Judgment - Islam (A.P.) v. Secretary of State for the Home Department Regina v. Immigration Appeal Tribunal and Another Ex Parte Shah (A.P.) (Conjoined Appeals)  continued

(back to preceding text)

The group cannot, therefore, consist of the victims of persecution who share no other common characteristic. Nor in my opinion can it consist of victims of persecution who do share a common characteristic if that is not the reason for their persecution. Non-causative characteristics are irrelevant. Battered wives do not form a social group because, if the group is limited to battered wives, it is defined by the persecution, while if it is extended to include all married women, those who are battered are not persecuted because they are members of the group. The Canadian Court of Appeal decided the contrary in Canada (Minister of Employment and Immigration) v. Mayers (1992) 97 D.L.R. (4th) 729, but I agree with McHugh J. in A. v. Minister for Immigration and Ethnic Affairs (at p. 358 note 120) that the decision cannot be supported.

It is in my opinion essential to bear in mind at all times that it is not enough for the applicant for asylum to establish that he or she is a member of a particular social group and is liable to persecution. The applicant must also establish that he or she is liable to persecution because he or she is a member of the group. The applicant must be the subject of attack, not for himself or herself alone, but because he or she is one of those jointly condemned in the eyes of their persecutors for possession of the characteristic which is common to the group.

With these preliminary observations I turn to consider the social group to which the present appellants claim to belong and for membership of which they fear persecution. As formulated by the appellants it consists of women in Pakistan who have been or who are liable to be accused of adultery or other conduct transgressing social norms and who are unprotected by their husbands or other male relatives. It is a subset of the set "women" (and of the subset "married women"). The third qualifying condition can be disposed of at once. The fact that the appellants have no one to protect them helps to show that their fear of persecution is well founded. But it does not help to define the social group to which they belong. I am content to assume that the appellants would not be persecuted if they had someone to protect them from attack. But they are not persecuted because they have no one to protect them; that is not the ground of persecution. The "but for" test of causation, which is always necessary but rarely sufficient, is beguilingly misleading in this context.

This qualifying condition was no doubt included because of an erroneous belief that all the members of the group must be equally liable to persecution. That is not the case. It is no answer to a claim for asylum that some members of the group may be able to escape persecution, either because they have powerful protectors or for geographical or other reasons. Such factors do not narrow the membership of the group, but go to the question whether the applicant's fear of persecution is well founded. Thus I would accept that homosexuals form a distinct social group. In a society which subjected practising homosexuals but not non-practising homosexuals to persecution the relevant social group would still consist of homosexuals, not of the subset practising homosexuals. A non-practising homosexual would have no difficulty in establishing that he was a member of a persecuted group. His only difficulty would be in establishing that his fear of persecution was well founded, having regard to the fact that he was not a practising homosexual. This would be a matter of evidence, but given the hostility encountered by all homosexuals in such a society and the obvious problems the applicant would have in satisfying his tormentors of his own sexual abstinence, I doubt that the difficulty would be a real one.

Whether the social group is taken to be that contended for by the appellants, however, or the wider one of Pakistani women who are perceived to have transgressed social norms, the result is the same. No cognisable social group exists independently of the social conditions on which the persecution is founded. The social group which the appellants identify is defined by the persecution, or more accurately (but just as fatally) by the discrimination which founds the persecution. It is an artificial construct called into being to meet the exigencies of the case.

The appellants contended for the subset because they recognised that the head set of "women" or "married women" would not do. Officially, I understand, Sharia law regards women as "separate but equal," a description which, I observe, was also applied, with scant regard for the truth, to apartheid in South Africa. The evidence clearly establishes that women in Pakistan are treated as inferior to men and subordinate to their husbands and that, by international standards, they are subject to serious and quite unacceptable discrimination on account of their sex. But persecution is not merely an aggravated form of discrimination; and even if women (or married women) constitute a particular social group it is not accurate to say that those women in Pakistan who are persecuted are persecuted because they are members of it. They are persecuted because they are thought to have transgressed social norms, not because they are women. There is no evidence that men who transgress the different social norms which apply to them are treated more favourably.

In the course of argument an illuminating instance was put forward. Suppose, in the early years of the Third Reich, Jews in Nazi Germany were required to wear a yellow star on pain of being sent to a concentration camp and murdered if they did not. Would they have failed to qualify as refugees on the ground that they were not liable to persecution on racial grounds, but because they were defying the law? Of course we know now that they should not have failed to qualify, because the law was not merely discriminatory but a necessary part of the intended persecution. Jews were required to wear a distinguishing badge in order to mark them out for persecution. At the time this would have been a matter of evidence; but given the absence of any other rational explanation for the law, the virulence of the state-inspired racial propaganda which formed the background against which it was enacted, and the wholly disproportionate penalty for disobedience, there should have been no difficulty in satisfying the requirements of the Convention even in the absence of other evidence of persecution (of which there was an abundance). I find the example instructive precisely because of the differences between it and the present case rather than the similarities.

I am accordingly not willing to accept, as a general proposition, the submission that those who are persecuted because they refuse to conform to discriminatory laws to which, as members a particular social group, they are subject, thereby qualify for refugee status. Such persons are discriminated against because they are members of the social group in question; but they are persecuted because they refuse to conform, not because they are members of the social group. Nor am I able to accept the submission that somehow this ceases to be the case where the persecution is sanctioned or tolerated by the state. As I have said, this goes to a different question, whether the applicant's fear of persecution is well founded. It is still necessary to establish that the persecution, whether or not sanctioned by the state, is for a Convention reason.

Of course, the fact that the persecution is sanctioned by the state may make it easier to categorise it as persecution on political grounds. In extreme cases, where persecution is employed as an instrument of state policy and is actively encouraged by the authorities, as in Nazi Germany, the distinction between discrimination and persecution may be a distinction without a difference. Persecution takes many different forms. Where the authorities perceive a particular social group to be hostile, they may persecute its members by openly withdrawing their protection and leaving them to the mercy of criminal elements. The fact that those who take advantage of the situation to use violence against members of the group do so for their own private purposes does not matter; the members should be regarded as the victims of official persecution by the state. To qualify for refugee status, however, they must still prove that the state authorities have withdrawn their protection for a Convention reason.

Such questions will depend on the evidence. The evidence in the present case is that the widespread discrimination against women in Pakistan is based on religious law, and the persecution of those who refuse to conform to social and religious norms, while in no sense required by religious law, is sanctioned or at least tolerated by the authorities. But these norms are not a pretext for persecution nor have they been recently imposed. They are deeply embedded in the society in which the appellants have been brought up and in which they live. Women who are perceived to have transgressed them are treated badly, particularly by their husbands, and the authorities do little to protect them. But this is not because they are women. They are persecuted as individuals for what each of them has done or is thought to have done. They are not jointly condemned as females or persecuted for what they are. The appellants need to establish that the reason that they are left unprotected by the authorities and are liable to be persecuted by their husbands is that they are women. In my opinion they have not done so.

I would dismiss the appeals.


Lords Parliament Commons Search Contact Us Index

© Parliamentary copyright 1999
Prepared 25 March 1999