Kozak v. Canada (Minister of Citizenship and Immigration)

Kozak v. Canada (Minister of Citizenship and Immigration) Court (s) Database Federal Court of Appeal Decisions Date 2006-03-27 Neutral citation 2006 FCA 124 File numbers A-419-04, A-420-04 Notes Reported Decision Decision Content Date: 20060327 Dockets: A-419-04 A-420-04 Citation: 2006 FCA 124 CORAM: DÉCARY J.A. EVANS J.A. SHARLOW J.A. BETWEEN: A-419-04 KOZAK GEZA, CSEPREGI ATTILA KOZAK GEZA (minor) and CSEPREGI SZILVIA Appellants and THE MINISTER OF CITIZENSHIP AND IMMIGRATION Respondent A-420-04 SMAJDA SANDOR, SMAJDA ZSOLT, SMAJDA SANDOR, GYULAVICS TIMEA, SMAJDA CLAUDIA and SMAJDA JOZEF Appellants and THE MINISTER OF CITIZENSHIP AND IMMIGRATION Respondent Heard at Toronto, Ontario, on November 22, 2005. Judgment delivered at Ottawa, Ontario, on March 27, 2006. REASONS FOR JUDGMENT BY: EVANS J.A. CONCURRED IN BY: DÉCARY J.A. SHARLOW J.A. Date: 20060327 Dockets: A-419-04 A-420-04 Citation: 2006 FCA 124 CORAM: DÉCARY J.A. EVANS J.A. SHARLOW J.A. BETWEEN: A-419-04 KOZAK GEZA, CSEPREGI ATTILA KOZAK GEZA (minor) and CSEPREGI SZILVIA Appellants and THE MINISTER OF CITIZENSHIP AND IMMIGRATION Respondent A-420-04 SMAJDA SANDOR, SMAJDA ZSOLT, SMAJDA SANDOR, GYULAVICS TIMEA, SMAJDA CLAUDIA and SMAJDA JOZEF Appellants and THE MINISTER OF CITIZENSHIP AND IMMIGRATION Respondent REASONS FOR JUDGMENT EVANS J.A. A. INTRODUCTION [1] Administrative agencies are often required to be procedurally innovative in order to handle a heavy case load effectively and to make the most efficient use of scarce resources. They are also required to observe the duty of fairness. The central issue in this case is whether the Immigration and Refugee Board breached that duty by creating a reasonable apprehension of bias in the way that it established a “lead case” format for determining the appellants’ refugee claims, and by not separating the functions of management and adjudication. [2] The appellants are Hungarian citizens who claim refugee status in Canada. They say that they have a well-founded fear of persecution in Hungary because of their Roma ethnicity. In particular, the appellants allege that they have been subject to persecution by gangs of racist skinheads, from whom state authorities, including the police, are unable or unwilling to protect them. [3] After conducting a hearing spread over fourteen days in October and November 1998 under the former Immigration Act, R.S.C. 1985, c. I-2, the Convention Refugee Determination Division (as it then was) of the Immigration and Refugee Board rejected the appellants' claims. The Board found that the appellants had failed to prove that they had a well-founded fear of persecution in Hungary and that state authorities would not, or could not, provide them adequate protection. [4] The Board relied on common evidence of contemporary country conditions in Hungary, but heard separate evidence relating to the particular incidents on which individual appellants relied. The Board preferred the written and expert oral testimony provided on behalf of the Minister on the situation facing Roma in Hungary, and on the availability of protection by the police and other state authorities. It found that the appellants exaggerated the seriousness of the problems that they had faced, which amounted only to discrimination, not persecution. [5] The appellants applied to the Federal Court for leave to make applications for judicial review to set aside the decisions, alleging that the Board’s process was unauthorized and unfair. They also argued that, in finding that adequate state protection would not be available to them in Hungary, the Board ignored important evidence supporting their claims, and thus committed reviewable error. The leave applications were granted. [6] However, their applications for judicial review were dismissed: although the family name of the principal claimant in one of the appeals is Kozak, the decision is reported as Geza v. Canada (Minister of Citizenship and Immigration), [2005] 3 F.C. R. 3, 2004 FC 1039. The Applications Judge certified the following question under paragraph 74(d) of the Immigration and Refugee Protection Act, S.C. 2000, c. 27 (“IRPA”): “Did the IRB have jurisdiction to conduct a “lead case” under the Immigration Act?” [7] In a document entitled, “Lead Case Backgrounder”, the Board stated that the purpose of a “lead case” is to identify a refugee claim in which to create a full evidential record on which the hearing panel could make informed findings of fact and provide a complete analysis of the relevant legal issues. [8] While not binding on other panels of the Board, the factual findings and legal conclusions in the lead case were said by the Board to be intended to provide guidance to future panels hearing similar cases. The lead case would thus promote consistent, informed, efficient, and expeditious decision-making. The appellants’ cases were the first and, to date, only, “lead case” organized and heard by the Board. [9] A “lead case” is different in at least two respects from a “jurisprudential guide”, another technique used by the Board to enhance the quality and consistency of decisions. A Board decision is identified as a jurisprudential guide after it has been rendered, while a “lead case” is planned and organized before the case is heard. In addition, a jurisprudential