Alternative Law Journal
Catherine Dauvergne and Jenni Millbank[*]
The evidentiary practices of tribunals are vital to asylum outcomes as it is in tribunals, rather than in the courts, that the vast majority of cases for asylum seekers are finally determined. In this article we argue that the evidentiary practices and procedures developed by the Australian Refugee Review Tribunal (RRT) are operating at such a routinely low standard that they contribute to decisions that are manifestly unfair and potentially wrong in law. Our conclusions are drawn from our detailed study of 331 refugee tribunal decisions made in Canada and Australia in response to asylum claims brought by lesbians and gay men. The comparative data from our study forms a back drop to the analysis in this article, which is focused specifically on Australian decisions to the end of 2002. One of our objectives is to argue for better grounded refugee decision-making in Australia.
In this article we examine the range of evidence considered by the RRT and discuss the unique and problematic role played by evidence provided by Australia's Department of Foreign Affairs and Trade (DFAT) and evidence taken from the Spartacus International Gay Guide. We then analyse inappropriate uses of evidence including: a markedly gender-blind approach to evidence; selective use of available evidence; and an indiscriminate approach to questions of the weight and relevance of evidence.
We examined decisions on sexuality-based claims from the refugee tribunals of Australia and Canada focusing on the six-year period from 1994-2000. The two countries have similar laws and processes for refugee determination and both were among the first countries to accept that lesbians and gay men were eligible for convention protection. Our study intended to evaluate how lesbian and gay asylum seekers were faring after a settled period of claims. We were surprised by the extent of difference that we found-Australia was consistently harsher to applicants than Canada, both in the reasoning employed in the decisions and in the trend of outcomes, which had a far lower success rate for both lesbian and gay applicants. Considering all claims, 35% of decisions were favourable to the applicant. In Australia only
22% of claims overall were successful, while in Canada the figure was more than double that, at 54%. In is noteworthy that in the years since our case pool ended, the success rate of applicants in Australia has not increased: in the years 2000-2002 it remained at 21.6%.
In a typical RRT hearing there are only three people in the room: the decision-maker, the claimant and an interpreter. The claimant tells her story and the Tribunal member decides whether she faces a 'well founded fear of persecution' and is therefore a refugee within the terms ofthe 1951 Convention Relating to the Status of Refugees as adopted by the Migration Act 1958 (Cth). Rather than rely on witnesses, documents or cross-examination, refugee decision-makers listen to the story told and translated and then make decisions about credibility largely by comparing that story to other information that they have about the place the story describes. This other information is known as 'independent evidence' because it is not provided by the claimant. It comes from diverse sources including the international and overseas local press, reports of non-governmental organisations, and government sources. A claimant is more likely to be believed if the story she tells fits within the parameters of the story presented in these other sources.
Given the complexities of the refugee determination setting, neither the RRT nor the Canadian tribunal are bound by the formal rules of evidence. This is essential to the logic of the hearing. Without this proviso, many aspects of claimants' stories, as well as some of the independent information, would not be allowed into the process. The removal of the often very technical strictures of the law of evidence facilitates the provision of decisions which are 'economical, informal and quick'. In the RRT, where lawyers are rarely present, adherence to the formal rules of evidence would be completely unworkable. Yet, in the absence of formal rules for dealing with evidence, the practice of the RRT has evolved in a highly inconsistent manner. This inconsistency, and the resulting problems in terms of the weight and relevance accorded to evidence, departs from the central principles underlying the laws of evidence.
We argue that the tribunal values of speed, economy and informality were encoded to best serve those with a stake in the decisions to be delivered, and were never intended to be separated from the values of fairness and justice. The same section which frees the Tribunal from the laws of evidence also mandates that the Tribunal act according to 'substantial justice and the merits of the case'. The Tribunal is required to provide written reasons which set out 'findings on material questions of fact' and which refer to the 'evidence' or other material on which those findings are based.
