CASES AND COMMENTS Horvath v. Secretary of State for the Home Department House of Lords 6 July 2000 Lord Hope of Craighead: My Lords, The appellant is a citizen of the republic of Slovakia. He comes from a village called Palin in the county of Michalovce, where he lived with his wife and child and other members of his family. He and his family are Roma, who are known colloquially as gypsies. The Roma, who are widely distributed across the country, constitute about 10 per cent of the population of Slovakia. They are a small minority in the village to which the appellant belongs. On 15 October 1997 he arrived in the United Kingdom with his wife and child and claimed asylum. He said that he feared persecution in Slovakia by skinheads, against whom the Slovak police were failing to provide protection for Roma. He also said that, along with other Roma, he had been unable to find work, that he had not been afforded the normal public facilities as to his marriage and schooling for his child and that in these respects he was being discriminated against. He maintained that he was afraid that if he and his family were returned to Slovakia they would again be attacked by skinheads as they were Roma, and that they would not get protection from the police. His application for asylum was refused by the Secretary of State. The Special Adjudicator did not find him to be a credible witness and dismissed his appeal. The Immigration Appeal Tribunal found that his assertions of fact were consistent with other evidence which was before them about the position of Roma in Slovakia, so they reversed the Special Adjudicator's finding on credibility. But they concluded that, while he had a well- founded fear of violence by skinheads, this did not amount to persecution because he had not shown that he was unable or, through fear of [2001] 1 AC 489; [2000] 3 WLR 379. Lord Hope of Craighead, Lord Browne-Wilkinson, Lord Lloyd of Berwick, Lord Clyde, Lord Hobhouse of Woodborough. International Journal of Refugee Law Vol. 13 No. 1/2 Oxford University Press 2001. All rights reserved Cases and Comments 175 persecution, unwilling to avail himself of the protection of the state. The Court of Appeal (Stuart-Smith, Ward and Hale LL.J.) dismissed his appeal against the determination of the tribunal: [2000] I.N.L.R. 15. The parties are agreed that the issues in this appeal all relate to the proper construction of article 1A(2) of the Geneva Convention relating to the Status of Refugees 1951. The problem to which these issues are directed arises from the fact that the appellant's claim to refugee status is based upon the alleged insufficiency of state protection against persecution by non-state agents. It is not part of his case that he has a well-founded fear of persecution by the state itself or by organs or agents of the state. His claim is based on his fear of violence by skinheads, who are not agents of the state, and on the alleged failure of the state through its police service to provide him with protection against their activities. He also based his claim on discrimination in the field of employment, the right to marry and education, but the tribunal concluded that any abuse of his rights in respect of these matters did not amount to persecution. The Court of Appeal held that the tribunal were fully entitled to reach that conclusion, and there has been no appeal against that part of its decision to this House. Your Lordships are concerned only with the allegation of failure by the state to protect the appellant against the activities of non-state agents. Article 1(A)(2) of the Convention (Cmd. 9171), as amended by the New York Protocol of 31 January 1967 (Cmnd. 3906), provides that the term `refugee' shall apply to any person who: owing to a well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country; or who, not having a nationality and being outside the country of his former habitual residence as a result of such events, is unable or, owing to such fear, is unwilling to return to it. The following issues arise in the determination of the question raised by the problem that the parties have identified in regard to the allegation of persecution by non-state agents: (1) does the word `persecution' denote merely sufficiently severe ill-treatment, or does it denote sufficiently severe ill-treatment against which the state fails to afford protection? (2) is a person `unwilling to avail himself of the protection' of the country of his nationality where he is unwilling to do so because of his fear of persecution by non-state agents despite the state's protection against those agents' activities, or must his fear be a fear of being persecuted there for availing himself of the state's protection? (3) what is the test for determining whether there is sufficient protection against persecution in the person's country of origin -- is it sufficient, to meet the standard required by the Cases and Comments176 Convention, that there is in that country a system of criminal law which makes violent attacks by the persecutors punishable and a reasonable willingness to enforce that law on the part of the law enforcement agencies? Or must the protection by the state be such that it cannot be said that the person has a well-founded fear? These three issues raise questions about the structure of article 1A(2) and about the meaning of words and phrases used in various parts of that article. The point is commonly made in regard to the Convention that it is not right to construe its language with the same precision as one would if it had been an Act of Parliament. The Convention is an international instrument. So, as my noble and learned friend Lord Lloyd of Berwick has observed, its choice of wording must be taken to have been the product of the inevitable process of negotiation and compromise: Adan v. Secretary of State for the Home Department [1999] 1 A.C. 293, 305B- C. And the general rule is that international treaties should, so far as possible, be construed uniformly by the national courts of all states. This point also suggests that the best guide to the meaning of the words used in the Convention is likely to be found by giving them a broad meaning in the light of the purposes which the Convention was designed to serve. It will be necessary to examine the wording of the article. But it may be helpful as a starting point to identify the relevant purpose or purposes. It seems to me that the Convention purpose which is of paramount importance for a solution of the problems raised by the present case is that which is to be found in the principle of surrogacy. The general purpose of the Convention is to enable the person who no longer has the benefit of protection against persecution for a Convention reason in his own country to turn for protection to the international community. As Lord Keith of Kinkel observed in Reg. v. Secretary of State for the Home Department, Ex parte Sivakumaran [1988] A.C.958, 992H-993A, its