(back to preceding text)

    The need for asylum and the obligation upon another state to provide it should then arise where the home state has failed in its duty of protection. The most obvious case of such failure is where the persecution in any of its various forms is the direct work of the state or the agents of the state. In para. 5.1 of the Joint Position 96/196/JHA defined by the Council of the European Union it is stated that "Persecution is generally the act of a state organ (central state or federal states, regional or local authorities) whatever its status in international law, or of parties or organisations controlling the state." Such official invasions of rights, whether or not involving physical violence, may well provide the typical examples of what may constitute persecution for the purposes of the Convention. But as the Joint Position itself recognises, persecution for those purposes may also occur where the immediate act of persecution is not that of the state or its agents. In para. 5.2 it is stated:

"Persecution by third parties will be considered to fall within the scope of the Geneva Convention where it is based on one of the grounds in article 1A of that Convention, is individual in nature and is encouraged or permitted by the authorities. Where the official authorities fail to act, such persecution should give rise to individual examination of each application for refugee status, in accordance with national judicial practice, in the light in particular of whether or not the failure to act was deliberate . . . ."

Professor Guy S. Goodwin-Gill (The Refugee in International Law, 2nd ed. (1996) p. 73) observes "Where the state is either unable or unwilling to satisfy the standard of due diligence in the provision of protection, the circumstances may equally found an international claim." The important consideration here to my mind is that the persecution is encouraged or permitted by the authorities or they are unable or unwilling to provide it. Even in cases where the state may not immediately initiate or direct the acts complained of, its encouragement, permission, toleration or helpless acceptance of the acts, may constitute a case of persecution. Thus the acts may be seen as constructively acts by the state and so be within the kind of acts which the Convention is concerned to cover. It is suggested that para. 5.2 is not intended to be definitive. But in so far as it seeks to express the necessary element of state participation, whether direct or indirect, active or passive, it seems to me to be expressing one basic ingredient in the concept of persecution for the purposes of the Convention. In the present case the activities immediately complained of are the activities not of any agents of the state but of third parties. The skinheads are a body independent of the state authorities. But if their oppressive behaviour was encouraged or permitted by the state authorities, or the state was unable or unwilling to provide protection, a case of persecution could be made out.

    A question arises, and it has been canvassed in some detail in the oral and written submissions before us, as to the level of protection which is to be expected of the home state. This was identified by the appellant as the third of three issues which he set out in his case. Priority was however given to it in the useful written submission which was provided on behalf of the Refugee Legal Centre, who regarded it as the principal issue in the appeal. I do not believe that any complete or comprehensive exposition can be devised which would precisely and comprehensively define the relevant level of protection. The use of words like "sufficiency" or "effectiveness", both of which may be seen as relative, does not provide a precise solution. Certainly no one would be entitled to an absolutely guaranteed immunity. That would be beyond any realistic practical expectation. Moreover it is relevant to note that in Osman v. United Kingdom [1998] 29 E.H.R.R. 245 the European Court of Human Rights recognised that account should be taken of the operational responsibilities and the constraints on the provision of police protection and accordingly the obligation to protect must not be so interpreted as to impose an impossible or disproportionate burden upon the authorities. At the least, as is noted in condition (iii) in rule 334 which I have quoted earlier, the person must be able to show that if he is not granted asylum he would be required to go to a country where his life and freedom would be threatened. There must be in place a system of domestic protection and machinery for the detection, prosecution and punishment of actings contrary to the purposes which the Convention requires to have protected. More importantly there must be an ability and a readiness to operate that machinery. But precisely where the line is drawn beyond that generality is necessarily a matter of the circumstances of each particular case.

    It seems to me that the formulation presented by Stuart-Smith L.J. in the Court of Appeal may well serve as a useful description of what is intended, where he said [2000] I.N.L.R. 15, 26, para. 22):

"In my judgment there must be in force in the country in question a criminal law which makes the violent attacks by the persecutors punishable by sentences commensurate with the gravity of the crimes. The victims as a class must not be exempt from the protection of the law. There must be a reasonable willingness by the law enforcement agencies, that is to say the police and courts, to detect, prosecute and punish offenders."