Our argument is that the intent of the legislation was never to discard the spirit of the law of evidence and the core commitment to fairness that it represents. In the next section, we canvass the role of independent evidence in decisions, consider the types of independent evidence used in RRT decisions and contrast these sources with those used in the Canadian tribunal.
The Australian Tribunal has a staff section devoted to gathering country information and also a library. It is significant, however, that in contrast to Canada, this information is not available to claimants or their advisers. While decision-makers are required to provide claimants with an opportunity to comment on adverse information that may be used in assessing their claims, claimants do not necessarily receive such information in advance of the hearing, and cannot use the resources of the Tribunal in building their own case. The requirement that adverse information be put to an applicant does not mean that the applicant will necessarily receive a copy of the documents in question. The practice of directing an applicant's attention to evidence adverse to their claim orally during the hearing was implicitly approved by the High Court of Australia in Muin. It is important to note that the procedural fairness mechanisms in place since 1998 requiring applicants to be provided with particulars of each specific piece of adverse evidence do not, in fact, assist the vast majority of the claimants we discuss in this article. In most of the cases from our pool, it was not particular evidence but generic information about country conditions that was adverse to the claimant's case and held to defeat it.
The inaccessibility of independent evidence used by the RRT clearly has the potential to disadvantage claimants. It is an obstacle both to those working to build a file of documentation in support of their claim generally and to applicants attempting to counter negative conclusions based on documentation they have not seen. As the RRT provides a review of negative decisions, a claimant is always in the position of arguing against the negative conclusions in their file - and the documents which ostensibly support such conclusions - from the outset.
While both the Canadian and Australian tribunals had roughly comparable usage of sources, such as international commercial and mainstream media reports, international human rights organisations and published academic papers; beyond these sources, comparability broke down. The RRT referred to cables sent by the DFAT in 48% of cases. The Canadian equivalent was never used in the Canadian Tribunal. The Australian Tribunal also referred to the Spartacus International Gay Guide in 24% of cases. This guide was never cited in Canada. Both DFAT and Spartacus are highly problematic sources of evidence and we address them in specific case studies below.
The Australian Tribunal used reports solicited from DFAT as an authoritative source of 'independent' evidence about country conditions in almost half of the cases in our pool. DFAT evidence was usually provided in the form of a short cable. It is noteworthy that such cables are not authoritative governmental documents. They do not express a government or departmental position on an issue; rather, they are written by low level diplomatic staff and are far more likely to reflect a personal rather than institutional view. The independence as well as the quality of this information was a significant concern, as its use had a clear impact on the outcomes of cases.
There is not formally an adversary in the refugee decision-making process, and therefore using government generated information does not, at a theoretical level, create any particular problems. On the other hand, as the present government has demonstrated considerable hostility towards onshore asylum seekers, there may be reasons to believe that evidence it generates will be less than objective. There are potentially two structural reasons why the Australian Tribunal has developed such an extensive reliance on the services of DFAT. The first is that this information is cheap, readily available, and can be easily targeted to a specific information request. Whereas sources such as well known human rights NGOs, the US State Department, or the international commercial press might not mention the specific situations of gay men and lesbians, DFAT cables do. Further the information is brief and directive, unlike more detailed comprehensive sources, and is therefore easy to use. Tribunal members face high case loads and considerable pressure to produce reasons quickly, and may therefore more readily rely on sources of evidence that are easily digested. Second, in many cases this information is already in the file when it reaches the Tribunal.
Any given RRT decision might refer to a number of cables spanning several years. DFAT cables were usually expressed as generalised statements about 'the situation' for 'homosexuals' in a given country. DFAT evidence often did not name the original source or authority, but would refer to a 'local lawyer', 'diplomat', or 'expert' as the basis for their claims. The level of generality in DFAT cables was at times breathtaking, for example, '[t]he level of discrimination against homosexuals in Ghana is probably not more than can be found in most western developed countries', or 'Lebanon does not have a culture of gay bashing per se', or there is 'no queer bashing' in Indonesia.