general purpose is to afford protection and fair treatment to those for whom neither is available in their own country. In Canada (Attorney-General) v. Ward (1993) 103 D.L.R. (4th) 1, 12 La Forest J. said: At the outset, it is useful to explore the rationale underlying the international refugee protection regime, for this permeates the interpretation of the various terms requiring examination. International refugee law was formulated to serve as a back-up to the protection one expects from the State of which an individual is a national. It was meant to come into play only in situations when that protection is unavailable, and then only in certain situations. This purpose has a direct bearing on the meaning that is to be given to the word `persecution' for the purposes of the Convention. As Professor James C. Hathaway, The Law of Refugee Status (Butterworths, 1991) p. 112 has explained, `persecution is most appropriately defined as the sustained or systemic failure of state protection in relation to one of the core Cases and Comments 177 entitlements which has been recognised by the international community.' At p. 135 he refers to the protection which the Convention provides as `surrogate or substitute protection', which is activated only upon the failure of protection by the home state. On this view the failure of state protection is central to the whole system. It also has a direct bearing on the test that is to be applied in order to answer the question whether the protection against persecution which is available in the country of his nationality is sufficiently lacking to enable the person to obtain protection internationally as a refugee. If the principle of surrogacy is applied, the criterion must be whether the alleged lack of protection is such as to indicate that the home state is unable or unwilling to discharge its duty to establish and operate a system for the protection against persecution of its own nationals. Although the matter does not arise for further discussion in this case, it may be worth noting that the literature suggests that states differ in their approach to the problem posed by persecution by non-state agents. As Laws L.J. explained in Reg. v. Secretary of State for the Home Department, Ex parte Adan [1999] 3 W.L.R. 1274, 1288­1289, France and Germany subscribe to the `accountability' theory, which limits the class of case in which a claimant may obtain refugee status to situations where the persecution alleged can be attributed to the state so that the status of refugee is not available, on the German view, where there is no effective state authority or, on the French view, the state authority is unable to provide protection. On the other hand a majority of the contracting states, including the United Kingdom, the United States and Canada, subscribe to the `protection' theory. After referring to Lord Lloyd's explanation of the substance of this theory in Adan v. Secretary of State for the Home Department [1999] 1 A.C. 293, 306B, where he said that the qualifications for refugee status are complete when for whatever reason the state in question is unable to afford protection against factions within the state, Laws L.J. then added this comment at p. 1289D-E: This accords with other jurisprudence in the English jurisdiction. Our courts recognise persecution by non-state agents for the purposes of the Convention in any case where the state is unwilling or unable to provide protection against it, and indeed whether or not there exist competent or effective governmental or state authorities in the country in question. This is what has been called the `protection' theory. It is, as we have said, shared by a majority of the states signatory to the Convention and the UNHCR. Fortunately the situation in Slovakia is not such as to give rise to the problems which may arise in other jurisdictions where there is no effective state authority or the state authority is unable to provide protection. The present case is relatively straightforward. The institutions of government are effective and operating in the Republic of Slovakia. The state provides Cases and Comments178 protection to its nationals by respecting the rule of law and it enforces its authority through the provision of a police force. But, as the tribunal said in paragraph 59 of its judgment, there is racial violence against the Roma perpetrated by skinheads. The police do not conduct proper investigation in all cases and there have been cases where their investigation has been very slow. But there was also evidence that the police have intervened to provide protection when they have been asked to do so and that stiff sentences are imposed at times for crimes that are racially motivated. The Tribunal's conclusion was that the violent attacks on Roma are isolated and random attacks by thugs. The first issue In the Court of Appeal there was a difference of view on the question where the alleged insufficiency of state protection against persecution by non-state agents fits in to the definition of `refugee' in article 1A(2) of the Convention. Stuart-Smith L.J. took as his starting point Lord Lloyd's division in Adan v. Secretary of State for the Home Department [1999] 1 A.C. 293, 304C-E of the persons treated by the article as refugees into four categories. The first two categories which Lord Lloyd identified relate to nationals who are outside their country owing to a well-founded fear of persecution for a Convention reason and who are either `unable' to avail themselves of the protection of their country or, owing to such fear, are `unwilling' to do so. Lord Lloyd then drew attention to the fact that in the case of each category two tests had to be satisfied. He described these tests as `the fear test' and `the protection test.' Stuart-Smith L.J. said [2000] I.N.L.R. 15, 23G, that he regarded this as a clear statement of principle that the fear test is separate and distinct from the protection test. Ward L.J., disagreeing with Stuart-Smith L.J. on this point, said, at p. 47B-C, that in his view a holistic approach had to be taken to the definition of `refugee.' The critical question was the stage at which the degree of state protection entered into the analysis. As he put it: The degree of state protection may indirectly be a factor in judging whether the fear is well-founded but this is not the only or best place for it in a proper analysis of the definition of refugee. What state protection is available is a fact to be considered in for the protection test, but the question of state protection is not confined to the `protection test'. Hale L.J. also favoured the holistic approach. At p. 52A, she said that her view was that the sufficiency or insufficiency of state protection against the acts of others might be relevant at three points in the argument: . . . if it is sufficient, the applicant's fear of persecution by others will not be `well-founded'; if it is insufficient, it may turn the