And in relation to the matter of unwillingness he pointed out that inefficiency and incompetence is not the same as unwillingness, that there may be various sound reasons why criminals may not be brought to justice, and that the corruption, sympathy or weakness of some individuals in the system of justice does not mean that the state is unwilling to afford protection. "It will require cogent evidence that the state which is able to afford protection is unwilling to do so, especially in the case of a democracy." The formulation does not claim to be exhaustive or comprehensive, but it seems to me to give helpful guidance.

    In the present case the tribunal formed the view that there was a sufficiency of state protection available. On one view of the case that finding in fact may be seen as conclusive. If the definition is reduced to its bare bones in the terms of requiring that the person be exposed to a reasonable risk of a violation of his Convention rights with an inadequacy of state protection the appellant must fail. But the analysis of the definition which has been explored in the present case has been more sophisticated. What has given rise to debate, and to the difference of opinion in the Court of Appeal, is the precise stage or stages at which in the course of the definition the element of state protection ought to be allowed in.

    The debate upon this question opens up a real risk of embarking upon the kind of precise analysis of the definition which seems to me, at least if taken to extremes, runs counter to the proper approach to be adopted to the construction of the Convention. But it is certainly correct to notice that the definition comprises two parts, both of which require to be satisfied. The first part requires that the person be outside the country of his nationality for a particular reason, namely a well-founded fear of being persecuted for what may be conveniently referred to as a Convention reason. This part is concerned with the reason why he is outside the country of his nationality. The second part is concerned with the possibility of the person availing himself of the protection of that country. It requires, as an additional requirement, that the person be unable, or owing to the well-founded fear already mentioned unwilling, to avail himself of that protection. The second part is thus concerned with the possibility of the person returning to his own state. Both parts are expressed in the present tense. Thus, as was held in Adan v Secretary of State for Home Department [1999] 1 A.C. 293, the well-founded fear must be a current fear, not an historic one. In that case use was made of the labels of a "fear test" and a "protection test" as a means of reference to the two respective parts into which the definition falls. While I would recognise the practical convenience of the use of labels as a form of reference, I feel that even that language may possibly colour the approach to the matter of construction and I would prefer simply to use the expressions "the first part" and "the second part" of the definition as a means of reference.

    So far as the first part is concerned it was laid down in Reg. v Secretary of State for the Home Department, Ex parte Sivakumaran [1988] A.C. 958 that the existence of a well-founded fear required the establishment of what was described by Lord Keith of Kinkel as "a reasonable degree of likelihood" (p. 994), by Lord Templeman as a "real and substantial danger"(p. 996) and by Lord Goff of Chieveley as "a real and substantial risk" (p. 1000). These are equivalent expressions and the test, while having a subjective element, is in that respect objective. In the present case the tribunal found as matter of fact that the appellant did have a well-founded fear. The remaining question concerns the object of his fear. The tribunal found that his fear was not that of persecution. The discussion in Sivakumaran bears on the construction of the reference to a well-founded fear. It does not resolve the question as to the constituent elements of "persecution."

    I have much sympathy with the view expressed by Simon Brown L.J. in Ravichandran v. Secretary of State for the Home Department [1996] Imm. A.R.. 97, 109 that "the question whether someone is at risk of persecution for a Convention reason should be looked at in the round and all the relevant circumstances brought into account." But in identifying the precise point of dispute in the present case it is necessary to isolate what is and what is not relevant. We are not immediately concerned with problems about the standard of proof. The consideration which was given to that issue in the Court of Appeal in the present case, and more recently in Karanakaran v. Secretary of State for the Home Department (25 January 2000), is not of immediate relevance to the present case and I shall say nothing about it. Nor are we concerned in this case with the nature or the quality of the activities which may or may not be included within the concept of persecution, or the various ways and means by which persecution may be inflicted. It appears that the word carries with it some element of persistence or continuity, to use Professor Hathaway's language (The Law of Refugee Status, p. 101) it is "sustained or systemic." But the term is left undefined so as to include a wide variety of types of behaviour. In relation to such questions the ordinary use of the word should provide sufficient guidance and its application will be a matter of the facts and circumstances of each particular case. There may be little purpose to be served by looking to such expressions as harassment or oppression which may be approximately synonymous. There is no doubt in the present case that the activities upon which the appellant founds are of such a nature and quality as would enable them to fall within the scope of the term. Further there is no doubt that the behaviour in question is prompted by considerations of race so that if there is persecution it is for a Convention reason. The ill-treatment of which the appellant is afraid is based on his being a Roma and the violation of his rights which he fears is on the ground of his race.

