on behalf of the defence. The GoR also submits, in broad terms, that when the MICT reached its decision, its proceedings were informed by more information than was possessed by Martin Witteveen when he gave his evidence before SDJ Arbuthnot.
185. The case of Uwinkindi arises again in relation to the "debate" between Witteveen and Arguin, as expanded by the material which became available after the conclusion of the hearing before SDJ Arbuthnot. We address this below, after analysing that debate, and the fresh material, in detail.
186. The next case addressed by SDJ Arbuthnot was that of Charles Bandora. He was extradited from Norway in 2011, following a hearing in the District Court with a one day hearing in June 2011 and a decision in July and a decision of the Appeal Court in September 2011. Bandora argued ECHR Article 6. A police superintendent gave evidence in reliance on a report he himself prepared, following a number of trips to Rwanda. Some other extraneous material (for example, reports from Human Rights Watch and Amnesty International) were before the Court. The first instance Norwegian decision in relation to fair trial was dealt with in two pages of their judgment. Making reference to the changes in the Rwandan legal system, the Court concluded there was "no longer" any ground for rejecting the application for extradition.
187. In paragraph 376, SDJ Arbuthnot records how the Court in Bandora gave great weight to the decision of the ICTR in Uwinkindi. She noted in particular that the police superintendent described how 149 witnesses had been seen and there was no indication that any were reluctant or fearful. She goes on to say "unfortunately the judgment is silent as to the precise number of defence witnesses interviewed or what they were told about giving evidence". The Norwegian First Instance Judgment was short and "a little lacking in detail". The matter went on appeal, and the Borgarting Court of Appeal focussed on whether the correct test of law was applied by the lower Court, concluding that it was.
188. Bandora arrived in Rwanda in March 2013 and appeared first in the Rwandan High Court in November 2013. In relation to his case as it developed in Rwanda, the Senior District Judge relied considerably on the evidence of Martin Witteveen. We have outlined his background earlier in this judgment. It was in dealing with Witteveen's evidence in the Bandora case, that SDJ Arbuthnot made clear her view of his capacity as a witness and the importance of his evidence overall. It is worth quoting the following paragraphs:

"386. It was clear from his evidence that this was a witness with great experience of investigations in Rwanda. I found him to be a credible, reliable and compelling witness. He was compelling as he was clearly an objective witness who told me what his recent experiences of the trial system in Rwanda were. His evidence was also useful as he had conducted many investigations in Rwanda and he explained to the court his experience of dealing with witnesses in genocide cases.

387. Mr Witteveen's first report was dated 19^th September 2014. This concerned his experience of interviewing witnesses and his views of the independence and impartiality of judges in Rwanda. He explains in the section "Scope of this report" that his written evidence is based solely on his experience and he tries not to rely on secondary evidence. He also makes it clear in Paragraph 16 that he is going to refrain from making observations on the human rights and political situation in Rwanda for two reasons, firstly it is not his expertise but secondly he endorses the GoR's approach that he does not see how observations on human rights and the political situation in Rwanda can help the court to decide whether VB and others will get a fair trial. With respect to a witness that I found persuasive in many ways, the High Court in Brown and others disagreed with him and so do I.

388. Overall the tenor of Mr Witteveen's first report was positive in relation to Article 6 and it was clear that he considered that any alleged genocidaire returned would have a fair trial. In terms of the trials of Bandora and the others returned to Rwanda, the second report he wrote on 3^rd June 2015 is more relevant. His second report was written at his own instigation and served by the GoR days before he was due to give evidence.

389. In the months since his first report in September 2014, he had attended a number of court proceedings in the five transfer cases. The purpose of the additional report was to provide an update on the topics he had addressed in his first report as well as to provide an expert opinion on the status and work of the defence attorneys in the transfer cases currently in the High Court in Kigali. He emphasised in his Paragraph 8 that his intention was not to criticise but to assist Rwanda in building their justice system.

390. His concern as expressed in Paragraph 14 onwards was in relation to the status and quality of the defence lawyers acting for the defendants transferred to Rwanda:

"I have a deep concern on the status and quality of the defence attorneys acting for their clients in the genocide transfer cases. In the cases I witnessed, none of the defence attorneys performed at a level that meets any international standard. In summary: in some cases there is currently no defence, either officially or materially, in other cases the defence attorneys act or acted substandard and even irresponsible (sic)".

391. His evidence is based on a number of visits to court hearings in relation to all five transferred men. Importantly his expert evidence is not only based on what he saw but also on the notes made by the legal officer of the Dutch Embassy in Rwanda who had attended almost all the trial sessions between September and December 2014. A local member of staff accompanied them and translated for them and made typed notes of the proceedings."

189. When considering Bandora's trial in Rwanda, the judge noted that he paid for two defence lawyers until his money ran out, and then applied for legal aid. The trial took place between September and December 2014. Twelve prosecution and fourteen defence witnesses gave evidence and Witteveen was very critical of the defence performance. Cross-examination of prosecution witnesses and examination in chief of defence witnesses was "chaotic" (judgment, paragraph 394); the defendant led much of the questioning without guidance or direction from his attorneys (judgment, paragraph 394); the lawyer's questioning was unstructured without strategy, often talking over each other and being interrupted (paragraph 395); Bandora and his lawyers repeatedly mentioned the name of a protected witness who was supposed to remain anonymous (paragraph 395).
190. A major problem was that the defence failed to explore matters of critical importance to the prosecution. Witteveen advanced the example of two prosecution witnesses, Hakizimana and Baziga, an example which clearly impressed SDJ Arbuthnot. Both these witnesses had incriminated Bandora in gacaca proceedings but then retracted their statements in the High Court. Each said they had been visited by prosecutors in prison and had been promised release if they testified against Bandora. They also said they had given evidence against Bandora in the gacaca appeal "because they were forced by businessmen to do this".
191. The GoR make a considerable attack on their own witness Witteveen's evidence on this point and on SDJ Arbuthnot's assessment of how the matter was handled. The suggestion is that there may be "various reasons" why defence counsel failed to explore the series of retractions and assertions about retractions: the detail of the allegations appears to be consistent since these witnesses gave accounts in similar terms to investigators and prosecutors in different years. The GoR suggests "this may well be why the defence counsel in Rwanda left the issue alone". In other words, the suggestion is of a considered and sensible tactical approach.
192. Counsel for Brown/Bajinya, dealing with this issue on behalf of all the Appellants, puts the matter differently. Hakizimana provided two witness statements in November 2010 and February 2011. He did not implicate Bandora in either of those statements. On 4 June 2013, Hakizimana provided a third statement in which he accused Bandora of ordering killings and collaborating with the Interahamwe. In March 2014, he was questioned about a letter he had written in 2008, in which he said he had given false evidence against Bandora at the gacaca proceedings. He said he had done so because he had been pressurised in prison. It is established that each of these witness statements had been taken by a prosecutor for the GoR, named Faustin Nkuzi. It was Nkuzi who was prosecuting in Court in September 2014 when Hakizimana gave evidence.
193. In his evidence in Court, Hakizimana failed to implicate Bandora. Under questioning from Faustin Nkuzi he said "those who are here today representing the public prosecution" had promised him release from prison if he implicated Bandora. This whole area was not explored by the prosecution, by defence counsel, or by the Court. In the judgment convicting Bandora, the trial Court makes no mention of this evidence and gave no explanation as to why it rejected the repeated evidence of Hakizimana that he had been pressurised into a wrongful implication of Bandora.
194. SDJ Arbuthnot comments (paragraph 397) that the GoR served material from the Bandora trial in rebuttal but "unfortunately it was not put to [their own witness] Mr Witteveen". The material shows "an evolution in the evidence of one of the witnesses, from saying he was pressurised to give false evidence against Bandora in 2008 … to accusing him of direct killings when interviewed by the prosecutor whom he said had induced him to falsely accuse Bandora". The judge also noted that the second significant witness, Baziga, "also said he had been induced into giving false evidence, his evidence follows a similar pattern to that of the first witness, Hakizimana" (paragraph 397). It clearly was important in SDJ Arbuthnot's mind that the GoR had not gone through this in detail with Martin Witteveen, if it was to test his judgment about the capacity of the defence counsel in the trial. As she observed in paragraph 398, Witteveen found it "incomprehensible that the defence attorneys did nothing with this information". In the ensuing paragraphs of her judgment she recited further evidence from Baziga and another witness as to allegations of bribery, leading to "virtually no enquiry". She also recited Witteveen's concern that only a limited number of witnesses gave evidence, even though the names of "many other possible witnesses were mentioned yet no attempt had been made by the defence to get them to Court". Witteveen also explained that there had been no effort made to contact witnesses abroad and the defence lawyers had "not even asked MiniJust for a budget to investigate the defence case". The judge accepted the evidence of Witteveen, summarising the conclusion in paragraph 402:

"402. I accept the evidence of Witteveen; he had seen some of the trial and for anything that he had missed he had been able to rely on the contemporaneous account of it made by someone from the Dutch embassy. According to Witteveen, the defence lawyers had neither the experience nor the skill to ensure that Mr Bandora received a fair trial. Nothing, however, that Witteveen said led me to conclude in that case that the judges were not independent or impartial nor that defence witnesses were reluctant to give evidence."

195. We return to this conclusion below. Clearly the GoR relies on the judge's remarks about the impartiality of the bench. Clearly, too, we keep in mind the fact that the Rwandan legal system is in essence a civil law system, and the role of attorneys may be rather different than would be expected in a common law Court. However, we record our concern here that this points to the obligation of the Court to investigate such obviously relevant material if the defence do not. We find it hard to accept the proposition that SDJ Arbuthnot was wrong to rely on Witteveen's conclusions about this trial, on the basis that the inaction of the defence was tactical rather than proceeding from incompetence. He himself comes from a civil law jurisdiction, made a judgment of these attorneys and the system which supported them, which was a general judgment proceeding from a good knowledge base.
196. It is also worth recording that Witteveen's evidence was moderate in its tone. The Appellant relies on the fact that Witteveen would not say directly that the representatives of Bandora were incompetent. He did not seek to overstate in any way yet leading counsel for Munyaneza emphasises the following question and answer from Witteveen:

"Q: If it is dire, if they are not taking obvious points, if they are walking out of cases in a huff, the situation is plainly that you will not get the effective representation of counsel? Would you accept that?

A: I agree."

197. Between paragraph 403 and 435 of her judgment, SDJ Arbuthnot dealt with the case of Munyagishari, another man charged with genocide, sent to Rwanda by the ICTR for trial. He was arrested in May 2011 and appeared before the ICTR in June. The hearing before the Referral Chamber included presentation by the defence of affidavits from sixteen anonymous witnesses who said they were afraid to testify in Rwanda. The Chamber found the defence affidavits "prima facie credible" (judgment, paragraph 409), but found it was beyond their role to determine whether the fears were well-founded. The Chamber was satisfied as to safeguards in Rwanda. The Referral Chamber also held (paragraph 410) that the legal aid system in Rwanda was adequate. They did so on the basis of availability of counsel and the existence of the legal aid payment scheme. The Chamber also found that the lack of independence of the judiciary in Rwanda was "speculation" and the Chamber relied on the Rwandan law and on the constitution.
198. They were referred to the events surrounding the trial of Victoire Ingabire, which they concluded were "unsubstantiated" (judgment, paragraph 411), a conclusion far from the picture before us. The Chamber also found that the appointment of international judges to try the case "was a realistic possibility". On that issue, SDJ Arbuthnot agreed with the GoR's witness, Dr Clark, "that such an appointment is very unlikely for the reasons he gave, one amongst which is national pride" (judgment, paragraph 411). The ICTR also drew comfort from the fact that monitoring would take place in relation to ICTR referred trial and there was at any stage the possibility of recall from Rwanda, both of which safeguards are absent in the case of extradition (judgment, paragraph 412). Further, SDJ Arbuthnot accepted the submission that the ICTR decision in this case (in common with the other cases) was taken without the "benefit of the quantity, quality and type of evidence that is before this Court" (judgment, paragraph 412).
199. One significant point is that the Referral Chamber only approved sending the case to Rwanda, on condition there was a written guarantee that Munyagishari would be assisted by counsel with previous international experience. This was appealed by the prosecution and their appeal was allowed. Munyagishari was transferred to Rwanda without such a guarantee. SDJ Arbuthnot found this particularly troubling (judgment, paragraph 415).
200. Paragraphs 416 and following of the judgment detail the prolonged and stuttering conduct of the case against Munyagishari in Rwanda. He arrived on 14 July 2013. At the time of the hearing in the instant case below, the Munyagishari proceedings were still at a preliminary stage. In June 2015 disputes about representation were still ongoing. Defence lawyers had not turned up for a hearing on 3 June. Witteveen gave evidence that there were no negotiations which might lead to a contract with defence lawyers and:

"Witteveen's view of Munyagishari's case is that by and large he is defending himself, his lawyer is in court usually quiet, apparently acting pro bono and the only contributions he makes are in relation to a few procedural issues and his complaint about the contract." (judgment, paragraph 433)

201. As a consequence, SDJ Arbuthnot "questioned the professionalism of those defending Munyagishari" (paragraph 437). Great delay had arisen from the refusal of the GoR to contemplate proceedings either being in French or to grant a translator. Munyagishari states that he cannot follow the case in Kinyarwanda. The case of Munyagishari led SDJ Arbuthnot to the following conclusion:

"438. When it comes to payment for the defence lawyers, I cannot say without more that the GoR is being intransigent but I consider that the GoR had not anticipated how expensive these sort of cases are to investigate and defend adequately. As the complaints to the ICTR multiplied and in the light of the comments of Witteveen, who was after all the GoR's own witness, the GoR have bowed finally to the pressure and changed the law in relation to investigators. This should have happened before when this issue was raised by other defendants such as Uwinkindi in 2011 at the ICTR and in 2012 in the High Court in Rwanda. The change in the law is so recent there is no evidence how the new law will be applied in practice. I have to put in the balance the unfortunate fact that there are numerous examples of the law in Rwanda being changed but being very slow to be put into practice (video-links are but one example).

439. I concluded that the initial reported budget the GoR had set aside for such cases was clearly inadequate and would not begin to cover the sort of investigation required by these genocide cases."

202. Between paragraphs 442 and 461 of the judgment the Senior District Judge deals with the case of Dr Leon Mugesera. As with others of the transferred cases, this defendant's proceedings were still current in Rwanda at the time of the hearing below. SDJ Arbuthnot noted that this case has been the subject of detailed written submissions from Ms Ellis QC, which have also been before us. Those submissions occupy twelve pages, are fully referenced and cross-referred. We have found them highly persuasive, as it is clear did SDJ Arbuthnot.
203. Mugesera faces trial on a range of genocide related offences, but it is not alleged that he took part in the actual commission of genocide. The centre of the allegations spring from a speech he made in Rwanda in 1992, two years before the large-scale killings. Dr Mugesera was deported to Rwanda from Canada. Assurances were given to the Canadian Court before his deportation. On 24 December 2009, Martin Ngoga gave an assurance that the case, if proceeding by deportation rather than extradition, "would qualify as a transfer of case from Canada", thus giving rise to the safeguards attendant on a transfer from the ICTR. In addition there were specific assurances about legal aid. A Note set out that the accused should apply to the judge for legal aid (judgment, paragraph 441).
204. Ms Ellis QC and the judgment set out the sequence of events over an extended period. Legal aid was refused to this Defendant. The GoR claims that the Defendant and his lawyers refused to fill in the right form. On this issue, SDJ Arbuthnot found:

"I cannot say from the evidence where the fault lies. It is quite clear that the most recent legal aid form was not provided to the accused until July 2014 when the case had been running since January 2013." (Judgment, paragraph 458)

205. The conclusion by the SDJ that this process involved a "competition in stubbornness" (paragraph 457) must be seen in that light. The SDJ went on to conclude:

"459. The other failings I note from the evidence I have read and heard is that the prosecution in Rwanda have failed to provide the documents required for his defence to [Mugesera's attorney] and crucially she was not told that the 2009 amendments to the Transfer Law would apply to a case of deportation. The failure too to pay for translation costs of the documentation also means that foreign lawyers would be unable to represent Mugesera. Kanas, the American lawyer originally instructed to defend Mugesera, explains in a statement read to this court the problems with the lack of funding. They could not find an expert to examine the recording of a speech which was given by Mugesera and is at the centre of the allegations against him; furthermore they could not hire investigators.

460. It seems to me that with the reluctant defence witnesses I have heard about from Ms Nerad the funding to allow for a professional investigation is crucial.

461. The evidence I have read and heard leave me with considerable reservations about whether the assurances given by the GoR to Canada are being respected and whether Dr Mugesera's trial is fair."

206. For our part, we have seen no satisfactory answer to these conclusions from the GoR.
207. SDJ Arbuthnot dealt more swiftly with the case of Mbarushimana. This Defendant was extradited from Denmark, arriving in Rwanda on 3 July 2014. He was presented with a list of counsel which was different from, and much shorter than, the list which formed part of the contract between the Ministry of Justice and the RBA. Over the ensuing year he had a number of hearings before the High Court. He was still in dispute as at June 2015 and was then not yet represented. SDJ Arbuthnot was unable to reach a conclusion as to whether the lack of representation was the fault of the government or the Defendant. She did observe that the judges in the case appeared to be dealing with this case in a fair and sensible way, granting adjournments. Counsel for Ugirashebuja made submissions in relation to this defendant and were able to add, from open source material in July 2016, that the prosecution had commenced opening arguments in the case but Mbarushimana had not accepted assigned counsel and appeal to the Rwandan Supreme Court was pending. This case is clearly the least significant of those considered in detail by the judge.

Fair Trial: Other Defence Evidence

208. Between paragraphs 467 and 492 of the judgment, SDJ Arbuthnot summarises a number of witnesses for the Respondents. We have already analysed the evidence of Dr Gahima and Dr Rudasingwa. The judge heard from Philippe Larochelle, an ICTR defence advocate who had defended in a Rwandan genocide case in Canada ending in 2013. The Court found coordinated fabrication by the prosecution witnesses. The scale of the case (and by implication the scale of such cases as these Respondents) was indicated by the activity of that defence team, which consisted of three lawyers and three investigators who had met more than 500 potential witnesses around the world. Mr Larochelle indicated that witnesses were intimidated on a wide scale, the intimidation ranging from speeches during annual commemorations of the genocide to police harassment. The defence had to conceal the identity of their witnesses. The Canadian prosecuting authorities were concerned at Rwandan involvement with prosecution witnesses.
209. Mylene Dimitri gave unchallenged evidence. She was defence counsel in different proceedings in Canada. She stated that the defence witnesses in her case were not assisted by the WPU. All arrangements for witnesses' travel from Rwanda were made by the defence team. Witnesses had to be collected from their villages in Rwanda and concealed from the authorities in transit. Witnesses were very frightened for their jobs, their family and their future. After giving evidence, four refused to return to Rwanda and disappeared. One who went back to Rwanda was visited by the authorities. Another had to face a gacaca retrial and "was locked up for genocide offences of which he had been earlier acquitted".
210. The witness John Philpot had worked extensively for the ICTR and had been critical of the Rwandan government. He was no longer able to go to Rwanda for fear of his own safety. Mr Philpot was asked about the story of the American attorney Erlinder, who had been held after arriving in Rwanda to assist Ingabire in her defence. Mr Philpot was apprehensive that this would prevent other foreign lawyers coming to Rwanda to provide robust defence representation.
211. Christopher Black was another defence lawyer at the ICTR. As a result of his actions he had received threats and was no longer able to go to Rwanda. He had received a formal warning from the Canadian Security Intelligence Service that Rwandan authorities wanted to kill him "and were going to try to do so".
212. Judith Rever is a Canadian journalist who has researched war crimes alleged to have been committed by the RPF. Whilst working in Belgium in 2014 she had been warned that the Rwandan embassy in Brussels posed a threat to her safety. During her time in Belgium the Belgian authorities had provided her with 24 hour security as a consequence of this threat.
213. The statement of a witness anonymised as EN-O was read into the evidence. He or she is a Rwandan attorney who refused to reveal their identity, on the basis that at the least they would lose work and at most could face imprisonment "for speaking against the government. In the past, attorneys, investigators and even judges have been arrested and detained just for going against the wishes of the government". S/he, too, spoke of fear on the part of witnesses of the government and of repercussions if they came forward.
214. Under this rubric, SDJ Arbuthnot touched on some of the evidence from the Appellant's expert, Dr Clark. He confirmed that he had spoken to the Minister of Justice Busingye and the Minister confirmed to him that executive interference in trials has continued. This conversation was in general terms and, as SDJ Arbuthnot put it, "unfortunately Dr Clark did not obtain more information".
215. Minister Busingye was scheduled to give evidence to the Court. He never came. SDJ Arbuthnot observed, however, that he did attend Westminster Magistrates' Court (the same Court) in relation to the arrest on an EAW of General Karenzi Karake in June 2015. SDJ Arbuthnot leaves hanging in the air the question why the Justice Minister was able to appear on a specific EAW but was unable to appear as scheduled in this case. It was Mr Busingye who, as President of the High Court at the time of the Brown case in 2008, told others that judges of his Court had been subject to attempts at influence by the executive (see Brown, paragraph 86).
216. Dr Clark also spoke as to his conversation with the previous Minister for Justice, Mr Karugarama, who also confirmed that politicians might ring judges concerning the outcome of cases, but said "they just ignored the call". Dr Clark went on to recite the evidence of another past President of the High Court, Mr Cyanzayire, who described the justice sector as "very prone to corruption". The Ombudsman, Mr Rutaremara, had commented that "in recent years there were far fewer complaints of political interference in trials" and Dr Clark cited the statement of Minister Busingye that since 2004, 40 judges had been dismissed for corruption or serious misconduct. However, there was no specific evidence to show how this related to genocide cases. As SDJ Arbuthnot said, "there is no evidence as to the sort of cases in which corruption or undue influence can occur and who might be the would-be corrupters" (judgment, paragraph 483).
217. In a short but striking passage, SDJ Arbuthnot dealt with anonymised evidence from those who had acted as judges in gacaca hearings. The question, considered later, of the quality of their evidence, is a separate matter. Here, it is the response of the state to their unwelcome conclusions which matters:

"484. In these proceedings I also heard evidence from judges who gave evidence that after the acquittals of CU and EN in Gacaca courts they were summoned and interrogated about the decisions they had reached. CU/2 and CU/3, for example, were both former Gacaca judges, CU/2 had been a judge at Mr CU's Gacaca proceedings in 2008. CU was exonerated after the hearing of numerous defence witnesses and the acquittal was then nullified. CU/3 said that after the acquittal, some Gacaca judges were arrested and interrogated."

218. Professor Reyntjens stated that there was "a well documented history of attempts to manipulate or fabricate evidence through either the use of inducements or through the threat or use of violence". His most recent example concerned falsified evidence against a UN coordinator said to have been organised by representatives of the Rwandan presidency. He dated that to September 2012, the significance of which is that it considerably post-dates the 2010 Brown judgment.
219. The Judge heard other allegations that evidence had been fabricated by Rwandan prosecutors. In paragraph 487/488, she cited the example of evidence from Peter Robinson, a former US federal prosecutor, now "an eminent and respected criminal defence attorney". Dr Clark provided a transcript of Robinson's account made at a seminar in 2010, evidencing pressure by the Rwandan prosecutor in a case on a potential witness for the defence. The threat was said to come from Rwandan military intelligence as well as the prosecutor and consisting of an assurance that if he testified for the defence "he would never be let out of prison".
220. In dealing with this evidence, SDJ Arbuthnot accepted that the judiciary was not independent until 2004, that there was a period of reform from about 2008 (in respect of which she cited Dr Clark and Professor Reyntjens) and she went on to consider the expert evidence of Professor Reyntjens and Professor Longman. The Judge recited that evidence between paragraphs 493 and 503. Both professors looked at the core claim by the GoR that there had been significant changes since the Brown decision. Professor Longman accepted (paragraph 498) that professionalism in the justice sector had increased "and that the quality of justice had improved, it was just that the politicisation of the system had increased". He described seeing "manipulation in every high profile trial". SDJ Arbuthnot did not reject this evidence, but noted that Professor Longman's account had two drawbacks, namely that he had not been to Rwanda since 2006 as he did not feel safe, and that he had never been to the Rwandan High Court: his evidence was generalised. Equally, Professor Reyntjens considered that there was gross manipulation by the GoR in some cases and was critical of the recent decisions by the ICTR to transfer defendants. The ICTR approach:

"…had placed too much emphasis on recent legislative changes, there was pressure on the ICTR to wind down and no other countries were willing to take these defendants."

221. Professor Reyntjens did agree that the quality of the judiciary had improved "tremendously" and he was not saying that "no-one could have a fair trial in Rwanda" (judgment, paragraph 503).
222. In paragraphs 505-513 of her judgment, SDJ Arbuthnot touched on the evidence of Dr Clark for the Appellant on the fair trial issue. Dr Clark had more recent experience in Rwanda than Professors Reyntjens or Longman (for the reasons they have explained). He had interviewed extensively individuals within the judicial system, including gacaca judges but no High Court judges. Dr Clark emphasised the importance of extradition and prosecution for the Rwandan government and SDJ Arbuthnot accepted from him that the GoR had shown "a marked dedication" to reforming the domestic judiciary. Against that background, Dr Clark set out to repudiate the findings in Brown and the reports of Reyntjens and Longman. He relied on the stream of ICTR Appeal Chamber decisions, and advanced the position that the Bizimungu case "was the only case of judicial interference" (judgment, paragraph 510).
223. Dr Clark's evidence about interference with the judiciary appears to us to have been rather ambiguous. As the Judge recited in paragraph 511, none of the sources interviewed by Dr Clark in April 2010 could cite concrete examples of political interference in genocide trials. However, in giving evidence before the Court in June 2015, Dr Clark agreed that the view of many of those he had interviewed was that there was continuing "subtle executive interference" with the judiciary. A number of individual witnesses to that effect were cited. A safeguard which many of his respondents considered would be helpful was the instruction of international defence counsel for high profile Rwandan suspects. However, despite the change in the Rules, it appears that had never happened. Moreover, in paragraph 513, SDJ Arbuthnot observed that although there are no concrete examples of interference provided by Dr Clark "he did not pursue the question perhaps as much as he could have done".
224. SDJ Arbuthnot then turned once more to the evidence of Martin Witteveen. We deal below with the question of the incomplete account of Witteveen's views which was before her. For the present we consider what Witteveen material she did have. SDJ Arbuthnot noted that at the time of Witteveen's first report in September 2014 he had been working full time within the NPPA for only three months, but he had previously spent a considerable period in Rwanda investigating two criminal cases. Witteveen's broad conclusion was that there was no evidence the judiciary were not independent when it came to genocide proceedings (judgment, paragraph 514). He did not consider that "ill-judged comments made by politicians in Rwanda would be any worse than many judges face from around the world" (judgment, paragraph 516). He, too, said he had never seen interference with specific criminal cases since the Bizimungu case and had not seen a loss of independence, neutrality or objectivity. He did not consider these respondents to be accused of "political crimes" (judgment, paragraph 516).
225. After touching briefly on the evidence of Ms Kabasinga and on the ICTR trial monitors, SDJ Arbuthnot gave her conclusions in relation to the independence of the judiciary in paragraphs 519 to 545 inclusive.
226. These paragraphs are carefully expressed and nuanced and are the subject of both appeal and cross-appeal. Mr Knowles QC for the GoR submits that insofar as SDJ Arbuthnot has found that there is a risk of interference in trials, she overstates the matter. He also submits that there is internal contradiction within these paragraphs, in particular between 531 and 545. The Respondents, on the other hand, submit that SDJ Arbuthnot was insufficiently robust in her findings and that the material before her should have led her to consider there was a much higher risk of executive interference with justice in genocide cases such as these.
227. What is certain is that the findings expressed in these paragraphs are the essential basis upon which the judge reached her final conclusions as to the very great importance of effective defence representation which in turn led her to find against extradition. Because of their length and importance, and so that this judgment can be read comprehensively, we attach these paragraphs from her judgment as Annex 3 to this judgment.
228. The starting point for the judge's thinking is that there is established evidence of executive interference in "political cases" (paragraph 522). The problem with which the judge was "wrestling" is whether that approach will affect these genocide cases (paragraph 523). The various Courts that have sent defendants back to Rwanda for trial have done so on the basis that genocide cases are not "political". Although she noted the significance that three of the Respondents were bourgmestres, and Brown/Bajinya allegedly a member of the Akazu, she found that he was not a "high profile" member of the Akazu. The Judge did not find there was credible evidence that the Respondents were "political opponents to the GoR in 2015". Nor was she persuaded that the extradition request was driven by political considerations (paragraph 526). She did accept that the Respondents, if returned, would be "high profile" defendants, and that seems established. She concluded that the "GoR will use the return of the RPs as examples to show that the State's justice system is recognised internationally as a system which can try defendants fairly" (judgment, paragraph 527).
229. The Judge concluded that she could rely on the monitors' reports from the five transferred cases as demonstrating, in relation to those decisions, that there was no recurring theme of complaint about the Judge's lack of independence or of partiality. She stated that she considered it "arguable that interference may not be picked up by the monitors, but the lack of independence or fairness would be" and said that she herself "was struck by the even handedness of the Judges when confronted with, for example, applications for adjournments by the defence" (judgment, paragraph 530). She accepted that there were continuing comments by the President and Ministers:

"…about the guilt of these RPs in the sort of autocratic state that I have found Rwanda to be. The GoR seems not to respect the presumption of innocence and that might put pressure on the judges if there was no international judge amongst them. I accept however the evidence from Professors Longman and Reyntjens, for example, that the quality of the judiciary has been increasing gradually in Rwanda and I have heard nothing which suggests that they are not able to ignore comments made by the President and others in non-political cases" (paragraph 531).

