In re B (FC) (Appellant) (2002) Regina v. Special Adjudicator (Respondent) ex parte Hoxha (FC) (Appellant) [2005] UKHL 19

House of Lords
[2005] UKHL 19
Between
In re B (2002) ReginaAppellantSpecial Adjudicator ex parte HoxhaRespondent
Before
Baroness Hale of RichmondLord Brown of Eaton-underheywoodLord Hope of CraigheadLord Nicholls of BirkenheadLord SteynDate 10 March 2005
[1]I have had the advantage of reading in draft the speeches of my noble and learned friends Lord Hope of Craighead, Baroness Hale of Richmond and Lord Brown of Eaton-under-Heywood. For the reasons they give I would dismiss these appeals. LORD STEYN My Lords,[2]I have read the opinions of my noble and learned friends Lord Hope of Craighead, Baroness Hale of Richmond and Lord Brown of Eaton-under-Heywood. I agree with their opinions. I would also dismiss the appeals. LORD HOPE OF CRAIGHEAD My Lords,[3]I have had the advantage of reading in draft the speech of my noble and learned friend, Lord Brown of Eaton-under-Heywood. I agree with it and with the observations of my noble and learned friend Baroness Hale of Richmond on the subsidiary issue. For the reasons that they have given I too would dismiss these appeals.[4]I adopt with gratitude Lord Brown's description of the background and his analysis of the authorities and other relevant materials. This permits me to deal briefly with the points that lie at the heart of the case.[5]The appellants claim that they have produced a compelling body of evidence which shows that the modern construction of the proviso to article 1C(5) of the Geneva Convention Relating to the Status of Refugees of 28 July 1951 (Cmd 9171) ("the Convention") is one which covers all refugees and that it is not limited in its application to refugees falling under article 1A(1) ("statutory refugees"). In my opinion however one has only to scratch the surface to see that this proposition is not based on any hard evidence that this indeed is what the proviso is being regarded as meaning, as a matter of legal obligation binding on all states parties to the Convention. There is a profound gap between what various commentators would like the proviso to mean and what it has actually been taken to mean in practice.[6]No-one questions the broad humanitarian principles which underlie the Convention. The social and humanitarian nature of the problem of refugees was expressly recognised in the preamble to the Convention. So too was the fact that it was the express wish of all states to do everything within their power to prevent the problem from becoming a cause of tension between them. The 1967 Protocol Relating to the Status of Refugees ("the Protocol") (Cmnd 3906) recognised that new situations had arisen since the Convention was adopted and that further provisions were needed as persons who had become refugees since 1 January 1951 might not fall within its scope. As the third paragraph of the preamble to the Protocol put it, it was desirable that equal status should be enjoyed by all refugees covered by the definition in the Convention irrespective of the dateline of 1 January 1951.[7]As a result of the amendments which it made to article 1A(2) of the Convention, these two instruments now provide the cornerstone of the international legal regime for the protection of refugees: see paragraph (3) of the preamble to Council Directive 2004/83/EC of 29 April 2004 on minimum standards for the qualification and status of third country nationals or stateless persons as refugees ("the Directive"). These are to be seen as living instruments, to which the broadest effect must be given to ensure that they continue to serve the humanitarian principles for whose purpose the Convention was entered into.[8]Care nevertheless needs to be taken, when analysing the evidence on which the appellants rely, to distinguish between the meaning of the words which article 1 of the Convention uses to identify those who are entitled to the status of refugee and the practices which contracting states have chosen to adopt in their discretion to give effect to these humanitarian principles. A large and liberal spirit is called for when a court is asked to say what the Convention means. But there are limits to this approach. The court must recognise the fundamental fact that the Convention is an agreement between states. The extent of the agreement to which the states committed themselves is to be found in the language which gives formal expression to their agreement. The language itself is the starting point: see Adan v Secretary of State for the Home Department [1999] 1 AC 293, 305D-E, per Lord Lloyd of Berwick.[9]As Lord Bingham of Cornhill said it in Brown v Stott [2003] 1 AC 681, 703E, it is generally to be assumed that the parties included the terms that they wished to include and on which they were able to agree, omitting other terms which they did not wish to include or on which they were unable to agree. Article 31(1) of the Vienna Convention on the Law of Treaties (1980) (Cmnd 7964) provides that a treaty must be interpreted in good faith in accordance with the ordinary meaning of the terms of the treaty in their context, and in the light of the object and purpose of the treaty. There is no warrant in this provision for reading into a treaty words that are not there. It is not open to a court, when it is performing its function, to expand the limits which the language of the treaty itself has set for it.