The DFAT evidence, when it was on the question of country conditions specific to sexuality, was almost universally negative to the applicant's case. As such, it appeared that DFAT country information played a role in the high level of negative outcomes for asylum seekers in Australia as compared to Canada. When DFAT evidence on sexuality was accepted, 89% of applicants were unsuccessful (a failure rate more than double the general failure rate in Australia over the period of our study). On the rare occasions when DFAT evidence on sexuality was rejected, the numbers of successful applicants actually exceeded those who were unsuccessful. Where DFAT evidence was at odds with country information supplied by other sources, such as NGOs, the Tribunal tended to prefer the evidence of DFAT.
DFAT cables frequently included conclusions of fact or law. Such conclusions were very problematic as they were often speculative and these statements were frequently adopted as authoritative in the RRT decisions. So, for example, a DFAT cable noted that there was widespread extortion of gay men in Lebanon, but concluded in a speculative fashion that this was not motivated by 'hatred of gays' but by greed. The Tribunal accepted that conclusion and held that widespread extortion was not evidence of persecution. The Tribunal did not analyse the evidence further, for example, as evincing a particularly vulnerability of the targeted group, or as evidence.of a likely failure of state protection. There are numerous other examples of a single speculative line in a DFAT cable being used as the basis for comprehensive conclusions on a country situation. A case concerning a lesbian applicant from Ghana illustrates how even brief statements were used to support very broad conclusions,
The Tribunal notes that although the Spartacus Gay Guide and the International Encyclopaedia of Sexuality only refer to gay men, the DFAT Cable was in response to a more general question about discrimination against male and female homosexuals in Ghana. In light of the generality of the questions put by DFAT, and since the nature of the answers is largely consistent with the information contained in the Spartacus Gay Guide and the International Encyclopaedia, the Tribunal believes that it is reasonable to infer that the observations contained in these three sources would be equally applicable to gay men as well as lesbians and there is no evidence before the Tribunal that lesbians are treated differently in Ghana from gay men.
In fact the DFAT cable in question only referred to 'homosexuals' and it was quite apparent that this did not include lesbians as it stated, 'We have received no reports of imprisonment/torture of gay men'.
DFAT cables were also, on occasion, simply wrong. For instance, a cable on Lebanon stated that prosecutions for homosexual sex under the criminal law were 'rare', but this was directly contradicted by Lebanese press reports. At times the information within a DFAT cable was clearly self-contradictory. For example DFAT evidence on Nepal was that, 'homosexuals receive the same treatment from police as other citizens' but also stated that there were no open or known homosexuals in Nepal, which makes equal treatment rather difficult. 
As all of the DFAT cables for a particular country would be sourced on the RRT electronic database (CISNET) there was considerable scope for choice as the cables were collated over the years. It is notable that DFAT cables on occasion contradicted each other, and so decision-makers were able to pick and choose among cables in compiling country evidence, with no indication in the text of the decision that any selective process was going on.
There were a number of areas in which the quality of information used by the RRT was dubious or its application inappropriate. The Spartacus Guide is the principal example of this. Use of the guide as a source of country evidence was strongly associated with negative decisions: of the decisions where it was used, 90% of applicants were unsuccessful.
The Spartacus International Gay Guide is a travel guide aimed at gay men. It contains a wealth of information about tourist venues, is illustrated with extensive advertising, and relies on a certain amount of reader input into the compilation of the material which is updated each year. The Spartacus guide is not focused, nor would one expect it to be, on human rights issues or legal issues.. It contains information clearly aimed at foreigners, and does not purport to describe the conditions faced by locals who live in the countries in question. Most decision-makers did not make informed use of the guide. It was, for example, used frequently as evidence of 'increased tolerance' and a 'flourishing gay scene' in Shanghai. Yet the 95/6 edition had a Shanghai listing that in fact included only a few outdoor cruising areas.