    The particular question which is raised in the present appeal is whether the word "persecution" requires to take account of the attitude of the state to the violence which the appellant fears. The appellant contends that persecution comprises simply the acts of violence in question and no account should be taken of what the state is doing or can do about them. A failure by the state to protect is then irrelevant. On this approach it may become easier for him to satisfy the first part of the definition. If considerations of the availability of state protection are excluded from the definition of "protection," it is sufficient for him to show that despite such efforts at protection as there may be there is a reasonable risk of his suffering an abuse of his rights. In other words it is enough for him to show that there is a reasonable risk both that he will suffer abuse and that he will not be protected from such abuse. On this approach it would appear that the level of state protection has to be such as to exclude any real risk of an abuse of Convention rights occurring. Indeed in the formulation put forward by the Refugee Legal Centre it was suggested that the protection should be such as to so reduce the risk to the applicant that his fear of persecution could not be said to be well-founded.

    The issue which the appellant put at the forefront of his case, and which was referred to as the first issue, accordingly was whether the matter of the protection afforded by the state is or is not a relevant ingredient in the concept of persecution. On this question the tribunal considered that it was relevant. Stuart-Smith L.J. took the view that persecution meant ill treatment pure and simple without any account taken of the state protection. The majority of the Court of Appeal took the opposite view. I consider that the majority were correct.

    It may seem at first sight attractive to analyse the definition into the two parts and see protection as belonging to the second part and not to the first part. Here particularly the use of the label of the "protection test" as applicable to the second part may prove dangerous. While the appellant argued that the element of protection had no part to play in the definition of persecution, he did accept that protection was relevant to the concept of a well-founded fear. But that admits that the element of protection may stray across from the second part of the definition to the first. Stuart-Smith L.J. took a more consistent approach in completely denying the admissibility of considerations of protection to the first part and admitting them only into the second. That appears to accord with the observation made by my noble and learned friend Lord Lloyd of Berwick in Adan (at p. 306A) where he envisaged that persons could satisfy the fear test, without regard to matters of protection, but not the protection test. If the second part truly was a comprehensive test about protection then the suggested label would be appropriate and the logical scheme which Stuart-Smith L.J. preferred would have the more to commend it. But in deciding whether a fear of persecution is well-founded it seems to me that account must be taken of the availability of the forces of the state to counter the fear. And if that is correct it is no longer possible to confine the consideration of protection to the so-called "protection test." For a fear to be well-founded it seems to me that all the circumstances relating to the fear have to be taken into account. In assessing the existence of a real risk of the violation of rights occurring anything which may bear on the likelihood of the incidence of the violation will be relevant. It is the applicant's fear which is in issue, and so matters particularly relating to him will be important. For example his prominence in society or political life, or anything else which might make him a particular target of persecution may be relevant. The history of past violations, the extent to which the applicant has personally been directly affected, either by being the victim of violence or the recipient of threats of violence, considerations of geographical location, of all the factors which might stimulate or facilitate a violation, will be among the circumstances to be taken into account. As also will factors which may discourage or deter or render a violation less likely. The political and legal situation in the country should be taken into account. And among those will be the element of the protection which the state affords. While state protection may not be the only factor bearing on the real risk of a violation occurring I do not see that it can be omitted from the considerations relevant to that issue.