[10]The structure of the definition of the term "refugee" in article 1 of the Convention was based on that of the Statute of the Office of the United Nations High Commissioner for Refugees, adopted by General Assembly Resolution 428(v) of 14 December 1950. This was the initial point of reference in formulating the definition in the Convention: Professor James Hathaway, The Law of Refugee Status (1991), p 66-69. Paragraph 6 of the Statute, which defines the competence of the High Commissioner, contains two subparagraphs. Paragraph A contains a definition of the persons to whom that competence was to extend in terms which were adopted by article 1A(1) and (2) of the Convention. It also contains a list of six circumstances in which his competence was to cease. Under this scheme the assumption of competence and the cessation of competence were kept separate. It was only if the High Commissioner had assumed competence that the question of cessation could arise.[11]This two stage approach was reproduced in article 1C of the Convention, but with some significant changes to the words used to describe the last two circumstances. Paragraph B of the Statute, on which article 1A(2) of the Convention was based, was in these terms:
"Any other person who is outside the country of his nationality, or if he has no nationality, the country of his former habitual residence, because he has or had well-founded fear of persecution by reason of his race, religion, nationality or political opinion and is unable or, because of such fear, is unwilling to avail himself of the protection of the government of the country of his nationality, or, if he has no nationality, to return to the country of his former habitual residence."
[emphasis added][12]As Professor Hathaway explains at p 68, the compromise that emerged from the drafting process when the definition for the purposes of the Convention was being formulated was to reject the past assessment of risk and to establish instead present or prospective assessment of risk as the norm for refugee protection. It was decided to honour the past persecution standard for persons who were within the scope of the agreements which had been entered into before 1 January 1951. But persons who were outside the scope of those agreements were to be required to demonstrate a current well founded fear of persecution in order to qualify for refugee status. The words "or had" which had been included in paragraph 6B of the Statute were omitted from article 1A(2) of the Convention. It is plain from the drafting history that this was no accident. The appellants are unable to establish a current well founded fear, so they are unable to bring themselves within the wording of article 1A(2).[13]A similar approach was taken to the cessation provisions which were derived from the Statute. As Lord Lloyd of Berwick observed in Adan v Secretary of State for the Home Department [1999] 1 AC 293, 306G, the cessation provision in article 1C(5) takes effect naturally when the refugee ceases to have a current well-founded fear. This is in symmetry with the definition in article 1A(2). The words "no longer", which were taken from the cessation provisions in paragraph 6A of the Statute, support that interpretation. On this approach the appellants are unable to bring themselves within the opening words of article 1C(5). This means that their case fails at the first hurdle before they reach the proviso which was added to that paragraph, whose its meaning lies at the heart of this case:
"provided that this paragraph shall not apply to a refugee falling under section A(1) of this article who is able to invoke compelling reasons arising out of previous persecution for refusing to avail himself of the protection of the country of nationality."
[emphasis added][14]The word "previous" makes it plain that in this context, in contrast to what was contemplated by the leading provision in article 1C(5), the test looks backwards. The question here is not whether the person has a current well-founded fear, but whether those who had previously been determined to be refugees under article 1A(1) - the statutory refugees - could justify their refusal by reference to what had happened in the past. A proviso in similar terms was inserted in article 1C(6).[15]Grahl-Madsen, The Status of Refugees in International Law (1966), vol 1, p 410 explains the reasoning behind this exception in this way:
"What the drafter of the Convention had in mind was the situation of refugees from Germany and Austria, who were unwilling to return to the scene of the atrocities which they and their kin had experienced, or to avail themselves of the protection of a country which had treated them so badly. The fact was appreciated that the persons in question might have developed a certain distrust of the country itself and a disinclination to be associated with it as its national."