The Spartacus Guide was used as a source of law in 26 cases. More than half of the time the guide was used as a reference to the criminal law. The Spartacus Guide was also on occasion used to counter other sources which gave a more detailed and legally focused account. For example, in one claim by a person from Costa Rica, the Tribunal used Spartacus to counter information from Amnesty International and the Third Pink Book:
The Tribunal notes that although The Third Pink Book published in 1993 states that police raids are irregularly reported in San Jose, a later gay guide Spartacus, says nothing about such raids and indicates that Costa Rican society mostly tolerates gays.
This conclusion is drawn from the single statement: 'In spite of the influence of the Catholic church, society mostly tolerates gays'. Most importantly, it ignores the informatior. found two pages later that, '[t]he police regularly carry out raids at the following cruising areas!' The guide was similarly used to counter information from the Third Pink Book in assessing claims from Ghana. Unlike Spartacus, the Third Pink Book does actually claim to document the social and legal situation of the countries that it covers.
While the Spartacus Guide provides lists of venues that may be of interest to tourists, its use by the Tribunal as a legal authority is inappropriate, particularly given the availability of other sources of evidence. The problem is therefore twofold; that the Spartacus Guide is used frequently despite its tenor and intended audience, and that it is often used incompletely, without averting to all the information it provides on a particular country or cruising area.
Further, we found inappropriate gender-blind use of the Spartacus Guide in assessing claims by lesbians. The Spartacus Guide was referred to as a source of country information in ten Australian cases concerning lesbian applicants. While the Tribunal referred to Spartacus as a 'guide for gay and lesbian travellers', the book itself is in fact subtitled 'international gay guide', refers only to 'gay' travellers throughout the 95/6 edition, and is filled with advertisement for male venues featuring nude or semi-nude men. There are no images of women, no mention of lesbian travellers specifically and no venue coding for women-only or lesbian-specific venues. Moreover the book is clearly slanted towards venues where male-male sex might be found around the world - so although it lists gay-friendly accommodation and coffee shops, most listings typically focus on gay bars, cruising areas and porn cinemas.
In all of the ten lesbian cases where the Spartacus Guide was used, it was clearly inapplicable to country conditions for lesbians and inappropriate to the applicant's individual claims. In every one of the ten cases the lesbian applicant was unsuccessful. In each of these cases, the Spartacus Guide was referred to as evidence that there were support groups, a gay scene, a visible gay presence, decreased hostility towards gays and lesbians, or increased tolerance of lesbians and gay men. This evidence was expressly used to support a conclusion that there was not a likelihood of future persecution of the applicant. Yet when the guidebook was read closely, it was clear that none of the cited material was actually applicable to lesbians.
For example, in a 1999 case from Colombia, the Tribunal used a listing of venues from Spartacus, all of which were coded as gay-only, including porn cinemas and bath houses, to find that the lesbian applicants (a couple) were objectively safe from persecution. While there is no logical reason to think that any lesbian should know where to find a gay men's bath house, the Tribunal also doubted the credibility of the claimants on the basis that they did not know of these venues.
The use of gender inappropriate evidence in the RRT was very widespread. Such evidentiary practices contributed strongly to the dramatically low success rate-a mere 7% -for lesbian applicants in the Tribunal.
Both the Spartacus Guide and the use of DFAT evidence demonstrate problems with the types of independent evidence used by the RRT. While there are better and worse uses of both sources of evidence, even the best use of each presents significant problems. In addition to problems of which evidence is chosen, our study also showed problems of how evidence was used once it had been accepted by the Tribunal.
It was disturbing that in a number of cases, the Australian Tribunal selectively used sources of information in such a way as to misrepresent them or present a misleading picture of the totality of available evidence. In several cases, Tribunal members quoted sources about the existence of clubs or cruising locations as evidence of a 'thriving' or 'visible' gay scene to disprove an objective likelihood of persecution. On comparison with the original source material, it was clear that the decision-maker had seen, but chosen not to record, accompanying warnings that such venues were dangerous or commonly subject to police harassment.