guide is normally intended to be persuasive on questions of law, and mixed law and fact. In contrast, it was intended that lead cases would also establish persuasive findings of fact on country conditions. See further, Policy on the Use of Jurisprudential Guides, Policy no. 2003-01 (Ottawa: Immigration and Refugee Board of Canada, March 21, 2003). [10] Counsel for the appellants says that, even if the Board had the legal authority in principle to adopt a lead case strategy in the interests of consistency, quality and efficiency in decision-making (which he does not admit), the circumstances surrounding the Board’s organization and conduct of the hearings of these lead cases gave rise to a reasonable apprehension of bias and of a lack of independence in the decision-makers. In particular, he argues, a reasonable person would conclude that the Board’s lead case strategy was designed to reduce the percentage of successful refugee claims by Hungarian Roma and to deter potential claimants from coming to Canada, thus precluding an unbiased adjudication of the appellants’ claims. [11] The appellants were heard together before both the Board, which provided separate reasons for dismissing their refugee claims, and the Applications Judge. We also heard the appeals in Court File No. A-419-04 (Geza) and Court File No. A-420-04 (Smajda) at the same time, since they raise many of the same legal and factual issues. These reasons apply to both appeals, and a copy will be inserted in each file. B. FACTUAL BACKGROUND [12] In this appeal, most of the primary facts about the origins of the lead cases and the consequent planning process, the hearing itself, and certain post-hearing events, are not seriously disputed. Nor is the applicable legal test of impartiality contested. The dispute in this appeal is essentially about whether the primary facts give rise to a reasonable apprehension of bias. [13] The record compiled by the Board in response to the appellants’ request for its certified record pursuant to rule 317 of the Federal Courts Rules contains little information about the facts alleged to give rise to a reasonable apprehension of bias. It was not an issue that the Board was asked to address when hearing the appellants’ claims. [14] However, counsel for the appellants supplemented the Board’s record with material obtained from the Board, after the appellants had been granted leave to make an application for judicial review. The material was obtained in response to requests made under the Access to Information Act, R.S.C. 1985, c. A-1. Much of it comprises e-mails among Board officials about the origins and organization of the “lead case”. The e-mails also reveal the involvement in the planning process of Board Member Vladimir Bubrin, the leader of the Board’s case management team for Europe, who was also a member of the two-person panel which heard the appellants’ refugee claims. Further information is contained in affidavits prepared for the applications for judicial review. All the material in the appeal book was before the Applications Judge. (i) origins of the lead case initiative [15] As evidence of the context in which the lead case initiative was developed, counsel referred to an e-mail, dated May 6, 1998, from Joan Steegstra, who was then the Operations Service Manager for the European case management team. In this e-mail to a colleague (Appeal Book, vol. 10, pp. 3128-29), Ms Steegstra follows up on a previous conversation with the recipient about the “growing intake of Hungarians”, and advises him that she has heard from a lawyer who had been involved in Czech Roma claims, that “there are 15,000 (yes, fifteen thousand) Hungarian Roma on their way to Canada.” She also says that the same lawyer asked her to confirm a press story that the Board had already rendered 28 positive decisions in Hungarian Roma cases in the first five months of 1998. She says that she declined to confirm the “specifics” but said that she did tell the lawyer that “the Hungarian intake is growing”. Mr Bubrin was on the distribution list for this e-mail. [16] E-mails dated May 7, 1998 (Appeal Book, vol. 10, p. 3140), suggest that, as of that date, media interest in Hungary seemed not to have caused an increase in the number of Roma refugee claimants arriving in Canada. However, one official noted that “the figures do not look so good” and that, although there had been no increase in April, “the numbers remain relatively high”. [17] These e-mails appear to have been exchanged among a management group at the Board, and officials of Citizenship and Immigration Canada (“CIC”). An affidavit from Robert Orr, who was appointed Director General of CIC in 2003, describes in general terms the contacts between CIC and the Refugee Protection Division (as it now is) of the Board: Appeal Book, vol. 12, pp. 3358-59. He said that contacts take place at the Headquarters level and involve Directors General and above. Discussions focus on policy issues of mutual interest, not individual cases. For example, CIC shares with the Board its information about trends and anticipated increases in the volume of claimants from a particular country. The broader implications of a particular decision may also be discussed at meetings at the regional level. Board members do not attend any of these meetings. [18] In further intra-Board e-mail