The drafting history indicates that, while the framers of the Convention had the opportunity to extend the benefit of this proviso to all refugees, a deliberate decision was taken to confine its application to the statutory refugees who had been identified in article 1A(1). A distinction was thus created between them and those identified in article 1A(2).[16]It can no doubt be said that the effect of this distinction was that these two classes of refugees were not to be treated equally. It can also be said that unequal treatment is inconsistent with the general humanitarian principle underlying the Convention and with the principle which is revealed by the prohibition on discrimination in article 3. This point was not overlooked by the United Kingdom delegate at the Geneva Conference, Mr Hoare. He stated that he regretted the limitation of the proviso to article 1A(1) "statutory" refugees, although he appreciated the motives that had prompted it. Nevertheless he accepted it in the interests of accommodating the concerns of other states. There is, then, no getting away from the plain words of the proviso. The only conclusion that can properly be drawn from its terms, having regard to their context and the drafting history, is that the contracting parties were not willing at the time the Convention was entered into to extend the benefit of the proviso to non-statutory refugees.[17]The appellants say that there was no need for article 1A(2) to spell out in terms that it contained a similar rule, based on the humanitarian principle, to that expressly identified in the proviso to article 1C(5) in the case of refuges under article 1A(1). This is because there were hardly any article 1A(2) refugees at the stage when the wording was being finalised. They also maintain that, if the Convention is interpreted consistently with that principle as a living instrument capable of changing with the modern world, the proviso is capable of being applied to all refugees. They point to various texts which urge states to adopt practices which give practical effect to that construction of it.[18]For example, paragraph 136 of the UNHCR Handbook on Procedures and Criteria for Determining Refugee Status (Geneva, 1979) states with reference to article 1C(5):
"The second paragraph of this clause contains an exception to the cessation provision contained in the first paragraph. It deals with the special situation where a person may have been subjected to very serious persecution in the past and will not therefore cease to be a refugee, even if fundamental changes have occurred in his country of origin. The reference to article 1A(1) indicates that the exception applies to 'statutory refugees'. At the time when the 1951 Convention was elaborated, these formed the majority of refugees. The exception, however, reflects a more general humanitarian principle, which could also be applied to refugees other than statutory refugees. It is frequently recognized that a person who - or whose family - has suffered under atrocious forms of persecution should not be expected to repatriate. Even though there may have been a change of regime in his country, this may not always produce a complete change in the attitude of the population, nor, in view of his past experiences, in the mind of the refugee."