In a series of decisions about claimants from China, the Tribunal referred to a DFAT cable from 1997 which stated,
'There is a sizeable gay community in Shanghai, with known meeting places'. In only one case did the Tribunal print the full response from DFAT which continued:
Shanghai's population and the large foreign presence makes it easier for homosexuals to get together. That said, it also makes it easier for the public security bureau (PSB) to identify and harass homosexuals. Harassment can include arrest, temporary detainment, physical violence and demands for bribery (see below).
(b) There is still widespread social prejudice against homosexuality which would make it extremely difficult for gay couples to live together openly.It is possible that a more discreet arrangement might be tolerated.
The cable answered 'yes' to questions about police raids of gay venues (c and d) and continued,
(e) Crackdown on drugs and other criminal activities have been given as the reason for police harassment of homosexuals. Homosexuality per se is not a crime. We understand that some of those arrested in crackdowns inNovember 1996 and Aprill997 were released after paying large sums to the PSB.
In numerous cases on China only part of the above cable was quoted. So, for example, it appeared as:
DFAT commented in 1997 that there was a sizeable gay community in Shanghai and that there are known meeting places for homosexuals. Social prejudices against homosexuality still persist and that homosexuals could still be subjected to harassment. [DFAT SH642, 30 July 1997, CX25265]
Several of these decisions held that there was little or no likelihood of arrest, although the cable itself documented that arrests had taken place. In some cases the cable was quoted only to support the proposition that there was a sizable gay population and that the likelihood of persecution was therefore objectively low.
The same pattern of selective and misleading representation of the country evidence appeared in cases concerning Bangladesh. In those cases, evidence of the head of an HIVIAIDS foundation in a fax from 1997 included information that there were very few gay identified men in Bangladesh, few or no cohabiting gay couples, that men who conform outwardly to social norms, 'most importantly by marrying and having children' can 'get away with male to male sex provided it is kept secret', and that harassment of men who had sex with men mostly took the form of extortion by hustlers and the local police but could also include being bashed. The author stated that he was not aware of any prosecutions or of any significant harassment from fundamentalist Islamic groups. This evidence was repeatedly cited as standing only for the view that there were no prosecutions and no harassment from Islamic groups.
We found numerous other cases in which material was either misquoted or misrepresented. In all of the cases where selective presentation of the evidence took place, a fair assessment of the totality of the source material did not demonstrate, as the decision-maker had held, an unlikelihood of persecution. Indeed it is arguable that some of the country information quoted in full stood for exactly the opposite, that persecution was possible or likely.
More commonly, the Australian RRT selected among different sources of evidence and referred to only some of them, without acknowledging either what sources were omitted or on what basis the chosen sources were to be preferred.
When we examined together all of the cases concerning applicants from Lebanon, there was a very sharp division between the successful and unsuccessful decisions, with completely different evidence relied upon in each group. In three of the negative decisions, the Tribunal relied on a DFAT cable as decisions that there was not a persecutory environment, yet none of the four positive decisions mentioned this cable. Conversely, three of the four positive Australian decisions cited a 1987 statement by the Lebanese Embassy that 'homosexuality is not accepted in Lebanon', although this evidence was not referred to anywhere in any of the four negative decisions.
One source of problems outlined above appears to be the use or overuse of the 'cut and paste' function by Tribunal members in writing their decisions. It was quite common in decisions on the same country for large slabs of text on country conditions to be repeated verbatim over a period of years. The cut and paste approach to country information is particularly questionable when country conditions are alleged to have changed, or, as in some cases, where the applicant's own experience was in direct contradiction to the evidence and no real attempt was made to reconcile them. This practice almost certainly contributed to misquoting and selective usage of evidence, as original sources were not being consulted or checked for accuracy and reliance was placed instead on a secondary precise of the evidence from other decisions.