    But the critical question is whether considerations of the involvement of the state enters the definition of "persecution." Here, as it seems to me, one comes very close to the kind of detailed dissection of the definition which I have criticised before. But the argument which was presented before us makes it necessary to pursue this fine analysis. Of course in the ordinary use of words and out of the context of the Convention persecution may well comprise simply acts of ill treatment. But it is in the context of the Convention that the matter has to be approached. As I have already observed the context in which the definition occurs, although not expressly so stated in the terms of the Convention, is that of the protection which the individual may expect from his or her home state. In that context it seems to me inevitable that the persecution to which the Convention refers is a persecution which takes account of the protection available. Of course where the state is itself through its agents the persecutor, the question does not require to arise. Active persecution by the state is the very reverse of protection. In that context it is sufficient to proceed simply upon dictionary definitions to stress the high standard of oppression which has to be found, as in Reg. v. Immigration Appeal Tribunal, Ex parte Jonah [1985] Imm. A.R. 7. So also in Demirkaya v. Secretary of State for the Home Department [1999] I.N.L.R. 441 where the complaint was of persecution by agents of the state attention could be concentrated upon the issue of the gravity of the oppression. It is in the context of persecution by third parties that the problem of protection becomes more significant.

    It is no part of the international scheme that people should qualify as refugees merely because private persons in their home state seek to interfere with their rights and freedoms. If there is a sufficiency of protection available to them in that state, then there should be no obligation on another state to afford a surrogate protection. The persecution with which the Convention is concerned is a persecution which is not countered by a sufficient protection. The responsibility to protect the citizen which is abrogated in a case of active state persecution is still relevant in assessing what may be seen as a constructive state persecution, where the ill-treatment by other citizens is encouraged or tolerated by the state without direct participation on its own part. Here the concept of encouragement or toleration on the one hand may be seen as the expressing the same thing as the failure by the state to provide adequate protection. A toleration which amounts to a constructive persecution by the state and the failure by the state to provide adequate protection may be the two sides of the same coin. It may be permissible to use the language of a failure in protection against the abuse as equivalent to an encouragement or toleration of the abuse or to an acquiescence in it.

    This view of what is intended by persecution in the context of the Convention seems to me to be consistent with the purpose of the Convention. In addition considerable support can be found for it. In the Handbook produced by the Office of the United Nations High Commissioner for Refugees, January 1992, which has the weight of accumulated practice behind it, it is stated, in para. 65:

"Where serious discriminatory or other offensive acts are committed by the local populace, they can be considered as persecution if they are knowingly tolerated by the authorities, or if the authorities refuse, or prove unable, to offer effective protection."

While the Joint Position defined by the Council of the European Union on 4 March 1996 (OJ 1996 L63/2) may only be a statement for purposes of guidance, and indeed is not so worded as to escape debate upon its meaning, the passage in para. 5.2 which I have already quoted seems to me to support the view that the attitude of the state, whether it be one of action or inaction, is relevant to the concept of persecution.

    As regards the academic writers, Professor Hathaway (op.cit.p. 105) observes "A well-founded fear of persecution exists when one reasonably anticipates that remaining in the country may result in a form of serious harm which government cannot or will not prevent . . . ." At p. 127 he states:

"Beyond these acts of commission carried out by entities with which the state is formally or implicitly linked, persecution may also consist of either the failure or inability of a government effectively to protect the basic human rights of its populace."

Geoffrey S. Gilbert, Right of Asylum: A Change of Direction, International and Comparative Law Quarterly, Vol. 32, 633, 645, under reference to the Canadian case In the Matter of McMullen 658 F 2nd 1312 (9th Circuit, 1981) states "persecution by a third party where the government offers no protection because of clandestine support or inability to control is just as valid." Professor Goodwin-Gill (The Refugee in International Law, 2nd ed. (1996), p. 67) states that:

"evidence of the lack of protection on either the internal or external level may create a presumption as to the likelihood of persecution and to the well-foundedness of any fear."

He also observes (p. 75) in relation to the case of flight from civil war that "it nevertheless remains for the applicant to show that he or she is unable to obtain the protection of the state, and to establish the requisite Convention link."