acts of others into persecution for a Convention reason; in particular it may supply the discriminatory element Cases and Comments 179 in the persecution meted out by others; again if it is insufficient, it may be the reason why the applicant is unable, or if amounts to persecution unwilling, to avail himself of the protection of his home state. I agree with the view of the majority. For my part, I would regard the analysis of the article which was provided by Lord Lloyd in the Adan case as being both helpful and instructive. It is an important reminder that there are indeed two tests that require to be satisfied. A person may satisfy the fear test because he has a well-founded fear of being persecuted, but yet may not be a `refugee' within the meaning of the article because he is unable to satisfy the protection test. But it seems to me that the two tests are nevertheless linked to each other by the concepts which are to be found by looking to the purposes of the Convention. The surrogacy principle which underlies the issue of state protection is at the root of the whole matter. There is no inconsistency between the separation of the definition into two different tests and the fact that each test is founded upon the same principle. I consider that it has a part to play in the application of both tests to the evidence. I would hold therefore that, in the context of an allegation of persecution by non-state agents, the word `persecution' implies a failure by the state to make protection available against the ill-treatment or violence which the person suffers at the hands of his persecutors. In a case where the allegation is of persecution by the state or its own agents the problem does not, of course, arise. There is a clear case for surrogate protection by the international community. But in the case of an allegation of persecution by non-state agents the failure of the state to provide the protection is nevertheless an essential element. It provides the bridge between persecution by the state and persecution by non-state agents which is necessary in the interests of the consistency of the whole scheme. It is important to note throughout that the humanitarian purposes of the Convention are limited by the tests set out in the article. As Dawson J. observed in A. v. Minister for Immigration and Ethnic Affairs [1998] I.N.L.R.1, 18F-G: No matter how devastating may be epidemic, natural disaster or famine, a person fleeing them is not a refugee within the terms of the Convention. And by incorporating the five Convention reasons the Convention plainly contemplates that there will even be persons fearing persecution who will not be able to gain asylum as refugees. At p. 19B-C he went on: No doubt many of those limits in the present context spring from the well- accepted fact that international refugee law was meant to serve as a `substitute' for national protection where the latter was not provided due to discrimination against persons on grounds of their civil and political status. It would therefore be wrong to depart from the demands of language and context by invoking the Cases and Comments180 humanitarian objectives of the Convention without appreciating the limits which the Convention itself places on the achievement of them. As Hale L.J. pointed out [2000] I.N.L.R. 15, 59B, it is remarkable that the literature to which reference has been made deals with the role of the state in relation to persecution by non-state agents largely in the context of the definition of persecution rather than in the context of the inability or unwillingness of the applicant to avail himself of its protection. Professor Hathaway, The Law of Refugee Status, makes it plain in his chapter on `Persecution' that in his view the intention of the drafters of the Convention was to restrict refugee status to situations where there was a risk of a type of injury that would be inconsistent with the basic duty of protection owed by the state to its own population. In the course of his discussion of this concept he says this, at p. 104: The existence of past or anticipated suffering alone, therefore, does not make one a refugee, unless the state has failed in relation to some duty to defend its citizenry against the particular form of harm anticipated. At pp. 104­105 he suggests that persecution may be defined as `the sustained or systemic violation of basic human rights demonstrative of a failure of state protection.' Guy S. Goodwin-Gill, The Refugee in International Law, (Clarendon Press, 1996), 2nd ed., pp. 70­71 notes that the concept of persecution is not limited to the actions of governments or their agents, and that persecution can result where protection is unavailable because governments are unable to suppress the activities of the perpetrators or unwilling or reluctant to do so or are colluding with those responsible. The link between the acts of violence and failure on the part of the state authorities is also indicated by the paragraph 65 of the UNHCR Handbook on Procedures and Criteria for Determining Refugee Status (re-edited Geneva 1992), which states: Persecution is normally related to action by the authorities of a country. It may also emanate from sections of the population that do not respect the standards established by the laws of the country concerned. A case in point may be religious intolerance, amounting to persecution, in a country otherwise secular, but where sizeable fractions of the population do not respect the religious beliefs of their neighbours. 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. Mr. Plender Q.C. for the appellant accepted that a holistic approach was appropriate when the definition in article 1A(2) was being applied to the facts. As he put it, all the circumstances must be considered in order to see whether the definition is satisfied. But he maintained that it did not follow that circumstances which were relevant to one test were also Cases and Comments 181 relevant to the other. He acknowledged that the issue of state protection was relevant to the question whether the applicant's fear of severe ill- treatment for a Convention reason was a well-founded fear. But he did not accept that it was the failure by the state to provide protection that converted severe ill-treatment into persecution. Adopting what Stuart- Smith L.J. said at p. 21G, he submitted that if severe ill-treatment by non-state agents was of sufficient gravity to amount to persecution it did not lose that quality because the state could offer adequate protection against it. He referred to various cases in which consideration had been given to the meaning of the word `persecute': e.g. Chan Yee Kin v. Minister for Immigration and Ethnic Affairs (1989) 169 C.L.R. 379, 388, per Mason C.J.; Damouni v. Minister for Immigration, Local Government and Ethnic Affairs (1989) 87 A.L.R. 97, 101, per French J. He also referred to the decision in Canada (Attorney-General) v. Ward (1993) 103 