The reproduction of country information was not confined to cases by the same decision-maker; we noted several cases where different decision-makers used large sections of text that were identically worded or used very similar wording with minor editing or variation, for example in numerous recent decisions on Iran and Bangladesh. This finding suggests an entrenched institutional practice whereby evidence is 'cut and pasted' into decisions. While this practice may reflect institutional imperatives, such as a high case load with a correspondingly short time in which to produce reasons, the practice of verbatim reproduction of evidence across decisions by different Tribunal members must at least raise the real apprehension that some of the decision makers in question have never at any prior time actually read the original evidence on which they are basing their decisions.
Acknowledging the considerable difficulties of weighing evidence to determine if an applicant faces a real chance of future persecution, we considered - and rejected - the argument that the RRT is doing the best it can with the information available. We did this by making two comparisons: with the Canadian Tribunal and with evidence available from the International Gay and Lesbian Human Rights Commission, a US-based non-government organisation formed in 1990. It was of considerable interest to us that, although the RRT claimed to have holdings oflnternational Gay and Lesbian Human Rights Commission (IGLHRC) country information packets, they were not referred to in any of the Australian cases in our pool. The evidence used by the RRT was often at odds with that used in the Canadian tribunal and with that available to the public in the IGLHRC country information packets. Given the significance of country information in determining the outcomes of refugee cases, it is strongly argued that greater effort should be made to access the most comprehensive information available.
Canada and Australia have very similar laws and processes for refugee determination. Both countries have been adjudicating refugee claims based on sexual orientation since the early 1990s. Yet the two countries manifest very different results, with Canada being more than twice as likely to accept refugee applicants on the basis of sexual orientation compared to Australia. The greater overall success rates of applicants in Canada and the vastly greater success rate of lesbian applicants in Canada, compared to Australia, can be explained in part by the extremely poor evidentiary practices of the Australian Tribunal.
There are a number of areas in which the RRT could make changes to its current practice, including:
• making the RRT Country Information database publicly accessible;
• introducing Refugee Protection Officers, as used in the Canadian Tribunal, to assist in putting evidence before the Tribunal;
• subjecting the RRT hearing itself to more stringent requirements of procedural fairness, including a duty to inquire;
• making all decisions of the RRT publicly available;
• making a greater effort to obtain and use evidence from organisations with specific expertise in human rights abuses based upon sexual orientation;
• ending the practice of using DFAT cables as sources of country evidence;
• ending the practice of using the Spartacus International
Gay Guide as a source of country evidence;
• making greater use of country information collected by the Canadian Tribunal;
• considering having RRT members specialise in 'sexuality' claims as they currently do in 'country' claims;
• providing training for RRT members, specifically on the issue of evidence in claims based on sexual orientation; and
• reaffirming a commitment to the principles of evidence and drawing up guidelines for assessing the relevance and weighing of sources of evidence.
In the majority of asylum applications in Australia, the Refugee Review Tribunal is the ultimate decision-maker. It is in the interests of justice as well as efficiency that its determinations be based on high quality evidence and sound evidentiary practices. Our study demonstrates that this has not been the case in Australia to date. Many improvements could be undertaken at the level of administrative practice to ensure better quality decision-making through the use of the best available evidence as well as consistency and accuracy in its application. We urge the Tribunal to immediately begin a process for implementing change, in the interests of fairness, Australian refugee law, and the international leadership role that Australia plays in developing refugee jurisprudence.
[*] Catherine Dauvergne teaches law at the University of British Columbia.
Jenni Millbank teaches law at the University of Sydney. Thanks to Tiffany Hambley for her research assistance on this paper. A far more detailed version of these arguments is published in August 2003 Federal Law Review.
This project was supported by a grant from the University of Sydney, Faculty of Law, Legal Scholarship Support Fund.