exchanges on May 27, 1998 (Appeal Book, vol. 11, p. 2888), Ms Steegstra states that “HQ is now interested in the ever-growing intake of Hungarian cases” and reports that, since the beginning of the year, “we have had 12 positives, but only 4 with reasons, two of which have been bench positive reasons.” In an e-mail sent earlier that day, Ms Steegstra had said that, since November 1997, there had been “13 positives and 2 negatives”. “HQ” appears to refer to the Board, not CIC. Mr Bubrin was on the distribution list for these e-mails. In an e-mail from the Deputy Chair, dated July 2, 1998, it is reported that, since the beginning of the year, approximately 65 Hungarian refugee claims had been decided by the Board, most of them from Roma, and that almost all had been in favour of the claimants. [19] In an e-mail addressed to Mr Bubrin on May 28, 1998 (Appeal Book, vol. 11, p. 3212), Mr Gregory James, a co-ordinating member of the Convention Refugee Determination Division, emphasised the importance of “fully elaborated reasons” for positive decisions in Hungarian Roma cases and expressed the view that “it is inevitable that public scrutiny will rise to levels similar to Czech cases.” He said that he would be recommending that a special hearing should be set up “for a real but archetypal case”, the processing of which would be expedited. Mr James proposed that the Minister participate at the hearing of the “lead case” and that “good counsel” should be involved. [20] Mr James noted in this e-mail that he had suggested a similar strategy to deal with Czech Roma claims and expressed disappointment that it had not been acted upon. He further stated that the Department (that is, CIC) is “not impressed with our decisions and keeps making noises that they have information that we didn’t consider”. A “full blown hearing” of a Hungarian Roma claim, with full participation by the Minister, he opined, may cause CIC “to put up rather than simply snipe from the sidelines.” Mr James concludes by saying that Board members hearing subsequent claims by Hungarian Roma would not be bound by the findings made in the lead case, “but they may find it persuasive and at least it should serve to clear the air.” [21] Counsel for the Minister said that it is clear from Mr James’ e-mail that the lead case strategy was an initiative of the Chair and the Deputy Chair of the Board, and did not emanate from CIC. (ii) planning the lead case [22] Ms Steegstra was responsible for the day-to-day management of the lead cases. She testified in her affidavit that Peter Wuebbolt was selected as the lawyer who would represent claimants, on the ground that he had more Hungarian Roma files before the Board than any other lawyer. It was contemplated that the lawyer selected to represent the claimants in the lead case would be able to obtain a level of funding from the Ontario Legal Aid Plan which would adequately reflect the hours involved in preparing for and attending a lead case hearing, and that, once a lawyer had been identified, the Deputy Chair should contact legal aid for this purpose: Appeal Book, vol. 11, p. 3071. [23] Ms Steegstra described the process by which the appellants’ claims were selected as the lead case. The aim was to select “very typical claims, not particularly complex or high profile.” As requested, Mr Wuebbolt brought into a Board meeting all his files and, after the main part of the meeting had ended, he and Ms Steegstra reviewed them, removing from consideration “particularly problematic files”. [24] Meanwhile, Mr Wuebbolt explained the lead case “scenario” to his clients and submitted a short list to Ms Steegstra and Mr Bubrin. Ms Steegstra made the final selection, “not someone involved in the hearing process.” It is clear that Mr Wuebbolt participated actively and willingly in the lead case process. [25] A Refugee Claims Officer was selected for the hearing, and the Minister of Citizenship and Immigration was invited to participate. The Minister’s representative at the hearing was selected by the CIC Manager in Toronto. [26] Refugee claims are normally decided on the basis of written material about country conditions available in the Board’s Documentation Centre. The most distinctive feature of the hearing of the lead case was that expert witnesses were to be called to testify about the situation facing Roma in Hungary, especially regarding the availability of protection against discrimination and persecution provided by state authorities, including the police. [27] Mr Wuebbolt selected two witnesses whom he would call to testify at the hearing on behalf of the appellants. The Minister’s representative selected four witnesses: Dr Holtzl, a representative of the Hungarian Government; Dr Kaltenbach, an Ombudsman for minority rights in Hungary; Mr Farkas, president of the National Roma Minority Self-Government; and Mr Biro, a journalist and, among other things, Chair of the Board of the European Roma Rights Centre. [28] The hearing panel originally comprised one Toronto-based Board Member, Mr Popatia, and one from Montreal, Ms Berger. The rationale for including a Montreal-based Member to sit on a panel to hear a case arising in Toronto was that, to the extent that there were differing regional perspectives on refugee claims made by Hungarian Roma, it would increase the persuasiveness of the lead case to