D.L.R. (4th) 1, in which it was held, in case where the persecution which was alleged emanated from a non-state agent, that state complicity in the persecution was not a valid prerequisite to a valid refugee claim. But I would be cautious about drawing conclusions from that case about the approach to be taken to the definition in article 1A(2) in view of the fact that this definition has been reproduced in a different form in section 2(1) of the Canadian Immigration Act 1976 by breaking it down into sub-paragraphs. In any event the fact that, at p. 15, La Forest J. quoted with approval the definition of `persecution' which appears in the UNHCR Handbook is a good indication that the point that was of concern to him was whether the complicity of the state was a necessary element. He appears to have been content to accept the point which is relevant to this case, that acts by non-state agents when combined with state inability to protect may constitute persecution. To sum up therefore on this issue, I consider that the obligation to afford refugee status arises only if the person's own state is unable or unwilling to discharge its own duty to protect its own nationals. I think that it follows that, in order to satisfy the fear test in a non-state agent case, the applicant for refugee status must show that the persecution which he fears consist of acts of violence or ill-treatment against which the state is unable or unwilling to provide protection. The applicant may have a well-founded fear of threats to his life due to famine or civil war or of isolated acts of violence or ill-treatment for a Convention reason which may be perpetrated against him. But the risk, however severe, and the fear, however well-founded, do not entitle him to the status of a refugee. The Convention has a more limited objective, the limits of which are identified by the list of Convention reasons and by the principle of surrogacy. The Tribunal said in paragraph 53 of its judgment that in its view it was the failure of the state to provide protection that converts the Cases and Comments182 discriminatory acts into persecution. On that approach, having considered the evidence, it decided that the appellant fell below the threshold which it believed was required for international protection in a case where the fear was of discriminatory acts and where it was alleged that there was not a sufficiency of protection from non-state agents. In paragraph 60 the Tribunal stated: `It is our view that his fear is not that of persecution.' For the reasons which I have given I consider that the Tribunal approached the matter in the right way, by examining the question as to the sufficiency of state protection at the first stage when they were considering whether the appellant's fear was of `persecution' within the meaning of the Convention. In the view of the conclusion which the Tribunal reached as to this part of the definition in article 1A(2), it was unnecessary for it to consider whether the second part of the definition was satisfied. But it is obvious that, as the appellant had failed to show that he had a well- founded fear of being `persecuted' for the purposes of the first part, he would be bound to fail the requirements of the second part also. The words `such fear' in that part assume that the fear which he has is a fear of being `persecuted.' The second and third issues I do not think that it necessary for the disposal of this appeal to dwell further on the matters that were discussed in regard to these two remaining issues. As regards the second issue, I wish merely to say that on the view which I have taken about the proper approach to the first issue it loses much of its significance. But it follows from that approach that, if the second part of the definition is to be satisfied, the applicant's fear must be a well-founded fear of being persecuted for availing himself of the state's protection. As regards the third issue, the answer to it also is to be found in the principle of surrogacy. The primary duty to provide the protection lies with the home state. It is its duty to establish and to operate a system of protection against the persecution of its own nationals. If that system is lacking the protection of the international community is available as a substitute. But the application of the surrogacy principle rests upon the assumption that, just as the substitute cannot achieve complete protection against isolated and random attacks, so also complete protection against such attacks is not to be expected of the home state. The standard to be applied is therefore not that which would eliminate all risk and would thus amount to a guarantee of protection in the home state. Rather it is a practical standard, which takes proper account of the duty which the state owes to all its own nationals. As Ward L.J. said at p. 44G, under reference to Professor Hathaway's observation in his book at p.105, it is axiomatic that we live in an imperfect world. Certain levels of ill-treatment may still occur even if steps to prevent this are taken by the state to Cases and Comments 183 which we look for our protection. I consider that the Tribunal in this case applied the right standard when they were considering the evidence. Conclusion Where the allegation is of persecution by non-state agents, the sufficiency of state protection is relevant to a consideration whether each of the two tests -- the `fear' test and the `protection' test -- is satisfied. The proper starting point, once the tribunal is satisfied that the applicant has a genuine and well-founded fear of serious violence or ill-treatment for a Convention reason, is to consider whether what he fears is `persecution' within the meaning of the Convention. At that stage the question whether the state is able and willing to afford protection is put directly in issue by a holistic approach to the definition which is based on the principle of surrogacy. I consider that the Tribunal was entitled to hold, on the evidence, that in the appellant's case the requirements of the definition were not satisfied. I would refuse the appeal. Lord Browne-Wilkinson: My Lords, I have had the advantage of reading in draft the speeches prepared by my noble and learned friends Lord Hope of Craighead and Lord Clyde. I agree with them and for the reasons which they give I would dismiss the appeal. Lord Lloyd of Berwick: My Lords, The appellant, Milan Horvath, is a citizen of Slovakia, and a member of the Roma community. He arrived in this country on 15 October 1997, together with his wife and child. He claimed asylum on the ground that he and his family, together with other gypsies in their neighbourhood, were being persecuted by skinheads. The appellant's application for asylum was rejected by the Secretary of State. There was an appeal to a Special Adjudicator. The Adjudicator heard the case on 26 March 1998. He found against the appellant on the ground that he was not a credible