© 2003 Catherine Dauvergne and Jenni Millbank (text)
© 2003 Stuart Roth (cartoon)
 See Jenni Millbank, 'Imagining Otherness: Refugee Claims on the Basis of Sexuality in Canada and Australia'  MelbULawRw 7; (2002) 26 Melbourne University Law Review 144; Catherine Dauvergne and Jenni Millbank, 'Before the High Court: Applicants S396/2002 and S395/2002, a gay refugee couple from Bangladesh'  SydLawRw 6; (2003) 25 Sydney Law Review 97.
 127 of the decisions studied were Canadian and 204 were Australian. We have also reviewed reported Australian cases to the end of2002 to assess whether significant changes had occurred.
 See Audrey Macklin, 'Cross-Border Shopping for Ideas: A Critical Review of United States, Canadian, and Australian Approaches to Gender-Related Asylum Claims' (1998) 13 Georgetown Immigration Law Journal 25. The most important distinction is that the refugee tribunal in Canada makes the decision at first instance, whereas the Australian Tribunal conducts a full merit review of a decision first made by a delegate of the Minister. This distinction is less significant than it seems at first because the acceptance rates at firs instance m Australia are very low (8%) and a majority of those who are rejected seek a merits review in the RRT where the decision is re-made.
 Lesbian claimants were dramatically under-represented in our data, with only 14% of the Canadian claims and 21% of the Australian claims brought by women. Comparing lesbian claimants in both countries is shocking: in Canada their success rate was ten times that of Australia. In Canada, lesbian claimants actually had a somewhat higher success rate than gay men (69% while gay men had a 52% success rate) while in Australia lesbian claimants were overwhelmingly unsuccessful (a mere 7% of lesbian claimants succeeded, compared to 26% of gay men). We argue that one of the most important reasons for the difference in outcomes in Australia and Canada is the evidentiary practices of the RRT.
 Section 420(1) of the Migration Act 1958 (Cth).
 Migration Act 1958s430(1).
 Muin v Refugee Review Tribunal; Lie v Refugee Review Tribunal  HCA 30.
 Migration Act 1958 (Cth) s 424A.
 In RRT Reference N98/24600 (Unreported, Russell, 26 November, 1998).
 In RRT Reference N95/10132 (Unreported, Griffin, 16 September 1997) and RRT Reference N95/09483 (Unreported, Mathlin, 26 November 1997).
 RRT Reference N97115062 (Unreported, Short, 17 November 1997); RRT Reference V97/07412 (Unreported, Haig, 24 December 1997) and RRT ReferenceN98/21139 (Unreported, Witton, 8 October 1998).
 RRT Reference N95/JOJ32 (Unreported, Griffin, 16 September 1997).
 RRT ReferenceN98/24600 (Unreported, Russell, 26 November 1998). For a more detailed discussion see Millbank, 'Imagining Otherness', above n 1.
 RRT ReferenceN95/09584 (Unreported, Blair, 31 October 1996); RRT Reference N95/09483 (Unreported, Mathlin, 26 November 1997) and RRT Reference N99/28440 (Unreported, Mathlin, 20 January 2000).
 RRT Reference N91/14745 (Unreported, Hardy, 14 July 1998); RRT Reference N97/14489 (Unreported, Gutman, 23 July 1998) and RRT Reference N98/23955 (Unreported, Gutman, 24 September 1998).
 RRT Reference N98/20912 (Unreported, Cristoffanini, 3 February 1999).
 RRT Reference N98/24600 (Unreported, Russell, 26 November 1998); and RRT Reference N98/24718 (Unreported, Russell, 19 March 1999).
 Aart Hendricks et al (eds), The Third Pink Book: A Global View of Lesbian and Gay Liberation and Oppression (1993).
 RRT Reference N97118897 (Unreported, Kelleghan, l3 November 1998). All of these sites are coded as, 'At your own risk. Dangerous place with risk of personal attack and police activity'.
 RRT Reference V98/09564 (Unreported, Vrachnas, 4 May 1999).