230. The Judge noted the "sea changes" relied upon by the GoR but concluded these changes "would be more significant if there was evidence they were affecting trials in practice" (judgment, paragraph 533). The only change which appears to be actually put in practice was the amendment to the Transfer Law of 2009, allowing three or more judges to try complex and important cases. The Judge noted (paragraphs 534, 535) that Dr Clark thought it unlikely that any application for an international judge to come and try a transfer or extradition case, pursuant to the legal amendment of 28 November 2011, would be granted since it would be unacceptable to Rwanda - "politically unpalatable" as Dr Clark had put it (paragraphs 434-5).
231. Whilst the Judge could accept the acquittal rate of 20 to 30 per cent in the High Court, as advanced by the GoR, there was no information about acquittal rates in genocide trials. SDJ Arbuthnot considered carefully the evidence of Dr Gahima and (as we have noted above) accepted that he was a measured, credible and impressive witness, described by Dr Clark as "a respected academic". The Judge also noted (paragraph 539) that the change of security of tenure to appointment for a determinate term ran counter to the normal principle that security of tenure should ensure independence and impartiality, but she found that "there is no evidence that the change to removal on the grounds of serious professional misconduct or incompetence is being used by the High Council of the Judiciary to remove judges for bringing in the wrong verdicts" (paragraph 539). The Judge noted that these cases would not be monitored, by contrast to those transferred from the ICTR. Such a mechanism would be a reassurance were it to exist (paragraph 541).
232. Having noted the evidence that although the judiciary was formally independent, it was in reality "subordinated to the will of the executive in all politically sensitive matters" (as described by the Bertelsmann Stiftung Transformation Index 2014 Report), the Senior District Judge moved from a more general picture to the specific picture of the genocide cases and focussed on the evidence of Witteveen, which she described as "particularly helpful". Mr Witteveen had been a witness to some of these hearings and had read a great deal about them from the notes of the legal officer of the Dutch embassy. She relied upon his evidence that he had not seen "any credible evidence that any authorities had interfered specifically in cases of genocide" (paragraph 546).
233. SDJ Arbuthnot reached her conclusion on this aspect of the case in paragraph 545:

"545. To conclude, I am drawing a distinction between the way the Rwandan High Court tries cases with a political flavour and the way they try genocide allegations. This is based on the clear evidence I have seen about the approach taken by the Specialised Chamber towards the five transferred genocide cases. Having considered all the evidence, I cannot exclude a risk of interference but judging from the transferred defendants the highest risk is from the pressure exerted by GoR ministers' comments in public and in the press. I consider any such risk would be reduced by a robust, able and experienced defence team with an ability to investigate the defence case and international monitoring of some sort. I consider that without both of these the RPs would be at a greater risk of judges behaving partially and being influenced by factors outside the evidence."

Our Conclusions on Independence of the Judiciary

234. After careful consideration, we reject the submission of both sides that the judge was wrong on this issue. We reject firmly the submission of the GoR that the judge should unequivocally have stated the judiciary was independent and that there was no risk of bias or interference in these cases, because they are not "political". The clearly authoritarian nature of the regime; the long and continuing history of the influence of political will over the justice system where the case is perceived to matter; the evidence of Gahima and others; the evidence of threats arising from criticisms of the regime, not in a political context but in the context of the justice system; the "Osman" warnings; the experiences of witnesses who gave evidence in genocide cases unfavourable to the prosecution, whether in Rwanda or abroad: all these point to a high level of risk of pressure within the system.
235. In our view, the fact that the ICTR has sent a number of cases back to Rwanda is important, but not conclusive. It is not clear that the ICTR had access to the amount and quality of evidence which was before the Court below, or is available to us. Moreover, the ICTR took the decisions in those cases in the knowledge that there was a procedure for revocation, and therefore an escape route if matters were shown to have gone wrong. There is no such escape route for the English Courts. Moreover, it seems an inevitable, but important, context for the decisions of the ICTR that, if trials were not to continue indefinitely before the ICTR, prolonging its existence far beyond the planned end-point for that jurisdiction, then the only destination for such cases (apart from other foreign Courts) would be Rwanda. We should not be understood to mean that the decisions of the ICTR were influenced by improper considerations. However, it would appear to us that this must inevitably have represented an important institutional consideration and the source of some momentum towards transferring cases to Rwanda. So far as other decisions in other jurisdictions are concerned, it is clear that the GoR has advanced the decisions of the ICTR with a powerful rhetorical flourish: if the ICTR has agreed to transfer cases, who are you to refuse? In an unpersuasive piece of advocacy, Mr Lewis QC for the GoR did so to us.
236. Another matter relevant to this conclusion arises from the "sea change" advanced by the GoR. It is a definite advance that three judges can sit in such cases. However, what is to be made of a change of law permitting a foreign judge to sit in such cases, which on the ground of national pride is never expected to be put into practice, as conceded by the GoR's own expert? At best this is an empty gesture, at worst an attempt to beguile.
237. The strongest points in favour of reposing confidence in the independence of trials are: (1) the general view that the judiciary are better trained and are possessed of integrity and (2) the eyes of the world will be on these cases if the Respondents are returned. The GoR has, objectively, a real interest in their trials proceeding fairly. We consider those to be positive factors worthy of consideration. However, they are moderated by one or two other considerations. Firstly, as the experts conceded and the judge recognised, any pressure behind the scenes will likely go undetected: indeed will be virtually undetectable. Secondly, objective self-interest may be a poor guide to official or political action in an authoritarian state. Short term political considerations, or a show of power, or mere prejudice, may easily supersede.
238. We also bear in mind the concerns of Martin Witteveen, as expressed in his memorandum of 18 June 2015, of which SDJ Arbuthnot was unaware, as we mention below. He expressed himself there to be unsure of the independence of the Rwandan judiciary ("To an extent … I even question the independence of the judges").
239. For these reasons, we incline to the view that the judge's conclusions on this issue veer towards being too sanguine as to the prospects of independent and unbiased trial. However, we do not go so far as to say she was wrong. Her decision was reasonable, based on the evidence before her, but the evidence before us has itself shifted the balance.
240. However, the great emphasis she placed on effective defence representation becomes entirely comprehensible. Given the risk of pressure or bias, in order to prevent a flagrant denial of justice, it is indeed vital that defendants charged with genocide-related offences are represented well, so that all necessary material is gathered and presented, that prosecution cases are fully challenged and proper arguments presented. The weight SDJ Arbuthnot places on representation and advocacy in this context is entirely correct. Without such representation, convenient verdicts will be easy to reach. With such representation, judges who wish to be fair will be given the material which may compel a correct but unpopular outcome.

Fair Trial:

Defence Witnesses

241. This topic is dealt with by SDJ Arbuthnot between paragraphs 546 and 592 of the judgment, with her conclusions set down in paragraphs 593 to 599. We are able to deal with this subject rather more briefly than the last. This topic really concerns whether witnesses will be too frightened of revenge or repercussions to appear in genocide trials, for the prosecution and perhaps particularly for the defence.
242. The Senior District Judge summarised the evidence given in both directions. As we have already indicated, there was a good deal of evidence to suggest that witnesses would be fearful of coming forward. The SDJ began by observing that the Respondent Mutabaruka already has nine defence witnesses whom he says he will be able to call if tried in Rwanda (paragraph 548). She noted there is no particular explanation as to why these witnesses are willing, if it is true that there is a general fear. The Appellant submits that there is much evidence that defence witnesses do come forward to give evidence at every level of Court in Rwanda, as they have done in cases before the ICTR. The ICTR has never found that the Rwandan authorities have systematically engaged in witness intimidation (paragraph 552).
243. In the earlier proceedings, the District Judge and then the Divisional Court agreed that both prosecution and defence witnesses had been attacked and killed, a point relied on in her submissions by Ms Ellis QC. She went on to point in the Tables to a number of the individuals threatened but, as SDJ Arbuthnot observed, the majority of those would appear to be political opponents or critics of President Kagame (judgment, paragraph 544). We have mentioned before that Professor Reyntjens was concerned he could not safely return to Rwanda. His concern crystallised in a conversation with a Rwandan ambassador in May 2012, when the Professor told the ambassador he would not return to Rwanda as he was afraid he would be arrested and prosecuted for revisionism. "The ambassador had said 'maybe you wouldn't be wrong'" (judgment, paragraph 554).
244. One witness who gave evidence before the Magistrates' Court in May 2014 (Eustache Nkerinka) reported that he was threatened after giving his evidence in this case.
245. In paragraphs 557 to 569 of her judgment, SDJ Arbuthnot dealt with the evidence of Scarlet Nerad. This witness gave a number of examples of witnesses concerned with their safety: prosecution witnesses too frightened to change their evidence and potential defence witnesses too fearful of assisting at all through concern as to reprisals. She also advanced evidence that witnesses who had given evidence for the defence in gacaca proceedings had suffered as a result. Ms Nerad gave evidence of four prosecution witnesses who now say pressure was put upon them to incriminate Mr Ugirashebuja. On the same point, Mr Fitzgerald QC gave another example from the papers. However, the Judge observed that Ms Nerad had been "most effectively cross-examined" producing a number of concerns about her evidence. In relation to proceedings against Ugirashebuja, it emerged that the acquittal of which Ms Nerad had spoken arose in relation to events in a different sector with different witnesses from those concerned within the extradition request. More seriously, the Judge found that Ms Nerad had a tendency to accept what she was told without question. A number of the instances of concern about evidence arose in deportation cases where no guarantees had been given, and thus the level of concern may have been inappropriate. It also became clear that Ms Nerad had not set out in any of the statements she had taken from anonymised witnesses that they had been told of various measures which could be taken to protect their identity, such as anonymisation, video linking and so forth.
246. SDJ Arbuthnot observed that:

"The main problem for the GoR is that evidentially … there is similar evidence of witnesses' fear as there was in 2009 before the High Court and conceivably the situation has worsened with witnesses' fears heightened by the GoR's perceived ability to abuse telecommunication networks. Of the 20 statements from new witnesses obtained in 2013 and 2014 in relation to CU's case, all of them have been anonymised, because the witness says he or she is frightened of being identified. A further five more had given statements for CU in 2007 and they remain frightened of the GoR …" (Judgment, paragraph 571)

247. Other examples of fearful witnesses were given, for example by the investigator Ralph Lake. The Judge accepted his account that various authorities in a specific prison were fearful of being blamed for assisting the defence in the provision of evidence. Senior officials were prepared to lie to hide things from those in Kigali in relation to these events, although the Senior District Judge took into account these events took place in 2007 (judgment, paragraph 576).
248. The Judge was impressed with the evidence of Fernand Batard, whose evidence is summarised in paragraph 578 to 581 and 586. Batard was an experienced investigator who had carried out investigations for the ICTR following the genocide. He accepted that "thousands of witnesses had testified in gacaca proceedings publicly, and hundreds of defence witnesses had given evidence at the ICTR and returned to Rwanda afterwards" (paragraph 579). Batard explained that each witness had different concerns and would be expressed differently. They were often very subjective. His overall conclusion was that witnesses were fearful, but the fears may often not be justified.
249. The Judge also rehearsed the evidence of Martin Witteveen on this topic. He had experience of recruiting defence witnesses to give evidence about cases, but outside Rwanda. He gave evidence that the Rwandan authorities gave assistance to investigators from Finland, Norway and Holland. His contribution was to suggest that in his experience witnesses were not frightened by what the GoR might do, but they were often concerned by how the local community and defendant's family might react. He considered there were times when witnesses influenced each other but he did not know of any example where they had been influenced by the authorities (judgment, paragraphs 587/8).
250. It is of particular interest that in relation to the five transferred cases, the judge found there was "not much information in relation to the defence case in the main because of a lack of effective defence representation". In Mugesera no defence witnesses were called. In Uwinkindi a handful of defence witnesses had been called, but at a time when the defendant was unrepresented. The majority of his defence witnesses lived abroad and no budget had been set aside for finding and interviewing witnesses in Rwanda or for preparation of witnesses abroad. In Bandora, on the evidence of Mr Witteveen, the defence had called 14 defence witnesses. In Munyagishari no funds for an investigator had yet been provided and no defence witnesses presented. In Mbarushimana the case was not yet begun, but there was as yet no budget for obtaining defence evidence. In the five transfer cases the video-link system has not been used "simply because the witnesses have not been identified as the funds have not been provided." (judgment, paragraph 591).
251. It was on the basis of that conspectus of the evidence that SDJ Arbuthnot reached her conclusions on this issue, which are set out in paragraph 593 to 599.
252. The Judge accepted that a number of witnesses have said to investigators they were too frightened to give evidence. She found they were expressing genuine fear. Many such were from rural backgrounds or of relatively low educational attainment. The Judge noted that "the reputation of the GoR at home and abroad … cannot be of assistance either", but she noted Witteveen's evidence that witnesses may be more frightened of local repercussions rather than national. The witnesses are afraid of a range of outcomes, from death, imprisonment, and torture to prosecution. The fact that some defence witnesses have been heard in Uwinkindi and Bandora, and others have agreed to give evidence in Mutabaruka's case, was of significance. The Uwinkindi Referral Chamber in 2011 had found that "in most of the then 36 genocide cases tried in the High Court, defence witnesses were available" but the Judge was able to note no detail of the numbers.
253. In paragraph 597 the Judge found that the following provisions would be provided by the judicial authorities in Rwanda for frightened witnesses:

•    A guarantee that defence witnesses could not be prosecuted for anything said or done in the course of a trial.

•    A facility to enable them to give evidence anonymously.

•    A facility to enable them to give evidence by video-link or by deposition in Rwanda or abroad. Video-link includes voice distortion and other methods of disguising a witness' identity.

•    The use of the WPU, which she found to be now fully functional, which can provide a safe house, assistance with travel and other protective measures. It is independent of the GoR and is run by the Registrars of the High Court and Supreme Court.

254. The final two paragraphs of her conclusions read as follows:

"598. There are many reports of witnesses being willing to give evidence in genocide trials in the High Court in Kigali. I consider that although there can never be any guarantee of safety at least some of the frightened defence witnesses are likely to give evidence for the defence were the defendants to be returned for trial. I do not accept that the defence in this case will be unable to marshall a sufficient number of defence witnesses if the defendants are properly represented with adequate assistance and the new provisions for video-link, anonymity, the WPU etc are explained to them which will enable them to give evidence in a protected environment.

599. I find that there is no real risk of a flagrant denial of fair trial in relation to the availability of defence witnesses as long as the RPs, if returned, are represented by able and effective representatives who are able to investigate and put together the case for the defence."

Our Conclusions Regarding Defence Witnesses

255. The GoR naturally supports the conclusions here, save for the pointer to the Judge's later conclusions, adverse to the GoR on adequate representation. The Respondents all submit the Judge's conclusions were over-optimistic and failed to accord sufficient weight to the evidence, particularly as to the delays, as well as the reputation of the Rwandan state, to the evidence of such witnesses as Professor Longman, and to the very limited extent to which defence evidence in transferred cases – high profile cases such as these – has been actually presented in Court in Rwanda.
256. In our view the matter needs a little analysis. The entry point must be: to what degree are these cases "high profile", and likely to attract attention of the authorities? What kind of attention is it likely to be? We have addressed this issue already in the previous section. Given the level of investment in these proceedings, we see it as inevitable the cases will be high profile, but a restraining factor may be the desire to gain international credibility by demonstrating fair trial.
257. That response will probably have a significant impact in two ways. Firstly, it is likely to restrain direct threat or intimidation. That may be so, but it is far from certain, for the reasons we have given above concerning decision-making in authoritarian regimes. Secondly, it is likely to mean that the safeguards designed to induce witnesses' fears, and listed by SDJ Arbuthnot in paragraph 597 of the judgment, are indeed put in place. They are steps which are verifiable by monitors and by reporting, but in particular by effective defence legal teams.
258. As recognised by the judge, it is inevitable that witnesses will be fearful. That is not the issue. The issue is whether the fear can be overcome to a degree sufficient to avoid a flagrant denial of justice through the absence of witnesses or the distortion of their evidence, and to do so in these cases, which even if not "political", must to some degree be regarded as "trophy" cases by the Rwandan authorities.
259. Here again we conclude that the Senior District Judge did not fall into error: her conclusion was reasonable, based on the evidence before her. Here too, our own view is that she was somewhat more sanguine or confident about the outcome than we would have been. But, she heard the evidence, she had well in mind the high bar required by the law, and her conclusion was within the reasonable range.
260. However, it is vital to understand the nature of her conclusion here. She was deciding that conditions could be put in place so that identified witnesses, at least in sufficient numbers, could be reassured, brought to Court and give their evidence under conditions which meant the evidence would not be distorted by fear. To that end, the safeguards identified are vital. So also is the presence of adequate representation, "able and effective representatives … to investigate and put together the case for the defence". The judge was stating that such representation was necessary to overcome the fear of witnesses who were identified, but she was also stressing the interlocking nature of her findings overall. The closer we have considered her conclusions, the deeper our respect has become for the underlying logic in her thinking.
261. It must also be emphasised that her conclusion here is emphatically not that witnesses will be found to be deployed: that is a question still outstanding. Here she concludes that witnesses, once found, may usually be got to give their evidence.

Fair Trial:

Defence Lawyers and Investigation of the Case; the "Witteveen/Arguin Debate"

262. As we have indicated, it was this critical point which brought the Judge to her conclusion adverse to extradition. It is also this area which has produced the most important further material, the additional material from Witteveen and the report from James Arguin. We intend to look at the material which was before the Court below, the judge's conclusions, the dispute about the introduction of further material, and then our conclusions. We emphasise the fact that SDJ Arbuthnot had the opportunity to see and assess Mr Witteveen, which we have not. Nor has any Court been able to see and assess Mr Arguin.

The Witteveen/Arguin Debate

263. For convenience we have characterised the differing points of view expressed by Mr Martin Witteveen and Mr James Arguin as the "Witteveen/Arguin debate". It is unfortunate that material parts of that debate have taken place "on paper" and not across the floor of a courtroom where the differing viewpoints could be tested. Most particularly, the debate was not conducted before the SDJ. This Court is left with the difficult task of evaluating the competing viewpoints.
264. We will give our conclusions in due course, but a broadly chronological narrative of how the expert debate developed is of importance.
265. The expert evidence relied upon by the GoR in the extradition proceedings in 2008/2009 was provided by Professor William Schabas. As the Divisional Court observed on a number of occasions, he was the only expert called in those proceedings who was prepared to say that the Appellants (as they were in those proceedings) would have a fair trial in Rwanda. To that extent his evidence was "of great importance for the GoR's case" (Brown, paragraph 108). Whilst his general distinction as an expert was not in issue, there is no doubt that the Court took the view that his evidence on that occasion was flawed and his status "as a dispassionate expert" was substantially undermined (Brown, paragraph 118).
266. In those circumstances it is not surprising that the GoR chose to rely upon different expert evidence in the present proceedings. As already indicated, two experts relied upon were Dr Clark and Mr Witteveen. Given Mr Witteveen's significance, and the fact that Dr Clark is not a lawyer or a judge, we consider his background and experience in a little more depth.
267. Mr Witteveen is a Dutch legal practitioner. His first report, dated 19 September 2014, detailed his background. He said that it involved two careers, one in his national jurisdiction from 1984 until 2004 and the second in the international sphere. In his national jurisdiction, all of his work, certainly since 1989, was as a prosecutor in the Court Prosecutor Service, mostly in the field of organised crime. In 2004 he moved into the international arena and, in the first instance, became an Investigation Team Leader in the Office of the Prosecutor of the International Criminal Court in The Hague. His work over the next four years largely centred on investigations into alleged war crimes and crimes against humanity in Northern Uganda. He conducted a good number of field missions in Uganda and had primary responsibility for "witness protection, organisation and networking."
268. In 2008 he became an investigation judge (not a trial judge) in The Hague District Court for International Crime. He described his role as follows:

"As investigation judge I conducted pre-trial investigations in criminal cases of international crimes for which the district court in The Hague is competent. During the pre-trial investigations, I heard numerous witnesses, mostly abroad during which the prosecution and the defence team were present and had the opportunity to examine the witnesses. During these years, I conducted investigations into two criminal cases of genocide in Rwanda, one criminal case against the leadership of the Tamil Tigers in the Netherlands, one criminal case of an Afghan general allegedly involved in atrocities during the 80s in Afghanistan as well as various cases of human trafficking around the world. I have conducted approximately 30 field missions to Rwanda, each mission one to two weeks in length …"

269. He remained in that position until 2012 when he was given permission by the authorities in the Netherlands to work for a Rule of Law Mission of the EU in Palestine, the purpose of which was to assist the Palestinian prosecution "in capacity building". He worked in that way until 2014.
270. Witteveen's main report of September 2014 contained a detailed analysis of what occurred in the two Rwandan cases to which he referred when describing his background. They concerned Yvonne Ntacyobatabara and Joseph Mpambara. Each was of Rwandan descent, the former becoming a Dutch citizen and the latter having been granted fugitive status in the Netherlands. Each was tried by The Hague District Court. Between October 2008 and March 2009 Mr Mpambara stood trial for war crimes and torture. He was found guilty of torture and sentenced to 20 years' imprisonment. That sentence was increased to life imprisonment on appeal in July 2011 when certain matters in respect of which he had been acquitted were replaced with convictions. So far as Ms Ntacyobatabara is concerned, her trial took place between October 2012 and March 2013 when she was convicted of incitement to genocide and sentenced to the maximum sentence available, namely, 6 years and 8 months.
271. Mr Witteveen conducted the pre-trial investigations as investigation judge in each case. That involved hearing witnesses and conducting other investigations. In relation to Ms Ntacyobatabara, he heard 72 witnesses, the "vast majority" of whom were heard in Kigali with the assistance of the NPPA. Eighteen of the witnesses were defence witnesses, four of whom were heard in Kigali. Those heard in Kigali were heard in a meeting room in the High Court, although one defence witness was, at her request, heard in a hotel room in Kigali. In relation to Mr Mpambara, Mr Witteveen heard 32 witnesses during the appeals phase, almost all of whom were "victim witnesses" and were heard in Kigali. Three witnesses were defence witnesses and since they were already convicted of various offences by the ICTR (and thus serving sentences in UN prisons), they were seen where they were serving their sentences.
272. What Mr Witteveen concluded about this aspect of his experience was as follows:

"In summary, the assistance rendered to me by the [National Public Prosecution Authority] NPPA during the Investigations and hearings in Rwanda, was complete, professional, of high standard and absent of any undue influences or interferences. The circumstances under which I could conduct hearings and other investigation activities created the best possible conditions in which I could pursue truth finding in these complex cases, which enabled the trial judges in the district court in The Hague to adjudicate the case and render their judgment. The NPPA has set a high standard and were exemplary for how other countries could also cooperate in these matters."