witness. In the adjudicator's view the appellant and his family had come to the United Kingdom, not by reason of any fear of persecution, but in order to improve their economic circumstances. The Immigration Appeal Tribunal, consisting of His Honour Judge Pearl and two other legally qualified members, found that the appellant's account of the facts was consistent with other evidence relating to the position of gypsies in Slovakia. So they reversed the Adjudicator's conclusion on credibility. But they went on to find that the appellant had failed to discharge the burden of proving that he was unable or unwilling, through fear of persecution, to avail himself of the protection of the state of Slovakia. So they dismissed his appeal. There was then an appeal to the Court of Appeal [2000] I.N.L.R. 15. The court unanimously dismissed the appeal. But there was a difference Cases and Comments184 of opinion between the members of the court as to the right approach to article 1A(2) of the Geneva Convention relating to the Status of Refugees (1951) (Cmd. 9171), and in particular whether the absence of state protection is a necessary ingredient in the definition of persecution. Article 1A(2) provides as follows: For the purposes of the present Convention, the term `refugee' shall apply to any person who . . . owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country . . . In Adan v. Secretary of State for the Home Department [1999] 1 A.C. 293 one of the questions for decision was whether an applicant for asylum has to show a present well-founded fear of persecution, or whether it is enough that he had a well-founded fear when he left his country of origin. In answering that question in favour of the former view, I found it convenient to regard article 1A(2) as comprising two separate tests -- the `fear' test and the `protection' test. For it had been argued on behalf of the applicant in that case that if the fear test were confined to present fear, then the protection test would be otiose. An applicant with a present fear of persecution would always be unable or unwilling, owing to such fear, to avail himself of the protection of his country. In order, therefore, to give the protection test some effect, it was said to be necessary to enlarge the scope of the fear test so as to include historic fear as well as present fear. But as I pointed out in the Adan case, that argument was based on a misunderstanding. It assumed that in every case persecution is by the state. It is now well established that that is not so. Persecution by groups or factions within a state may qualify the victim for refugee status, provided the other requirements of article 1A(2) are satisfied. At pp. 305­306 I said: If category (1) [i.e. nationals who are outside their country owing to a well- founded fear of persecution for a Convention reason, and are unable to avail themselves of the protection of their country] were confined to refugees who are subject to state persecution, then I can well see that such persons would, ex hypothesi, be unable to avail themselves of state protection. On that view the words would indeed serve no purpose. But category (1) is not so confined. It also includes the important class of those who are sometimes called `third party refugees' i.e. those who are subject to persecution by factions within the state. If the state in question can make protection available to such persons, there is no reason why they should qualify for refugee status. They would have satisfied the fear test, but not the protection test. Why should another country offer asylum to such persons when they can avail themselves of the protection of their own country? But if, for whatever reason the state in question is unable to afford protection against factions within the state, then the qualifications for refugee status are complete. Both tests would be satisfied. Cases and Comments 185 I accept of course that in the end there is only one question, namely, whether the applicant has brought himself within the definition of refugee in article 1A(2) of the Convention. But in order to answer that question, I held that it was permissible as a matter of language, and helpful as a matter of analysis, to divide the question into two. If the applicant fails to show that he has a well-founded fear of `persecution' according to the ordinary meaning of that word, then the question whether he is unable or unwilling to avail himself of the protection of his country of origin does not arise. In the Court of Appeal it was common ground between the parties that the question of state protection has nothing to do with whether what the applicant fears is `persecution' within the meaning of article 1A(2) of the Convention. In principle this must be right. For it has been settled law since the decision of Nolan J. in Reg. v. Immigration Appeal Tribunal, Ex parte Jonah [1985] Imm. A.R. 7, 13 that persecution should be given its ordinary dictionary meaning. So far as I know the correctness of that decision has not been challenged. Indeed, in the course of his argument before your Lordships' counsel for the Secretary of State conceded that the ordinary meaning of the word persecution does not involve a failure of state protection. But he submitted that in the present context the word bears a different, and more limited, meaning. For my own part I can find nothing in the immediate context which colours the meaning of the word. As for the wider context, article 33(1) provides: No Contracting State shall expel or return (`refouler') a refugee in any manner whatsoever to the frontiers of territories where his life or freedom would be threatened on account of his race, religion, nationality, membership of a particular social group or political opinion. This is the converse of article 1A(2). Although it is generally accepted that persecution is not confined to threats to life or freedom, there is nothing in article 33 to suggest that `persecution' in article 1A(2) bears anything other than its ordinary meaning. Nor is there any hint that the failure of state protection is an ingredient in the meaning of the word. To the same effect is the UNHCR Handbook, paragraph 51: There is no universally accepted definition of `persecution,' and various attempts to formulate such a definition have met with little success. From article 33 of the 1951 Convention, it may be inferred that a threat to life or freedom on account of race, religion, nationality, political opinion or membership of a particular social group is always persecution. Other serious violations of human rights -- for the same reasons -- would also constitute persecution. As for the purpose or purposes underlying the Convention, we were not referred to anything in the travaux préparatoires which throws any light on the question. I agree that the idea of `substitute' or `surrogate' Cases and Comments186 protection which was first developed by Professor Hathaway in The Law of Refugee