Persecution based on treatment of female draftees
34 The applicant claimed that she would face a real chance of persecution if forced to undertake military service on return to Eritrea by reason of being a woman. This claim was made and considered separately from her claims to fear persecution based on draft evasion.
35 First, the applicant gave evidence that women who undertake military service are raped and abused sexually and left with children to cope with. She stated any woman with such a child will be shunned by her own family. She stated she had witnessed many cases of young girls returning home after becoming pregnant while in military training camps. She also stated it was widely known amongst the community that the girls had been subjected to sexual abuse by the officials at the training camp. It is worth noting that these statements were given against a background of evidence, extracted below, of "state-tolerated and state-sanctioned gender discrimination", to employ a phrase of Lord Steyn's from Regina v Immigration Appeal Tribunal; Ex parte Shah [1999] 2 AC 629 at 646.
36 Secondly, a witness who visited Eritrea the previous year also gave evidence before the Tribunal. She gave evidence that on that occasion she had spoken with women who had become pregnant in Sawa (the military training camp to which the applicant would be assigned). She stated that the women to whom she spoke were unwilling to talk of their experiences apart from referring to the power and position of the officials at the camp.
37 Thirdly, the Tribunal also had before it country information, the US State Report extracted above, which reported instances of parents of draft age girls, fatally resisting their daughters being "rounded-up" for military service and noted reports of sexual abuse and harassment of female draftees. It was this document which contained evidence of the institutional discrimination against women in Eritrean society as follows:
"The Government has not taken a firm public stance against domestic violence and generally has ignored the problem. Violence against women was pervasive. Spousal abuse is a crime; however, spousal abuse, especially wife beating, was common. Domestic violence seldom was discussed openly by women because of societal pressures. Such incidents more commonly were addressed, if at all, within families or by religious clergy. It was estimated that more than 65 percent of women in the Asmara area were the victims of domestic violence during the year. The Government response to domestic violence was hindered by a lack of training, inadequate funding, and societal attitudes.
Rape is a crime; however, no specific information was available on its prevalence in the country."
The reports of sexual abuse of female draftees and parents risking death to avoid having daughters drafted fell to be assessed against this social and cultural milieu.
38 Fourthly, the applicant also relied on an article published in "The Age" newspaper on 12 May 2002 by Xavier La Canna entitled "When rape is a requirement of military service."
39 That article noted that Sawa is 315 kilometres North-West of Eritrea's capital, Asmara. It purported to report commentary from the former Eritrean ambassador to Sweden, Hebret Berhe, and United Nations staff, all of whom were said to be aware of rape in military camps and the fact that it was not isolated but was in the words of the former Swedish ambassador "wholesale". The article also purported to quote a former Eritrean diplomat who had been granted asylum in Australia as follows:
"Always beautiful girls are the target of officers . . . They (the women) are always pressured by punishment, and given privileges if they agree [to sexual relations],".
40 That article also reported that the Eritrean ambassador to Australia said the article contained false information. The Tribunal noted that there was a rebuttal of the article by the Eritrean Embassy on 12 December 2002 and that The Age had subsequently withdrawn the article from its website "for bias". There is no evidence before this Court as to what facts are covered by the Tribunal's reference to bias. The Tribunal did not indicate whether it gave no, or any, weight to this article.
41 Fifthly, there was other country information to which the Tribunal said it made reference, including a Human Rights Watch World Report for 2003, Amnesty International, the U.K. Country Information and Policy Unit of the Home Office Asylum and Appeals Policy Directorate Report on Eritrea and a report from the Immigration and Refugee Board of Canada. The Tribunal noted none of those reports of international organisations made mention of the prospect of women being discriminated against in military service. It might be observed in passing that the Amnesty International Human Rights Watch Report for 2002, which constituted part of the evidence relied on by the applicant and which is contained in the Court Book, was a report on Eritrea's violations of the human rights of Government critics, especially focussing on persons in detention for criticising the Government. The subject matter of this report is neither national service generally nor female draftees in particular. It was not surprising it did not deal with the subject matter which the Tribunal was considering.
42 At the hearing before me "country information" entitled "Amnesty International Report - 2003 - Eritrea" said to be relevant to this issue was sought to be tendered. The respondent's counsel objected to this course. It was not clear from the face of the document whether this was in fact evidence to which the Tribunal was referring or fresh evidence. For reasons which will become clear it will not be necessary for me to rule on that application.
43 There is no evidence before me of any document answering the other descriptions of country information of international organisations to which the Tribunal referred.
44 The Tribunal's conclusions on this aspect of the applicant's claims were as follows:
"The Tribunal is prepared to accept that there may be instances of sexual harassment and abuse of women in military training as noted in the US Department of State report and by the applicant's witness. However, the Tribunal is unable to accept, on the evidence before it, that the incidence of such actions is sufficient as to suggest the applicant faces a real chance of being persecuted should she undertake military service or return to Eritrea, for reasons of being a woman."