[emphasis added][19]The reasoning in this passage reflects a long standing concern on the part of the UNHCR that the cessation clauses in the Convention are being taken too literally. On 11 October 1991 its Executive Committee, in its General Conclusion on International Protection (Doc No 65 (XLII) 1991), para (q), underlined the possibility of the use of this clause more generally where compelling reasons might, for certain individuals, support the continuation of their refugee status. But in footnote 8 to para 24 of its Note on the Cessation Clauses dated 30 May 1997 the Standing Committee of the Executive Committee of the High Commissioner's Programme (EC/47/SC/CRP.30) accepted that the proviso expressly covered only those refugees falling under article 1A(1). The only qualification which it contains is that para 136 of the Handbook "suggests" that the exception reflects a more general humanitarian principle and that it "could" also be applied to refugees other than those in article 1A(1). The point is not developed further. I take this to be an acknowledgment that there is nothing in the Convention that actually obliges contracting states to apply the proviso to other refugees.[20]An opportunity to address this issue arose in 1967 when the Protocol was being entered into. The preamble to the Protocol indicates that there was a desire to bring those affected by refugee situations that had arisen since the Convention within its scope and that equal status should be enjoyed by all refugees covered by the definition in the Convention. But the preamble to article 1C(5) was left untouched. I think that we must take it that, although they were aware that events had moved on since 1951, the states parties were still not willing to agree to a relaxation of the limitation that had been expressly written into the proviso.[21]A further opportunity to address this issue arose when the European Commission was framing its proposal for a Council Directive on minimum standards for the qualification and status of third country nationals and stateless persons as refugees. Its purpose was to lay down a common definition of the concept of "refugee" as contained in the Convention, and to provide a minimum standard of protection for those who fell outside that definition to complement the Convention in all member states. Article 13 of the proposal, which was in the same terms as paragraph 1 of article 11 in what is now Council Directive 2004/83/EC, describes the circumstances in which a third country national or a stateless person is to cease to be a refugee. It repeats the language of article 1C of the Convention, except that it omits the provisos to paragraphs 5 and 6 of that article.[22]The Commission's commentary on the proposal contains this paragraph with reference paragraph (e) of article 13, which is the equivalent to article 1C(5):
"The Member State invoking this cessation clause should ensure that an appropriate status , preserving previously acquired rights, is granted to persons who are unwilling to leave the country for compelling reasons arising out of previous persecution or experiences of serious and unjustified harm, as well as to persons who cannot be expected to leave the Member State due to a long stay resulting in strong family, social and economic links in that country."
[emphasis added][23]The proposal having thus been initiated by the Commission, it was subject to scrutiny by the legislatures of member states in accordance with the principle of subsidiarity enshrined in article 5 of the EC Treaty and with article 3 of the Protocol on the role of national parliaments in the European Union. As part of this process it was examined by the House of Lords Select Committee on the European Union in 2002: see its report, Defining Refugee Status and those in need of International Protection : Session 2001-02, HL Paper 156. The inquiry was conducted by Sub-Committee E under the Chairmanship of Lord Scott of Foscote. Evidence was taken, both written and oral. Comments were received in the course of that process from, among others, Mr Goodwin-Gill, UNCHR and the European Council on Refugees and Exiles (ECRE). A striking feature of this inquiry, in the light of the arguments that have been advanced in this case, is the absence of any suggestion from Mr Goodwin-Gill or from UNHCR that paragraph (e) of article 13 should be qualified by a proviso in relation to those who had ceased to be refugees as defined by the Directive in the same way as, in relation to article 1A(1) refugees, article 1C(5) of the Convention was qualified.[24]Mr Goodwin-Gill made this comment on article 13 in para 15 of his memorandum:
"While the commentary to article 13(e) (cessation of status by reason of change of circumstances in the country of origin) recognizes the special situation of person who have compelling reasons arising out of previous persecution or other serious and unjustified harm, their continuing protection should be formally included in the Directive. The numbers will always be small, and this interpretation has already been adopted, sometimes by legislation, in states aware of the reality of the refugee experience."
He stops short of saying that it is the general practice of states to interpret the proviso to article 1C(5) in this way and that a proviso to this effect was needed in order to reflect the existence of this general practice in the Directive. In para 37 of its memorandum the UNHCR expressed its pleasure at the fact that article 13 had taken in the cessation clauses of the Convention and that it placed the burden of proving the cessation of refugee status on the state asserting it. In the following paragraph it added its only other comment. This was raising a different point:
"UNHCR would, however, recommend that the commentary on this article be amended to reflect the generally accepted position that, in certain circumstances, the refugee may be able to obtain or renew his or her national passport without forfeiting his or her refugee status."
[25]In its comment on article 13 the ECRE did raise the issue about inserting a proviso to article 13(e):
"ECRE agrees with the provisions of this article which in most aspects reflect those of article 1C of the Refugee Convention. We would propose that a provision is added to paragraph (e) to ensure that Member States exempt from the application of article 13(e) refugees who are able to invoke compelling reasons arising out of previous persecution for refusing to avail themselves of the protection of the country of nationality or persons who cannot be expected to leave the Member State due to a long stay resulting in strong family, social and economic links in the country. This is referred to in the Commission's Commentary on article 13 and should be reflected in the main text of the proposal."