273. He enlarged on this evidence in his oral evidence, but there can be little doubt that his account suggested that the Rwandan authorities cooperated with the Dutch prosecution authorities in enabling those proceedings in The Hague to take place. That was his direct experience. He reported on (and gave evidence about) information he had received from other sources concerning the attitude of the Rwandan authorities.
274. From 24 June 2014 Witteveen was assigned as an adviser on International Crimes to the NPPA in Rwanda. He was assigned to deal with genocide cases. His time in this capacity was funded by the Dutch government. The initial assignment was for one year and renewable thereafter. He served for only one year in this position.
275. In his oral evidence he said that within the NPPA there are two separate units, one being the GFTU which conducts investigations and tracking of fugitives and deals with international matters such as preparing extradition requests, and the other being the International Crimes Unit ('ICU') which litigates in court. Although he was based in an office in the GFTU in Kigali, and the ICU was based in the headquarters of the NPPA about one mile away, he said that he was "fully embedded within the NPPA" and had full and unlimited access to anyone he wished to speak to. The Head of the GFTU at the material time was Mr John Bosco Siboyintore, to whom we have referred previously and with whom Mr Witteveen worked closely.
276. We should add also that since 2011 Mr Witteveen has served on the board of an NGO called International Justice Mission ('IJM') which aims to help national authorities in fighting violent crime (including human trafficking and bonded labour) through developing criminal investigations. Mr Witteveen's essential viewpoint is that of someone who has investigated and prosecuted very serious crimes in the international sphere. His first report ends with the ringing words "Impunity is not an option". This echoes the well-known ethos of the ICC articulated in the expression that it is engaged in "a global fight to end impunity".
277. By the time Mr Witteveen wrote his first report for the present proceedings he had been permanently based in Rwanda for less than three months, although he had been involved in the field missions and investigations referred to above. Mr Witteveen's first report was supportive of the proposition that the Respondents would receive a fair trial if returned to Rwanda. Since Witteveen was not in place until June 2014, the report could not have been available when the proceedings before the SDJ commenced on 3 March 2014. Mr Witteveen's first report was disclosed at some stage and he was scheduled to give evidence on 8 June 2015. Shortly before that (on 3 June 2015) he produced an 'Additional' report in which he raised his concerns about the quality of the defence representation available in Rwanda. Both reports formed the focus of his evidence and cross-examination before the SDJ on 8-10 June 2015. As is already apparent, it was the Additional Report and the evidence he gave concerning defence representation that was critical to the SDJ's decision on the fair trial issue.
278. It emerged subsequently that, after Mr Witteveen had drafted the Additional Report, he sent it to his former NPPA colleague, Mr Siboyintore. He (Mr Siboyintore) suggested that it should not be submitted to the SDJ and that he (Mr Witteveen) should await seeing what questions were asked in Court. However, Mr Witteveen clearly disagreed, and decided that he should submit the Additional Report to the Court "after considerable thought". He acknowledged subsequently in his memorandum that expressing the opinion set out in this Additional Report would "backfire" on him.
279. That was the state of the documentary evidence when Witteveen gave evidence orally in the Magistrates' Court. SDJ Arbuthnot placed great importance on his additional report. She singled out for quotation paragraphs 61 and onwards of that additional report on the necessity for high quality and professional investigations and stated:

"Part of this professionalism and these standards is the necessity to have defence attorneys who possess the knowledge, experience and the resources to conduct investigations for the defence, including the capabilities to conduct investigations abroad." (Judgment, paragraph 603)

280. Witteveen's experiences in the Uwinkindi and Munyagishari cases, his contacts with the monitors from the ICTR, and his close reading of the notes with his legal colleague at the Dutch Embassy gave him a clear view. Uwinkindi had been "without any defence" (additional report, paragraph 21). We have looked at these cases above, but Witteveen's views on them all deserve to be considered together. In relation to Munyagishari, his trial was stalled due to his defence lawyer's refusal to accept the latest financial terms. Witteveen found it remarkable that Mugesera's lawyer had maintained "complete silence" during the proceedings and throughout the evidence of 23 prosecution witnesses (judgment, paragraph 605). As to Bandora's trial, which was concluded, Witteveen described the lawyer's handling of witnesses in such a way that led the Judge to describe it as "chaotic and clearly unprofessional … Their questioning of their own witnesses was very short, irrelevant and repetitious and the protected witness' name was mentioned". Bandora's lawyers had made no attempt to get a budget to investigate the defence case (judgment, paragraph 606). Witteveen pointed out to the Judge that neither the Defendants, nor the majority of their lawyers, had "any trust in the GoR institutions" which led to complacency or confrontational attitudes (amended report, paragraph 20; judgment, paragraph 607).
281. The Judge was impressed by Witteveen's description that the defence was "by far the weakest link in the justice sector in Rwanda" but pointed out that, unlike the NPPA and the judiciary, they "had hardly received any capacity building from outside donors".
282. In cross-examination, Witteveen accepted that the skills of investigators are different from the skills of lawyers, but expressed his concern that the lawyers in Rwanda, as commercial practitioners, would not be able to "accommodate such investigations in practice. Even if they are willing to do the work required they do not have the right skills, capacity, support or time to conduct intensive investigation required" (judgment, paragraph 609). The Judge went on to express her own view that that was "even more the case, when they are being paid 15 million RwFr for the whole case including appeals" (judgment, paragraph 609).
283. Witteveen explained to the Court the importance of establishing a defence in genocide cases, given their importance and scope, where the only investigation as a basis for putting an alternative scenario to the Judge comes from the defence team. His experience of prosecution files was that they were "basic". A typical investigation took two weeks, during which time about 10 to 15 witnesses are interviewed. Witteveen confirmed that he had never seen any more investigations carried out by the police after the transfer of a defendant (judgment, paragraph 610/11).
284. SDJ Arbuthnot was concerned by the reports of Ms Buff, the ICTR monitor. Outside the transfer system cases, cases in the Rwandan judicial system did not usually involve any cross-examination of witnesses.
285. The Judge linked the evidence from Witteveen and Ms Buff to the evidence concerning the Defendant Uwinkindi, noting that "from the earliest days of Uwinkindi's proceedings in Rwanda the defence were asking for investigators". They were not provided. Ms Buff also informed the Court that in the ICTR cases, 74 per cent of the defence witnesses who gave evidence lived outside Rwanda, pointing to the importance of evidence gathering outside the country (judgment, paragraph 615).
286. Presenting the problem to the Rwandan High Court achieved no solution. The Court referred the request for investigation funding to the Ministry of Justice. Nothing happened. Eventually on 11 October 2013, more than a year after the request, the High Court ruled that no investigators should be appointed, but that Uwinkindi's counsel should find witnesses on his behalf (judgement, paragraph 616).
287. The Judge recorded the submissions of the GoR in respect of their own witness's evidence. The GoR essentially treated Witteveen as a hostile witness. They submitted to her that he did not have all of the available material concerning the defence in this litigation due to the "limited sessions he observed". The Judge rejected this, being confident that he had done much more than observe just a few sessions in court. Again the GoR submitted that there were alternative explanations for Witteveen's observations. The Judge rejected that. Finally, counsel for the GoR submitted to the Judge that Witteveen had expressly disavowed that his concerns met the high threshold necessary for proper objection under Article 6. The Judge firmly rejected that submission: Witteveen was placing his evidence at the disposal of the Court and leaving the Court to decide if the evidence and his opinions based on the evidence met the Article 6 test (judgment, paragraph 618).
288. SDJ Arbuthnot then expressed her conclusions on this issue carefully in paragraphs 619 to 631 of her judgment. Given the importance of these conclusions in the case, we reproduce those paragraphs here:

"619. There have been thousands of genocide trials in Rwanda at all levels of the judicial system and at present it would seem that that country can ill afford to put in the sort of resources required to investigate properly each case, especially for the defence but probably not for the prosecution either. One of my concerns is whether it is appropriate that this court should look at the prospects of a fair trial using as a benchmark the very high standards expected in the European context. I suspect that by Rwandan standards, the five transferred defendants are having a fair trial. They are occasionally represented and whether they are or not, can address the court. They can ask questions of the prosecution witnesses and although they cannot call defence witnesses from abroad at least they can call some witnesses from Rwanda. The authorities may well question with some justification why it should be that defendants transferred from abroad should have the resources to defend themselves that local defendants lack. I remind myself that defendants in Gacaca proceedings did not have lawyers and were tried by people of integrity with little or no legal training in the local communities. I agree with Mr Witteveen in the conclusion in his first report that with genocide allegations impunity is not an option. Nevertheless my responsibility is to conclude from the evidence whether if returned there would be a real risk of a flagrant denial of justice in relation to the prosecution of these men.

620. Witteveen was an objective witness who unlike any other had witnessed the trials of the transfer cases and considered the monitors' reports. Although of course counsel for Rwanda are right when they say he had seen only a limited number hearings but he had read the notes provided by his colleague as well as all the other evidence in relation to the conduct of the trials. The evidence he gave about how shocked he was by what he had witnessed of the defence representation of Bandora was striking and vivid. He had such "deep concern" and "profound doubts" (cross-examination 9.6.15) about the quality of defence representation that he felt duty bound to draft his Additional Report and wanted to give the court a true picture of what was going on. One can only imagine what a difficult situation he must have found himself in.

621. On a more positive note, I also saw some evidence in the ICTR reports, in the early days of Uwinkindi's proceedings, that the defence lawyers even with their pay continually being fought over, seemed capable of arguing the points that they should, usually procedural ones. It is the preparation and presentation of the defence case which thus far has been such a failure.

622. Witteveen does not blame the defence community in Rwanda for their lack of experience or ability but rather points out that whilst the prosecution (NPPA) and the judiciary have received extensive help in capacity building from donors, the Rwandan Bar Association has received virtually none.

623. I feel reluctant to consider the rates of pay fixed by the President of the Kigali Bar Association who after all knows the local conditions and what the cost of living in Rwanda is, which this court does not. Nevertheless I did consider that the officers of the KBA who negotiated the rates of pay for the defence lawyers with MiniJust did not understand the demanding nature of even an adequate defence approach to such cases and had never considered the amount of preparation required. It was clear from MiniJust's approach that it had completely underestimated the time it would take to defend such cases when it had decided on the original fees of 30K RwFr per hour per lawyer. MiniJust was concerned this was open to abuse and since then it has gradually reduced the fees which has led to the disputes. It is mark of the lack of professionalism of the lawyers that they have allowed the disputes to overshadow the work that should have been taking place to defend the transferred men who face such serious charges with long sentences if they are convicted.

624. I have said above that the early lawyers in Uwinkindi were clearly able to argue procedural points; Ms Kabasinga in the Rebuttal Material provided the brief curriculum vitae of the two later lawyers appointed for Uwinkindi. In the Supreme Court, Uwinkindi argued that one of the two new counsel had been found by another court not to have the ability to plead a genocide case whilst the second lawyer had no relevant experience. I take into account that they may not have included their experience of genocide cases in their cvs but on the face of it I had to agree with Uwinkindi they did not seem to have the experience that is needed in such cases.

625. At the end of 63 days of evidence in this case, I have seen the sort of work needed to be carried out to investigate alleged genocide cases. If the cases of VB and EN are typical of such cases, and I have no reason to think they are not, there is much important evidence, potentially undermining of the prosecution case, that can be gleaned from Gacaca proceedings and other cases where the prosecution witnesses have given evidence.

626. In VB's case, instructed lawyers and investigators have poured over a mass of Gacaca material to find out what witnesses have said on earlier occasions, some of these witnesses were defendants in trials. Although VB has carried out some of this work, the other RPs if returned to Rwanda will have to do so too. Another example, this time not in the case of VB but in the case of CMU, is the witness Sabine Hategekimana who accepts lying before the ICTR in another case. Proper investigations into that matter will have to be undertaken which will include obtaining statements taken by ICTR investigators.

627. The importance of defence preparation and investigation is shown by the example found in the case against VB where witnesses transpose accusations against one man to the RP. Miss Malcolm contends a similar situation has arisen in the case against CMU, in that whereas Alfred Musema is accused of orchestrating killings at Bisesero, the same witnesses now suggest CMU is the perpetrator. Miss Malcolm and Mr Weeks have produced schedules which analyse the evidence of two ICTR witnesses, Kabagora and Ntagara in their submissions. They show a number of differences in their accounts at different times. All these investigations I accept will have to take place conducted by lawyers or investigators who have experience, knowledge, application and the appropriate funding.

628. This leads to the second concern I have in relation to these RPs which is the lack of funding for the identification and locating of witnesses in particular abroad. Without such funding and without defence counsel with the ability to identify, locate, contact and interview such witnesses themselves or without an investigator to do it for them, it is difficult to see how Uwinkindi or any other defendant will have a defence case to put before the court.

629. In August 2015 a new law was passed allowing for applications for defence funding for investigations in genocide transfer cases. Uwinkindi had raised this as a problem at the ICTR in 2011 and in the Rwandan High Court in 2012. It took four years for the law to be changed and I suspect it had a lot to do with Uwinkindi's ability to refer his case back to MICT. This change of course opens the way for funding to be granted. It is too early to say how this will be applied in practice. I noted that the Uwinkindi budget for investigations abroad was US $ 64,595 (about £43,000), a vast sum. Any investigation budget for witnesses abroad inevitably will dwarf amounts spent on the lawyers and the investigations in Rwanda. In the cases of these five RPs some of the investigation work has been done and it may well be the costs will be lower, nevertheless the GoR will have to spend a great deal on investigations and it remains to be seen whether they will commit that sort of money to defence cases.

630. From all the evidence I have read and heard I concur with Witteveen's Final Conclusions in his Additional Report, he is certain that the facts in genocide cases can be established but "only under the condition of high quality and professional investigations, applying internationally accepted standards. Part of this professionalism and these standards is the necessity to have defence attorneys who possess the knowledge, experience and the resources to conduct investigations for the defence, including the capabilities to conduct investigations abroad". He has profound doubts that the defence lawyers assigned to the transfer cases can do that. His solution is still to extradite but to provide a lawyer with the right experience to work alongside Rwandan lawyers who would be provided with appropriate funds to conduct investigations. In his view (and in my view) "it ensures the necessary adequate defence capabilities for the defendant that meets the required standard and guarantees not only procedurally fair trial but also a fairness to the trial". Unfortunately this court does not have the power to order extradition on conditions and I am applying a different test.

631. I find that if extradited, as things presently stand, the defendants would be denied the effective representation of counsel in cases which so obviously call for effective and skilled representation by suitably experienced and resourced defence lawyers. It is too early to say that sufficient funding for defence investigations in relation to witnesses abroad will be provided. These defendants are legally aided in this country and will be indigent in Rwanda. I have seen in this case what the effective representation by counsel can achieve. Without such representation and funding, the High Court in Rwanda would be presented with the prosecution case and the RPs would find it impossible to present their side of what happened. I find the RPs would be exposed to a real risk of a flagrant denial of justice and a breach of Article 6."

289. Shortly, we will look at the process by which the further views of Mr Witteveen and the views of Mr Arguin emerged. That further material we take into account in our overall conclusions. Before doing so we should make it clear that, given the material before SDJ Arbuthnot, we do not consider that her conclusions set out above were in error. If no more material had emerged we would have affirmed those conclusions.
290. What did not emerge during the evidence Mr Witteveen gave, (as we have said) between 8-10 June 2015, was that he entertained other reservations about what was occurring in the trial system in Rwanda. He sent a Memo headed 'CONFIDENTIAL – NOT FOR DISTRIBUTION' to Mr Arguin, the Chief of the Appeals and Legal Advisory Division of the MICT. It was dated 18 June 2015 and was thus composed in the week following the testimony given before the SDJ. It also emerged from a further memo from Witteveen of 1 June 2016 that the Memo of 18 June 2015 was prompted by comments made by Arguin to Siboyintore on Witteveen's Additional Report, which Witteveen saw when he returned to Rwanda. Witteveen has said that he knew Arguin dealt from Arusha with the referral cases, though he had never seen him in Rwanda. Witteveen felt that Arguin should know his (Witteveen's) opinion in relation to the handling of genocide cases once they reached Rwanda. That was the background to the formulation of the Confidential Memo of 18 June 2015.
291. The process by which Mr Witteveen's additional concerns became known is of significance. The Respondents are concerned that the full level of concern was never disclosed when it should have been. For that reason, we analyse the sequence of communications.
292. In Mr Arguin's report of 22 May 2016 he says that he forwarded Witteveen's June 2015 memorandum to the NPPA. He did not indicate precisely when he did so or to whom, but the inference to be drawn from his report is that it was done soon after it was received. Indeed, he subsequently confirmed that he sent it to Siboyintore and to Ms Kabasinga (to whom we have referred previously), as soon as he received it in June 2015.
293. On this issue we received material and submissions post-dating the hearing before us. Ms Kabasinga was then the Senior Legal Adviser to the NPPA's ICU. She told Mr Siboyintore as recently as January this year (2017) that she did indeed receive the email from Mr Arguin, to which the memo was attached, on 19 June 2015. We have seen an undated witness statement from 2017, made by Ms Kabasinga, in which she says that she received an email from Mr Witteveen on 28 May 2016, with his additional statement or report attached. Her account is that she replied to him informing him that she would not read his further statement "because I was a witness in the proceedings and did not feel it was appropriate". She also confirms that she received Witteveen's email with his memo of 19 June 2015, on the day it was sent to Arguin, because Arguin forwarded it to her. She did read this, although she is unclear when she did so. She forwarded this to Richard Muhumuza, then Prosecutor General of Rwanda. This was because she regarded it as the duty of such an expert as Mr Witteveen to "raise concerns internally and suggest solutions". She says she never discussed that with Mr Siboyintore. It does not appear that she took any action about it or sought to discuss it with Siboyintore at the time. Siboyintore has said that he did not see the memo at all until 6 December 2016, during the course of the proceedings before us. He has signed a witness statement dated 24 January 2017 confirming that he did not see the memo until December 2016. His account is that he has searched, but can find no record of having received the relevant email from Mr Arguin in June 2015. He is unable to explain why it was not received. He accepts that he was in regular contact by email with Arguin.
294. Siboyintore has said in his statement (reiterating something said earlier by him to the CPS) that Mr Witteveen's memo "crossed [his] mind when [he] was reading the report that [Mr Arguin] had provided in support of the Dutch Appeal Case" and that he (Siboyintore) "developed a curiosity to know what Witteveen, a former colleague at our Office and a friend, had written in his confidential memo". He has said that Mr Arguin "shared" with him his report dated 22 May 2016 and that he (Mr Siboyintore) emailed Mr Arguin on 7 June 2016 asking if he would "share with me" the confidential memo. Mr Siboyintore has exhibited a copy of that email, sent to Mr Arguin's personal email address, to his recent witness statement. However, Siboyintore has said that Arguin did not reply to the email or send him the memo and, as we have said, that he did not see it until 6 December 2016 during the course of the proceedings before us. Mr Siboyintore does not suggest that he pursued Mr Arguin further about the memo despite the request he made in June 2016.
295. This catalogue of events is, to say the least, surprising. Whatever may be the precise sequence, it means that an important document, with real potential implications for extradition proceedings of genocide suspects to Rwanda generally, attached to an email sent by Arguin to the Head of the GFTU, was not received by him (or drawn to his attention by the Senior Legal Adviser to the ICU to whom it was also sent) at the time it was sent. It also seems clear that when he became aware of it a year later, his email to Mr Arguin requesting a copy was ignored – and the request was not followed up. The critical further doubts of a vital witness were not communicated and not disclosed.
296. The decision of the SDJ was not promulgated until December 2015. The impact of Mr Witteveen's Additional Report, and of the evidence he gave, was not known for several months. We understand that Mr Siboyintore was present in court when the evidence was given. Nonetheless, it requires little imagination to appreciate that what Mr Witteveen said would not have been welcome to anyone or any body whose objective was to have genocide cases transferred for trial to Rwanda.
297. Witteveen's evidence was given shortly before his contract was due for renewal. In his further memorandum of 1 June 2016, Witteveen tells us that in June 2015 the Prosecutor General (Mr Ngoga) "was very positive" about his work and was preparing for the extension of the contract. Ngoga expressed his disappointment at Mr Witteveen's leaving (which appears to have been a decision taken by Mr Witteveen himself). However, according to Mr Witteveen, by September 2015 Mr Ngoga had become "very critical" of the way he had functioned. Ngoga wrote a letter to the Dutch Embassy in Kigali in that vein. It is hard to see what could have prompted such a change of view save for the memorandum of 18 June. This has given rise to serious concern. Writing on 1 June 2016, Mr Witteveen said this:

"Three different sources in Rwanda have strongly advised me not to return to Rwanda for my safety, if I would be granted a visa at all. These sources say some Rwandan authorities have called me a traitor."

This introduces a very worrying note into these proceedings.

298. We now return to what Mr Witteveen said in his Memo of 18 June 2015 to Arguin. It was not seen by the CPS until 11 July 2016. It runs to 48 paragraphs. Because of the importance of Witteveen's evidence, we think that we have no alternative but to quote a number of those paragraphs verbatim. Before we do so, it is to be noted that in the first paragraph of the Memo he indicates that he is ending his one-year contract in a few days' time and that he had submitted "a lengthy assessment report which contains dozens of recommendations how the work of the GFTU can be improved." He summarised his view in paragraph 3 of the Memo as follows:

"In summary, I support extraditions to Rwanda and, based partly on my own experiences in Rwanda, dating back to 2008 as well as experiences of others, dismiss allegations that government authorities intervene in cases, unduly influence witnesses, genocide cases are political of nature and defendants run security risks after being transferred. It is my observation that Rwanda has a functioning justice system for genocide transfer cases. However, based on my observations during my year in Rwanda, in my additional expert report, I have criticized the quality and performance of the defence attorneys in the five cases I observed. It is my assessment that in four of the five cases there is either no defence, formally or materially, and/or the defence is of substandard quality."

299. He continued thus:

"4. It was my deliberate choice in the additional expert report to focus on the quality of the defence in the transfer cases and not criticize government authorities. First because I see it as the first and foremost reason for an impediment to fair trial in these cases. Secondly, no one else will present this view because no one has an interest.

5. Remarkably, none of the defence attorneys in court in London questioned me about the role of government authorities in respect of the functioning of the defence attorneys in the transfer cases. Some came very close but at the end did not ask the right questions. Still there are observations, I made, which I believe are relevant. The purpose of this memo is to describe them. The main caveat is that many observations are objective facts but the inferences I make can be questioned, although are difficult to refute. It is my profound conviction that, in order to know the realities in Rwanda, one will have to dig a little deeper into the layers of truth. I have only scratched the surface, which is still much more than what others have been doing who judge on the functioning of the justice system in Rwanda, especially those in other jurisdictions, whose sources are mainly Rwanda government sources."

300. Having made the observation that "in four of the five cases there is no defence and that the defence attorneys in Bandora were substandard and did not present an effective defence", he ventured the following views:

"7. I do maintain my opinion in my additional report and my testimony in court that I am convinced that there are no defence attorneys who can perform a credible defence investigation, certainly not abroad. This is not necessarily a criticism to the attorneys, but an obvious fact as lawyers not very often have the knowledge, expertise, experience and skills to perform such an investigation. That is why defence teams hire qualified investigators. But under Rwanda [case] law, a defence investigator cannot be appointed and the lawyers will have to personally conduct this task. It is totally unrealistic.

8. More importantly, I have little doubt left that, if these qualified lawyers exist in Rwanda, you will not see them in the transfer genocide cases. Simply because these lawyers will challenge both the authorities in general as well as the court. These lawyers will not accept everything the Minister has determined in his policies, they will fight for sufficient resources to perform a credible defence investigation, they will delay proceedings with their motions and filings before the court. And this is not accepted. Consequently, they will be removed from the case as happened in Uwinkindi and will happen soon in Munyagishari. More 'cooperative' lawyers will be appointed."

301. He went on to say that the 4-year history of what happened in Uwinkindi demonstrated that Rwanda was not "ready for these cases, certainly not in respect of defence issues." He elaborated thus:

"Even after four years, there are no final policies, regulations and, very important, an organization that executes these regulations professionally. Even as we speak, drafts of contracts with defence attorneys are discussed, clauses are taken out or taken in, the fees issue has not been concluded definitely and defence attorneys are moved out of the case and replaced. One court orders that the defendant be submitted a list to choose his defence attorney, another court now finds that there is no right to choose a lawyer from a list when indigent and rules the RBA needs to appoint a lawyer. Whether this is a purely legal decision or is motivated by the need to speed up trials and sideline lawyers who are critical who have to be sidelined as non-cooperative is an open question. The decision of the Minister to keep the execution of legal aid to himself is in my view not a very wise decision as it makes the whole thing too political. It is him against the defence lawyers. The usual language that everything is solved now and the defence attorneys can concentrate on their cases, no longer suffices."

302. He expressed concerns about Ministerial influence in Uwikindi by asking the following question:

"When the Minister unilaterally ended the contract with the attorneys in the middle of the crucial part of the trial and considering he had already rejected a budget proposal submitted by the defence to conduct an investigation abroad, what else could the defence attorneys do and bring it to the court?"

303. He repeated his concerns that the good and competent lawyers that exist in Rwanda will not become involved in genocide cases and said this:

"… the competent, experienced lawyers … who have operated in the ICTR theatre, will not be seen in these cases."

304. Given the position he had taken in his initial report and his evidence the next few paragraphs in the Memo make concerning reading:

"17. The next thing I am worried about is the influence the Minister and the Ministry has had on the proceedings. To an extent that I even question the independence of the judges. (Emphasis added.) I have seen a number of events that I believe are a real concern in this respect.

18. First of all I believe it is clear that the origin of the recent difficulties in the proceeding is with the Minister. I believe it was him personally who decided that the proceedings were taking too long and the cases should be concluded. Based on the minutes of the meetings with the defence attorneys and other documents, there can be no doubt he decided to act. A question of course is: what were his motives and why he didn't leave it to the judges. He claims it was the limited budget and the fact that it got depleted by the defence attorneys delaying tactics. But I put that into question. Based on his legal aid policy, promulgated in September of 2014 he had budgeted 100.000.000 Rwandan Francs each year for four years for hiring lawyers for the transfer cases. If he pays each defence attorney 1 million a month, he can hire eight lawyers for these cases to stay within the budget. If the Minister spends 15 million Francs in a case and the case takes five years, he spends 3 million Francs per case per year. To spend 100.000.000 in a year he needs 33 cases, while there are only five. So, his assertion that he depleted the budget is not very convincing.

19. Secondly, the Minister has always blamed the defence attorneys for the delays in court. But the reality of course is that the judges decided on postponements during all these years. Sometimes these postponements were triggered by requests of the defence attorneys, but not always. In Uwinkindi the composition of the chamber changed in 2014 leading to an additional postponement. Many times the courts have been extremely indecisive in matters that should not take long to decide. Has the Minister started to get critical to the judges?

20. Whatever the reason, there has been a clear pattern since the end of 2014 where the judges started to accelerate the proceedings. Firstly in Uwinkindi of course where the judges have quickly rejected every motion and request by the defence and decided to proceed with the case without defence. When the defence attorneys decided to appeal the decisions of the High Court in January, I was told that the High Court would wait for the Supreme Court's ruling, as they did in the past. But I found out end of February that the High Court had changed its position and was now proceeding with the case. To my surprise they went on hearing witnesses, both prosecution witnesses as well as defence witnesses without any presence of defence attorneys. The judges had taken three years to take their time in deciding all kinds of preliminary matters and now, all of a sudden, they finished the witnesses in just a couple of days and without defence attorneys.

21. Then the court decided to close the case by requesting the prosecution to present its closing arguments. When the day, which was set for that, arrived, the prosecution requested a postponement of two weeks or so as they had not finalized the preparations. The postponement was given and when the next date arrived, the court decided to postpone the case till the Supreme Court had ruled in appeal. I have tried to find out why the court had changed position but I couldn't get a clear answer.

22. What had changed was that the President of the MICT had decided to put a full chamber of the deferral request. This immediately sparked tension in the prosecution in Rwanda and I have no doubt this also caused the change in position of at least the Prosecution fearing they would be criticized of hearing witnesses without any defence. If the court had not allowed the prosecution a slight delay in their closing arguments, Uwinkindi's case would have been closed. Clearly, the recent decision to have the new defence attorneys represent Uwinkindi and the possible re-hearing of the witnesses have come under pressure and was by no means a voluntary step."

305. We have highlighted the passage at the conclusion of Witteveen's paragraph 22 because Mr Tim Moloney QC, on behalf of Charles Munyaneza ('CM'), placed reliance upon it.
306. Mr Witteveen referred to the Bandora case as another example of the speeding up of a case where a defendant was represented by lawyers where, he observed, there was no such hastening where defendants are unrepresented. His conclusion was as follows:

"In conclusion: I believe there are clear indications that external factors have brought the judges of the High Court, at least in Uwinkindi and Bandora, to speed up the trial at the cost of credible defence. It will never be known whether the judges received an instruction to speed up or whether they did not need an instruction and knew what to do. The change in position in Uwinkindi is clearly the result of the MICT decision, not of a respect for defence rights."

307. We would, of course, observe in passing that speeding up the legal process, even in cases involving allegations as serious as genocide, cannot of itself be an indication that there is something inherently unfair in that process: attempting to speed up criminal trials is something with which we are familiar. If, however, such judicial drive is derived from something other than efficiency, that is a different matter. After expressing the conclusion to which we have referred, Mr Witteveen poses the question of "what could be behind this speeding up of proceedings and removal of attorneys". He expresses concern that the perception of "the justice system" is that defences to allegations of genocide should be controlled and limited to such an extent that it impedes a fair trial. He expressed himself thus:

"Some of these issues have been raised during the referrals in Munyagishari and Uwinkindi before the ICTR but rejected as unfounded or speculative. I believe they might be true after all."

308. He explained why in the following paragraphs:

"31. Generally, all defence attorneys claimed it is very hard to defend genocide defendants especially from abroad as society in general but their family, partners, clients and the like in particular do not accept they defend a defendant on genocide. They tell they lose clients. This seems relevant as private lawyers make their income on commercial cases rather than criminal cases. You wonder what will happen to defence attorneys who will be successful and get their clients an acquittal.

32. In the contracts that the defence attorneys signed there was a clause to the effect that the Minister could unilaterally end the contract in case the defence attorney openly criticized the Minister or the Ministry. This clause has been removed now from the new contracts offered by the Ministry but they may still be valid under the existing contracts such as in Bandora. Apart from what the clause says and whether the clause is valid or not, I continuously sense that defence attorneys are conscious not to criticize. The clause does not make much difference.