Status at pp. 124­128, (and which I gratefully adopted in the passage which I have already quoted from my speech in the Adan case), is a useful concept. It describes one of the conditions for refugee status. But I am unable to see how it helps on the definition of persecution. Thus the principle of surrogate protection finds its proper place in the second half of article 1A(2). If there is a failure of protection by the country of origin, the applicant will be unable to avail himself of that country's protection. But I can see no reason, let alone any need, to introduce the idea into the first half of the clause. Indeed, to do so could only lead to unnecessary complications. Take two countries each of which has a similar mix of skinheads and gypsies. Assume that the degree of violence is greater in one country than the other, but that the degree of protection is less. How is the fact-finding tribunal to balance these factors in determining whether the violence amounts to persecution? It is the severity and persistence of the means adopted, whether by the state itself, or factions within the state, which turns discrimination into persecution; not the absence of state protection. It is surely simpler, and therefore better from every point of view, not least that of an appellate court considering an appeal on a question of law, that the fact-finding tribunal should first assess the ill-treatment, and answer the question whether it amounts to persecution for a Convention reason, and then, as a separate question, evaluate the protection available to the applicant. I can see no advantage in running these two questions together. It follows that I agree with Mr. Plender's submission that the absence of state protection is not a relevant ingredient in the definition of persecution. To adopt a phrase used in a different context by La Forest J. in Canada (Attorney-General) v. Ward (1993) 103 D.L.R. (4th) 1, 20: `There is . . . no need for a judicial gloss.' It follows also that I agree with every word of the following passage from the judgment of Stuart-Smith L.J. [2000] I.N.L.R. 15, 20­21: [10] It is apparent that there are five conditions that the applicant must satisfy to establish his status as a refugee, namely that: (1) He is out of the country of his nationality because he has a fear of ill-treatment. (2) The ill-treatment that he fears is of a sufficiently grave nature as to amount to persecution. (3) His fear of persecution is well-founded. (4) The persecution is for a Convention reason. (5) He is unable, or owing to fear of the persecution, is unwilling to avail himself of the protection of that country. [11] These are separate and discrete tests each of which must be satisfied. Logic and convenience suggest that the fact-finding tribunal should address each question sequentially. Some issues may not be in dispute; Cases and Comments 187 equally in some cases there may be a short and obvious answer to the application on one particular ground. But I can see no reason or advantage in importing into the consideration of one issue, matters which logically fall to be considered under another. On the contrary it seems to me to be likely to lead to confusion. Thus as a matter of principle consideration of the question whether the applicant has shown that he is unable, or through fear of the persecution is unwilling to avail himself of the state's protection, which in turn involves a consideration of the state's ability and willingness to afford protection -- which I will call the protection test -- properly concerns the fifth test and should not be confused with the first three questions, which can broadly together be described as the `fear test'. [12] Again as a matter of principle it seems to me that the protection test has nothing to do with the second question, namely whether the ill- treatment which the applicant has suffered and fears in the future amounts to persecution. It is now well established that the word `persecution' is to be given its ordinary dictionary meaning of `to pursue with malignancy or injurious action especially to oppress for holding a heretical belief' (see per Nolan J. in Reg. v. Immigration Appeal Tribunal, Ex parte Jonah [1985] Imm. A.R. 7, 13). Equally it is well established that the persecution can be at the hands of non-state agents such as neighbours, family (as in Reg. v. Immigration Appeal Tribunal, Ex parte Shah [1999] 2 A.C. 629) or hostile factions (Adan v. Secretary of State for the Home Department [1999] 1 A.C. 293). I see no reason why, if the ill-treatment received at the hands of such perpetrators is of sufficient gravity to amount to persecution, it should cease to have that quality because the state can offer adequate protection against it. The majority of the Court of Appeal took a different view. Ward L.J. accepted an alternative submission of counsel for the Secretary of State, contrary to his primary submission, that the ability of the state to provide protection is part of what is meant by persecution. The reason for this alternative submission was, no doubt, that it was the approach adopted, wrongly in my view, by the I.A.T. In Ward L.J's view the ability of the state to provide protection comes in as a `necessary ingredient' at every stage of the analysis. He called this the `holistic' approach: p. 47. But apart from helpful references to a number of authorities he does not spell out his reasons for not confining the availability of protection to the `protection test.' Hale L.J. agreed with Ward L.J. that the state's role in providing protection is relevant to the `conception' of persecution itself. She gave a number of reasons of which the most powerful to my mind is the fifth at p. 57, para. 17. If there are thugs about perpetrating serious acts of maltreatment against the population as a whole, but the state offers protection only to some of its citizens, and not to others, in my view those citizens are being persecuted in just the sort Cases and Comments188 of way that merits the surrogate protection of other states under the Convention. But if the failure of state protection were relevant only to the fifth question [i.e. the inability or unwillingness of the applicant to avail himself of the protection of his country], it is difficult to see how the necessary link with discrimination can be made in such cases. This is, of course, a variation of the case of the Jewish shopkeeper described by Lord Hoffmann in Reg. v. Immigration Appeal Tribunal, Ex parte Shah [1999] 2 A.C. 629, 654 to which I will return later. I agree with Hale L.J. that the activities of the gang of thugs in her example could not amount to persecution for a Convention reason, since their activities are directed against the population as a whole. But the failure of the state to provide protection to some but not all the victims does not change the nature of those activities; nor could it provide the missing element of discrimination, unless one assumes