    As regards case-law, in Sandralingham v. Secretary of State for the Home Department [1996] Imm. A.R. 97 Staughton L.J. at p. 114 stated that persecution is "persistent and serious ill-treatment without just cause by the state, or from which the state can provide protection but chooses not to do so." The appellant pointed out that that observation was made in the context of a case concerned with persecution by the state; but the observation is expressed in quite general terms. Moreover that point of distinction cannot be made in respect of Reg. v. Immigration Appeal Tribunal, Ex parte Shah [1999] 2 A.C. 629, where, at p. 653, Lord Hoffmann noted the two elements which in that case comprised the persecution which the appellants feared - the personal threats of violence to them by, principally, their husbands, and the inability or unwillingness of the state to protect them. These two elements had to be combined to constitute persecution within the meaning of the Convention. He quoted and adopted the concise formula "Persecution = Serious harm + The Failure of State Protection." The appellant sought to find some support for his thesis in Canada (Attorney-General) v. Ward, 103 D.L.R. (4th) 1. It was there held that state complicity is not a necessary component of persecution; the inability of the state to protect may by itself be enough. But in that case the court was considering a formulation of the definition of refugee which does not exactly correspond with article 1A(2) and I do not consider that guidance on this point can safely be taken from it.

    The appellant argued that the view of the majority of the Court of Appeal produced an anomalous result. Counsel referred to a recent tribunal decision in Kovac v. Secretary of State for the Home Department [15 February 2000] where the tribunal observed that but for that view they would have treated the issue of the likelihood of protection simply as an aspect of assessing the real risk of persecution. "Otherwise it seems to us that one will be returning a refugee to a country in which ex hypothesis there is a serious risk of persecution." This apparent difficulty leads one back to a consideration of the level of protection required for the purposes of the Convention. If the matter of protection is treated simply as an aspect of assessing the existence of a real risk of an abuse of rights, asylum would be granted even although there was, in the way which I have already sought to describe, a reasonable level of state protection. But that would be contrary to the basic intention of the Convention. The sufficiency of state protection is not measured by the existence of a real risk of an abuse of rights but by the availability of a system for the protection of the citizen and a reasonable willingness by the state to operate it.

    It seems to me that on the contrary the appellant's approach gives rise to anomaly. If consideration of the state's attitude is excluded from the definition of persecution and considerations of protection in the first part are confined to the well-foundedness of the fear, then it would seem that some cases which ought to justify asylum would be excluded. The persecution must be for a Convention reason. But it is not difficult to conceive of cases where a person might be persecuted by other citizens for reasons of private gain which involve no element of Convention rights. If the state was motivated by considerations which were contrary to the Convention rights to tolerate such activity and deliberately refrain from protecting the person, such a case would appear not to be covered by the approach promoted by the appellant. That does not seem to be sound.

    I turn finally to what was referred to as the second issue in the appeal. This relates to the second part of the definition and concerns a point of construction which was propounded by Stuart-Smith L.J. in the present case. The second part of the definition comprises two branches, both of which must be given some content. The first branch is readily understood. If the state gives him no protection then the person will be unable to avail himself of the protection of the state. The obvious case is where the state is itself the persecutor. Other examples may be where the state is at war, or undergoing some internal disturbance which prevents it rendering any effective protection. The second branch requires that the applicant be unwilling owing to the well-founded fear referred to in the first part to avail himself of the protection of the state. That envisages that there is some protection available to him. But a careful distinction is made between cases of inability and unwillingness; in the latter, but not in the former, the unwillingness must be owing to the well-founded fear. This provision thus serves to exclude from the definition anyone who, while he has a well-founded fear of persecution, is unwilling to avail himself of such state protection as there may be for a reason other than that fear, such as for example a simple preference to enjoy the economic advantages of the host state. The focus of the second part is upon the possibility of the person availing himself of protection from his home state. It complements the first part and it seems to me that there is sufficient content in it without further elaboration.

    However, the respondent has suggested, following the view of Stuart-Smith L.J. that content should be found for the second branch of the second part in the requirement that the unwillingness must be due to a fear that if the person avails himself of the protection he will be persecuted for so availing himself. But that unwillingness will not be an unwillingness which is owing to the well-founded fear mentioned in the first part. It would be an unwillingness owing to a further fear, namely of persecution for seeking state protection. The reference to "such fear" is a reference to the whole fear described in the first part. That is not just a well-founded fear, but a well-founded fear of persecution on a Convention ground. The purpose of the reference is to require that the fear be of such persecution, as well of course as having the quality of being well-founded.

    I would dismiss the appeal.