include on the panel Board Members from Toronto and Montreal. [29] However, in an e-mail dated September 18, 1998 (Appeal Book, vol. 11, p. 3096), Mr Bubrin stated that he would sit on the panel in place of Mr Popatia, who he had withdrawn for “personal reasons”. In an e-mail, dated October 26, 1998, Ms Steegstra wrote that Mr Bubrin was concerned that he remain “distanced” from the lead cases since he was going to sit on the panel that would hear them: Appeal Book vol. 11, p. 3098. Accordingly, Mr Bubrin had asked not to be included in further communications about the lead cases, a suggestion with which the Deputy Chair concurred (ibid.), saying: it is a good idea to provide some distance between the panel and Board management to avoid any possible inference of improper influence over the panel’s deliberations. [30] Finally, it should be noted that the Board planned the lead cases without either publicity, or consultation with interested NGOs or members of the immigration and refugee law bar, other than Mr Wuebbolt. The Board’s first public explanation of the lead case strategy appeared in the “Lead Case Backgrounder”, which was not published until March 1999, a month after the appellants had filed their applications for leave to commence an application for judicial review challenging the procedural fairness of the Board’s rejection of their refugee claims. (iii) the lead cases’ hearing [31] As already noted, while the Board normally assesses country conditions solely on the basis of information available in the Board’s Documentation Centre, the lead cases were distinguished by the presence at the hearing of the expert witnesses called by the Minister’s representative. They testified on the country conditions prevailing in Hungary for Roma, and were cross-examined by Mr Wuebbolt, counsel for the appellants. [32] Counsel who appeared for the appellants in the appeal to this Court, Mr Galati, made two complaints about the procedure at the Board hearing. First, he said, the appellants were excluded from the hearing room during the testimony of Dr Holtzl, and had to view this part of the proceeding on a monitor in another room. Second, he submitted that the translation of some of the expert testimony was inadequate. (iv) events after the hearing of the lead cases [33] The reasons of the Board in Smajda were signed by Ms Berger and concurred in by Mr Bubrin, while Mr Bubrin signed the reasons in the Kozak case, and Ms Berger concurred. The reasons in both cases are dated January 20, 1999. The formal orders of the Board dismissing the appellants’ refugee claims were signed on behalf of the Board on January 21, 1999. [34] Counsel for the appellants referred to evidence of the publicity that the lead cases had received in the media in Hungary, both before and after the decisions in the appellants’ cases had been released by the Board. Mr Roger Rodrigues, a lawyer practising with Mr Galati, attached as an exhibit to his affidavit, dated July 11, 2000, a copy of the January 1999 issue of Roma Rights: Appeal Book, vol. 10, pp. 2787-88. This publication stated that, on January 5 and 7, 1999, two leading daily newspapers in Hungary had carried stories about two decisions by the Board on refugee claims by Hungarian Roma. One newspaper described the decisions as “precedent-setting” and stated that they meant that Hungarian Roma refugee claims would not be accepted in Canada. [35] In addition, each of the principal claimants of the Kozak and Smajda families swore affidavits to the effect that, while searching the internet on January 18, 1999, they had come across the website of a Hungarian newspaper which stated that a Canadian decision concerning refugee claims by two families of Hungarian Roma was expected “in a couple of days”: Appeal Book, vol. 10, pp. 2726-27 and 2739-40. [36] These affiants also referred to reports in two online Hungarian newspapers of January 22, 1999: Appeal Book, vol. 10, pp. 2727 and 2740. One newspaper announced a “precedent decision” involving refugee claims by two Hungarian Roma families. The identifying characteristics contained in the reports of the families concerned indicate that they refer to the appellants. On the same day, Hungary’s largest circulation newspaper carried a story under the headline, “Canada Does Not Grant Asylum to Hungarian Roma”: Appeal Book, vol. 10, p. 2788. [37] The Board’s decisions denying the appellants’ claims were signed on January 20, 1999, and released to the appellants the next day. Counsel suggested that a reasonable person might think that these press reports referred to the appellants’ cases and that, even before the decisions in the lead cases were signed, officials of the Board or CIC had started to ensure that they were known in Hungary, in order to deter Hungarian Roma from coming to Canada to claim refugee status. [38] Counsel for the appellants also pointed out that the percentage of positive decisions rendered by panels of the Board in Hungarian Roma claims dropped dramatically in the six months after the publication of the reasons in the lead cases: Appeal Book, vol. 10, pp. 2807 and 2810-11. Affidavits sworn by refugee lawyers and by officers from the Roma Advocacy Centre in Toronto state that, following the release of the reasons in the lead cases, the Board routinely rejected Hungarian Roma claims on the basis of the