33. In Mugesera, the defendant was given the opportunity to comment on the prosecution witnesses presented in the case. Mugesera said that one victim witness had been lying and called the witness a liar. The prosecutor in the case immediately reprimanded Mugesera that he could not call the witness a liar. The judge joined in criticizing Mugesera and used softer language to make Mugesera clear he could not use this kind of language to qualify the witness. Apparently victim witnesses do not lie.

34. In the case of Bandora the defence attorneys said it was a shame that the prosecutor had brought witnesses to court who testified they were forced by the prosecutors to wrongfully accuse the defendant. Again, the prosecutors objected to the wording "a shame" and again, the judges, more softly joined in that criticism. At the end, the defence attorney openly apologized in court to the prosecutors.

35. A crucial element of the trial in Bandora was in October 2014 when two prosecution witnesses retracted their statements claiming they had been visited by the prosecutors in the court room who requested him to wrongfully accuse the defendant. The statements by these two witnesses did not attract much attention and the defence never requested anything to the court in examining what exactly had happened, although it seemed relevant. I questioned the defence attorney afterwards why he had not requested that this incident were to be investigated. The defence attorney replied he thought it was sufficient for him the witnesses had retracted their statements. Although it is hard to prove, there is certainly a likelihood, in light of the other observations, that it is not expected from defence attorneys to request a further examination of alleged conduct by the prosecution. Surprisingly, in the verdict of the High Court in Bandora, the incident was reduced to mentioning the witnesses had retracted their statements. The fact that the witnesses claimed the prosecutors had requested them to falsely accuse the defendant did not make it to the verdict.

36. Recently, Mugesera [who is definitely more qualified to litigate his case than his lawyer, probably after he became experienced after ten years of litigation in Canada] commented on prosecution witnesses and requested the court to call an investigator as a witness because he believed that several witnesses had coordinated their statements. The judges refused and said that they could not do that and referred him to the prosecutor."

309. Witteveen appears to be speaking of a number of different matters that add up, as he would suggest, to an inadequate defence system. They can be summarised as follows:

i) an unwillingness on the part of competent lawyers to undertake defence cases;

ii) an attempt to put pressure (in the form of a provision in the contract) on those who were prepared to take on such cases not to criticise the Minister or the Ministry;

iii) an unwillingness on the part of the judges to permit prosecution witnesses to be called liars;

iv) an unwillingness on the part of the judges to permit criticism by the defence of the way the prosecution case has been presented, with a consequent unwillingness on the part of the defence advocates to do so.

310. He continues that theme by going on to question "the role of the RBA and their independence." He noted that it had not, at least initially, objected to the contractual provision referred to in (b) above and had not come to the defence of Mr Gashabana "while there was all the reason to do so." He drew attention to the choice by the President of the RBA of a new lawyer for Uwinkindi "with clearly less sufficient experience than dozens of other lawyers …, but likely a cooperative attorney."
311. He then referred to the issue of the ability of Uwinkindi to put forward a defence that put a different complexion of the genocide. This was summarised in [162] of the decision of the Referral Chamber of the ICTR dated 28 June 2011 as follows:

"The Defence submits that the Accused wishes to argue in his defence that the mass graves found in July and August 1994 at Kayenzi Church were not filled with Tutsi bodies as a result of his actions, but that those graves were filled with Hutu victims of the RPF. The Defence contends that running this particular line of defence is "simply untenable" in Rwanda's current political-legal climate. It concludes that despite the immunity added to the Transfer Law in 2009 it is unlikely that a potential witness would be willing to present evidence on the role of the RPF in killings in Rwanda and that it is equally unlikely that any defence counsel in Rwanda would agree to represent a client putting forward such a politically sensitive defence. Citing the arrest of Peter Erlinder, ICDAA argues that notwithstanding the immunity granted under Article 13 of the Transfer Law, Rwandan courts will still be able to curtail certain lines of defence by initiating contempt proceedings against lawyers or witnesses voicing politically sensitive views."

312. The Chamber took the view that the presumption of independence and impartiality of the Rwandan judiciary (a presumption that required a high threshold to be surmounted for it to be set aside) and the provisions of Article 13 (which in relation to transferred cases provides that "no person shall be criminally liable for anything said or done in the course of a trial"), together with its expectation that:

"if in the course of the trial in Rwanda the Accused, his counsel or any witnesses on his behalf make a statement amounting to a denial of the genocide, he or she shall not be prosecuted for contempt or perjury"

and the existence of the monitoring and revocation mechanisms, were answers to the concerns advanced by Uwinkindi.

313. Mr Witteveen's view on this issue, as expressed in his Memorandum, was as follows:

"Whether such a defence is really untenable in Rwanda we will probably never know, because it is highly unlikely the current defence attorneys or any defence attorney will make that argument. There is little doubt that no witness will take any chance by supporting the events as described by Uwinkindi."

314. We observe that Mr Arguin was part of the prosecution team that presented the case to the Referral Chamber for Uwinkindi's referral to the Rwandan Court.
315. Mr Witteveen's memorandum then returned to the theme that there never would be any credible defence investigation because it would be too expensive in time and resources. He referred to Mr Gashabana as "the only attorney who has requested a budget." He commented that although "the budget was very detailed and included cities where he intended to go, the budget has been refused as unrealistic, but it remains unclear what was unrealistic about the budget." He asserted that the Minister "just does not want to spend a whole lot of money on this and the challenges posed by these defence investigations are simply too big to handle." He recorded that Mr Gashabana clearly told him that "he would have to plead the case without witnesses from abroad." Mr Witteveen also indicated that he sensed reluctance on the part of Mr Gashabana (and another lawyer, Mr Nyibizi) to speak to him about any difficulties they experienced.
316. We can conclude this review of Mr Witteveen's memorandum by quoting the final four paragraphs:

"45. What looms is a situation where defendants are convicted without [sufficient] evidence and, through strict control and direction by the Ministry, embraced by the judges, capable and experienced defence attorneys are sidelined and replaced by handpicked lawyers who do not have any trust from their clients, are not conducting any credible defence investigation and are cooperative with the court. From the outside it will look consistent with fair trial. In fact it is flawed.

46. I do not claim that all is bad intent. Certainly on behalf of the prosecution service I have seen a sincere desire to comply with fair trial rights. What is the matter, I believe, is that this is all new to Rwandan institutions and it takes time to learn. There seems to be an innate attitude averse to anything that criticizes, confronts or challenges what government institutions present. The fact that these are genocide cases makes it more complex.

47. I do not take the position that extraditions should be stopped or referrals revoked. To the contrary, there is only one way which is the way ahead. But Rwanda needs pushes from external persons or institutions and we have the obligation to provide for and contribute to these pushes. Therefore I strongly advocate for forms of hybrid solutions. I believe this has been a missed chance years ago, when there were better opportunities to create hybrid systems in Rwanda. The case of the War Crimes Chamber in Bosnia and ICTY's role in it, could serve as an example. But there is still time to amend what was not fixed earlier. I have therefore recommended to have foreign defence attorneys team up with local attorneys and defend the cases with external donor funds. Equally important is it to have foreign judges sit in the High Court chamber, which was foreseen when Rwanda adjusted their laws to facilitate that and what the ICTR, when referring the cases, expected to happen [para 114 Uwinkindi referral decision].

48. I believe that my analysis and suggestions are in the interest of Rwanda and the broader aspects of justice. That is what I seek to do."

317. We will defer our views about the impact of this memorandum until we express our conclusions as a whole. However, there can be no doubt at all that the SDJ would have focussed on these observations, given the role that Mr Witteveen played in the proceedings and the significance she attached to his evidence. As we have observed, she made her decision in ignorance of what he had said.

The Disclosure Issue: The Role of the CPS

318. We step aside briefly from the main issue to consider disclosure. It is important to acknowledge that the CPS are not to be criticised for failing to bring these views to the attention of the SDJ, because they were unaware of those views until July 2016. We did ask for further information concerning the apparent failure of the Rwandan authorities to inform the CPS about these views, given that the proceedings before the SDJ were still continuing. Evidence continued through June and July 2015, the closing submissions of each side were made in August and September 2015 and the SDJ was preparing her judgment during October to December 2015. We made an order on 7 December 2016 requiring a signed witness statement from "any relevant authority" within the GoR by 4 pm on 16 December 2016, setting out when Witteveen's memo was first received. We required a full explanation of why the memo was not sent to the CPS promptly. We have set out some of the results above.
319. It is right to say that the initial statement served in purported compliance with this order was from Miss Kate Leonard, a barrister who is a specialist prosecutor with the International Justice and Organised Crime Division of the CPS. She has had conduct of these extradition proceedings on behalf of the GoR since 2013. Complaint was initially made (in our view, with some justification) that this did not comply with our order. However, matters were rectified subsequently. It seems that the first that the CPS knew of the existence of any further document emanating from Mr Witteveen was when Mr Frank Brazell, the solicitor acting for Dr Brown, emailed Miss Leonard saying that he had become aware of the existence of a further report/statement from Mr Witteveen, that had been produced to the Dutch authorities for the use in the extradition proceedings before the Dutch Courts. Miss Leonard's enquiries of the GoR resulted in an indication that the GoR had no knowledge of any such document. At that time the Dutch prosecutors also confirmed that no such document been produced or used in the Dutch proceedings. However, in early February 2016 it emerged that the Dutch prosecutors were in possession of the confidential memo from Mr Witteveen, but since it had been classified as "confidential", it could not be disclosed. Mr Witteveen subsequently confirmed to the CPS that he had indeed prepared the memo, but said that it was not intended for use in court proceedings. Whilst he was willing for the memo (and indeed his response to Mr Arguin's report) to be disclosed, it required the consent of the Dutch prosecutors before this could occur. That consent was forthcoming in due course. Mr Arguin also agreed that his report should be disclosed, but only when the United Nations Office of Legal Affairs ('OLA') had given its consent. That consent was given on 3 September 2016.
320. Miss Leonard has confirmed that the CPS had received Mr Arguin's report on 15 June 2016 and Mr Witteveen's memos on 11 July 2016, but said that she did not send them to the GoR at that stage because the CPS did not have the permission to do so of OLA or of the Dutch Prosecution authority. Eventually, the requisite permissions were received and "[all] three documents were added to the appeal bundles, encrypted on disc and served on the defence and Court on 29 September 2016." Miss Leonard says that "[due] to technical difficulties with the encryption process and the time involved in preparing the bundles for service, [the documentation] was not prepared and sent to Rwanda at this time". This chimes with what Mr Siboyintore has said. It is unclear to us to what extent those representing the Respondents in these proceedings were alerted to the inclusion of these documents in the bundles, but reference to the documents was made by Counsel in their Skeleton Arguments and in oral submissions. Thus it follows that at some stage in about October/November 2016 they became aware of Mr Witteveen's further written contributions.
321. We can take this general issue of disclosure no further except to observe that it is most unfortunate that important views, bearing on crucial issues in these proceedings, have taken so long to emerge and have done so in a somewhat piecemeal fashion. It is surprising that such important material was apparently not seen by Mr Siboyintore until so late.

The Witteveen/Arguin Debate (Resumed)

322. Returning to the chronology of events after the preparation of Mr Witteveen's Memo and before the SDJ's judgment, in the meantime there were further proceedings of relevance in Uwinkindi's case. It appears that, following the reporting to the President of the MICT of Uwinkindi's comments expressed in the Monitoring Report dated March 2015, consideration began to be given within the MICT procedures to an application for revocation of the referral. The formal application appears to have been made on 5 August 2015. The relevant dates are given in paragraphs 2-6 of the decision of the Trial Chamber given on 22 October 2015.
323. No oral arguments were heard in those proceedings and the matter was considered on the basis of competing written submissions. A number of matters were raised by Uwinkindi and on his behalf. Those of most relevance to the current proceedings were the suggestions (i) that the actions of the Minister of Justice denied Uwinkindi the choice of counsel, (ii) that insufficient funds were available to conduct a proper defence investigation and (iii) that his right to a fair trial was violated by the decision of the High Court to fine his counsel and by showing bias towards the prosecution.
324. Although the MICT Trial Chamber made no reference in its decision to Mr Witteveen's concerns about the Uwinkindi case, or to the substance of his concerns about defence representation generally in genocide cases in Rwanda, as expressed in his report of 3 June 2015, it does appear that the report was included in the written submissions made to the Trial Chamber and was raised in the arguments put before it: see paragraph 40 of the decision of the MICT Appeals Chamber dated 4 October 2016. The Trial Chamber was criticised by the Appeals Chamber for not referring to the report, aspects of which were said to have "clear relevance", but it was concluded by the Appeals Chamber that "consideration of the report as a whole does not substantiate Uwinkindi's contentions as to inequality of arms, and specifically, the inadequacy of the means available for the presentation of the defence in his case": see paragraph 56. We bear in mind that the further views of Mr Witteveen expressed in the confidential Memo were not available either to the Trial Chamber or the Appeals Chamber.
325. Mr Arguin was part of the prosecution team seeking a rejection of the Uwinkindi revocation application. The team did not, it would seem, disclose either the Memo in those proceedings or the substance of Mr Witteveen's views as expressed in the Memo. In his later report of 22 May 2016, Mr Arguin says that he "directed the prosecution's response" to this application and others by Uwinkindi and Munyagishari. In that report he addressed the disclosure of the Memo in the Uwinkindi proceedings:

"The memorandum was not disclosable under ICTR Rule 68 because it was not exculpatory in any sense. The ICTR and MICT did not recognize Mr Witteveen as an expert and, thus, his unsolicited memorandum containing only his personal opinions or beliefs was not regarded as evidence …. Uwinkindi, in all events, was well aware of the pending United Kingdom extradition proceedings as he attached various filings from that case, including copies of Mr. Witteveen's various reports, to his submissions in support of revocation. He did not, however, substantively develop any of the points raised in Mr.Witteveen's reports in his revocation request, despite the opportunity to do so."

326. Whether the document was disclosable in the Uwinkindi proceedings is, of course, not the test as to whether it was disclosable in the present proceedings. There can be no doubt that it was disclosable in these proceedings. As we have noted above, Witteveen's memorandum was received by Florida Kabasinga attached to an email of 19 June 2015, many months before SDJ Arbuthnot's decision was promulgated.
327. The Respondents in the present proceedings made a joint submission to the SDJ on the 11 November 2015 in relation to the Uwinkindi MICT proceedings, making the point (amongst others) that there was no reference in the decision to Mr Witteveen's concerns about the quality of representation as reflected in his Additional Report. Their submission, of course, made no reference to the confidential Memo because none of the legal teams was aware of it.
328. Whilst the appeal in Uwinkindi was proceeding, there were two decisions of the District Court of The Hague in which, in reliance on Mr Witteveen's report of 3 June 2015, extradition of two alleged genocidaires was stayed or prevented in preliminary relief hearings. In each of those cases the Court at first instance noted that the MICT Trial Chamber "did not include [the] report in its assessment and appears not to have had it." By the time those two cases reached the Court of Appeal in The Hague in July 2016 it was accepted that the MICT Trial Chamber had been aware of the report. In the Report of the appeal in the case of Jean Claude Iyamuremye dated 5 July 2016, for example, it is recorded that "[also] submitted on … appeal was a further memorandum from Witteveen of June 2016 entitled 'Why extradition to Rwanda is perhaps not such a good idea…'", along with the confidential Memo, Mr Arguin's report and Mr Witteveen's response to the Arguin report dated 1 June 2016. The first of these additional documents is not in evidence in the present proceedings, but its title suggests that it takes the stance that extradition to Rwanda may not be appropriate. The other documents referred to are now in evidence in these proceedings, albeit they were not so before the SDJ.
329. The report of the Dutch Court of Appeal's decision does not indicate which party to the appeal "submitted" the Memo (or indeed the other documents referred to), but it is to be noted that it was shortly after this appeal decision (and the decision in more or less identical terms in the other in the case of Jean Baptiste Mugimba) was promulgated, that the two Witteveen Memos and Mr Arguin's report were received by the CPS.
330. That hearing in the Court of Appeal in The Hague was conducted "on the papers" and the Court itself acknowledged (at [2.3] of its decision) that since the proceedings were "preliminary relief proceedings" there were "only limited options for investigating the facts". It also made clear that it started from the assumption that "in a constant series of rulings and most recently in the ruling of the MICT of 20 October 2015 in the Uwinkindi case, the ICTR and the MICT have ruled that the suspects transferred by the ICTR to Rwanda will receive a fair trial." The Dutch Court of Appeal continued by saying that these rulings of the ICTR and MICT "show that these tribunals have addressed the arguments of the suspects in detail and that they have dismissed the objections against the legal procedure under the Transfer Law in comprehensively reasoned judgments".
331. It is, with respect, entirely understandable that the Court of Appeal in The Hague, when considering the preliminary matters before it, should be influenced by what it perceived to be the "comprehensively reasoned judgments" of the ICTR and MICT and that those judgments should be "taken very seriously". However, that does not (or should not) preclude a closer examination of what was available for consideration by those "international legal forums" and to consider whether the material newly before us that was not available to those fora should be seen as altering the perspective to be applied to the decisions in those cases. Equally, whilst we pay due regard to the reasoning of the Court of Appeal in The Hague, we are not precluded from considering what conclusions we ought to draw from all the material that we have, even where it is the same as that Court had before it.
332. We will revert to our analysis of this material when we have summarised what Arguin said in his May 2016 report and what Witteveen said in reply.
333. Arguin's report was prepared for the purposes of the hearing of the appeals in Holland. It was not available to the first instance Courts, but was available to the Court of Appeal in The Hague. The report was very critical of Mr Witteveen's Additional Report and the contents of the confidential memo.
334. Arguin described Witteveen's Confidential Memo as "entirely unsolicited" and says that Witteveen did not consult him with regard to his assessment or conclusions. It will be recalled that Witteveen's account is that he sent the Memo because he had read some comments of Arguin on his (Mr Witteveen's) Additional Report.
335. Arguin's first substantive criticism is that Witteveen did not refer to the ICTR's "extensive referral proceedings and the robust mechanisms to safeguard in practice fair trial proceedings in Rwanda." He said that Mr Witteveen referred "only in passing and without adequate consideration … [to] the onerous standard that the ICTR Trial and Appeals Chambers applied in satisfying themselves that the accused in the referred cases will receive fair trials and that there has been no violation of the conditions imposed on referral." He suggested that Mr Witteveen did not consider adequately the safeguards afforded by monitoring and revocation. Mr Arguin referred to the provision made by the Security Council for the continuation of monitoring by the MICT of any indictments referred to by the ICTR to the Rwandan Courts.
336. We will endeavour to set against each of the various criticisms made the response of Mr Witteveen as set out in his further report dated 1 June 2016 prepared for the proceedings in the Court of Appeal in The Hague at the request of the defence in those cases. Interestingly, he indicated that he prepared this response reluctantly, after he had been informed that the lawyers representing the government of the Netherlands did not intend to seek his comments.
337. Mr Witteveen's response starts with the assertion that Arguin's reference to "the robust monitoring system the ICTR has put in place" is a "gross over-statement". He amplifies that assertion as follows:

"29. First of all, apart from the monitors appointed by the Registrar, the Prosecutor appointed his own monitor, who observed the proceedings and presumably reported to the Prosecutor. Arguin does not mention him and there may be reasons for this.

30. I have met the OTP-monitor twice in July and August of 2014. He was a retired Tanzanian judge. I found him quite professional and open. We had two discussions. I am not in a position to recount what he expressed to me. The fact of the matter is that I have not seen him again in Rwanda. Admittedly, I may have missed him as I have not been in all of the trial proceedings, but all the proceedings I was in, he was not there. I wonder what he reported and discussed with the Prosecutor and why we did not see him back. Did the Prosecutor not value his opinion?

31. The history of the official ICTR monitors has been marred with difficulties and issues. Throughout the years there have been many changes in the monitors. Not all monitors had a prosecutor or judge background or had experience with criminal trial proceedings themselves and none of the monitors have performed their tasks for any substantive time. Their reporting was at best superficial.

32. In February or March 2015 a new team of monitors were appointed, who were all members of the Kenya branch of the International Commission of Jurists. The monitors worked as a team from Nairobi and they were a huge improvement right from the start. I believe the quality of the reporting immediately improved. Whether that caused the President of the MICT to put a full chamber on the motion for deferral by Uwinkindi in April 2015, I do not know, but it was a remarkable, immediate change.

33. Yet, the monitors were bound by many limitations, as were the previous monitors as Arguin describes. In essence, they could only speak with the parties and other directly involved officials and they could not perform any analysis of what they found or express views.

34. Arguin tries to explain this as natural and logical, but I find it awkward. And even if you accept their limited mandate, it still does not explain why the monitors never described factually what I and others saw happening in the trial proceedings – e.g. as I have described from the Bandora case as well as other trial proceedings. I find it inexcusable that the monitors did not report on the apparent flaws of the defence in the various cases as they missed out on what was at the heart of the problem with the defence attorneys.

35. Even the trainee for the Dutch Embassy, who had no trial experience herself at all but sat on many of these trial proceedings, particularly in the Bandora case, with me, recorded exactly what was said and what the defence attorneys failed to do and say. She meticulously reported all of that to her Ambassador. Till today, I fail to understand why the monitors of the ICTR [albeit the 'old' monitors], who sat in the same proceedings, did not report any of that."

338. In short, therefore, he challenges the assertion that the monitoring is "robust" and suggests that, in the early stages, the monitors did not have the requisite experience and that their reporting was "superficial". Although he acknowledged that the monitoring changed for the better in February/March 2015, he said the parameters within which the monitors could report were very narrow.
339. In relation to Witteveen's views on the adequacy of the defence representation in Rwanda, Arguin says that Witteveen "gave little or no consideration to the comprehensive standards applicable to ICTR referral proceedings" and that this gap in the analysis, as he described it, is likely to have contributed to several fundamental "misimpressions about the Rwandan justice sector" which he then detailed. It is necessary for us to summarise the various points that he made in this connection.
340. Arguin said the following in relation to the first "misimpression":

"One fundamental misimpression is … that Rwanda has adopted an accusatorial or common law system for trials in transferred cases. Rwanda has traditionally been a civil law jurisdiction. Legal reforms instituted since 1994 have introduced aspects of common law practice, including expanded reliance on judicial precedent and cross-examination as a means of establishing truth. However, the Rwandan legal system cannot be categorized as either purely civil or common law; rather, it is a hybrid system that retains elements of both civil and common law practices."

341. He elaborated on this under a sub-heading of 'The conduct of the trial' as follows:

"37. Another aspect of Rwanda's hybrid legal system relates to the conduct of trials. Rwandan judges play a much more active role in questioning witnesses than in most common law systems. Although the Transfer Law applicable to referred and extradited cases allows for cross-examination of witnesses, it does not specify how cross-examination will be conducted. In keeping with its civil law traditions, many Rwandan judges conduct the questioning of witnesses, generally after seeking input from the prosecution and defence on the types of questions that should be asked. Increasingly, Rwandan judges are providing counsel with more leeway to conduct their own cross-examinations as has been seen in all of the transferred cases, including Uwinkindi and Munyagishari.

38. It also is not uncommon in Rwandan judicial proceedings for the accused to play an active role in the conduct of their defence. Accused persons are not restricted to merely watching the proceedings and listening to the arguments of their counsel as they are in many common law systems. Instead, accused persons are free to directly address the judges and respond to points made by the prosecution or testimony of witnesses. Counsel are present during these submissions and, when necessary, intervene to make any clarifications or additions that may be required to protect their client's interests.

39. Several of the criticisms Mr. Witteveen cites arise from his misimpression that trials in transferred cases are purely accusatorial in nature. He criticizes the adequacy of investigations without acknowledging the role of the judicial police in collecting both inculpatory and exculpatory evidence. He criticizes what he regards as the active role judges play in questioning witnesses without recognizing that this is a common feature in many civil law jurisdictions and without acknowledging Rwanda's efforts to expand the role of counsel in conducting cross­examinations. He also criticizes the accused's ability to make direct submissions without recognizing that, in Rwanda, this is a hallmark of the right to direct one's own defence. All of these criticisms contribute to Mr. Witteveen's overall assessment that Rwandan defence counsel are ineffectual but none of them adequately accounts for the essentially hybrid nature of Rwandan trial proceedings. Any assessment of fair trial rights in Rwanda must recognize this fundamental distinction."

342. Before indicating Witteveen's response, we observe that we consider it would be surprising if Witteveen, with all his experience of criminal proceedings, both within the civil system that operates in the Netherlands and in other international jurisdictions, misunderstood what he was watching, when he attended the various parts of the proceedings in Rwanda that formed the focus of his Additional Report. Witteveen's brief response was that Arguin's attempt "to save these defence attorneys, e.g. how they sit in silence next to their clients, by pointing at the hybrid system of Rwanda's justice system and the lack of any cross examination or defence, suggesting it could all be part of their strategy, which we are unaware of" was "unconvincing". He said:

"The question is not whether the Rwandan justice system is accusatorial or hybrid as [Mr Arguin] explains, but whether there is a balanced system where evidence for the defence and alternative scenarios for the prosecution case are presented …."

343. He referred back to those parts of his Additional Report which reflected on the inability or unwillingness of the defence to challenge the prosecution case in the accusatorial system that he said applied to the transfer cases that the SDJ accepted.
344. It is to be recalled that SDJ Arbuthnot considered that this general criticism of the defence in these cases was borne out by the material from one of the monitors, Ms Buff, whose reports she described as "detailed and impressive".
345. As indicated, Arguin said that Witteveen had wrongly questioned the role of the judicial police. Arguin said this:

"30. … Under Rwandan law, the judicial police are required to gather evidence both for and against the accused. While the judicial police are not investigating judges, they nevertheless perform the same function as investigating judges in other civil law jurisdictions. Like investigating judges, the judicial police are charged with preparing a dossier for each case that includes all of the evidence - both inculpatory and exculpatory - relating to the charges.

31. The judicial police also must consider requests from the prosecution and defence to conduct any additional investigations the parties may deem relevant. These investigations may be conducted either in Rwanda or abroad and the information generated from these additional investigations will be added to the official dossier. If the judicial police decline to conduct additional investigations requested by the defence, the defence may request the court or prosecutor to order them to do so.

32. The dossiers compiled by the judicial police are substantial. In the Uwinkindi and Munyagishari cases, the dossiers included the complete ICTR investigation files plus additional materials obtained by the judicial police in the course of their supplemental investigations. These dossiers each contained several thousand pages of documents, including dozens of witness statements. The dossiers prepared in relation to other transferred cases are likewise substantial. The Mugesera dossier, for instance, includes over 1,280 pages of documents obtained in Rwanda, plus two compact discs prepared by Canadian authorities in relation to Mugesera's extradition that contain approximately 5,000 additional pages of material. Mbarushimana's dossier is not yet complete but already contains approximately 500 pages, including documents obtained during the Danish extradition proceedings. The Bandora dossier was nearly 500 pages long. Although the size of a case file does not alone guarantee a competent investigation, the substantial dossiers compiled by the judicial police in transferred cases cannot fairly be described as "limited" or "rather basic" as Mr. Witteveen suggests.

33. One difficulty observed in relation to the cases referred to Rwanda for trial was that the accused and some of their counsel ignored the role of the judicial police. They attempted to bypass the judicial police by seeking independent defence investigations, without first demonstrating that the judicial police were unable or unwilling to conduct the investigation. The High Court rejected these requests not because it did not respect the right to an effective defence but because investigations in Rwanda are usually conducted by the judicial police."

346. Witteveen responds to this at some length. We will endeavour to summarise. He describes the judicial police as "one of these misnomers in Rwanda's judicial system" and says that "there is nothing judicial about the judicial police." He asserts that there is "just the Rwanda National Police, the RNP, with a regular Criminal Investigation Division ... [and that prosecutors] generally, do not instruct or oversee police investigations." He continued thus:

"For genocide investigations the NPPA uses police officers from the RNP who are physically located in the … GFTU, headed by Mr. Siboyintore. This gives the NPPA a slight advantage from the position where the RNP investigates genocide cases from their own offices."