Because the Tribunal made a finding about the incidence of sexual harassment and abuse of women in military training, I do not take the use of the auxiliary verb "may" to imply any improbability in the accounts; no adverse findings were made in respect of witnesses credit on these issues.
45 It is convenient to isolate issues relevant to this aspect of the applicant's claims which were not seriously contested before me. In accepting the oral evidence and the US Department of State Report the Tribunal accepted as fact that there may be many cases of young girls returning home pregnant from military camp and that the local community knew this was the result of sexual abuse and/or rape by "officials" at military camp (sexual abuse in that context connoted consensual relations which originated in some exploitation of the hierarchy of power as between officials and female draftees).
46 The Tribunal accepted the applicant was a member of a "particular social group" as that phrase is used in Art 1A(2) of the Convention. In this case, the particular social group was female draftees identifiable by two characteristics common to the group, namely being female and liable to the draft. Those two common attributes were independent of any shared fear of persecution and distinguished the group from society at large. On the evidence, it was possible to treat female draftees as a social group by reference to the legal, social and cultural norms prevalent in Eritrean society. The Tribunal's analysis of the facts was thus correct as they fell within the principles to be applied in determining a "particular social group" established in Applicant S at [36]. Further, it was not disputed that the conduct feared was feared "for reasons of" being directed at that particular social group: Ram v Minister for Immigration & Ethnic Affairs (1995) FCR 565; Applicant A v Minister for Immigration & Ethnic Affairs (1997) 190 CLR 225 at 257 per McHugh J..
47 Rape or sexual abuse by an official or a military superior and impregnation whilst on national service is distinguishable from rape occurring as a random incident of civil disturbance or unrest. cf Minister for Immigration & Multicultural & Indigenous Affairs v Ibrahim (2000) 204 CLR 1. Rape, sexual abuse and impregnation whilst on military service is capable of being characterised as "serious harm" within the meaning of ss 91R(1)(b) and 2(b) and (c) of the Act and depending on the evidence, such acts are also capable of constituting "systematic and discriminatory" conduct within the meaning of s 91R(1)(c). Any unwillingness by the State of Eritrea to protect young female draftees from rape, sexual abuse or impregnation by military superiors is relevant when determining whether a fear of persecution is well‑founded: see Khawar; see also Minister for Immigration & Multicultural & Indigenous Affairs v Respondents S152/2003 (2004) HCA 18; 205 ALR 487. It was not argued that the conduct feared did not constitute serious harm or that it was trivial or that it might have minimum impact on the applicant.
48 The issues discussed above did not appear to be in contest. The Tribunal having proceeded on a basis consistent with the propositions and authorities set out above. The contested issue for me was whether there was any jurisdictional error in the Tribunal's finding that it was not satisfied on the evidence before it that the "incidence" of sexual harassment and abuse, rape and impregnation of female draftees by officials was "sufficient" to suggest the applicant faces a real chance of being persecuted should she undertake military service on return.
49 The applicant's representative submitted that the Tribunal failed to properly consider all the relevant evidence in assessing mistreatment of women in the military. Against the backdrop of a submission that the Tribunal asked itself the wrong question, it was submitted that the Tribunal's analysis of this part of the applicant's claim constituted a failure to consider all the evidence and constituted a misrepresentation of the material giving rise to jurisdictional error.
50 The argument advanced on behalf of the respondent was narrow and simple. It was that the Tribunal had formed an adverse conclusion on the evidence as to the "sufficiency" of the "incidence" of sexual abuse, rape and impregnation of female draftees in respect of a fear of persecution based on that conduct and accordingly there was no jurisdictional error. This argument appeared to me to be based on treating the Tribunal's assessment as a qualitative assessment of the conduct feared, that is as a question of fact rather than a question of law: Minister for Immigration & Multicultural & Indigenous Affairs v Kord [2002] FCA 334 at [3] per Heerey J. It was not argued on behalf of the respondent that sexual abuse, rape and impregnation by officials or superiors, while on national service, could not constitute persecution of female draftees with the meaning of s 91R of the Act. This was a responsible position adopted on behalf of the Minister. On this aspect, it was submitted for the respondents that the Tribunal does not commit a jurisdictional error because it makes erroneous findings of fact, attributes weight to some pieces of evidence and not others or adopts unsound and questionable reasoning and finding error in such circumstances amounts to impermissible merits review: Rajalingam; Minister for Immigration & Multicultural Affairs v Al‑Miahi [2001] FCA 744 at 34.
51 There is a line of authority from Chan's case to Ibrahim which deals with a difficult aspect of the definition of refugee as set out in Article 1A(2) of the Convention and as also covered in s 91R of the Act which was inserted into the Act under the Migration Legislation Amendment Act (No. 6) 2001 (Cth). The line of authority deals with the correct meaning of persecution by reference to conduct which is "systematic and discriminatory".