 Eg RRT Reference V98/08938 (Unreported, Kissane, 2 November 1998) does include the first quote but omits the reference to police raids and the second quote; RRT Reference N97/19670 (Unreported, Holmes, 4 September 1998); RRT Reference N99/27818 (Unreported, Kelleghan, 29 June 1999); RRT Reference N99/26435 (Unreported, Hoysted, 31 May 1999); RRT Reference N98/25578 (Unreported, O'Brien, 2 March 1999); RRT Reference N97/19671 (Unreported, Kelleghan, 25 January 1999) and RRT Reference N97/20446 (Unreported, Zelinka, 11 March 1999) paraphrase elements.
 RRT Reference N98/26197 (Unreported, Gutman, II May 1999); RRT Reference N98121640 (Unreported, Klintworth, 21 December 1998); RRT Reference N98/21178 (Unreported, Klintworth, 4 December 1998) and RRT Reference N98/21639 (Unreported, Klintworth, 21 December 1998).
 See eg RRT Reference N97120090 (Unreported, Short, 8 March 1999); RRT Reference N97/14768 (Unreported, Thomson, 29 April1998).
 See RRT Reference N99/28009 (Unreported, Smidt, 19 June 2000) repeated verbatim in RRT Reference N00/36301 (Unreported, Rosser, 24 December 2001).
 See eg RRT Reference N98/21005 (Unreported, Thomson, 25 May 2000); RRT Reference N95109552 (Unreported, Woodward, 4 September 1998); RRT Reference N98/20994 (Unreported, Rosser, 4 May 1998). In another case, it was cited only as evidence that there were places men could meet for sex: RRT Reference N98/21362 (Unreported, Kelleghan, 28 March 2002). Note that most of the country information utilised was around five years old. More recent and more detailed evidence in a NAZ Foundation study of 124 Bangladeshi men who have sex with men documented, in direct contradiction to the Tribunal's repeated findings that Bangladesh is tolerant of male homosexual behaviour, widespread violence and harassment: see NAZ Foundation 'Social Justice, Human Rights and MSM', Briefing Paper No 7, 200i, online at <http://www.nazfoundint.com/home.html>
Papers, Essays & Reports, Briefing Papers, at 1December 2002. See also the range of information collated in International Gay and Lesbian Human Rights Commission (IGLHRC), Current Update Packet: Bangladesh, (2001).
 A DFAT cable quoting an unnamed 'highly-reliable lawyer' who alleged that while homosexuals were technically 'not tolerated', 'to my knowledge they do not suffer discrimination, ostracism, violence or death threats' was relied on in RRT Reference N98122311 (Unreported, Zelinka, 22 September 1998); RRT Reference N97/19504 (Unreported, Zelinka, 28 September 1998) and RRT Reference N97/18897 (Unreported, Kelleghan, l3 November 1998).
 RRT Reference N94106450 (Unreported, Fergus, 26 July 1996); RRT Reference N95109584 (Unreported, Blair, 31 October 1996); RRT Reference N95/09483 (Unreported, Mathlin, 26 November 1997); RRT Reference N99/28440 (Unreported, Mathlin, 20 January 2000). Although three of these cases predate the negative one, the cable was demonstrably available as it dates from 1995 and 1s on the RRT database: DFAT cable no. BI500123 of25 September 1995 CISNET CXII474.
 RRT Reference N95109584 (Unreported, Blair, 31 October 1996); RRT Reference N95/09483 (Unreported, Mathlin, 26 November 1997); RRT Reference N99/28440 (Unreported, Mathlin, 20 January 2000).
 Over a dozen decisions concerning applicants from China used virtually identical country information over a period of four years from 1996 to 1999. For discussion see Millbank, 'Imagining Otherness', above n 1.
 The IGLHRC is a non-governmental, non-profit organization that seeks to defend and advance the human rights of all peoples and communities subject to discrimination or abuse on the basis of sex ) orientation, gender identity, or HIV status. See IGLRHC website <http://www.iglhrc.org> at 16 December 2002.
 Mark Mantle, RRT Sydney, Head of Country Information Section, interview, 19 February 2002.