347. Based upon his personal experience in the GFTU, Mr Witteveen said that the quality of the police investigators was "variable" and that most "were limited in their skills and experience" and that there had been a regular turnover of staff. He said that "critically" the prosecutors who litigated cases in Court "had no contact with the investigators whatsoever" and simply had to work with whatever was submitted to them. He then adds the following:

"I have never seen, read or heard that any investigator collected or was even interested in collecting exculpatory evidence or was instructed to do so [emphasis added]. They were pressured to produce a genocide investigation in the timeframe of one or two weeks at best and collect as many incriminating witnesses' statements as they possibly could. [Anecdotally], I know they dismissed or ignored exculpatory evidence when they encountered it."

348. His response to the assertion that the files contained many documents was as follows:

"… I have seen many files produced by investigators but none was more than some tens of pages with witnesses' statements no longer than two to five pages per statement. When the transfer cases contain hundreds of pages of documents and statements, it is because the jurisdiction that seeks to transfer defendants to Rwanda for trial, produces these documents and statements for Rwanda's use. It still puzzles me whether the prosecutors, who litigate the cases, use these foreign made documents."

349. He concludes by saying that Arguin's reliance on the role of the judicial police "is the best example how he [Arguin] uses a theoretical framework, based on law and regulations, to present a reality, while the reality is totally different." He says that the suggestion that defence attorneys can refer defence witnesses to the judicial police to be heard on their behalf "cannot be taken seriously" and that he does not believe that "a defence witness will ever accept to be heard by Rwandan Police officers."
350. We will return to our conclusions about this in due course, but we should refer to another feature of Arguin's criticism of Witteveen. It relates to defence investigations. Arguin says that a practice direction issued by the Chief Justice on 6 August 2015 "preserves the role of the judicial police in criminal investigations and clarifies the procedure to be followed when an accused seeks to conduct investigations beyond those conducted by the judicial police." He described the position as follows:

"To obtain public funding for defence investigations, the accused must first submit an ex parte request to the High Court, which examines the reasonableness of the request, taking into account among other things the availability of alternative means of obtaining the evidence by video-link or phone call. Once the High Court is persuaded that the request is reasonable and in the interests of justice, it will issue an order to the Ministry of Justice to extend the required funds. In addition, the practice direction provides guidance on the types of expenses that will be covered, e.g., economy class tickets and daily subsistence allowance."

351. Arguin suggests that the practice is "similar to procedures used by other jurisdictions, including the ICTR, to control the expenditure of public funds [and] avoids situations where the defence make sweeping requests for additional investigations that are little more than fishing expeditions." He refers to what happened in Uwinkindi as illustrative of the problem and said that "no jurisdiction in which I have [practised], including the ICTR, would have approved [the] request to essentially travel the world to meet unspecified witnesses without some threshold showing establishing the relevance or need for the evidence."
352. Witteveen does not appear to respond to this directly, but he repeats his view that the system in Rwanda is such that there is "insufficient time and money to develop any credible defence" and that Arguin's position "does not describe a reality".
353. Arguin attacked Witteveen's contentions about the independence of the RBA. He says that the RBA is established as an independent entity under Rwandan law and is responsible for maintaining a list of qualified advocates and ensuring discipline within the profession. The RBA maintains a list of those it considers qualified to represent those made the subject of transferred cases. The flat rate fee each would receive is 15 million Rwandan francs (which we understand to have been around £12,500 at the time of the judgment below.). This fee is said to be free of tax and does not prevent the taking on by the advocate of other cases. It excludes the cost of various administrative services and of any additional investigations that may be permitted under the Chief Justice's Practice Direction. He says that there are 68 counsel who have more than ten years' experience and he mentions specifically the advocate assigned to the Uwinkindi case, Mr Joseph Ngabonziza, a former judge of the Rwandan Military Court, who has experience of defending those accused of genocide and related crimes.
354. He also says that the independence of the RBA is demonstrated by its objection to a provision in a draft legal aid contract that "appeared to authorise the removal of counsel for criticising the government." Its objection, he says, was sustained and the offending provision removed.
355. Witteveen responds by repeating what he had said in his original report and memo and says this about the deletion of the contractual provision to which we have referred:

"[Arguin] defends the Bar Association by pointing to the fact the Ministry deleted a provision in the contracts with the defence attorneys which prohibited them to openly criticize the Minister and the Ministry, which [he] attributes to the Bar Association. I do not understand that. It was the Bar Association [that] has accepted this and other shameful provisions in the contracts in the first place. It was external pressure and critics that made the change."

356. Arguin, a member of the prosecution team in Uwinkindi, says that Mr Witteveen's concerns about what happened in that case are misplaced. He describes the conduct of counsel (Mr Gashabana) in the absenting of himself from the trial over the legal aid structure as representing "gross misconduct". The Court was right to sanction him and to adjourn the proceedings until the RBA had appointed a replacement. Any problems thereafter were entirely due, he said, to Uwinkindi's failure to cooperate with the successor counsel. He drew attention to the decision of the MICT which was that Uwinkindi had "unjustifiably refused to cooperate with his newly appointed counsel." Arguin said this:

"In light of the record established by the MICT trial chamber, Mr. Witteveen's assertion that Uwinkindi was "without any defence" since January 2015 is plainly mistaken. Counsel always was available to assist Uwinkindi; he simply refused to accept their services. Nevertheless, counsel did their best to protect his interests and obtained appropriate relief from the trial chamber, including re­opening the examination of prosecution witnesses and obtaining more time for trial preparation, including review of the case file Uwinkmdi refused to provide counsel. Regrettably, these events occurred after Mr. Witteveen submitted his report and, thus, were not available to guide his assessment of the competence of Rwandan defence lawyers, whose performance must, of course, be assessed on an individual not collective basis."

357. Witteveen's response was as follows:

"66. [Arguin] tries to defend the court in its position to oust the defence attorneys in Uwinkindi in January 2015 and continue with the trial proceedings and hearing all the witnesses in just a day or so, including defence witnesses, without Uwinkindi being represented. [He] all blames it on the defence attorneys and Uwinkindi and presents it as if every court would have handled the case like it was handled.

67. I contest that. I do not state that the situation was easy, but I remain that it was indicative of the questionable situation as I have tried to describe it. To finalize the case without defence attorneys and to hear all the witnesses, including defence witnesses, in a day or two, without any defence attorney present is not defendable. The case would have been concluded by the court if the Prosecution would not have requested to reverse the situation under pressure of the scrutiny of the international community and the motion for deferral in Uwinkindi in the MICT which had been assigned to a full chamber.

68. The best indicator that the wrong decision had been taken and the court lost its sense for fair trial rights, is the fact that it was the Prosecution who took the initiative to reverse the situation and apparently, there was no mechanism within the court to do that. I have tried to describe in my Memo to [Mr Arguin] why I believe the Court did not apply common sense themselves and what caused the court to proceed with the case as it did."

358. His final words in his response were as follows:

"69. At the end of the day, I ask the question why the cases in our respective jurisdictions must be transferred Rwanda. Even when that is the preferred option, but if there are questions about fair trial rights and other questions, if cases are not transferred to Rwanda, it does not mean those cases will not be adjudicated. Each of our jurisdictions are competent to try these cases and we have the resources and the experience to do that.

70. My expert opinion has been developed over time, after I got to know the situation in Rwanda. More importantly, my opinion is highly influenced by the fact that in these cases, defendants are charged with genocide, without doubt the most serious crimes of all crimes the world has come to deal with. For each of these defendants, there is much at stake and each defendant is facing a lifelong imprisonment.

71. If that is the situation, does that not oblige us apply the utmost care and consider whether we should try these cases in our home jurisdictions?"

Our Conclusions on this Debate

359. This debate is only one part of what we have to consider in these appeals. However, it is an important part of the picture. We have set out much of the debate because it needs to be publicly understood and because our analysis of it may be of significance. Our task is to see what impact, if any, it has upon the question of whether the SDJ was wrong to reach the conclusion she did, bearing in mind that she did not have the benefit of considering this material. If Arguin's report was accepted in its entirety, there would be support for an argument that she ought to have decided the case differently. If, however, the broad thrust of Mr Witteveen's position in the debate is reliable along with his concerns, together with those the SDJ heard and accepted, then they support the conclusion that there is a risk of a flagrant denial of justice if the Respondents were returned to Rwanda.
360. As the principle in Wiejak v Olzztyn Circuit Court prescribes, we begin from the starting point that the SDJ's findings of fact should be accepted. She arrived at the material findings of fact by substantially accepting the evidence of Witteveen. Whilst, of course, we have to look at the position afresh given the availability now of Arguin's views, it would be wrong to ignore the fact that the SDJ found Witteveen a compelling and reliable witness. Had she seen his further written contributions, we think it likely that she would have seen what he said as further evidence of his objectivity. For our part, we would accept that what he has said has required courage. It would be a wholly unjustified conclusion to accept Mr Lewis's contention that his words were those of an embittered man. Our reading of what he has said suggests that what he has said has been said more out of sadness than anger or embitterment.
361. The additional material from Witteveen appears to us to add to his authority. He has been able to establish his experience, and has a closer knowledge of how the system has actually operated in Rwanda in recent times than any other expert. Moreover, he has given evidence which was clearly difficult given his position. It has made his continued connection in the area problematic, putting the matter neutrally. Yet in breaking with the obvious constraints of his being an employee (in effect) of the Appellant, he has not taken an extreme position in opposition to the GoR. Rather he has maintained a nuanced and balanced position. The Courts expect clarity, reason and moral courage in an expert witness. It appears to us that Witteveen has shown those qualities here, responding to his growing acquaintance with the facts and his developing understanding of the true problems facing Defendants in Rwanda in such cases.
362. Given that we have not heard from Arguin, we are hesitant about rejecting in its entirety what he has said, particularly given that other Courts have acted upon it, albeit they too have reached their conclusions without his evidence being tested. However, as we have observed, he does come to the issues that have been raised from a particular viewpoint. We find it difficult to accept that his approach has not been influenced to a degree by that viewpoint. It is noteworthy that he could have been put forward as the GoR's witness in these proceedings. He was not.
363. We have considered Arguin's views with care, without of course having been able to form an assessment of him ourselves or in his case having the benefit of an assessment from the Senior District Judge. We take the view that his knowledge of the situation on the ground in Rwanda is simply not demonstrated to the extent of that of Witteveen's. We are concerned that there is a tone of defensiveness about some of his views. We are inclined, from the content of his report compared to the thrust of the evidence about the transferred cases and their progress, to accept that Witteveen is right: Arguin is relying on the theory of how things should be, rather than giving us a picture of how they really are.
364. Therefore, we accept the picture painted by Martin Witteveen, augmented not weakened by the material which has become available since the hearing below.

Conclusions on "Fair Trial"

365. We have addressed the critical issues sequentially in the necessary detail, and indicated our responses. The threads can be drawn together in a relatively straightforward fashion.
366. It is very highly desirable that those accused of genocide, where a proper prima facie case exists, should be tried in the country where those appalling criminal acts took place. We fully recognise that as a very important objective. Trial in international Courts, or the Courts of another country, represents second best, for all manner of reasons.
367. We also recognise fully that there should be no impunity for those guilty of such terrible crimes as are alleged here. Impunity should be avoided by any possible means within the law.
368. The "high bar" set by the law, as we have analysed it above, reflects those principles. If the objection to extradition were merely that the quality of criminal process was less than perfect, less than desirable, then such objection, whether viewed as a moral question, or judged against the test set by the law, would mean that extradition would proceed. However, whether viewed as a moral question or tested by the law, if the extraditees were to return to face trial which represented a flagrant denial of justice, then the case is altered. And the judgment on that question cannot be distorted by the desirability of trial in the country where the events took place, much less by such questions as Rwandan national pride, or the aim of building capacity or improving justice in Rwanda, much less of relieving international institutions of a longstanding expensive burden. We think it unlikely that returning defendants to face charges in an inadequate criminal justice system would tend to improve matters there: rather, it would be likely to reduce the pressure to change and improve.
369. In any event, such considerations are irrelevant either way. The question is more basic. If there is a real risk of a flagrant denial of justice, that means there is a real risk of the innocent being convicted. To extradite in the face of such a risk, even if motivated by a desire to repatriate the criminal process to the country where it should properly be conducted, would be no more and no less a wrong than it would be to permit a serious miscarriage of justice here.
370. As we have stated above, the evidence suggests that Rwanda has, if anything, become more of an illiberal and authoritarian state than was the case in 2008/2009. As was SDJ Arbuthnot, we are struck by the fact that these renewed requests from the GoR, relying on improvements in the legal system, come from a state which, in very recent times, has instigated political killings, and has led British police to warn Rwandan nationals living in Britain of credible plans to kill them on the part of that state.
371. We are also struck by the conjunction of hostile remarks by President Kagame and others, made during the currency of criminal proceedings, and the reported threats to Professor Reyntjens, Professor Longman, to other lawyers and now to Martin Witteveen, in other words threats not confined to political opponents, but affecting experts and lawyers critical of the criminal justice system in Rwanda. We are not in a position to evaluate these threats in detail. However, the deaths and disappearances which have been established, set beside the detention and mistreatment of defence lawyers such as Peter Erlinder, mean that there is sufficient material to show a real risk of political pressure and political interference in the justice system in Rwanda. It is for these reasons that we have concluded that the view of SDJ Arbuthnot on this issue is not "wrong" but may be, if anything, rather too sanguine, particularly in the light of the new material available to us.
372. We have to consider carefully if these cases are such as to attract such pressure on intervention. The GoR argues that they are not to be compared with the cases such as that of Ingabire and others: straightforward political opponents of President Kagame. We accept the distinction, but we must consider how far it operates. We understand the argument that it is in the best interests of the Rwandan authorities to conduct, or permit these trials to be conducted, fairly, since that is the best means to ensuring future extradition of those who have fled. Yet the evidence is not encouraging that that objective reasoning is sufficient to guarantee the outcome. We consider that there is a risk, unless clear conditions of guarantee are established, of interference and pressure in these cases. These Respondents are not in general active political opponents of the regime. We address the individual arguments below. However, they are high-profile defendants, as all agree. They are, with one exception, people with involvement in authority during the period of the genocide. It is easy to see how, even if not regarded as first-rank political targets, one or more of them might be subjects of real interest to the authorities and thus the focus of pressure on the judiciary.
373. As to the independence of the Rwandan judiciary, we note carefully the context. The evidence suggests that judges are not appointed unless they have party membership of the RPF. Their appointments have moved from being indefinite to appointment for definite terms. The Rwandan executive can achieve the dismissal of serving judges: it has done so in recent times in respect of around 40 individual judges. We are not in a position to say whether the suggested misconduct or corruption was established in these cases: it may be so. But the capacity of the executive to get rid of judges is established. In such circumstances, there can be little doubt that judges will feel exposed.
374. The evidence is that judges in the past have ignored or resisted pressure. However, as will be obvious, there will rarely be evidence when judges have acceded to pressure. There are many indications that judges would wish to act independently, but there are also indications, including those illustrated by the later Report and Memorandum of Witteveen, when the conduct of trials has been disturbing. Our conclusion here, once again, is that SDJ Arbuthnot is not "wrong". However, some of the material we have seen was not available to her. Our view is that the evidence points to some risk, depending on the evidence before them and the safeguards in play, that judges might yield to pressure from the Rwandan authorities.
375. We have concluded, likewise, that SDJ Arbuthnot was broadly correct in her conclusions as to the preparedness of witnesses to appear. It seems likely that there will be a high degree of apprehension, and that some would simply decline to appear. However, there is evidence that many witnesses have appeared in genocide trials, in the gacaca and conventional Courts. We consider that the advent of the WPU should improve conditions for witnesses, and it is to be hoped that will communicate itself to them. However, this must be a qualified conclusion, depending heavily on the actual conditions of trial, and on the critical issue of the defence capacity to fund, proof and deploy the witnesses in the first place.
376. We do not consider the judicial police to be a credible mechanism for obtaining and deploying defence witnesses. The overwhelming evidence, despite any theoretical function, is that they are not organised or tasked to perform that role, and nor do they do so in practice. We touch on this theme again below, when considering defence capacity.
377. Which point brings us to the critical point in the judgment of SDJ Arbuthnot. The closer we have read the evidence in this case, the firmer has become our agreement with the judge below that defence capacity is the vital element, the capstone, of the case. Whilst in the context of our court system adequate representation is of course important, other safeguards in the system such as responsible unbiased prosecution, witness protection, unchallenged and complete judicial independence taken together, mean that inadequate defence may be compensated for and a reasonable quality of justice delivered overall. Even in that context, it is well established that miscarriages of justice will occur, where defence representation is inadequate. However, in the context of Rwanda, with the difficulties and weaknesses we have identified, the presence or absence of effective defence is absolutely central. We are completely of one mind with the judge below on that point.
378. For the reasons we have identified, and need not repeat, the arrangements for defence in Rwanda are clearly inadequate. They would be inadequate even if the remainder of the criminal justice system was acceptable and the concerns which arise were not present. In an authoritarian state, where judicial independence is institutionally weak and has been compromised in the past, where there is established fear by witnesses, not all of which can be effectively countered, the existing arrangements are quite insufficient to ensure a reasonable fairness in the proceedings. The story of the existing post 2009 genocide cases, the "transferred" case, gives rise to real concern. We cannot rule out a degree of tactical complaint and manoeuvring by those defendants, but objectively the conditions are such as to give ample opportunity for complaint. Here as in other aspects of the case, we pay great regard to Martin Witteveen. We consider that SDJ Arbuthnot was right to rely on his evidence, and that evidence is heavily fortified on this issue by material coming too late for her consideration but which is before us. We cannot find reassurance from the evidence of James Arguin.
379. Focussing on the correct legal test, we have considered carefully whether the problems identified are sufficient to establish a real risk of a truly serious or flagrant denial of justice. We find, as matters now stand, that they are sufficient to do so. We also find, with the remarks of Lord Phillips in RB (Algeria) in mind, that the result of such a fundamental breach of the principles of a fair trial might be likely to lead to serious miscarriages of justice, in this context where, in the one completed transfer case, there has been a conviction for genocide, leading to a prison sentence of 30 years, itself regarded as lenient.
380. For these reasons, subject to points we indicate below, the appeals will fail.
381. It also follows that, in relation to ECHR Article 6 and the "fair trial" issue, the cross-appeals also fail.
382. We turn below to considerations affecting individual Respondents. Before we do so one further matter arises in respect of the "fair trial" issue. We have emphasised more than once the desirability that those facing a prima facie case should be tried in Rwanda, if this is possible within the law. In brief submissions, Mr Knowles has raised the prospect that further or additional undertakings by the Appellant might allay the concerns of the Court. There are precedents for doing so: see, for example, the approach taken in Florea v Romania [2014] EWHC 2528 (Admin) and Giese v Government of the USA [2015] EWHC 2733 (Admin). Given the seriousness of the offences alleged here, we are prepared to permit the Appellant a final opportunity to seek to assure the Court that credible and verifiable conditions will be in place, to overcome the legal bar to extradition upheld above. It should by no means be assumed that such assurances will successfully overcome the bar to extradition given the historical failure of the Appellant to cooperate in prosecutions in England, the serious concerns articulated earlier in this judgment, the length of time that has passed and the inherent difficulty of being able to rely on assurances. However, it seems to us right that the opportunity should be afforded to the Appellant to persuade us that satisfactory assurances can be given. Given the very long history here, there can be no question of any long delay before the matter is finally resolved.
383. On the facts of this case, any successful future assurances or guarantees would have to be detailed, formal and underpinned with significant diplomatic weight. They would have to include at least (1) adequate funding for investigation and development of defence cases and for representation in Court by experienced and properly resourced advocates, (2) assurance of admission to the Rwandan Bar for suitably qualified and experienced foreign lawyers as defence counsel, where desired, and (3) inclusion of at least one non-Rwandan judge in any trial, such judge to be suitably experienced and independent of any connection with the Government of Rwanda (for example an existing judge of another relevant international Court or tribunal).
384. As we have indicated, we intend to give the Appellant the opportunity to consider such detailed further undertakings, dealing with those points we have identified and any others considered appropriate. The Appellant must indicate by 4pm on Friday 18 August 2017 whether they wish to take up this opportunity. If so, detailed proposals must be filed and served by 4pm on 8 September 2017. The relevant Respondents, in respect of whom other bars to extradition have not been established, may if they choose respond in writing by 4pm on 29 September 2017. If such submissions are made, we will consider them on paper and if necessary reconvene a hearing to address the matters raised.

Cross-appeal by Vincent Brown/Bajinya

385. As indicated above, five grounds of appeal were advanced on behalf of Brown/Bajinya against the reasoning of the SDJ. During the course of argument two of those grounds became subsumed in two of the other grounds and so three substantive matters fall for consideration.

Ground 1 – Section 81, Extradition Act 2003 – "extraneous considerations"

386. Extradition may be barred by reason of the existence of "extraneous considerations": section 79(1)(b). Such considerations are defined in section 81 which is set out in Annex 1.
387. The test for establishing whether a bar under section 81 is made out is whether there is a 'reasonable chance' or 'reasonable grounds for thinking' or a 'serious possibility' that such events will occur: Fernandez v Government of Singapore [1971] 1 WLR 987, as applied in Hilali v Central Court of Criminal Proceedings No.5 of Madrid and another [2006] EWHC 1239 (Admin). It is contended that this is a lower threshold than establishing a real risk of a "flagrant denial of justice", which is the test under Article 6.
388. Whilst the tests may be phrased differently, we are not persuaded that the operation of the test under section 81 will result in any significant difference of outcome, certainly in this case, and, as did the Divisional Court in 2009, we consider that the issue can be subsumed within the general Article 6 considerations.
389. For completeness we should say that the factual foundation underlying Mr Jones's argument lies in the allegation made against Brown/Bajinya (substantially contested by him) that he was a member of the Akazu and a close associate of President Habyarimana (see Annex 2, paragraph 9, below). It is said on Brown/Bajinya's behalf that this accords him "high political status" and means that he is being pursued for extradition because of his "political opinions" and/or would be prejudiced at his trial because of those political opinions. The nature of the case against him is summarised at paragraph 5 above (where paragraph 44 of the SDJ's judgment is repeated). That case mirrors the allegations made in the request for his extradition.
390. As we have indicated, the Divisional Court in 2009 saw any issue of the nature arising under section 81(1)(b) as one that was subsumed within the Article 6 issue: see paragraphs 23 and 32 of the 2009 judgment, paragraph 32 reading as follows:

"… we do not consider that anything is added by the distinct submission of prejudice at the appellants' trial within the meaning of s.81(b) of the 2003 Act. We have already referred to the potential overlap between the scope of s.81(b) and that of Article 6. The appellants' contention that, being Hutu, they will suffer prejudice if they are consigned to the High Court of Rwanda is in reality a theme of their general case that they will not be fairly tried."

391. Nonetheless, Mr Jones submits that the conclusions of the Divisional Court on the fairness of any trial concerning Brown/Bajinya are relevant to this discrete issue upon which he seeks to rely. He draws attention to the Divisional Court's reliance on what Professor Reyntjens said in his report when he said that Brown/Bajinya "cannot expect to receive a fair trial in Rwanda given the nature of the charges against him and the political dimension to them" and that he is "said to have been a close associate of President Habyarimana".
392. We have accepted that the 2009 decision is a starting-point for consideration of the issues in the present proceedings, though not strictly binding as such. The SDJ considered the issue in the light of the evidence before her and that was the correct approach. She summarised the argument advanced on Brown/Bajinya's behalf in this context as follows:

"VB contends that the GoR approach is that all 1994 Hutu officials were involved in the genocide and the prosecution has been made against him because of his high profile in the diaspora."

393. Her conclusion (which covers similar arguments advanced by other respondents) can be found in the following paragraphs of her judgment:

"92. All the RPs argue that they are prejudiced by their ethnicity. I do not find there is any evidence of prejudice against Hutus in prosecutions for genocide related offences. The Gacaca courts were introduced to speed up the process of trial for the many thousands held in custody. The Gacaca laws enable those convicted to be released into the community and to complete community work, many of those were Hutu. Likewise the abolition of the death penalty in 2007 would have affected in particular Hutus charged with genocide.

93. …

94. I do not find that any role that VB has in the diaspora today is such that it would lead to a prosecution for alleged genocide in 1994.

95. A number of the submissions I heard I will look at in greater detail when I come to consider the Article 6 fair trial arguments as there is a considerable overlap between the two. As regards section 81(a), I conclude that there is insufficient evidence for this court to find there is a reasonable chance, a serious possibility that the request for the RPs' extradition (though purporting to be made on account of the extradition offence) is in fact made for other purposes including their political opinions or by reason of their ethnicity. As can be seen later in this judgment I have found there is a prima facie case and I do not find that the case has been constructed by the GoR to punish and imprison these men just because they … had positions of power in 1994.

96. As regards section 81(b), there is a considerable overlap with Article 6 which I look at later in the judgment. I adopt the approach of the Divisional Court in Brown and others at Paragraph 32 where the Court did not consider that anything was added by the distinct submissions of prejudice at the RPs trial with-in section 81(b); that contention is in reality a theme of their general case that they will suffer prejudice if they are consigned to the High Court of Rwanda where they will not be fairly tried. I do not find that there is a reasonable chance or a serious possibility that they might be prejudiced at their trials with-in the meaning of section 81(b) by reason of their political opinions or ethnicity. I find their extradition is not barred by reason of extraneous considerations."

394. Mr Jones submits that her conclusions that there was no evidence of prejudice against Hutu and no reasonable chance of prejudice due to political opinions were "unsustainable". He also draws attention inter alia to what he says was the unchallenged evidence of Professor Reyntjens concerning the relevance of membership of the Akazu, the agreed note of which was as follows:

"Calling someone part of Akazu is something special – means that those who say it consider them as being party or privy to the central organisation of the genocide. Although this would not be fair to a number of people who I would consider part of the Akazu but played no part in the genocide. Calling a person a member of Akazu is a serious indictment."

395. He had said in his report, echoing what he had said in the earlier proceedings, as follows:

"In my opinion, the present Rwandan judicial system, embedded as it is in the Rwandan political system, is not capable of putting into effect the guarantees necessary in the present case. [VB] cannot expect to receive a fair trial in Rwanda given the nature of the charges against him, the allegation that he is a leading Hutu activist and the political context in which the charges are formulated. There is no prospect that a judge, operating under the current arrangements in Rwanda, will be able to act independently of the current pervasive RPF control. His or her decision will be subject to the will of the Rwanda government."

396. In the same passage of evidence as that referred to above, Professor Reyntjens was asked if he was saying that Brown/Bajinya cannot have a fair trial in Rwanda because he was a member of the Akazu. His response was:

"I have never said that no one who is extradited can possibly have a fair trial. I have expressed concerns about fair trial in this case and other cases. It is simply impossible to say that there is any certainty that these men or anyone extradited will not have a fair trial, but I have said likely will not have fair trial."

397. The Appellant makes the observation that Professor Reyntjens did not say that Brown/Bajinya was "at particular and specific risk of an unfair trial by reason of allegedly being an Akazu", but that he relied simply on his general opinion that the Respondents will not get a fair trial.
398. The evidence of Mr Witteveen is also of relevance in this context. This passage of cross-examination by Mr Jones reveals his view:

Q: It is alleged he was a member of the Akazu. If there is an allegation that a person is a member of the Akazu, does that not point to Government of Rwanda as having regarded this man as having been in a very important political position?

A: Yes, I think so.

Q: Doesn't it point to their current attitude to him, that they identify him as such? Doesn't it point to them regarding his former activities, as important?