that the word `persecution' includes partial acquiescence by the state in non-discriminatory persecution by others. But this begs the question. To my mind it is most unlikely that the framers of the Convention had any such unusual case in mind, or that they intended to cover what the noble and learned Lord Clyde aptly refers to as `constructive' persecution by the state. If so then it is not for us to create a link between the activities of the thugs and discrimination by the state, so as to extend the scope of the Convention by judicial interpretation, any more than we should limit the meaning of persecution by introducing into the definition of persecution the concept of state protection. We were referred to a number of authorities other than Ex parte Jonah and the Adan case. I do not find any of these of much assistance. Ex parte Shah is an important decision on the meaning of `particular social group' in article 1(A)(2). It was decided that women in Pakistan were such a group. Since there was no dispute as to any of the other ingredients in the definition of refugee status, it followed that the applicants qualified as refugees. Counsel for the Secretary of State relied on the passage in Lord Hoffmann's speech [1999] 2 A.C. 629, 653 in which he referred to the two elements which need to be combined to constitute persecution within the meaning of the Convention. He approved a formula taken from Gender Guidelines for the Determination of Asylum Claims in the U.K. (published by the Refugee Woman's Legal Group in July 1998) that `Persecution = Serious Harm + the Failure of State Protection.' But it was common ground in the Court of Appeal in the present case that Lord Hoffmann's observation was not necessary for his decision in that case, and counsel for the Secretary of State did not seek to persuade us otherwise. I note also that there is no reflection of the same approach in the leading speech of Lord Steyn. For my part I would agree with a formula `Persecution for a Convention Reason + Failure of State Protection = Refugee Status'. But I cannot agree the formula as it stands. Cases and Comments 189 As for the case of the Jewish shopkeeper, the question with which Lord Hoffmann was dealing was whether the shopkeeper was being persecuted on the ground of race, or whether he was being persecuted on the ground that a competitor wanted to drive him out of business. This is, as he pointed out, a question of causation. The example taken by Lord Hoffmann was intended to illustrate the point that questions of causation will often depend on the context in which they arise. The example was not, I think, intended to throw any light on the meaning of the word `persecution.' Counsel for the Secretary of State also referred to certain passages in chapter 4 of Professor Hathaway's book on Refugee Status. Although the title of that chapter is Persecution, a glance at the contents shows that it is not so limited. The author is as much concerned with the requirements for refugee status, as with the meaning of persecution. This is nowhere clearer than in the quotation at p. 129 from what Professor Hathaway regarded as `a landmark decision' in Rajudeen v. Minister of Employment and Immigration (1985) 55 N.R. 129. Counsel also relied strongly on the last sentence of paragraph 65 of the UNHCR Handbook which reads: 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. I agree that this sentence supports his case. But I note that there is no equivalent sentence in paragraph 51, which is the paragraph dealing specifically with the definition of persecution. The meaning of persecution cannot vary according to whether it applies to state persecution or third party persecution. I suggest that the last sentence of paragraph 65 would be more accurate if it read: Where serious discriminatory or other offensive acts are committed by the local populace, they can be considered as persecution, and therefore justify a claim for refugee status, if they are knowingly tolerated by the authorities . . . In any event there is a danger in regarding the Handbook as if it had the same force as the Convention itself. In the end it comes back to the language of article 1A(2). Of course one must give full weight to the context, and the purposes which the framers of the Convention had in mind. But even so, I cannot agree with the majority of the Court of Appeal that `persecution' as a word has anything other than its ordinary meaning. On the important issue which divided the Court of Appeal, I would therefore accept Mr. Plender's argument for the appellant as correct. Unfortunately, however, it does not get Mr. Plender's clients very far. Cases and Comments190 Assuming in their favour (as Stuart-Smith L.J. was prepared to hold) that they have a well-founded fear of persecution for a Convention reason, and that they thereby satisfied the fear test, they still have to satisfy the protection test. Have they shown that they are unable to avail themselves of the protection of the state of Slovakia? The answer must be no. I agree with the test proposed by Stuart-Smith L.J., at pp. 25­26, paras. 20­23, in which the other members of the Court of Appeal concurred. On the findings of the tribunal in paragraphs 60 and 61, we can infer that the authorities in Slovakia are able and willing to provide protection to the required standard, and that gypsies, as a class, are not exempt from that protection. This finding is not, of course, in any way inconsistent with the finding that the applicants had a well-found fear of persecution. As Stuart-Smith L.J. pointed out, there are parts of London or New York where one may indeed have a well-found fear of being attacked in the street. But that does not mean that there is not an efficient police force and an impartial judiciary. As for the second part of the protection test, there will not be many cases in which an applicant who is able to avail himself of the protection of his country of origin, will succeed on the ground that he is unwilling to do so. Here the applicant's case, as it appears from his written statement, is that he regards the local police as ineffective and indifferent: see paragraphs, 11, 12 and 19. But he is not the sole judge of that. The test is objective. The tribunal has found as a fact that the available protection satisfies the Convention standard. There are no special circumstances which would enable the applicant to succeed on the second branch of the protection test, having failed on the first. I would dismiss the appeal. Lord Clyde: My Lords, The appellant is a Roma national of Slovakia. He has claimed asylum in the United Kingdom on the grounds of a fear of violence at the hands of persons known as skinheads who were perpetrating acts of violence against the Roma people. He provided the Immigration Appeal Tribunal with examples of the treatment which members of the Roma minority in Slovakia were receiving from the skinheads, the details of which do not