52 In Ibrahim, McHugh J explains that the phrase "systematic conduct" has its origins in a decision of Wilcox J, in Periannan Murugasu v Minister for Immigration and Ethnic Affairs (unreported; Federal Court of Australia; 28 July 1987). His Honour went on to state at [95] and [99/100]:
"The use of the term "systematic conduct" has proved unfortunate. Tribunals have read it as meaning that there can be no persecution for the purpose of the Convention unless there was a systematic course of conduct by the oppressor. That was not what I meant by using that expression in Chan. I used it as a synonym for non‑random, and I think in Murugasu Wilcox J intended its use in the same way . . .
It is an error to suggest that the use of the expression "systematic conduct" in either Murugasu or Chan was intended to require, as a matter of law, that an applicant had to fear organised or methodical conduct, akin to the atrocities committed by the Nazis in the Second World War. Selective harassment, which discriminates against a person for a Convention reason, is inherent in the notion of persecution. Unsystematic or random acts are non‑selective. It is therefore not a prerequisite to obtaining refugee status that a person fears being persecuted on a number of occasions or "must show a series of co‑ordinated acts directed at him or her which can be said to be not isolated but systematic." The fear of a single act of harm done for a Convention reason will satisfy the Convention definition of persecution if it is so oppressive that the individual cannot be expected to tolerate it so that refusal to return to the country of the applicant's nationality is the understandable choice of that person.
Given the misunderstanding that has arisen from using the term "systematic conduct", it is probably better to refrain from using it in a Convention context. But if it is to be used, those who use it should make it clear that they are referring to "non‑random" acts; otherwise, they run the risk of making a legal error."
53 There can be no doubt that a single act or rape or sexual abuse resulting in impregnation of a female draftee by a military officer (resulting in being shunned thereafter by family) when many such examples have occurred before with female draftees, is "systematic" in that it is non‑random and so oppressive that the applicant could not be expected to tolerate it. "Systematic" has quite separate shades of meaning. It can mean habitual or regular; equally it can mean deliberate or pre‑meditated. A Full Court of this court has recognised this: Minister for Immigration & Multicultural Affairs v Hamad (1999) 87 FCR 294 at 297 [17]. The epithet "non‑random" is apposite to cover the different shades of meaning. Accordingly, to determine conduct is not "sufficient" for the purposes of the Chan test because the "incidence" is not sufficiently widespread can result in error where the seriousness of the harm is not in dispute. Where harassment can be described as minimal or low level, it can be appropriate to ask whether the incidence of such harassment is sufficient to constitute serious harm or "significant detriment" as expressed by Mason CJ in Chan at 389. Questions of whether the extent of harassment can be characterised as persecution in the Convention sense are questions of fact: see for example NABB at [16]. However, cases turning on factual matters such as the qualitative assessment of harassment to determine whether or not certain harassment amounts to serious harm are distinguishable from cases dealing with whether harm, the seriousness of which is not challenged, constitutes "systematic conduct" for the purposes of the meaning of persecution. Where the harassment shown on the facts to be serious harm falls to be assessed as to whether it is systematic conduct, legal error can occur if the applicant is required to show anything more than that the seriously harmful conduct feared is deliberate or pre‑meditated, that is motivated. It is not necessary for an applicant to show that the seriously harmful conduct has occurred on a scale which might answer to the description of an atrocity. This is particularly so when a single instance of the feared harm will be oppressive to the applicant.
54 Having been satisfied that rape, sexual abuse and impregnation by military officers was committed against female draftees, including at Sawa, the camp to which the applicant would be assigned, and having accepted country information which cited incidents of parents being killed whilst resisting the drafting of their daughters, the Tribunal should have asked whether the conduct in question was deliberate or pre‑meditated, then applied the Chan test, which it had correctly described, to the applicant. Such an approach may have led to a different result. Instead, it asked a question relevant to a qualitative assessment of whether the harm was serious, namely it asked whether the incidence of rape, sexual abuse and impregnation by military officers occurred on a significant scale to constitute persecution.
55 In a decision otherwise free of jurisdictional error, the Tribunal misdirected itself by not asking whether rape, sexual abuse and impregnation by military officers (of which facts it was satisfied) was deliberate or pre‑meditated conduct, exposure to which the applicant could not be expected to tolerate. The Tribunal was attempting to apply what has been a most problematic aspect of the relevant tests. It asked itself the wrong question which affected its exercise of power. It therefore made an error of law of the kind which was referred to in the joint judgment of McHugh, Gummow and Hayne JJ (with whom Gleeson CJ expressed agreement) in Minister for Immigration & Multicultural & Indigenous Affairs v Yusuf [2001] HCA 30; (2001) 206 CLR 323 at [82].