This recommendation went further than the Commission's Commentary, which had referred instead to ensuring that "an appropriate status" was granted to such persons. But it was not backed up by the assertion that the practice of states showed that this was how the proviso to article 1C(5) of the Convention was currently being interpreted It was not adopted when the Directive was finalised, as one would have expected if there had been evidence to this effect from other member states. The terms of the Directive, in its final form, must be taken to reflect the common position of member states as to the meaning and effect of the Convention. This evidence is wholly inconsistent with the appellants' argument that it is the general practice of states parties to construe the proviso to article 1C(5) as applying to persons who had acquired the status of refugees under article 1A(2) of the Convention.[26]I would hold therefore that the answer to the question whether the appellants are entitled to the status of refugee within the meaning of article 1 of the Convention, and to all the rights that flow from that status, must be in the negative. As the Court of Appeal said [2003] 1 WLR 241, para 49, the evidence does not establish a clear and widespread state practice sufficient to override the express words of limitation in the proviso. But this does not mean that it would not have been open to the Secretary of State in his discretion to allow the appellants to remain in this country. In cases where a person is or may be particularly vulnerable by reason of the continuing effects of the persecution that he has suffered in the past, and is thus less able to cope with the conditions which he would have to face in his country of origin, he may be given leave to remain here on compassionate grounds. This is the way in which the United Kingdom gives effect to the humanitarian principle. BARONESS HALE OF RICHMOND My Lords,[27]A subsidiary issue in this appeal is whether a person may fall within the definition of a refugee in article 1A(2) of the Convention in circumstances where he has a fear of the continuing effects of persecution inflicted upon him in the past. Both appellants suffered appalling ill-treatment at the hands of Serbian soldiers or policemen. In Mr Hoxha's case it is said that as a result he continues to fear life in Kosovo where he will be destitute, lack medical treatment, accommodation, employment, and the ability to earn a living. In the case of the B family, it is said that their younger son is still suffering mental and physical problems as a result of the knife attack upon him and that the whole family will face ostracism from their own community as a result of the wife's rape in front of so many villagers and her husband's determination to stand by her.[28]The Court of Appeal observed (paras 53 and 54) that it was implicit in the appellants' reliance on the continuing effects of past persecution that what might happen to them on their return to Kosovo did not amount to persecution. As it is well established that the fear of persecution and its well-founded nature have to be current, their argument could not succeed. If it could, there would have been no need for the proviso to article 1C(5) because all those with compelling reasons arising out of past persecution not to return would still have qualified as refugees.[29]Put in the broad way that the issue was framed before us, the answer must be 'no'. An understandable unwillingness to return based upon the continuing effects of past persecution is not enough. There must be a current fear of persecution for a Convention reason upon return: see Adan v Secretary of State for the Home Department [1999] 1 AC 293. But of course the persecution suffered in the past is relevant to whether a person has a current well-founded fear of persecution. Generally the past persecution will lead to the fear of similar persecution on return but that need not always be the case.[30]Hence there is a rather different case, which was touched on but not fully developed both in the Court of Appeal and before us, in relation to Mrs B and her family. This is that earlier persecution of one sort may lead to later persecution of a different sort. All four members of the B family suffered persecution at the hands of the Serb police because they were Kosovan Albanians and Mr B was suspected of involvement with the KLA. But the persecution of Mrs B was expressed in a different way from the persecution of her husband and sons. She was raped in front of her husband, her sons and twenty to thirty of their neighbours. As Rodger Haines QC notes in his paper on "Gender-related persecution", prepared for the UNHCR's San Remo expert roundtable in 2001 (see Feller, Turk and Nicholson, Refugee Protection in International Law, UNHCR's Global Consultations on International Protection , (2003), Chapter 5.1, p 336):
"Women are particularly vulnerable to persecution by sexual violence as a weapon of war."