A: Yes, I think so.

Q: In your view, ought a court to take that into account in assessing whether Brown/Bajinya can have a fair trial?

A: I don't understand your position. Everything is important to consider but as I have explained I do not see why a fair trial, with all comments I have made, would not be achieved, and why a high quality attorney would not be able to investigate properly.

399. As it seems to us, the decision of the SDJ, reflected in the paragraphs of her judgment quoted above, was not wrong in the light of the evidence to which we have referred briefly save that it may also have been a little more sanguine about how alleged membership of the Akazu would be perceived by the GoR than we would be in the light of the additional material we have seen. Simply asserting, as does the Appellant, that the issue of whether Brown/Bajinya was a member of the Akazu is a matter of fact is an insufficient answer to the point advanced on his behalf. However, what Mr Witteveen was saying in the passage to which we have referred is that with high quality representation the significance of the allegation that Brown/Bajinya was a member of the Akazu could be dealt with fairly. In the light of what is now known about Mr Witteveen's enlarged views, the added precaution of the addition of an independent judge to the panel would, in our judgment, be necessary to operate as an additional safeguard.
400. At all events, having analysed this ground of cross-appeal, we do not see it as adding anything to our analysis under Article 6.

Ground 2 – section 82 – the passage of time

401. Mr Jones submits that the SDJ was wrong to reject the argument that it would be unjust and oppressive to extradite Brown/Bajinya, relying on section 82 (see Annex 1). A large part of this argument is based upon what, in his Skeleton Argument, Mr Jones had characterised as "the forum argument" which, in short, is that the allegations of genocide against Brown/Bajinya should have been investigated and, if appropriate, tried in this country, an option that has been available since 2010.
402. The SDJ's decision can be found in the following paragraphs of her judgment:

"101. The RPs have not shown on the balance of probabilities a risk of prejudice in their defence by the passage of time other than the sort of difficulties facing any defendant who is being tried in relation to allegations that are historical. The prosecution and defence witnesses are still available and subject to my findings in relation to Article 6 the RPs are able to defend themselves. The work carried out by the investigators working on their behalf means that they are probably better equipped to defend themselves than many others facing similar charges in Rwanda.

102. There was delay between 1994 and the last proceedings whilst Rwanda attempted to piece together its justice system. The failure by the GoR to allow investigation by the British police post 2009 prior to a possible prosecution here does not mean the delay between 2009 and 2015 is such that I find extradition of either VB or EN would be unjust. It is understandable that Rwanda wishes to try these allegations in their own country; the alleged offences took place in Rwanda and that is where the prosecution witnesses are.

103. As to oppression, any attempt at extradition is followed by a long and hard process which causes anxiety to not only the RP but also his family. I accept VB has lived in London openly with his family since the last proceedings. I accept that he was in custody for two years and three months between December 2006 and April 2009 and for over six weeks before he was granted bail during this set of proceedings. I do not find he would have had a sense of security after the last proceedings ended. He was never told that the GoR would not pursue another extradition request. I have taken into account too as I am required to the gravity of the allegations. I do not find any culpable delay on the part of the GoR nor do I find that the RP's circumstances have changed significantly since 2009. I do take into account that both VB and EN's family will be greatly affected if he is extradited to Rwanda. In all the circumstances I find that extradition of VB or EN is not oppressive such that it comes within section 82 …."

403. The Divisional Court in 2009 dealt with the position concerning the passage of time until then: see paragraphs 137 to 138. We do not think that there is any need to add to that analysis and Mr Jones' principal target was not in relation to that period, but to the period since.
404. In his Skeleton Argument he contended thus:

"The further delay since the 2009 judgement has been caused entirely by the requesting state's refusal to allow the Metropolitan Police to investigate the allegations with a view to prosecution in the UK. The Judge failed to consider at all the failure of GoR to cooperate with the UK without meaningful explanation, in contrast to the various jurisdictions with which it was contemporaneously willing to assist in ensuring prosecution of Rwandans under universal jurisdiction in the foreign state."

405. The position is that as from 2010 it would have been possible for Brown/Bajinya (and others accused of genocide) to be tried in the UK: section 70, Coroners and Justice Act 2009. Brown/Bajinya has wanted to be tried, if he is to be tried, in the UK. Representations were made to the CPS by him and on his behalf on this basis and our attention has been drawn to various communications relating to this. However, a quotation from a letter from the Head of the Special Crime and Counter Terrorism Division dated 17 October 2013 is sufficient to confirm the position:

"Following the High Court decision in 2009 that Dr Brown and three others should not be extradited and the introduction of the amendments to the International Criminal Court Act made by Section 70 of the Coroners and Justice Act 2009 an investigation was commenced in 2010 by the Counter Terrorism Command of the Metropolitan Police into alleged war crimes committed by Dr Brown and the three others. A Letter of Request for mutual legal assistance was sent to the Rwandan Authorities by the CPS to establish if the Rwandan Authorities would assist the Metropolitan Police in their investigation. The CPS received a reply from the Prosecutor General stating that the Rwandans were not prepared to "cede jurisdiction to the UK authorities" and would not provide copies of their evidence to the police. Without the co-operation of the Rwandan Authorities an effective investigation could not be carried out and therefore a domestic prosecution in the UK is not possible."

406. It is contended by Mr Jones that the GoR has been taking an unreasonable stance in relation to the possibility of prosecution in the UK, particularly in light of its co-operation with other countries that have taken jurisdiction for dealing with offences of genocide. The domestic jurisdictions in Holland, Germany, Finland, Sweden, Belgium, France, Canada, the United States of America and Norway have each undertaken the prosecution of Rwandan genocide cases under the principle of universal jurisdiction. It is argued that it is to be inferred that the GoR co-operated in those cases. No reason for the decision of the GoR not to co-operate in the prosecution of Brown/Bajinya and others in the UK has been given. Ms Ellis, on behalf of EN, has argued that "such a vehement rejection of a similar request by UK authorities suggests that particular political importance has accorded to this case by the executive for reasons other than the interests of justice." It is also argued that there is no concept of "ceding jurisdiction" to a domestic Court. Mr Jones referred to the CPS "Guidelines on the handling of cases where the jurisdiction to prosecute is shared with prosecution authorities overseas." He says that the title of these guidelines "recognises correctly that jurisdiction is not something inherent in one state, which it can "cede" to another state." States, he submits, "have concurrent jurisdiction over crimes such as genocide."
407. We accept that, strictly speaking, Mr Jones is right about that, but any prosecution that is to be presented fairly requires both the prosecution case and the defence case to be deployed properly. If the GoR will not co-operate in providing the evidence for the prosecution case, that overall objective cannot be achieved. Views may differ about whether the GoR is acting reasonably in taking that position in this case or whether there is a political motive behind it, but it is that country's entitlement to argue that Brown/Bajinya (and the others) should be tried in Rwanda where, all things being equal, it would be best and most advantageous for the trial to take place. Equally, as Mr Knowles says, if the GoR cooperated with the UK authorities, ultimately the decision about whether to prosecute would be left to the independent prosecuting authority (the CPS) in the UK and would thus be out of the control of the Rwandan authorities.
408. Mr Jones has sought to place this aspect of Brown/Bajinya's case under the heading of "delay" on the basis that the GoR's stance means that any putative trial in Rwanda would not be complete for three or more years after his extradition (say, in 2018), such a timescale being consistent with those transferred cases presently progressing through the Rwandan High Court, whereas any trial in the UK could have been completed much earlier.
409. We accept that it is at least arguable that the position taken by the GoR will add some period of delay before any trial in Rwanda would take place if extradition is ordered, compared with the position had there been proceedings in the UK, but it is not clear how significant any such delay would be: there would have been considerable preparation before any trial in the UK was ready and arrangements made for witnesses from Rwanda and elsewhere to attend.
410. However, at the end of the day, the question is whether the passage of time renders it unjust or oppressive to extradite Brown/Bajinya. We consider that the SDJ's rejection of this argument was not merely one to which she was entitled to come, but was indeed correct.

Ground 3 – no sufficient case to answer – section 84(5)

411. The issue raised is whether "there is evidence which would be sufficient to make a case requiring an answer by [Brown/Bajinya] if the proceedings were the summary trial of an information against him": section 84(1) and (5).
412. The SDJ summarised the position as she saw it thus:

"110. In 2009 [District Judge] found a prima facie case against [VB] and this finding was accepted by the Divisional Court. I accept [Mr Jones'] argument that the evidence is not entirely the same as it was in 2007-8. The [judicial authority] has withdrawn some evidence and there is new evidence obtained since.

111. I have set out Mr Jones's submissions and the evidence relied on in an appendix to this judgment. In summary there are a number of witnesses relied upon by the GoR. They have given contradictory accounts which are set out in full in the appendix. Mr Jones relied heavily and justifiably on the fact that none of these witnesses who had been witnesses for the prosecution or indeed defendants in a number of other proceedings, had ever mentioned VB. I have concluded that this is a weak case against Dr Brown, that all of the witnesses have been undermined to a lesser or greater extent but that the true weight of the case against VB cannot be judged without hearing the witnesses on both sides give evidence and be tested in cross-examination. The evidence is not worthless. I find there is evidence which would be sufficient to make a case requiring an answer from him."

413. Mr Jones challenges that decision and submits that on a proper analysis of the evidence, no reasonable tribunal, properly directed, could convict and the SDJ would have so concluded if the correct legal test had been applied. He submits that the evidence taken as a whole is so weak, inherently unreliable, mutually and internally inconsistent that no reasonable tribunal, properly directed, could convict.
414. The SDJ had been presented with an Appendix to the closing submissions on behalf of Brown/Bajinya running to some 175 pages in which the evidence concerning his alleged involvement with the genocide was analysed in close detail. The SDJ herself scrutinised this analysis in 133 paragraphs. She considered all the evidence, including the evidence adduced by Brown/Bajinya, and her conclusion, as set out in paragraph 133 of the Appendix to her judgment was as follows:

"My overall conclusion is that this is a weak case against [VB], that all of the witnesses who say they saw him instigating and encouraging killings have been undermined to a lesser or greater extent but that the true weight of the case cannot be judged without hearing the witnesses give evidence and be tested in cross examination. I find there is a prima facie case against the RP. I do not find the evidence is worthless. What I do find though is that for Dr Brown to have a fair trial he will need to have an experienced defence team including an investigator to be able to marshal a great deal of material for cross examination of the prosecution witnesses and many reluctant witnesses for his defence."

415. Ultimately, an issue such as this involves the making of a judgment about the state of the evidence as it appears to the decision-maker to be. We do not consider that the SDJ's assessment of the state of the evidence can be said to be wrong. Indeed, reflecting on her analysis of the evidence, it seems to us that her judgment was entirely correct: there are weaknesses in the case against Brown/Bajinya which, if exposed competently, may well lead to sufficient doubt about his alleged involvement to result in an acquittal. However, that is for the trial, not for the tribunal considering extradition. Nonetheless, it adds emphasis to the need for a competent defence team to ensure that Brown/Bajinya has a fair trial.
416. We do not consider that the SDJ was wrong to reject this submission or to reject the submission based upon the alleged fabrication of evidence against Brown/Bajinya. If that was the case, it will be revealed by a competently presented challenge to the prosecution case and a competently presented defence case.

Ground 4 - extradition incompatible with Convention rights - section 87(2)

417. This ground was not abandoned as such, but the considerations arising were said by Mr Jones to be subsumed within the "extraneous considerations" issue dealt with earlier.
418. We say nothing further about this. In our view, Mr Jones was correct to say that the issues could be dealt with satisfactorily under that heading.

Ground 5 - The Forum argument

419. The points made under Ground 2 concerning the implications of the GoR's non-cooperation with a prosecution in the UK were made as a separate ground of cross-appeal, but were subsumed in the argument under Ground 2. We need say nothing further about those points.

The Cross-Appeal of Charles Munyaneza

420. Mr Moloney indicated that he would adopt the arguments of Mr Jones on matters raised in the cross-appeal of Munyaneza. Those matters have been dealt with previously and there is no need to repeat our overall conclusions save to emphasise that, in our view, the SDJ was entirely right in relation to the issue of whether there was a prima facie case in respect of Munyaneza.

The Cross-Appeal of Emmanuel Nteziryayo

421. The Skeleton Argument of Ms Ellis and Ms Evans deal with a number of points raised in the cross-appeal on behalf of Nteziryayo. Some raise broadly the same issues as those raised by Mr Jones on behalf of Brown/Bajinya (although sometimes labelled somewhat differently) and we propose to concentrate on those that raise mattes that are specifically raised by Nteziryayo.

"Political trial"

422. Our general response to the suggestion that the SDJ was wrong to approach these cases on the basis that they do not come within the category of "political" cases has been dealt with previously. The summary of the allegations made against Nteziryayo appear in paragraph 5 above. They include the allegations that he was a bourgmestre (which is not disputed as a matter of fact) and that he was a member of the MRND, both suggested to be matters that would mean that his "political affiliation" would be such as to make him a target of an unjustified claim that he was a party to the genocide.
423. In paragraph 93 of her judgment the SDJ said this:

"As to whether the RPs who are bourgmestres and have other positions of authority, would be prejudiced by the roles they played in 1994, I rely on the evidence of Martin Witteveen [who said] that in all his observations of the trials he had not seen any sign that anyone who had a leadership role in 1994 was presumed guilty. In the many exhibited ICTR monitors' reports examining the trials of Uwinkindi and Munyagishari there is no evidence of prejudice in relation to the local positions they held in 1994."

424. As indicated elsewhere, whilst we think this is somewhat too sanguine a view, albeit not wrong, it is something that could be addressed by a competently presented defence case before an independent tribunal.

"Worthless evidence"/no prima facie case

425. Ms Ellis submits that the evidence upon which the GoR would seek to rely is "demonstrably worthless" and would, in any event, be such that on a proper analysis no reasonable tribunal properly directed could convict Nteziryayo. She relies, in particular, upon what she says are "the irreconcilable differences between the statements presented for the extradition proceedings with statements from a number of the same witnesses, which are totally exculpatory and were provided, sometimes under oath and voluntarily, many years earlier." In other words, the suggestion is that these witnesses have changed their accounts subsequently to accommodate the desire of the GoR to convict Nteziryayo.
426. She submits that since the Divisional Court case in 2009 the evaluation of the case against him should change. Her submission and the response of the SDJ can be discerned from the following two paragraphs of the SDJ's judgment:

"124. In 2009 the Divisional Court concluded that in relation to the GoR's evidence found in a large number of witness statements "taken at face value there can be no doubt that the material in these statements was sufficient to make a case requiring an answer from each of the appellants" …. [Ms Ellis and Ms Evans] argue that in the light of the new material they have obtained, where much of the material is exculpatory, the court no longer has to take 'at face value' the prima facie statements. Some of the new material consists in previous statements made by the GoR's witnesses which are inconsistent with the statements relied on by the Ngoga. Miss Ellis argues that the veracity of the evidence is undermined.

125. I have set out in full Miss Ellis' arguments and evidence in the appendix to this judgment. I find that the witnesses are undermined to a lesser or greater extent but should be cross examined to determine how much weight a court should attach to their evidence. The defence evidence is strong and should be put before a court. I do not find the prosecution evidence is worthless (subject to what I have said in the appendix in relation to named prosecution witnesses). I find in short there is a prima facie case."

427. It will be apparent that Ms Ellis challenges that conclusion.
428. Appendix 3 to Ms Ellis' submission to the SDJ contains a formidable and persuasive "deconstruction" of much of the evidence relied upon in support of the extradition of Nteziryayo. As the SDJ said, a number of the witnesses are "undermined". It contains also a good deal of evidence which, at face value, supports Nteziryayo's case that he was not involved in any of the genocidal acts alleged against him.
429. As with Brown/Bajinya, we consider that the view taken by the SDJ reflected a "judgment" on the evaluation of the evidence, both prosecution and defence, that was open to her. What is required at a trial, if it takes place, is an effective defence on Nteziryayo's behalf which is capable of exposing the suggested weaknesses of the prosecution evidence and advancing the strengths of the defence evidence in front of a tribunal that is not susceptible to influence by the GoR.
430. We consider that the SDJ was correct to reject this aspect of the case advanced by Nteziryayo notwithstanding the persuasive nature of much of the material relied upon on his behalf.

Double jeopardy

431. The rule against double jeopardy is set out in Annex 1: section 80 of the Act.
432. The nature of the argument advanced and the SDJ's response can be seen from the following paragraphs of the SDJ's judgment in which she quotes from Mr Ngoga's affidavit (see paragraph 45 above):

"71. …. EN was convicted and sentenced in Tare I Gacaca proceedings on 30th October 2008. He was charged with the massacre of Tutsi at Nyamigina, setting up of roadblocks, inciting people to commit genocide and issuing identity papers on the basis of ethnicity. He was convicted of talking people into killing in different places, chairing meetings aimed at committing genocide and setting up roadblocks.

72. The Tare I decision was set aside by the Gacaca Appeal Court on 22nd December 2008 on the basis that the court "did not have jurisdiction to hear the case" (Ngoga's Affidavit Bundle 1 Paragraph 207).

73. A second case was heard in Tare II which ended in his acquittal on 23rd October 2008 for lack of sufficient evidence to suggest he committed the alleged crimes (Paragraph 208 ibid). This too was appealed and the acquittal was annulled on the basis that Tare II a trial "had already commenced in another jurisdiction" and "the Gacaca Court of Appeal has orders that the verdict given by the Sector Gacaca Court be annulled, with the case to follow the normal process in the prosecutor's office that had started to prosecute it".

74. In December 2008, five of the judges in Tare II were interviewed as to how they had acquitted EN and to find out who had leaked the verdict of the court. The judges were held in custody whilst they were being questioned.

75. Miss Ellis makes the point in her submissions that the records of these Gacaca hearings are incomplete and contradictory. I accept she is right. I have concerns about all Gacaca proceedings on the basis that they are clearly not Article 6 compliant, a position that the GoR also adopts. I find there is no evidence however that the conviction and acquittal have not been annulled lawfully. His extradition is not barred in relation to these proceedings."

433. We will return to that reasoning below, but news of these developments emerged at or about the time of the hearing before the Divisional Court. The hearing took place in December 2008 and the judgment was handed down on 8 April 2009. The Divisional Court said this:

"146. Towards the end of the hearing evidence emerged that CU and EN had been tried in their absence before gacaca courts. In the case of CU an acquittal was declared by the relevant Gacaca Appeal Court to be a nullity on the ground of lack of jurisdiction. In the case of EN there had been a conviction of certain offences at one Gacaca Court, and an acquittal of other apparently similar offences in another Gacaca Court. The evidence was still emerging as the hearing before us concluded. Professor Schabas told the judge that there was no possibility whatsoever that the appellants could be tried by the gacaca courts. It is not clear whether this surprising turn of events was simply a case of the left hand not knowing what the right was doing, or an indication of something more sinister. Had we been minded to reach a different conclusion on the fair trial issue it would have been necessary to explore the implications of these gacaca proceedings in more detail.

147. Further material relating to the gacaca proceedings and also to a number of other issues that had been raised during the hearing was supplied to us after the hearing had concluded. It is unnecessary to refer to that material since none of it casts any doubt on our conclusions … in respect of the principal issue in the appeals."

434. The chronology of the events set out in the SDJ's judgment is at first reading a little confusing. We will set out again what we understand to be the chronology, but it has to be said that some of the material emanating from Rwanda about these events (particularly about the gacaca trial that led to a conviction and the appellate process thereafter) is itself confusing:

September 2007 – May 2008: extradition proceedings before DJ Evans

6 June 2008: decision of DJ Evans in favour of extradition

1 August 2008: Secretary of State's decision to extradite

? August/September: appeal to Divisional Court launched

9 October 2008: first day of gacaca trial in Tare II (in absentia)

16 October 2008: second day of gacaca trial in Tare II

23 October 2008: third day of gacaca trial in Tare II at which EN was acquitted of all charges (setting up road blocks and distributing guns)

30 October 2008: a further gacaca trial took place in Tare I at which EN was convicted on the basis of the evidence of one witness (the charges were of massacring Tutsi at Nyamigina, setting up roadblocks with police, inciting people to commit genocide and issuing identity papers on the basis of ethnicity).

15 – 18 December 2008: five of the gacaca judges in Tare II were interviewed and asked about the reasons for EN's acquittal. The interviews were conducted by prosecutors from the NPPA and a Public Investigator from the Head Office of the Parquet General of the Republic in the GFTU. There is evidence that the judges were imprisoned during this period.

18 December 2008: gacaca court of appeal for Tare II annuls the acquittal, the following reasons being given:

"After considering that the Defendant is being prosecuted in the ordinary jurisdiction, the Gacaca Court of Appeal has orders that the verdict given by the Sector Gacaca Court be annulled, with the case to follow the normal process in the ordinary jurisdiction that had started to prosecute it."

22 December 2008: gacaca court of appeal for Tare I annuls the conviction, the following reason being given:

"After considering that the Defendant is being prosecuted in the ordinary jurisdiction, the Gacaca Court of Appeal has ordered that the verdict given by the Sector Gacaca Court be annulled, with the case to follow the normal process in the ordinary jurisdiction that had started to prosecute it."

435. We are told, and it has not been disputed, that during the first extradition proceedings in 2007/2008 it was said on behalf of the GoR that none of the Respondents would be the subject of gacaca proceedings at any time, an assurance reflected in what the Divisional Court recorded Professor Schabas as saying.
436. The Divisional Court in 2009 questioned whether these events afforded an indication of something "more sinister" than simply an instance of the "left hand not knowing what the right was doing". That is not, strictly speaking, the issue before us, but it is very difficult to resist the inference that steps were being taken by the prosecuting authorities in Rwanda to shore up its position vis-à-vis Nteziryayo either because of the recommendation for extradition or because of the known existence of a pending appeal to the Divisional Court. In the submissions made to the SDJ on Nteziryayo's behalf, Ms Ellis and Ms Evans contended that:

"had it not come to the attention of [the Divisional Court] in December 2008 that the gacaca proceedings had taken place, then, had the extradition application been successful, Nteziryayo would have been taken to prison on his return as a convicted prisoner."

We think there is some force in that submission.

437. However, the issue for double jeopardy purposes is whether, as Ms Ellis submits, the SDJ was wrong in law to hold that the acquittal and conviction had been annulled lawfully. If neither had been set aside lawfully, they would still be effective and, it is argued, Nteziryayo would still stand convicted of the charges which reflect those for which his extradition is sought in these proceedings.
438. The contention advanced by Ms Ellis is that, according to the provisions of the relevant law, the gacaca courts had jurisdiction to try 'leaders' charged with genocide as a result of an amendment to the Organic Law in May 2008. To that extent, within Rwandan law, she argues that the proceedings that led to Nteziryayo's acquittal were valid. If that is correct, the rule against double jeopardy bites and Nteziryayo should not be extradited. Furthermore, if we understand the argument correctly, since the gacaca proceedings are plainly not Article 6 compliant, there could be no basis for extraditing him to serve the sentence imposed. If both acquittal and conviction have been lawfully annulled, there is no impediment to extradition on the basis of double jeopardy.
439. The GoR places reliance in this context on Article 2 of the Transfer Law 2007 which, so far as material, reads as follows:

"Notwithstanding any other law to the contrary, the High Court of the Republic shall be the competent court to conduct on the first instance the trial of cases transferred to Rwanda as provided for by this organic law …."

440. The GoR has sought to argue that by virtue of this provision there was no jurisdiction to try Nteziryayo in the gacaca jurisdiction and, accordingly, the Appeal Courts were correct to annul both decisions because it was the High Court that had the jurisdiction in his case, not the gacaca courts. Ms Ellis submits that Article 2 on its face is only relevant to someone who has been "transferred" to Rwanda and that since Nteziryayo had not been "transferred" in October 2008, he was not subject to Article 2. Strictly speaking, an order for his transfer (by way of extradition) had been made, but its operation was suspended pending the appeal to the Divisional Court which, in due course, quashed the order.
441. Ms Ellis argues that the SDJ was wrong to say that there was no evidence that the gacaca rulings were not annulled lawfully: she says that it is plain, as a matter of Rwandan law, that the reason given by the Appeal Courts for annulling the rulings was flawed.
442. In this respect, we consider that Ms Ellis's argument is correct. The evaluation of the issue of double jeopardy in Nteziryayo's case does not depend on determining whether there was evidence that showed the lawfulness or otherwise of the annulment: it depended on interpreting the provisions governing the gacaca courts to which we have referred. In our judgment, the gacaca court did have jurisdiction to determine Nteziryayo's guilt or innocence according to Rwandan law at the time and, accordingly, the gacaca Court of Appeal had no power to declare that the acquittal was a nullity through lack of jurisdiction.
443. That being so, either through the operation of the double jeopardy provision per se, or by reference to the principle of abuse of process, Nteziryayo should not be extradited unless there is some other provision of Rwandan law that clears the path for this to occur. The only path upon which Mr Knowles seeks to rely is Article 8 of the Organic Law to which we refer below at paragraph 457, but which we repeat here for convenience:

"Article 8: Trial of an Extradited Person Sentenced by Gacaca Courts

A person extradited to be tried in Rwanda and who has been sentenced by Gacaca Courts should be tried by a competent courthouse provided by this organic law.

However, the decision of the Gacaca Courts shall first be nullified by that Court."

444. This provision on its face plainly applies only to those who have been sentenced (and thus convicted) and not to those who have been acquitted lawfully. In our view, it has no role to play in Nteziryayo's case.
445. In our judgment, Nteziryayo is entitled to be discharged on this ground.

Article 8 ECHR

446. To the extent that this remained an issue, it falls away in the light of our decision on double jeopardy. However, for completeness and for the avoidance of doubt, we do not consider that Article 8 considerations could possibly operate to prevent extradition in a case involving allegations of genocide.

Celestin Mutabaruka: Double Jeopardy and Abuse of Process

447. As we have indicated above (see paragraph 11), the Senior District Judge summarised the allegations against Mutabaruka in paragraphs 48-51 of her judgment. For convenience, we reiterate that he is alleged to have been involved in killings at or near the Presbyterian Church at Gatare in April 1994 and then in multiple murders on Muyira Hill in Bisesero in May 1994. The allegations are dealt with at some length in the deposition of Martin Ngoga at paragraphs 113-181.
448. At paragraphs 210-213, Martin Ngoga touches on gacaca proceedings against Mutabaruka. At paragraph 210 he indicates that Mutabaruka was acquitted before the gacaca court in October/November 2007. Mr Ngoga goes on to suggest "these decisions have been vacated by the higher Courts or higher authority" (Ngoga, paragraph 211).
449. The Senior District Judge addressed the question of double jeopardy, arising under Section 80 of the Extradition Act 2003 and as affecting Mutabaruka, between paragraphs 58-70, and then in paragraph 665 of her judgment. In paragraph 61, the Senior District Judge explains:

"On 13th November 2007, CMU was acquitted in relation to events in Gatare but that acquittal was overturned by the Gacaca Court of Appeal on 22^nd January 2009. At another Gacaca court in Bisesero, between the acquittal and appeal in relation to the Gatare events, CMU was convicted and sentenced to 30 years' imprisonment on 21^st January 2008. Both sets of proceedings are in relation to offences for which his extradition is sought...."