lead cases, without any independent assessment of the evidence of country conditions: Appeal Book, vol. 10, pp. 2755-74. [39] Thus, the percentage of positive decisions dropped from 71% in December 1998 to 27% in the three months after the decisions in the lead cases were released and to 9% in the next thee months. In the first six months of 2002, the percentage of positive decisions in Toronto was 6%, well below the average for this period in Montreal, Calgary and Vancouver. Based on these statistics and affidavits, counsel submitted that a reasonable person would think that the impact of the lead cases in Toronto confirmed that they had been designed to reduce the number of positive decisions in the future. [40] To the extent that the lead cases were intended to establish a firm factual and legal foundation for the future determination of Hungarian Roma claims (especially on the questions of whether claimants faced persecution or discrimination, and the adequacy of state protection), they did not prove very successful. Several Hungarian Roma claimants, whose claims the Board had rejected on the basis of the findings of fact in the lead cases, successfully applied for judicial review. [41] Thus, in a number of cases, the Federal Court held that the Board had relied too heavily on the country condition findings described in the “Hungarian lead case information package”, which contained more than 75 documents and transcripts of the oral evidence given by the six expert witnesses at the lead case hearings, without making its own analysis of the evidence before it. See, for example, Polgari v. Canada (Minister of Citizenship and Immigration), 2001 FCT 626; Sarkozi v. Canada (Minister of Citizenship and Immigration), 2001 FCT 649; Balogh v. Canada (Minister of Citizenship and Immigration) 2002 FCT 809; Mohacsi v. Canada (Minister of Citizenship and Immigration), [2003] 4 F.C. 771, 2003 FCT 429 (F.C.). In other cases respecting Hungarian Roma refugee claims, the Court has upheld negative decisions by the Board on the adequacy of state protection: see, for example, Racz v. Canada (Minister of Citizenship and Immigration), 2004 FC 1293, where some of the conflicting authorities are listed. C. THE FEDERAL COURT’S DECISION [42] The Applications Judge in this case rejected all the grounds of review advanced by the appellants. He concluded that they had failed to establish that: the “lead case” initiative was not authorized by the Immigration Act; its origin and execution created a reasonable apprehension of bias; the appellants were denied procedural rights at the hearing by their temporary exclusion from the hearing room and by inadequacies in the translation of oral evidence; and the Board committed reviewable errors in concluding that Hungarian authorities denied them adequate state protection from persecution. Nonetheless, having found that “special reasons” existed, he awarded costs to the appellants. [43] Of particular relevance to this appeal is the Judge’s conclusion that the Board’s motive for the lead case initiative was to help to ensure consistent decisions in the determination of refugee claims by Hungarian Roma. He found also that the e-mail evidence relied on by the appellants was insufficient to establish a reasonable apprehension of bias on the ground of a pre-disposition on the part of the hearing panel, as a result of the prior organization of the lead cases by members of the Board’s management, including Mr Bubrin, and the contacts between the Board and the CIC regarding Hungarian Roma claims. D. ANALYSIS Two preliminary matters (a) standard of review [44] Whether a tribunal’s decision was made in breach of the duty of procedural fairness, including the requirement of impartiality, is determined by a reviewing court on a standard of correctness. [45] In determining whether an administrative tribunal committed a reviewable error, an appellate court should normally put itself in the shoes of the Applications Judge: the focus of an appellate court thus generally remains on the reasons and actions of the agency under review, rather than on the reasons of the Applications Judge. [46] In this case, I have reached a different conclusion from the Applications Judge on whether the facts give rise to a reasonable apprehension of bias. In my respectful view, he did not recognize that, since Charter rights are at stake in refugee proceedings before the Board, an independent adjudicative body, the reasonable apprehension of bias standard is particularly demanding. [47] Contrary to the Judge’s conclusion, the appellants may establish a reasonable apprehension of bias without proving the motivation of the Board in orchestrating the lead cases. In my respectful view, it is sufficient that a reasonable person could conclude from a review of the evidence as a whole that the Board’s motive was such as to make it more likely than not that the hearing panel was not impartial. (b) the certified question [48] In view of the conclusion to which I have come, I need only address two substantive issues in these reasons. First, did the circumstances preceding the hearing and determination of the appellants’ refugee claims by the Board give rise to a reasonable apprehension of bias or lack of independence on the part of the panel members who heard and decided them? Second, if they did, was the