require to be repeated here. He also complained of certain acts of a discriminatory character on the part of the state or its agents, but these were held by the tribunal to be not of sufficient severity to amount to persecution. The factual background to the case is accordingly one of the risk of acts of intimidation and physical violence against the appellant. The question in the appeal is whether he qualifies as a refugee for the purposes of the Geneva Convention relating to the Status of Refugees 1951 as modified by the Protocol of 1967. The point arises in the context of Rule 334 of the Statement of Changes in Immigration Rules (H.C. 395) (1994) which at the relevant time provided the conditions under which an asylum applicant Cases and Comments 191 would be granted asylum in the United Kingdom. The conditions included as condition (ii) the requirement that the person be a refugee as defined by the Convention and Protocol. The critical part of the Convention with which we are concerned is the definition of refugees contained in article 1A(2). This refers to a person who owing to well-founded fear of being persecuted for reasons of race, religion, nationality, membership of a particular social group or political opinion, is outside the country of his nationality and is unable or, owing to such fear, is unwilling to avail himself of the protection of that country . . . It may be noted that condition (iii) in rule 334 also required that refusing his application would result in his being required to go . . . in breach of the Convention and Protocol, to a country in which his life and freedom would be threatened on account of his race, religion, nationality, political opinion or membership of a particular social group. We are concerned here with the construction of an international convention. The approach to be adopted must be appropriate to that situation. Regard must be given to the purpose of the Convention and the object which it seeks to serve. While the language of the article has to be respected, any pre-occupation with the precise words may fail to meet the broad intent of the Convention and any detailed analysis of its component elements may distract and divert attention from the essential purpose of what is sought to be achieved. As my noble and learned friend Lord Lloyd of Berwick observed in Adan v. Secretary of State for the Home Department [1999] 1 A.C. 293, 305. It follows that one is more likely to arrive at the true construction of article 1A(2) by seeking a meaning which makes sense in the light of the Convention as a whole, and the purposes which the framers of the Convention were seeking to achieve, rather than by concentrating exclusively on the language. A broad approach is what is needed, rather than a narrow linguistic approach. The dangers of over-sophistication in the construction and application of the Convention are real and significant. Prolonged debate about the niceties of the language may readily lead to delay in the processing of what in the interests of everyone should be a relatively expeditious process. Of course there may often be difficult points of fact to be resolved and uncertainties in matters of fact which may not immediately be open to a clear answer. But it is obviously undesirable to heap onto the shoulders of the adjudicators and the members of the tribunals who already have a heavy burden of work an additional complexity in the unravelling of legal issues on the precise construction of the particular words used in the Convention. The Convention was worked out and agreed between states and it is at a state level that it has to be understood. As its preamble records it is Cases and Comments192 prompted by concern for the enjoyment by refugees of the fundamental rights and freedoms propounded in the Charter of the United Nations and the Universal Declaration of Human Rights of 10 December 1948 (U.N. Doc. A/811) without discrimination. What it seeks to achieve is the preservation of those rights and freedoms for individuals where they are denied them in their own state. Another state is to provide a surrogate protection where protection is not available in the home state. The Convention assumes that every state has the obligation to protect its own nationals. But it recognises that circumstances may occur where that protection may be inadequate. The purpose of the Convention is to secure that a refugee may in the surrogate state enjoy the rights and freedoms to which all are entitled without discrimination and which he cannot enjoy in his own state. It is essentially against the background of that consideration of the protection which the individual may expect from his home state that the definition has to be understood. As Professor Hathaway observes in The Law of Refugee Status (pp. 103­104). the intention of the drafters was not to protect persons against any and all forms of even serious harm, but was rather to restrict refugee recognition to situations in which there was a risk of a type of injury that would be inconsistent with the basic duty of protection owed by a state to its own population. The same point was made by Lord Keith of Kinkel in Reg. v. Secretary of State for the Home Department, Ex parte Sivakumaran [1988] 1 A.C. 958, 992­993 where he observed that `the general purpose of the Convention is surely to afford protection and fair treatment to those for whom neither is available in their own country,' and by La Forest J. in Canada (Attorney- General) v. Ward (1993) 103 D.L.R (4th) 1, 12 where he said: International refugee law was formulated to serve as a back-up to the protection one expects from the state of which an individual is a national. It was meant to come into play only in situations when that protection is unavailable, and then only in certain situations. The international community intended that persecuted individuals be required to approach their home state for protection before the responsibility of other states becomes engaged. 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 Cases and Comments 193 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 Cases and Comments194 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. Cases and Comments 195 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 Cases and Comments196 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 Cases and Comments 197 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 Cases and Comments198 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 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. Cases and Comments 199 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 Cases and Comments200 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. Cases and Comments 201 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. Lord Hobhouse of Woodborough: My Lords, I agree with my noble and learned friend Lord Hope of Craighead and for the reasons which he has given that this appeal should be dismissed.