450. Before the Senior District Judge, the Appellant argued (see Ngoga, paragraph 210, et seq) that the relevant gacaca court had no jurisdiction to try Mutabaruka in relation to the Gatare matters. The argument before the Senior District Judge was that these decisions do not "attract the principle of non bis in idem".
451. The Senior District Judge noted (paragraph 63) that the GoR had given no explanation for the appeal processes and provided no evidence to support the contentions made. The Senior District Judge went on to make the following findings of fact:

"It is quite clear to me from the fact that Mr Ngoga had no idea that CMU had been convicted at the Bisesero Gacaca that the NPPA did not have oversight or indeed knowledge of the various Gacaca proceedings in relation to these RPs. The Bisesero conviction was discovered by accident. There is no evidence which would enable me to find that the Gacaca Court of Appeal in CMU's case was acting unlawfully. No reasons are given for the Gacaca Court of Appeal's vacating the acquittal but these community courts did not give full reasons for any of their decisions." (paragraph 64)

452. In paragraph 65, SDJ Arbuthnot went on to observe that in relation to the Bisesero gacaca conviction, the GoR "never responded to a statement served on it by the defence team for CMU". The relevant evidence came from CMU's solicitor, Mr Rackstraw, which appended a translation of the only available record of the relevant gacaca proceedings and records that after examining the charges in relation to Bisesero, Mutabaruka was convicted "placed in category 2 and sentenced to 30 years' imprisonment and to pay back 15 cows, the equivalent to 3 million RwFr". In paragraph 68, the Senior District Judge confirms that this conviction was never dealt with in any evidence from Mr Ngoga, or any other witness on behalf of the Respondent. Recording her conclusion in paragraph 70, SDJ Arbuthnot finds that in the absence of evidence of the lawful quashing or annulling of the Bisesero conviction:

"…the case against CMU is caught by Section 80 and the rule against double jeopardy and he is entitled to be discharged in relation to this extradition request."

In paragraph 665 of her judgment, SDJ Arbuthnot ordered Mutabaruka's discharge in relation to any extradition request.

453. The GoR seek to appeal this conclusion. Before considering the merits of any such appeal, Ms Malcolm QC for Mutabaruka takes a procedural point. She argues that we have no jurisdiction to entertain the appeal on this ground. Section 106 of the Extradition Act 2003 details the powers of this Court to entertain and address an appeal under Section 105 of the Act. Section 106(5) sets out the conditions under which such an appeal may be entertained and reads:

"(5) The conditions are that—

(a) an issue is raised that was not raised at the extradition hearing or evidence is available that was not available at the extradition hearing;

(b) the issue or evidence would have resulted in the judge deciding the relevant question differently;

(c) if he had decided the question in that way, he would not have been required to order the person's discharge."

454. Ms Malcolm says the approach of the GoR to this aspect of the appeal does not begin to satisfy the conditions laid down. In order to examine that proposition, we must first establish how the matter is put by the Appellant.
455. The Appellant advances three grounds: firstly, that there was insufficient evidence that the gacaca convictions arose in relation to the same conduct for which extradition is sought. It is said that the unofficial translation of a record (i.e. the Appendix to the statement of Mr Rackstraw) is an insufficient record of conviction. Further it is said that the record of the gacaca conviction (with its paucity of details):

"resulted in the entire conduct for which CMU is sought being barred by reason of double jeopardy."

Amplifying the latter point, the Appellant seeks to submit that the Gatare allegations were entirely separate from the massacres at Bisesero, and any finding that extradition was barred in respect of those matters was an error.

456. In respect of the Bisesero matters, it is said that the Court below "did not have sufficient information about the alleged conviction of CMU to determine that it was the same conduct alleged in the request". The Appellant submits that there is a "fundamental difference between the two allegations even from the paucity of information offered". The Appellant goes on to argue that the gacaca court, for reasons set down in paragraphs 216-227 of Ngoga's deposition, is "not a Court of competent jurisdiction". This was said to have been recognised by the SDJ at paragraph 67. It would therefore follow that the conviction and sentence is defunct, as it cannot be enforced in law.
457. The third argument is that a specific legal provision in Rwanda prohibits extraditees from being dealt with by gacaca courts. This is Article 8 of "Organic Law No 04/2012/OL of 15.06.2012: Terminating Gacaca Courts and Determining Mechanisms for Solving Issues which were under their Jurisdiction". Article 8 reads:

"Article 8: Trial of an Extradited Person Sentenced by Gacaca Courts

A person extradited to be tried in Rwanda and who has been sentenced by Gacaca Courts should be tried by a competent court as provided by this organic law.

However, the decision of the Gacaca Courts shall first be nullified by that Court."

458. Based on that provision, the Appellant argues that anyone extradited to Rwanda can only be dealt with by the High Court for the matters outlined in the request and that it will be:

"A matter for the High Court to quash the Gacaca conviction because the Gacaca jurisdiction no longer operates. The High Court will be in a position to quash the conviction when they are seized of CMU's case, i.e., following his surrender to Rwanda."

459. In a Further Note from the Appellant, the argument is amplified. It is said that it is "uncontroversial that the Extradition Act 2003 requires the judge to focus on the conduct alleged in the extradition request". Where different episodes arise, the Judge must focus on each episode separately and consider if the specific episode of conduct amounts to an extradition offence. The Extradition Act was modified by the Extradition Act 2003 (Multiple Offences) Order 2003 (SI 2003/3150). Specifically, Article 23(4) amends Section 78(4) so as to stipulate that the extradition judge:

"(4) … must decide whether –

…

(b) each offence specified in the request is an extradition offence:"

460. Article 24 amends Section 79 of the Act, so that when considering any of the relevant bars to extradition, including the rule against double jeopardy:

"79(3) If the judge decides any of the questions in subsection (1) in the affirmative in relation to any offence, he must order the person's discharge in relation to that offence only."

The Appellant emphasises the "crucial" words in Section 79(3). Hence, it is said that even if the Senior District Judge was right that Mutabaruka's extradition for the gacaca conviction in relation to the Bisesero Hills conduct was barred, she was wrong to discharge him in relation to all of the conduct in the request.

461. Having identified the arguments and issues sought to be raised by the Appellant on this aspect of the case, we revert to considering the jurisdictional point advanced by Mutabaruka.
462. Ms Malcolm starts from the observation that the Appellant "failed to argue any of the matters now raised in …. its skeleton argument in the Court below". The Appellant should be debarred from doing so now, because these matters do not constitute "an issue that was not raised at the extradition hearing" nor "evidence available that was not available at the extradition hearing", within Section 106(5). Ms Malcolm argues that the approach adopted in relation to the parallel provisions under Section 104 of the 2003 Act, governing appeals against an order for extradition, should constitute a parallel approach. The principles laid down in the well-known case of Szombathely City Court v Fenivesy should be applied mutatis mutandis. The policy behind Section 104 and 106 is clear: the parties should not be permitted without good reason to raise issues or introduce evidence on appeal without having first litigated them below.
463. The outstanding Bisesero conviction and 30 years' sentence evidenced by Mr Rackstraw in the Appendix to his statement, should have been the subject of disclosure by the Appellant at the time of the extradition request, or as soon thereafter as it was brought to the attention of the GoR. In fact, this material has never been formally disclosed, despite a statement made in the course of the hearing before the Senior District Judge in June 2014 that the GoR "were going to make some urgent enquiries". The explanation is that the representatives of the Appellant missed the point at the time of drafting their closing submissions, and so far as it goes that is accepted by Mutabaruka's representatives.
464. However, Ms Malcolm submits this was not a single error. She argues that the Appellant failed on multiple occasions to address the question of double jeopardy in the case of Mutabaruka. The chronology is as follows:

i) The issue was raised in the Statement of Issues on 27 June 2013.

ii) The issue was raised again in Mutabaruka's request for disclosure on 29 July 2013.

iii) The Bisesero conviction was mentioned on various dates in proceedings in the lower court as a matter of administration and "housekeeping".

iv) Evidence was called by Mutabaruka bearing on the issue on 18 June 2014 (papers, volume 12/5971: 6124/36).

v) The matter was argued on Mutabaruka's behalf in closing submissions on 17 August 2015.

vi) The matter was further argued in the reply to the Appellant's submissions on 5 October 2015.

vii) On 25 September 2015 the Senior District Judge made an explicit request of the Appellant as to whether they intended to press this issue.

viii) An additional period of more than a year from that reminder passed with an additional ten months since the Perfected Grounds of Appeal were filed and served on behalf of the Appellant.

ix) Three orders from Ouseley J, directing the service of any fresh material, drew no response from the Appellant by way of fresh material.

x) The first attempt to introduce any document from the Appellant on this issue is said to have been served on the first day of the appeal hearing before us.

xi) In the premises there has been no attempt to comply with Criminal Procedure Rules paragraph 50.20(3) or 50.20(6).

465. In our view, Mr Knowles (who did not appear below), try as he might, cannot contradict the sequence of multiple failures to address this issue on the part of his clients, initially by way of appropriate disclosure, and subsequently by introducing any substantive evidence or argument before SDJ Arbuthnot. We consider Ms Malcolm is correct to submit there "has been no attempt to comply with the [Criminal Procedure] Rules". Whilst we have generally been liberal in our approach to fresh material in this case, since it is so voluminous, and since matters have continued to emerge, we have reached the view here that the egregious failures on this issue are such that by no liberality of approach can the threshold for jurisdiction be passed in respect of material now sought to be introduced. This matter must therefore be decided on the material available below.
466. In relation to matters of law, the correct approach may be different. Here, Mr Knowles relies on authority. In the case of The Queen (on the application of Soltysiak) v Judicial Authority of Poland [2011] EWHC 1338 (Admin), the question arose in relation to the parallel provisions under Section 104 of the 2003 Act as to whether a fresh argument of law, overlooked before the District Judge, might be raised in the appeal to the Divisional Court. Bean J (as he then was) considered two earlier authorities, Hoholm v Norway [2009] EWHC 1513 (Admin) and Mehtab Khan v The United States of America [2010] EWHC 127 (Admin), where apparently contradictory outcomes were reached on such a question. In Hoholm, the Court permitted a fresh point of law to be raised on appeal, the matter having been overlooked below. In Mehtab Khan, the Divisional Court held:

"… that the new point was a bad one; so in my view their observations are not part of the ratio of the decision. I therefore consider that I am bound by Hoholm. In any event, I consider that Hoholm was correctly decided." (Soltysiak, paragraph 11)

467. Bean J went on to consider that he was:

"… in no doubt that it is open to an Appellant to argue on appeal to this Court that the offence for which extradition is sought is not an extradition offence as defined by Section 64, even if that point of law was not spotted before the Magistrates' Court. I say so both as a matter of interpretation of Section 27 and following the decision in Hoholm. I am also in no doubt that until and unless this issue is resolved authoritatively, possibly by a three judge Divisional Court or perhaps by a still higher Court, it is the duty of advocates to ensure that this Court, when it is sought to raise a point of law not taken below, is made aware of both Hoholm and Khan." (Soltysiak, paragraph 18)

468. We accept the views expressed in Hoholm and Soltysiak. We consider there was a serious sequence of failures on the part of the Appellant to address this issue, in disclosure, in the introduction of evidence and in argument below. As we have said, we propose to admit or consider no evidence on this question from the GoR which was not before the Senior District Judge. However, we do propose to consider arguments of law advanced within the confines of the material before the Court below.
469. On the evidence as it was below, the two sets of allegations were those relating to Gatare in April 1994 and Bisesero in May 1994. Without fresh evidence, there is nothing to disturb that factual conclusion of SDJ Arbuthnot.
470. Mr Knowles's argument is that the unofficial transcript of a record of conviction is insufficient evidence of the gacaca convictions. However, we do not see why that is so. Such evidence was advanced for that purpose and remained uncontradicted before the Senior District Judge. In our view her factual conclusions were proper.
471. We therefore proceed on the factual basis that Mutabaruka was acquitted before the Nyarwungo gacaca court in relation to the Gatare allegations, that that acquittal was in relation substantially to the same allegations in respect of which extradition is now sought, and that it was substantially the same set of allegations in respect of which the acquittal was overturned before the gacaca Court of Appeal in January 2009. On the evidence before SDJ Arbuthnot (and us) we see no basis for the kind of distinction reached in Kulibaba v USA [2014] EWHC 178 (Admin).
472. We accept that the amended Sections 78(4) and 79(3) apply to these proceedings. Therefore the double jeopardy/abuse of process bar must be approached with each offence in mind. However, the factual position below also covers the Bisesero conviction and sentence. Therefore, absent a successful argument that gacaca proceedings are to be discounted for the purposes of double jeopardy, the proceedings here are concerned with both aspects of the offending in respect of which extradition is sought. In the one instance he has been acquitted and his acquittal overturned, in the other he has been convicted in his absence and faces a 30 year sentence.
473. We have borne in mind the principles laid down in Fofana and Another v Deputy Prosecutor Thubin Tribunal de Grande Instance de Meaux, France [2006] EWHC 744 (Admin). In that case, the Divisional Court emphasised:

"18. In summary the authorities establish two circumstance in English law that offend the principle of double jeopardy:

i) Following an acquittal or conviction for an offence, which is the same in fact and law – autrefois acquit or convict; and

ii) following a trial for any offence which was founded on "the same or substantially the same facts", where the court would normally consider it right to stay the prosecution as an abuse of process and/or unless the prosecution can show "special circumstances" why another trial should take place.

19. In Connelly, their Lordships reached this position in practical, though not unanimously in formal, terms by, in the main, confining the notion of double jeopardy to the narrow pleas in bar of autrefois acquit or convict, but allowing for a wider discretionary bar through the medium of the protection afforded by the court's jurisdiction to stay a prosecution as an abuse of process. In Humphreys, where their Lordships sanctioned a prosecution for perjury based on the same facts plus evidence of perjury by the defendant at an earlier failed prosecution for a driving offence, Lord Hailsham of St Marylebone indicated the second broader discretionary bar in the following passage at 41D-E:

"(10) Except where the formal pleas of autrefois acquit or convict are admissible, when it is the practice to empanel a jury, it is the duty of the court to examine the facts of the first trial in case of any dispute, and in any case it is the duty of the court to rule as a matter of law on the legal consequences deriving from such facts. In any case it is, therefore, for the court to determine whether on the facts found there is as a matter of law, a double jeopardy involved in the later proceedings and to direct a jury accordingly.""

474. It is to be noted that Fofana post-dated the 2003 amendment of the Extradition Act. It is also to be noted that the same approach is adopted to this bar whether it arises in relation to a European Arrest Warrant and Part 1 territory or not: see for example Bohning v Government of the United States of America [2005] EWHC 2613 (Admin).
475. Mr Knowles argues these points are immaterial since, as Mr Ngoga asserted in relation to the Gatare matter in his deposition at paragraph 210, "the gacaca had no jurisdiction to hear this case". This is hard to follow, given that they did hear both matters. Mr Ngoga went on to argue that "these gacaca decisions have been vacated by the higher courts or higher authority" (deposition, paragraph 11). It is not clear what is meant by "higher authority". Nor was there evidence below to substantiate this claim.
476. There is uncontroverted evidence that many prisoners have been tried, convicted and served lengthy sentences from the gacaca courts. It is also instructive, as Ms Malcolm points out, that in the case of Uwinkindi, there was explicit evidence before the ICTR that his gacaca convictions had been vacated: see the Uwinkindi ICTR judgment and the deposition of Ngoga at paragraph 213. The implication is clear that, absent such vacation the principle of ne bis in idem would have been applicable in relation to those gacaca convictions.
477. We also bear in mind the cases of Mudahinyuka and Mukeshimana, about whom Ms Nerad gave evidence below. Each was convicted and sentenced by gacaca courts in their absence. Each was required to serve those sentences once expelled from the USA and returned to Rwanda.
478. The Appellant argues that the Senior District Judge "recognised … at paragraph 67" that the gacaca court is "not a Court of competent jurisdiction". Here too we prefer the view of Ms Malcolm. She submits that SDJ Arbuthnot did no such thing. What the Senior District Judge did say was that the gacaca court "does not meet the requirements for Article 6 and as a result … Mutabaruka could never be returned to serve the sentence imposed by gacaca". As Ms Malcolm submits, the fact of wholesale breaches of the standards of fair trial does not amount to a lack of jurisdictional competence.
479. We now turn to the third argument advanced by Mr Knowles, namely that the effect of Article 8 of Organic Law No 04/2012/OL is that the gacaca verdicts will be nullified by the Rwandan High Court, and Mutabaruka tried exclusively by the High Court.
480. The starting point here is that the gacaca verdicts (and sentence) have not been nullified. At present, they stand. For reasons already given, this is not a part of this case where fresh evidence can properly be received. As a matter of record, there is no suggestion that the verdicts have been nullified. As the Appellants concede in their principal written submissions, the decision below was based on the absence of evidence that the gacaca conviction had been annulled or quashed. The Appellant's response to this is to point to the Organic Law No 04/2012/OL and the evidence of Dr Clark that the gacaca conviction will be quashed or annulled. Part of Ms Malcolm's reply is to cite the criticisms made by the GoR at other points in these proceedings, to the effect that it is unsafe to rely on "expert" opinion on the law from a non-legal expert (such as Dr Clark). Reliance here is also rendered more difficult since, in effect, the Rwandan Government is relying on a future decision of the judicial branch, a promise Ms Malcolm says they should not be in a position to enforce.
481. Mr Knowles submits that the Organic law was enacted "precisely to avoid" this issue. The law is plain and unambiguous. In saying the Court will grant the application, the Prosecutor General was saying no more than that the Court will apply Rwandan law. The assurance from the GoR is acceptable in principle: see the similar examples to be found in In Re Peci (Unreported) [1999] EWHC J1105-6 and Lodhi v Governor of Brixton Prison [2001] EWHC 178 (Admin). Mr Knowles submits there is no need to test the assurances in evidence, but the assurance is in any event effective. Mr Knowles advances a number of reasons for this:

"But if they are examined then they bear out that this assurance is effective. Othman is at A2/Tab 31. As to the criteria: (a) the assurance has been disclosed to the Court; (b) the assurance is specific; (c) the Prosecutor General can bind the receiving state (see above); (d) it has been issued by the central government; (e) the proposed treatment is legal in Rwanda; (f) Rwanda has no track record of not abiding by assurances given. In relation to Mr Mudahinyuka's case, the assurances were given for the purposes of an extradition request which never materialised because he was deported, and so they ceased to have effect; (g) compliance can be objectively verified."

482. We have considered these submissions but they are subject to three objections or problems. Firstly, they really rest upon (or even consist of) fresh evidence. We are not prepared to accept such evidence in this area for the reasons we have given. Secondly, there is a real concern about assurances from the GoR, for reasons sufficiently articulated earlier in this judgment. Thirdly, the bald fact remains that Mutabaruka has been tried, received verdicts (and a sentence) by Courts treated as competent within Rwandan law, and by Courts whose very substantial sentences have been carried into execution.
483. Looking at the matter overall, we find that the basis for this bar is made out. In reaching that conclusion we follow the approach in Fofana. This is a case where the broader principle of abuse of process applies.
484. We consider that the Senior District Judge was correct in finding that Mutabaruka should not be extradited because of double jeopardy and/or abuse of process, pursuant to Section 80 of the 2003 Act.

Celestin Mutabaruka: Section 81 Extradition Act – Extraneous Considerations

485. SDJ Arbuthnot dealt with this aspect of Mutabaruka's case in paragraphs 86 to 92, and 95 to 96 of her judgment. The claim advanced on behalf of Mutabaruka is that he will be at increased risk of a political and prejudicial trial because of his political involvement. It is said that in 1993/1994, and now, he is and was a politician. His party is in opposition to the GoR.
486. The Senior District Judge accepted that Mutabaruka helped set up a political party known by the name UNISODEC, which was allied to President Habyarimana. The judgment deals with that history in paragraph 87. More recently, the case for Mutabaruka is that he set up a new party called "Rwanda Rise and Shine" ["RaS"]. SDJ Arbuthnot accepted his earlier political involvement, but considered that there was "weakness in the evidence that he is currently a political force in Rwandan politics or in the Diaspora". In paragraphs 89 to 91 the Judge gave her conclusions. Broadly she considered that whilst Mutabaruka might be involved in RaS, essentially this "fairly new small organization" was insignificant and there was no evidence "that would lead to me to believe that Kagame would find [Mutabaruka] worthy of eliminating from the political scene".
487. In common with the Divisional Court in 2008/9, and with her approach to the other Respondents where Section 81 issues or political engagement have been raised, SDJ Arbuthnot found there was insufficient evidence to found a serious possibility that the extradition requests were being made for other purposes than the offending named (paragraph 95) and found that there was no reasonable chance or serious possibility that these extraneous factors might lead to any additional prejudice at Mutabaruka's trial (paragraph 96).
488. Ms Malcolm QC submits that the SDJ failed to take sufficient account of the evidence of Mutabaruka's political links and activities and, in short, underestimated the risk relevant to Section 81(b)/Article 6. She submits that it is significant the Rwandan Arrest Warrants were issued by the Prosecutor General and were therefore not subject to any judicial scrutiny or oversight. The extradition request is an executive act. The deposition of Simon Ngoga links UNISODEC directly to the MRND, for which it is said to have acted as "little more than a cat's paw". Ms Malcolm suggests there is little evidence for this, but the expression of the opinion may be sufficient warning of a risk to fair trial. Although, she submits, that in fact UNISODEC was a non-sectarian party, seeking to reduce ethnic tension, the stereotype or misrepresentation of the nature of the party may be the significant factor. Ms Malcolm submits that the evidence of Adelbert Rugeruza and Azariah Shiyranbere demonstrates that, following his move to the United Kingdom, Mutabaruka remained politically active, although the activities were limited. She relies upon the fact, as noted by SDJ Arbuthnot, that the witness Rene Mugenzi was aware of RaS as the metamorphosis or successor to UNISODEC. The submission is in part that the judge was insufficiently sensitive to the practical realities for opposition parties to the Kigali regime. Caution as to the adverse consequences of opposition mean that activities may be muted.
489. Much of the evidence already addressed in the course of our judgment underscores the determination of the Rwandan regime to stamp out opposition.
490. Ms Malcolm submits that Peter Mutabaruka's evidence demonstrates that RaS is of growing influence in Tanzania, a country where opposition to and conflict with the GoR is established. The further submission is that SDJ Arbuthnot "failed to analyse the parallels between this case and those of Patrick Karageya and Faustin Nyamwasa", members of another opposition party based in South Africa, both of whom were tried and convicted of national security offences in their absence. Karageya was subsequently murdered, as we have noted elsewhere.
491. Ms Malcolm also relies upon the evidence of the witness Patrice Niyonteze, who suggests in his witness statement that Mutabaruka "is accused because of his position", distinguishing his own position as someone not able to influence or mobilise people and therefore not accused.
492. The reply to these specific points on behalf of the Appellant can be summarised as follows. The evidence of Mutabaruka's son, Peter Mutabaruka, and his long-standing friend Rugeruza was "entirely self-serving and deeply unconvincing". The conclusions of the Senior District Judge were correct. Cross-examination of Rugeruza demonstrated that Rwanda Arise and Shine was not a continuation of UNISODEC, and indeed was not formed until, at the earliest, around 2012. Hence the creation of "an alleged party of opposition" post-dates the investigation into Mutabaruka, and calls into question the genuineness of the political activity. Moreover there is no "cogent evidence" that the party operates in any recognisable sense. Rugeruza was said to be part of the party structure, but could not explain the party's aims. Peter Mutabaruka's evidence also exposed the party as a "sham", given his incapacity to answer as to ordinary details of the party's membership, location and activity. His answers were too thin to be credible, even allowing for caution as to individual names. Peter Mutabaruka did confirm that the RaS had only operated publicly over the last two to three years, further undermining the suggestion that this extradition is for political reasons. In short, the reliance on RaS is reliance upon a party "created as a cynical ploy to provide CMU with an additional defence to these proceedings".
493. Taking all of this evidence and these arguments into account, we do not conclude that the Senior District Judge was in error or wrong in her conclusions. On the contrary, her overall conclusion that there was no sufficiently compelling evidence that Mutabaruka was a significant political figure is undisturbed by the submissions made. For those reasons, we conclude that this aspect of the case adds nothing to the concerns about fair trial expressed elsewhere in this judgment.

Celestin Ugirashebuja: Abuse of Process

494. The claim of abuse of process in relation to Ugirashebuja's criminal proceedings before the gacaca courts was the subject of no substantial oral submissions before us. The matter was dealt with by the Senior District Judge between paragraphs 658 and 664 of her judgment. She heard a number of live witnesses on this issue, in particular two anonymised witnesses: CU/2 and CU/3. Each was a former gacaca judge, CU/2 having been one of the judges acquitting Ugirashebuja.
495. As the Senior District Judge pointed out, Ugirashebuja was tried by the gacaca court in Kigoma and acquitted after hearing prosecution and defence witnesses. Quite extensive papers relating to these proceedings have been exhibited and SDJ Arbuthnot heard reasonably full oral evidence via video link from Rwanda. The essence of the submissions and evidence to SDJ Arbuthnot from the Respondent Ugirashebuja was that the gacaca acquittal was probably reached and that the quashing of that acquittal by the gacaca court of appeal was procured by the GoR, through intermediaries. In this area of the case, SDJ Arbuthnot had a clear advantage over us since there was extensive oral evidence and she was able to make an assessment of the witnesses. As she made clear, she reached the view in this instance that:

"the brief reasons the court [gacaca court of appeal] gave for the nullification were adequate bearing in mind the type of lay court it is. I will not give such weight to the evidence of CU/2 and CU/3 that would enable me to find the Gacaca Court of Appeal nullification was unlawful" (judgment, paragraph 662)

496. We have read with care the written material, witness statements and transcripts of evidence of these witnesses. We are not able to conclude that the decision of SDJ Arbuthnot on this point was wrong. However, a further point arises. In paragraph 663 of her judgment, she noted that there was a distinction between the evidence before the Rwoga gacaca hearing and the Kigoma allegations, which were quite discrete. None of the 16 witnesses relied on in the request before the Senior District Judge were witnesses against Ugirashebuja in the Rwoga gacaca hearing. She noted that the Claimant's witness, Ms Nerad, stated that the Rwoga gacaca hearing related to different allegations in a different sector, made by different witnesses. There may have been some "small overlap". In paragraph 663, in addition to the lack of significant overlap, SDJ Arbuthnot simply found that she could not conclude there were reasonable grounds for "believing the conduct alleged by CU may have occurred. I do not find an abuse in these circumstances".
497. Although the events in these gacaca proceedings are often hard to follow in detail, from the evidence and transcripts available, we cannot conclude that the decision by SDJ Arbuthnot on this aspect of the case was wrong. In this instance the lack of overlap, combined with the apparently comprehensible reasons for the gacaca quashing, means, taken together, that we cannot find the basis for an abuse of process in relation to Ugirashebuja on the request for extradition presented by the Appellant. Hence, on this aspect of the case, the cross-appeal is dismissed.

Summary of Conclusions

498. For the reasons given, we find bars to extradition in the cases of Mutabaruka and Nteziryayo. In each case extradition would represent a breach of Section 80 of the Extradition Act 2003. In the case of Nteziryayo, the cross-appeal succeeds on that point.
499. In respect of all Respondents, we consider the Senior District Judge was correct in her conclusion that, if extradited, they would be at risk of a flagrant denial of fair trial. We conclude as of the date of this judgment that remains the case. As indicated, we leave open a final opportunity, if the Appellant wishes to take it, to seek to persuade the Court that conditions will be put in place sufficient to overcome that bar to extradition.

ANNEX 1

EXTRADITION ACT 2003

79 Bars to extradition

(1) If the judge is required to proceed under this section he must decide whether the person's extradition to the category 2 territory is barred by reason of—

(a) the rule against double jeopardy;

(b) extraneous considerations;

(c) the passage of time;

(d) hostage-taking considerations.

80 Rule against double jeopardy

A person's extradition to a category 2 territory is barred by reason of the rule against double jeopardy if (and only if) it appears that he would be entitled to be discharged under any rule of law relating to previous acquittal or conviction if he were charged with the extradition offence in the part of the United Kingdom where the judge exercises his jurisdiction.