appellants’ right to relief waived by their and Mr Wuebbolt’s failure to raise the issue of bias earlier? The third issue to be addressed concerns the Applications Judge’s award of costs to the appellants. [49] Accordingly, I do not find it necessary to answer the question certified by the Applications Judge, namely, “Did the IRB have jurisdiction to conduct a “lead case” under the Immigration Act?” The Minister relied on the following provision in the Immigration Act as legal authority for the lead case strategy: 65(3) The Chairperson may, after consulting with the Deputy Chairperson and the Assistant Deputy Chairpersons of the Refugee Division and the Appeal Division and the coordinating members of the Refugee Division, issue guidelines to assist the members of the Refugee Division and Appeal Division in carrying out their duties under this Act. 65(3) Le président, après consultation du vice-président et des vice-présidents adjoints de la section du statut et de la section d'appel et des membres coordonnateurs de la section du statut, peut, par écrit, donner des directives aux membres de ces sections en vue de les assister dans l'exécution de leurs fonctions. This provision has been replaced by subsection 159(1) of the IRPA: 159. (1) The Chairperson is, by virtue of holding that office, a member of each Division of the Board and is the chief executive officer of the Board. In that capacity, the Chairperson … (h) may issue guidelines in writing to members of the Board and identify decisions of the Board as jurisprudential guides, after consulting with the Deputy Chairpersons and the Director General of the Immigration Division, to assist members in carrying out their duties; … 159. (1) Le président est le premier dirigeant de la Commission ainsi que membre d'office des quatre sections; à ce titre : sections autres que la Section de l'immigration; […] h) après consultation des vice-présidents et du directeur général de la Section de l'immigration et en vue d'aider les commissaires dans l'exécution de leurs fonctions, il donne des directives écrites aux commissaires et précise les décisions de la Commission qui serviront de guide jurisprudentiel; […] [50] The Board issued no guideline under the Immigration Act respecting a lead case format for claims by Hungarian Roma. However, I do not decide whether the Board required explicit statutory authority for its lead case initiative. Differences in the underlined wording of these two statutory provisions make it inappropriate to opine needlessly on the meaning of either the repealed provision or its replacement. Issue 1: Were the lead case decisions vitiated by bias or lack of independence? (i) the applicable test [51] Applying the familiar test of bias articulated by de Grandpré J. in Committee for Justice and Liberty v. Canada (National Energy Board), [1978] 1 S.C.R. 369 at 394, the Applications Judge concluded that the appellants had not established that the circumstances surrounding the origin, planning and execution of the lead cases gave rise to a reasonable apprehension of bias. Accordingly, he concluded that the adoption of procedures by an administrative agency to promote consistent decision-making was warranted, provided that the independence of future panels was not thereby compromised. [52] Although trite, the definition of bias bears repetition. A tribunal’s decision is liable to be set aside for bias if a reasonable person, who was reasonably informed of the facts and had thought the matter through in a practical manner, would conclude on a balance of probabilities that the decision maker was not impartial. A similar test determines whether a tribunal is independent. Three preliminary considerations may be added to this general proposition. [53] First, the standard of impartiality expected of a particular administrative decision-maker depends on context and is to be measured by reference to the factors identified by L’Heureux-Dubé J. in Baker v. Canada (Minister of Citizenship and Immigration), [1999] 2 S.C.R. 817 at para. 47. The independence of the Board, its adjudicative procedure and functions, and the fact that its decisions affect the Charter rights of claimants, indicate that the content of the duty of fairness owed by the Board, including the duty of impartiality, falls at the high end of the continuum of procedural fairness. [54] The reasonable person in the rule against bias is not to be equated with either the losing parties or the unduly suspicious. However, the high standard of impartiality and independence applicable to the Board will be reflected in the determination of whether the appellants have established a reasonable apprehension of bias. [55] Second, the Board is charged with a uniquely difficult mandate of administrative adjudication. For instance, throughout the 1990s, the Board carried a very heavy caseload and had a large membership. Its approximately 200 members sat across Canada in panels of two. The Board had to keep abreast of the often rapidly changing human rights conditions in the places from which refugee claimants came, and was subject to unpredictable and sudden increases in claimants from particular countries. In addition, because of the often sensitive nature of its subject-matter, the Board operated in the glare of the political and public attention attracted both by individual decisions and more systemic issues, such as the time taken to render decisions and the backlog of cases waiting to be heard. [56] In view of these challenges, the Board has had to devise means of maintaining and enhancing the consistency and quality of its decisions, which is of critical importance to its ability to perform its statutory functions and to retain its legitimacy. To this end, the Board’s procedure should not be confined in a model of due process that draws exclusively on the judicial paradigm and discourages innovation. Nonetheless, procedures designed to increase quality and consistency cannot be adopted at the expense of the duty of each panel to afford to the claimant before it a high degree of impartiality and independence. [57] Third, the legal notion of bias connotes circumstances that give rise to a belief by a reasonable and informed observer that the decision-maker has been influenced by some extraneous or improper consideration. Similarly, a belief that a decision-maker is not independent goes to the improper surrender of freedom as to how disputes should be decided. In determining propriety, the legitimate interests of the agency in the overall quality of its decisions cannot be ignored. (ii) applying the test to the facts [58] I cannot point to a single fact which, on it own, is sufficient to establish bias. There is, for example, no evidence of a statement by a senior Board official or member that the purpose of the lead cases was to reduce the number of positive decisions in Hungarian Roma cases and to deter potential claimants, although there are references early in the planning stage to the high rate of positive decisions previously rendered, to CIC’s concerns about this, and to public opinion. [59] I would identify Mr Bubrin’s participation in the hearings as particularly unfortunate, given the leading role that he had had in planning and organizing with Board management the lead cases. The fact that the Board did not involve members of the refugee Bar or interested NGOs in the planning process for this novel initiative, and released no explanatory public statement until after the appellants had sought leave in the Federal Court to apply for judicial review, also contributed to creating a cloud of suspicion. [60] Nonetheless, despite the absence of a “smoking gun”, I have concluded on the basis of the entire factual matrix of this case that a reasonable person who had considered every aspect of the matter and had thought it through carefully, would think that the hearing panel was biased and was not acting independently when it rejected the appellants’ claims for refugee status. [61] Reading the e-mails exchanged among members of the senior management in the early stages, a person could reasonably conclude that the lead case strategy was not only designed to bring consistency to future decisions and to increase their accuracy, but also to reduce the number of positive decisions that otherwise might be rendered in favour of the 15,000 Hungarian Roma claimants expected to arrive in 1998, and to reduce the number of potential claimants. [62] As counsel for the appellants pointed out, the Board’s previous decisions on similar claims had been overwhelmingly positive, with little evidence of inconsistency. The e-mails also reveal that, on the basis of meetings with CIC officials, members of the Board’s management believed that CIC had concerns about the high number of positive decisions, which they believed resulted from the inaccuracy of the information on which the Board had based decisions regarding claims by Hungarian Roma, and were looking for a procedure that would satisfy those concern. [63] When considered in the context described above, the Board’s selection of both the lawyer and the cases to serve as the “lead cases”, without any wider consultation with the immigration and refugee Bar, would also trouble the reasonable observer. The Board’s selection of the lawyer and of the lead cases may be seen as part of Board management’s response to the concerns of CIC about the Board’s previous positive decisions and its future handling of a large number of Hungarian Roma claims. [64] Mr Bubrin’s decision to sit on the hearing panel provided a link between the adjudication of the appellants’ claims, and the activities on the part of the Board’s management, including Mr Bubrin as head of the European case management team, in initiating and planning the lead cases. Indeed, recognizing the delicacy of his position, Mr Bubrin requested after he had decided to sit on the panel that he be not included in e-mails regarding the lead cases. [65] To summarize, given the high standard of impartiality to which the Board is held in its adjudicative capacity, a reasonable person might well have concluded on the basis of the above that the panel hearing the appellants’ claims was not impartial. This is because one of its two panel members may have been predisposed towards denying the appellants’ claims since he had played a leading role in an exercise that may seem to have been partly motivated by a desire by CIC and the Board to produce an authoritative, if non-binding legal and factual “precedent”, particularly on the adequacy of state protection, which would be used to reduce the percentage of positive decisions in claims for refugee status by Hungarian Roma. The panel may