81 Extraneous considerations

A person's extradition to a category 2 territory is barred by reason of extraneous considerations if (and only if) it appears that—

(a) the request for his extradition (though purporting to be made on account of the extradition offence) is in fact made for the purpose of prosecuting or punishing him on account of his race, religion, nationality, gender, sexual orientation or political opinions, or

(b) if extradited he might be prejudiced at his trial or punished, detained or restricted in his personal liberty by reason of his race, religion, nationality, gender, sexual orientation or political opinions.

82 Passage of time

A person's extradition to a category 2 territory is barred by reason of the passage of time if (and only if) it appears that it would be unjust or oppressive to extradite him by reason of the passage of time since he is alleged to have—

(a) committed the extradition offence (where he is accused of its commission), or

(b) become unlawfully at large (where he is alleged to have been convicted of it).

84 Case where person has not been convicted

(1) If the judge is required to proceed under this section he must decide whether there is evidence which would be sufficient to make a case requiring an answer by the person if the proceedings were the summary trial of an information against him.

(2) In deciding the question in subsection (1) the judge may treat a statement made by a person in a document as admissible evidence of a fact if—

(a) the statement is made by the person to a police officer or another person charged with the duty of investigating offences or charging offenders, and

(b) direct oral evidence by the person of the fact would be admissible.

(3) In deciding whether to treat a statement made by a person in a document as admissible evidence of a fact, the judge must in particular have regard—

(a) to the nature and source of the document;

(b) to whether or not, having regard to the nature and source of the document and to any other circumstances that appear to the judge to be relevant, it is likely that the document is authentic;

(c) to the extent to which the statement appears to supply evidence which would not be readily available if the statement were not treated as being admissible evidence of the fact;

(d) to the relevance of the evidence that the statement appears to supply to any issue likely to have to be determined by the judge in deciding the question in subsection (1);

(e) to any risk that the admission or exclusion of the statement will result in unfairness to the person whose extradition is sought, having regard in particular to whether it is likely to be possible to controvert the statement if the person making it does not attend to give oral evidence in the proceedings.

(4) A summary in a document of a statement made by a person must be treated as a statement made by the person in the document for the purposes of subsection (2).

(5) If the judge decides the question in subsection (1) in the negative he must order the person's discharge.

87 Human rights

(1) If the judge is required to proceed under this section (by virtue of section 84, 85 or 86) he must decide whether the person's extradition would be compatible with the Convention rights within the meaning of the Human Rights Act 1998 (c. 42).

(2) If the judge decides the question in subsection (1) in the negative he must order the person's discharge.

(3) If the judge decides that question in the affirmative he must send the case to the Secretary of State for his decision whether the person is to be extradited.

95 Speciality

(1) The Secretary of State must not order a person's extradition to a category 2 territory if there are no speciality arrangements with the category 2 territory.

(2) But subsection (1) does not apply if the person consented to his extradition under section 127 before his case was sent to the Secretary of State.

(3) There are speciality arrangements with a category 2 territory if (and only if) under the law of that territory or arrangements made between it and the United Kingdom a person who is extradited to the territory from the United Kingdom may be dealt with in the territory for an offence committed before his extradition only if—

(a) the offence is one falling within subsection (4), or

(b) he is first given an opportunity to leave the territory.

(4) The offences are—

(a) the offence in respect of which the person is extradited;

(b) an extradition offence disclosed by the same facts as that offence, other than one in respect of which a sentence of death could be imposed;

(c) an extradition offence in respect of which the Secretary of State consents to the person being dealt with;

(d) an offence in respect of which the person waives the right that he would have (but for this paragraph) not to be dealt with for the offence.

(5) Arrangements made with a category 2 territory which is a Commonwealth country or a British overseas territory may be made for a particular case or more generally.

(6) A certificate issued by or under the authority of the Secretary of State confirming the existence of arrangements with a category 2 territory which is a Commonwealth country or a British overseas territory and stating the terms of the arrangements is conclusive evidence of those matters.

ANNEX 2

Extract From Brown v Government of Rwanda [2009] EWHC 770 (Admin)

BACKGROUND

7. We should first say a little about the state of affairs in Rwanda before the genocide which took place there in 1994. As will be apparent this is nothing but the barest outline, intended only to provide some context for the events giving rise to the issues we must decide.

8. Before colonisation, Rwanda's social structure included three groups, the Hutu, the Tutsis and the Twa. The Twa, who were pygmies, formed no more than a small percentage of the population. The majority of the people were Hutu. The monarchy, and many of the chiefs, were Tutsi. Rwanda gained full independence in 1962. Before that, in 1959, political unrest led to a great deal of violence. The first victims were Hutu. Thousands of Tutsi were killed. There ensued a cyclical pattern of violence involving the two groups. An election gave an overwhelming majority to Hutu political parties. The Tutsi monarch fled abroad. In 1961, after a referendum, the Tutsi monarchy was abolished and Rwanda became a republic. In 1961 and 1962, Tutsi guerrilla groups staged attacks into Rwanda from outside the country. Hutu within Rwanda responded. Thousands were killed.

9. We may go forward to 1975, when after a political coup President Juvenal Habyarimana, a Hutu, established a one party system. His political party was the MRND. Every Rwandan became a member, like it or not. But the Tutsi population were not proportionately represented in the political and social life of the country. The Habyarimana regime was hostile not only to the Tutsi, but also to Hutu who did not originate from the north-west of Rwanda where Habyarimana was based. Habyarimana surrounded himself with persons from that region. They were popularly known as the "Akazu". In 1990 an attack was launched from Uganda by displaced Tutsi who had formed the Rwandan Patriotic Front ("RPF").

10. At length domestic and international pressure persuaded President Habyarimana to accept a multi-party system in principle, implemented by a new constitution promulgated on 10 June 1991 which established four further political parties. Meanwhile Tutsi exiles launched incursions into Rwanda under the banner of the Rwandan Patriotic Army ("RPA"). Violent incidents ensued. In early 1992 the President began the training of youth members of the MRND to form militias known as the Interahamwe. The Interahamwe later massacred Tutsi, and committed other crimes which largely went unpunished. The division between Hutu and Tutsi widened. In March 1992, a group of Hutu hard-liners founded a new radical political party, the CDR, which was more extremist than Habyarimana himself.

11. We should make some reference to the office of bourgmestre, which was held by three of the appellants. Until the time of the genocide Rwanda was divided into eleven prefectures, each headed by a prefet. The prefectures were further divided into communes; and the bourgmestre was in effect the mayor of the commune. He had many public functions and considerable legal power and authority. A decree of 20 October 1959, originally passed by the colonial powers but still good law in 1994, gave the bourgmestre power to order the evacuation, removal or internment of persons in a state of emergency. He had judicial functions, and was also a trusted representative of the President; as such he had a series of unofficial powers and duties. He was a figure of great importance in the daily life of ordinary people, who would look to him for protection. The Trial Chamber of the International Criminal Tribunal for Rwanda (the ICTR: it has an important place in the arguments before us, and we will explain its provenance and jurisdiction below) found in its first judgment, in the case of Akayesu (delivered on 2 September 1998), that:

"In Rwanda, the bourgmestre is the most powerful figure in the commune. His de facto authority in the area is significantly greater than that which is conferred upon him de jure."

12. As we have said, three of the appellants were bourgmestres. CM was the bourgmestre of the Kinyamakara commune, EN of Mudasomwa commune, and CU of the Kigoma commune. VB was based in the Rugenge prefecture in Kigali. He moved there in about 1990, having previously been based in the Gitarama prefecture. He was not a bourgmestre, but is said to have been a close associate of President Habyarimana and a member of the Akazu.

THE GENOCIDE

13. The events of the 1994 genocide were to be authoritatively described in some detail by the ICTR in Akayesu. We give these extracts:

"106. … On 6 April 1994, President Habyarimana and other heads of State of the region met in Dar-es-Salaam (Tanzania) to discuss the implementation of the peace accords [sc. which had been signed earlier]. The aircraft carrying President Habyarimana and the Burundian President, Ntaryamirai, who were returning from the meeting, crashed around 8:30 pm near Kigali airport. All aboard were killed.

107. The Rwandan army and the militia immediately erected roadblocks around the city of Kigali. Before dawn on April 7 1994, in various parts of the country, the Presidential Guard and the militia started killing the Tutsi as well as Hutu known to be in favour of the Arusha Accords and power-sharing between the Tutsi and the Hutu. Among the first victims were a number of ministers of the coalition government, including its Prime Minister, Agathe Uwilingiyimana (MDR), the president of the Supreme Court and virtually the entire leadership of the parti social démocrate (PSD). The constitutional vacuum thus created cleared the way for the establishment of the self-proclaimed Hutu-power interim government, mainly under the aegis of retired Colonel Théoneste Bagosora.

108. Soldiers of the Rwandan Armed Forces (FAR) executed ten Belgian blue helmets, thereby provoking the withdrawal of the Belgian contingent which formed the core of UNAMIR. On April 21 1994, the UN Security Council decided to reduce the peace-keeping force to 450 troops.

109. In the afternoon of 7 April 1994, RPF troops left their quarters in Kigali and their zone in the north, to resume open war against the Rwandan Armed Forces. Its troops from the north moved south, crossing the demilitarized zone, and entered the city of Kigali on April 12 1994, thus forcing the interim government to flee to Gitarama.

110. On April 12 1994, after public authorities announced over Radio Rwanda that 'we need to unite against the enemy, the only enemy and this is the enemy that we have always known...it's the enemy who wants to reinstate the former feudal monarchy', it became clear that the Tutsi were the primary targets. During the week of 14 to 21 April 1994, the killing campaign reached its peak. The President of the interim government, the Prime Minister and some key ministers travelled to Butare and Gikongoro, and that marked the beginning of killings in these regions which had hitherto been peaceful. Thousands of people, sometimes encouraged or directed by local administrative officials, on the promise of safety, gathered unsuspectingly in churches, schools, hospitals and local government buildings. In reality, this was a trap intended to lead to the rapid extermination of a large number of people.

111. The killing of Tutsi which henceforth spared neither women nor children, continued up to 18 July 1994, when the RPF triumphantly entered Kigali. The estimated total number of victims in the conflict varies from 500,000 to 1,000,000 or more."

14. The ICTR proceeded to consider whether the massacres which took place in Rwanda between April and July 1994 fell within the definition of genocide contained in the 1951 Convention on the Prevention and Punishment of the Crime of Genocide ("the Genocide Convention"), which had been acceded to by Rwanda in 1975. In doing so the ICTR gave further details of the facts:

"114. Even though the number of victims is yet to be known with accuracy, no one can reasonably refute the fact that widespread killings were perpetrated throughout Rwanda in 1994.

115. Indeed, this is confirmed by the many testimonies heard by this Chamber. The testimony of Dr. Zachariah who appeared before this Chamber on 16 and 17 January 1997 is enlightening in this regard. Dr. Zachariah was a physician who at the time of the events was working for a non-governmental organisation, 'Médecins sans frontières'. In 1994 he was based in Butare and travelled over a good part of Rwanda upto its border with Burundi. He described in great detail the heaps of bodies which he saw everywhere, on the roads, on the footpaths and in rivers and, particularly, the manner in which all these people had been killed. At the church in Butare, at the Gahidi mission, he saw many wounded persons in the hospital who, according to him, were all Tutsi and who, apparently, had sustained wounds inflicted with machetes to the face, the neck, and also to the ankle, at the Achilles' tendon, to prevent them from fleeing. The testimony given by Major-General Dallaire, former Commander of the United Nations Assistance Mission for Rwanda (UNAMIR) at the time of the events alleged in the Indictment, who was called by the defence, is of a similar vein. Major-General Dallaire spoke of troops of the Rwandan Armed Forces and of the Presidential Guard going into houses in Kigali that had been previously identified in order to kill. He also talked about the terrible murders in Kabgayi, very near Gitarama, where the interim Government was based and of the reports he received from observers throughout the country which mentioned killings in Gisenyi, Cyangugu and Kibongo.

116. The British cameraman, Simon Cox, took photographs of bodies in many churches in Remera, Biambi, Shangi, between Cyangugu and Kibuye, and in Bisesero. He mentioned identity cards strewn on the ground, all of which were marked 'Tutsi'. Consequently, in view of these widespread killings the victims of which were mainly Tutsi, the Chamber is of the opinion that the first requirement for there to be genocide has been met, the killing and causing serious bodily harm to members of a group.

117. The second requirement is that these killings and serious bodily harm, as is the case in this instance, be committed with the intent to destroy, in whole or in part, a particular group targeted as such.

118. In the opinion of the Chamber, there is no doubt that considering their undeniable scale, their systematic nature and their atrociousness, the massacres were aimed at exterminating the group that was targeted. Many facts show that the intention of the perpetrators of these killings was to cause the complete disappearance of the Tutsi. In this connection, Alison Desforges, an expert witness, in her testimony before this Chamber on 25 February 1997, stated as follows: 'on the basis of the statements made by certain political leaders, on the basis of songs and slogans popular among the Interahamwe, I believe that these people had the intention of completely wiping out the Tutsi from Rwanda so that - as they said on certain occasions - their children, later on, would not know what a Tutsi looked like, unless they referred to history books'. Moreover, this testimony given by Dr. Desforges was confirmed by two prosecution witnesses, witness KK and witness OO, who testified separately before the Tribunal that one Silas Kubwimana had said during a public meeting chaired by the accused himself that all the Tutsi had to be killed so that someday Hutu children would not know what a Tutsi looked like.

119. Furthermore, as mentioned above, Dr. Zachariah also testified that the Achilles tendons of many wounded persons were cut to prevent them from fleeing. In the opinion of the Chamber, this demonstrates the resolve of the perpetrators of these massacres not to spare any Tutsi. Their plan called for doing whatever was possible to prevent any Tutsi from escaping and, thus, to destroy the whole group. Witness OO further told the Chamber that during the same meeting, a certain Ruvugama, who was then a Member of Parliament, had stated that he would rest only when no single Tutsi is left in Rwanda.

120. Dr. Alison Desforges testified that many Tutsi bodies were often systematically thrown into the Nyabarongo river, a tributary of the Nile. Indeed, this has been corroborated by several images shown to the Chamber throughout the trial. She explained that the underlying intention of this act was to 'send the Tutsi back to their place of origin', to 'make them return to Abyssinia', in keeping with the allegation that the Tutsi are foreigners in Rwanda, where they are supposed to have settled following their arrival from the Nilotic regions.

121. Other testimonies heard, especially that of Major-General Dallaire, also show that there was an intention to wipe out the Tutsi group in its entirety, since even newborn babies were not spared. Even pregnant women, including those of Hutu origin, were killed on the grounds that the foetuses in their wombs were fathered by Tutsi men, for in a patrilineal society like Rwanda, the child belongs to the father's group of origin. In this regard, it is worthwhile noting the testimony of witness PP, heard by the Chamber on 11 April 1997, who mentioned a statement made publicly by the accused to the effect that if a Hutu woman were impregnated by a Tutsi man, the Hutu woman had to be found in order 'for the pregnancy to be aborted'."

ANNEX 3

Extract from the Judgment of SDJ Arbuthnot

Conclusions in relation to the independence of the judiciary

500. The Divisional Court in Brown and others in 2008 and 2009 when it considered the independence of the judiciary relied on the Bizimungu case allied with the conclusions about the state of the polity in Rwanda. The High Court also relied on the HRW report of 2008 and the evidence of Professors Reyntjens, Sands and Schabas and in particular on the acceptance by the GoR's expert Professor Schabas that there was probably executive interference in the Bizimungu case to find evidence of specific positive incidents of judicial interference.
501. In 2014 to 2015, I must consider the implications of the trial of Ingabire which HRW found to be unfair. I bear in mind of course that like Bizimungu, Ingabire was a politician. She was planning on standing against President Kagame in the Presidential elections in 2010. Her case was not a transfer or extradition case. She was not tried by a Specialised Chamber of the High Court; Mutabazi another case relied upon by the defence, was tried by a military tribunal. Most if not all the examples of unfair trials in Rwanda are in relation to political opponents of the GoR.
502. The difference between her case and Bizimungu's is that there is no evidence of executive interference in her trial although there is other concerning evidence which I have set out above, evidence that led Amnesty to find that she had an unfair trial.
503. The ICTR Referral Chamber and Appeals Chamber and courts in other countries subsequently and the expert called by the GoR in these proceedings have drawn a distinction between what might be termed political trials and non-political cases. None of those courts have accepted that there is evidence of executive interference in the judiciary trying allegations of genocide. They have implicitly accepted that there is interference with the judiciary in political cases.
504. I find that opponents of the GoR do not appear to have fair trials in Rwanda. The problem I have been wrestling with is whether the finding that the High Court does not try fairly those 'political' defendants means that these RPs will necessarily not be tried by independent and impartial judges sitting in this Specialized Chamber trying international and cross border crimes.
505. The various courts that have sent defendants back to Rwanda for trial and the GoR in this case too, rely on the argument that genocide cases are not political. The argument from the defence is to the contrary. The defence say the RPs had high profiles in 1994 and there is a current political will to have them convicted. In Munyakazi, a pre 2009 case and a defendant who was not returned to Rwanda, I noted that the ICTR Appeals Chamber described genocide cases as "politically sensitive" in the judgment at Paragraph 26.
506. I found that the RPs positions in 1994 were not that they were senior politicians on the national stage but within the local areas they were men of standing or importance in the community. In his affidavit Ngoga has explained the significance of the bourgmestres. Three of the RPs were bourgmestres. There is some limited evidence that one of the RPs was a member of the Akazu, I find if he was, he was not a high profile one and the final RP was a director of an important company in the local area.
507. I do not find there is credible evidence that the RPs are political opponents to the GoR in 2015 and I do not find there is any persuasive evidence that the extradition request is driven by political considerations.
508. I find that the trials of these RPs, if they are returned, nevertheless will be high profile. The expert witnesses, Dr Clark and Professors Reyntjens and Longman consider that they would be. Mr Ngoga agreed that this extradition case has a high profile in Rwanda and there was a strong adverse reaction in Rwanda when the High Court in London in 2009 overruled the 2008 district judge's order referring the case to the Secretary of State. I would anticipate that an extradition order made in this country would be widely reported in Rwanda and such media attention would continue through any trial in the High Court in Rwanda. The GoR will use the return of the RPs as examples to show that the State's justice system is recognised internationally as a system which can try defendants fairly.
509. What Brown and others lacked in 2008 and 2009 was the significant evidence of the trials of the five men transferred or extradited to Rwanda. This evidence has enabled me to distinguish between 'political' trials on the one hand and trials of those accused of genocide.
510. I have read the extensive monitors' reports and summarised them above. I have also been provided with extensive summaries of decisions given by the High Court in relation to the five transferred, in particular Uwinkindi. I have had the benefit of Witteveen's comments about the trials. I do not agree with the contention of Miss Ellis at her Submissions Page 32, Paragraph 74, that an analysis of the proceedings of the five transferred cases "provides further evidence of a lack of judicial independence and impartiality in the face of deep unfairness to the defence".
511. I rely on the fact that in all the many pages of monitors' reports there is no recurring theme of complaint about the judges' lack of independence. In relation to the decisions of the judges there is no evidence in those records of any partiality or anything untowards. I suppose it is arguable that interference might not be picked up by the monitors but a lack of independence or fairness would be. I appreciate the following comment relates to procedural issues only but I was struck by the even handedness of the judges when confronted with, for example, applications for adjournments by the defence. The ICTR monitors make it clear that any defence request for an adjournment is considered carefully by the court and reasonably dealt with (usually the adjournment is granted). Apart from the problem of representation, and it is difficult to know what the court could do about that, the court's approach seems to be fair.
512. I am concerned, of course, about the effect on judges of comments made by the President and ministers about the guilt of these RPs in the sort of autocratic State that I have found Rwanda to be. The GoR seems not to respect the presumption of innocence and that might put pressure on the judges if there was no international judge amongst them. I accepted however the evidence from Professors Longman and Reyntjens, for example, that the quality of the judiciary has been increasing gradually in Rwanda and I have heard nothing which suggests that they are not able to ignore comments made by the President and others in non political cases. Dr Clark's evidence was that the most important developments had taken place since Brown and others and a number of reforms had been prompted by criticisms made by the ICTR and other jurisdictions.
513. The Divisional Court considered it was difficult to consider the independence of judges in the absence of a consideration of the State from which they come. I have set out my findings about the State at Paragraphs 221 onwards. I also, however, noted Dr Clark's evidence that while he accepted there may be justified concerns about political developments in Rwanda, they "do not inherently impinge on the delivery of justice for genocide suspects…we must recognize that the Rwandan judiciary to a large extent has a life of its own and must be evaluated in terms of judicial performance and concrete evidence from trials…The Rwandan judiciary has displayed a marked dedication to ensuring the fair trial of extradited suspects and taken concrete measures to fulfil the "checklist" criteria.". His view was echoed by Martin Witteveen who had observed a number of hearings and had investigated the independence of the judiciary. He had no concerns.
514. The 'sea changes' relied upon by the GoR which I have set out at Paragraph 11-29 inclusive at the beginning of this judgment, would be more significant if there was evidence they were affecting trials in practice. As far as the independence of the judiciary is concerned the only change which appears to have bedded in is the amendment to the Transfer Law in 2009 that allows three or more judges to be designated to try complex and important cases. All five of the transferred or extradited defendants in Rwanda either have been tried or are being tried by three judges in the Specialised Chamber of the High Court. I find their number, of course, makes interference less likely.
515. One of the new provisions relied upon by Mr Ngoga is the amendment to Rwandan law of 28^th November 2011 allowing for international judges to come and try transfer and extradition cases alongside Rwandan judges. This could be at the request of the accused, his lawyer or the prosecution from Rwanda or elsewhere. The President of the Supreme Court would decide on the application. Although Ngoga said in his affidavit that there was no reason why the application would not be granted, Dr Clark thought it unlikely.
516. Rwanda is rightly proud of the improvements to its judicial system since the majority of judges and lawyers were killed in 1994. I am not so sure that an independent country like Rwanda would accept a non Rwandan judge coming in to try one of their genocide cases. I did not accept Ngoga's evidence and agreed with Dr Clark that Rwanda is a proud country and would no doubt feel patronized if they had to accommodate a foreign judge in these circumstances. The GoR would find such a move "politically unpalatable because of a spirit of self-reliance and independence" as Dr Clark put it. In any event there is no evidence that any of the returned accused have been aware of the change in the law or if aware that they have applied for an international judge, let alone one being granted. It must be said were a judge with international experience to sit on any of the RPs trials that would reduce still further any remaining concerns I have in relation to the independence of the judiciary.
517. There was no real reason for me to doubt the GoR's evidence that the acquittal rate is around 20 to 30% in the High Court whilst in the gacaca Courts it had been around 25%. I accepted that of 238 High Court decisions that were reviewed by the Supreme Court, 219 were upheld (evidence conveniently summarized in the Prosecution Closing Submissions at Page 32, Paragraph 120). Unfortunately there was no information about the acquittal rate in genocide trials. Dr Clark, however, did report on a positive note that of a number of genocide defendants that HRW was concerned about in their 2008 report, Law and Reality, three including Biseruka and Twagrimungu were later acquitted.
518. I accept too that the judiciary are governed by a number of ethical codes as well as have annual performance evaluations and take an oath to discharge their duties responsibly. Reforms since 2003 ensure that candidates for the judiciary take an exam and have to have a law degree and six years' post qualification experience for the High Court. The three highest office holders in the Supreme Court and the two highest in the High Court are appointed by Presidential Order but this is approved by the Senate.
519. On the other hand, the evidence of the exiled past Prosecutor General and Vice-President of the Rwandan Supreme Court, Dr Gahima, was clear, a large number of current judges are active RPF members. He explained the official narrative is that all those who were involved in senior leadership are guilty of genocide. Although he left the county in 2004 he had been responsible for putting the list of judges together after the introduction of the new constitution and knew the way the system of appointments worked. He was a measured, credible and impressive witness and Dr Clark accepted he was a respected academic.
520. Edward Fitzgerald QC relies on an amendment to the Rwandan Constitution which he says may affect the independence of the judges and this is that since 2008 judges of the High and Supreme Courts are no longer appointed for life but are appointed for a determinate term which is renewable by the High Council of the Judiciary. This had not emerged by the time of the appeal to the Divisional Court in Brown and others (Appendix F, Paragraphs 80 onwards - Decisions on Transfer to Rwanda since 2006). I accept that security of tenure should ensure independence and impartiality but I find there is no evidence that the change to removal on the grounds of serious professional misconduct or incompetence is being used by the High Council of the Judiciary to remove judges for bringing in the wrong verdicts.
521. I noted on a positive note that transparency and fair decisions would be encouraged by the requirement that provided for all hearings to be in public. The requirement to give reasons for decisions is also protective so it was unfortunate that the defence were unable to find very many cases reported on the internet. Miss Ellis at Annex 1 to EN's Reply to GoR Submissions, sets out a chronology of requests made to the GoR in relation to statistics but more particularly judgments emanating from the High Court of Rwanda in genocide cases. There is fortunately a mass of material now in relation to the five transferred men, particularly in the Rebuttal Material produced by Ms Kabasinga. Before that only four judgments were found by Miss Ellis and her team. I found it concerning that a witness for the GoR could find the judgments with apparent ease whilst the thorough research by the defence team could not.
522. I noted that ICTR cases are monitored and there is a mechanism for the return of the defendant in certain circumstances. Dr Clark was of the view that observation of cases by a wide range of bodies, such as HRW etc has a positive effect. It sends out a message that the judiciary must operate in a fair and transparent manner. I am sure that is helpful for all the parties to keep such cases in the international eye. I would feel more confident in this case if there were such a mechanism in the extradition arrangements with Rwanda.
523. Amongst the NGO evidence I was particularly concerned by the Bertelsmann Stiftung Transformation Index 2014 Report which covered a period from January 2011 to January 2013, I have set out what it says at Paragraph 190 of this judgment but I will repeat it here: although the judiciary is formally independent, "in reality it is subordinated to the will of the executive in all politically sensitive matters". It describes a biased judicial system which enables prosecutions for offences such as genocide ideology to take place and political verdicts result particularly in relation in trials of opposition leaders and critical journalists. The comments various political authorities think it appropriate to make in relation to on-going genocide trials lead me to have concerns that if returned these defendants would be facing the same.
524. The Divisional Court in Brown and others wanted evidence from genocide cases being tried in the Rwandan High Court. I have heard that evidence. The evidence of trials taking place is more current, contemporaneous and detailed than the evidence from experts and others who through no fault of their own have been unable to go to Rwanda for a number of years. Whether it was via the ICTR monitors' reports, through the judgments of the courts provided by Ms Kabasinga or through Mr Witteveen's evidence which was particularly helpful, I have been able to see how the judiciary involved in similar high profile genocide cases of transferred defendants are making their decisions.
525. Most significantly, Mr Witteveen had been a witness to some of these hearings. What he had not seen he had read about in the notes the legal officer of the Dutch Embassy had made. He had had a great deal of experience investigating allegations of genocide before his current tenure in the NPPA. He had not seen any credible evidence that any authorities had interfered specifically in cases of genocide, nor had he seen any evidence that judges had lost their independence, neutrality and objectivity.
526. To conclude, I am drawing a distinction between the way the Rwandan High Court tries cases with a political flavour and the way they try genocide allegations. This is based on the clear evidence I have seen about the approach taken by the Specialised Chamber towards the five transferred genocide cases. Having considered all the evidence, I cannot exclude a risk of interference but judging from the transferred defendants the highest risk is from the pressure exerted by GoR ministers' comments in public and in the press. I consider any such risk would be reduced by a robust, able and experienced defence team with an ability to investigate the defence case and international monitoring of some sort. I consider that without both of these the RPs would be at a greater risk of judges behaving partially and being influenced by factors outside the evidence.