reasonably be seen to have been insufficiently independent from Board management and thus tainted by the Board’s motivation for the leading case strategy. Support for a belief that the lead case strategy was motivated by a desire to deter potential claimants is the apparent leak to the Hungarian media of the negative decisions before they were released, and the ensuing publicity calculated to deter Roma from leaving for Canada in order to claim refugee status. Issue 2: Did the appellants waive their right to judicial review? [66] Parties are not normally able to complain of a breach of the duty of procedural fairness by an administrative tribunal if they did not raise it at the earliest reasonable moment. A party cannot wait until it has lost before crying foul. In this case, however, the appellants could not have known of the circumstances surrounding the origin and planning of the lead cases, including the involvement of a panel member, Mr Bubrin. Indeed, the Board provided the appellants with no explanation of the format and potential significance of the “lead case”. [67] Nonetheless, the Minister argues that the appellants’ counsel before the Board, Mr Wuebbolt, was fully aware of the situation and had participated willingly and actively in the planning process for the lead cases, and therefore had sufficient knowledge to enable him to object before the decisions were rendered. [68] While parties to litigation are generally bound by the conduct of their counsel, in the circumstances of this case it does not serve the interests of justice to apply this general rule. While Mr Wuebbolt gave some kind of briefing to the appellants on their hearing and obtained their consent to the process, he was not aware of all the facts that gave rise to the reasonable apprehension of bias. In particular, he was not privy to most of the intra-Board e-mails referred to in these reasons. Nor is there reason to believe that he was aware of the apparent “leak” to the Hungarian news media about the Board’s decisions prior to their release. Second, Mr Wuebbolt’s involvement with the Board’s selection of the lead cases, and other aspects of the process, may have caused him to lose sight of his role as the appellants’ counsel. Issue 3: Should the Applications Judge’s award of costs to the appellants be set aside? [69] Costs may only be awarded in immigration and refugee matters if “for special reasons” the Court so orders: Federal Court Immigration and Refugee Protection Rules, SOR /93-22, section 22. Nonetheless, and despite the fact that he dismissed the appellants’ application for judicial review, the Applications Judge awarded the appellants their costs. The Judge held (at para. 77) that “the novel and recognized contentious nature of the lead case at the time that it was brought” constituted “special reasons” warranting an award of costs. [70] Given the broad discretion exercisable over costs, and the Judge’s awareness of the fact that costs are only awarded in immigration and refugee matters in limited circumstances, I am not persuaded that, on the unusual facts of this case, the Judge erred in principle in awarding costs to the appellants. Having so concluded, I need not determine the appellants’ motion to strike from the Minister’s memorandum of fact and law the paragraphs challenging the Applications Judge’s award of costs. E. CONCLUSIONS [71] For these reasons, I would allow the appeals, set aside the order of the Applications Judge except for the award of costs, allow the applications for judicial review, set aside the decisions of the Board, and remit the matters to the Board, differently constituted, for re-determination. For the reasons given by the Applications Judge, as well as the extra-record material obtained by counsel in order to establish that the process culminating in the decisions in the lead cases was flawed, the appellants should be awarded costs on this appeal pursuant to section 22. [72] I do not know how many negative decisions have been rendered subsequently by the Board in which the “lead case information package” was used. Nor do I know how many of these decisions have been upheld on judicial review. I would only note that the decision in the present appeals does not necessarily mean that the factual conclusions in the lead cases are unreliable, or that subsequent decisions which have relied, to any extent, on the findings in the lead cases are thereby vitiated. “John M. Evans” J.A. “I agree Robert Décary J.A.” “I agree K. Sharlow J.A.” FEDERAL COURT OF APPEAL NAMES OF COUNSEL AND SOLICITORS OF RECORD DOCKET: A-419-04 STYLE OF CAUSE: KOZAK GEZA, CSEPREGI ATTILA, KOZAK GEZA (minor) and CSEPREGI SZILVIA v. THE MINISTER OF CITIZENSHIP AND IMMIGRATION DOCKET: A-420-04 STYLE OF CAUSE: SMAJDA SANDOR, SMAJDA ZSOLT, SMAJDA SANDOR, GYULAVICS TIMEA, SMAJDA CLAUDIA and SMAJDA JOZEF v. THE MINISTER OF CITIZENSHIP AND IMMIGRATION PLACE OF HEARING: TORONTO, ONTARIO DATE OF HEARING: NOVEMBER 22, 2005 REASONS FOR JUDGMENT BY: EVANS J.A. CONCURRED IN BY: DÉCARY AND SHARLOW JJ.A. DATED: MARCH 27, 2006 APPEARANCES BY: Mr. Rocco Galati FOR THE APPELLANTS Mr. John Loncar Mr. Gordon Lee FOR THE RESPONDENT SOLICITORS OF RECORD: Galati, Rodrigues, Azevedo & Associates Barrister and Solicitor Toronto, Ontario FOR THE APPELLANTS John H. Sims, Q.C. Deputy Attorney General of Canada FOR THE RESPONDENT