He goes on to quote Heaven Crawley, Refugees and Gender: Law and Process , 2001, pp 89-90:
"During war, women's bodies become highly symbolic and the physical territory for a broader political struggle in which sexual violence including rape is used as a military strategy to humiliate and demoralise an opponent; women's bodies become the battleground for 'pay-backs', they symbolise the dominance of one group over another . . . It is important to recognise that sexual violence and rape may be an actual weapon or a strategy of war itself, rather than just an expression or consequence. In the context of armed conflict or civil war, the rape of women is also about gaining control over other men and the group (national, ethnic, political) of which they are a part."
[31]Following the San Remo roundtable, the UNHCR published Guidelines on Gender-Related Persecution within the context of Article 1A(2) of the 1951 Convention and/or its 1967 Protocol relating to the Status of Refugees (HCRR/GIP/02/01, 7 May 2002). These make the same point at para 24:
" . . . the persecutor may choose to destroy the ethnic identity and/or prosperity of a racial group by killing, maiming or incarcerating the men, while the women may be viewed as propagating the ethnic or racial identity and persecuted in a different way, such as through sexual violence or control of reproduction."
[32]If sexual violence is used in this way, the consequences, not only for the woman herself but also for her family, may be long-lasting and profound. This is particularly so if she comes from a community which adds to the earlier suffering she has endured the pain, hardship and indignity of rejection and ostracism from her own people. There are many cultures in which a woman suffers almost as much from the attitudes of those around her to the degradation she has suffered as she did from the original assault. The UNHCR Guidelines recognise that punishment for transgression of unacceptable social norms imposed upon women is capable of amounting to persecution.[33]Nonetheless, it seemed to the Court of Appeal in this case (para 53) that "the appellants are right not to seek to allege that what would happen to the applicants on their return would itself amount to persecution, since that entails 'acts of violence or ill-treatment' of a sufficiently grave nature: see Horvath v Secretary of State for the Home Department [2001] 1 AC 489, per Lord Hope of Craighead, at p 499H and per Lord Lloyd of Berwick, at p 504C to D."[34]However, whether feared ill-treatment is sufficiently grave to amount to persecution has to be seen in the context of each individual case. Gender is an important component of that context. The San Remo roundtable concluded that there was no need to add sex or gender to the Convention grounds, because "the text, object and purpose on the Convention require a gender-inclusive and gender-sensitive interpretation" (see Feller, Turk and Nicholson, op cit , Chapter 5.2, "Summary Conclusions: gender-related persecution", p 351).[35]The UNHCR Handbook (1992) states (at para 51) that there is no universally accepted definition of persecution, although a threat to life or liberty or other serious violations of human rights for a Convention reason would amount to persecution. Further:
"52. Whether other prejudicial actions or threats would amount to persecution will depend on the circumstances of each case, including the subjective element to which reference has already been made . . . The subjective character of fear of persecution requires an evaluation of the opinions and feelings of the person concerned. It is also in the light of such opinions and feelings that any actual or anticipated measures against him must necessarily be viewed. Due to variations in the psychological make-up of individuals and in the circumstances of each case, interpretations of what amounts to persecution are bound to vary. 53. In addition, an applicant may have been subjected to various measures not in themselves amounting to persecution (e.g. discrimination in different forms), in some cases combined with other adverse factors (e.g. general atmosphere of insecurity in the country of origin). In such situations, the various elements involved may, if taken together, produce an effect on the mind of the applicant that can reasonably justify a claim to well-founded fear of persecution on 'cumulative grounds'."
The Handbook's guidance on discrimination is developed in the 2002 Guidelines on Gender-related Persecution :
"14. While it is generally agreed that 'mere' discrimination may not, in the normal course, amount to persecution in and of itself, a pattern of discrimination or less favourable treatment could, on cumulative grounds, amount to persecution and warrant international protection. It would, for instance, amount to persecution if measures of discrimination lead to consequences of a substantially prejudicial nature for the person concerned… 15. Significant to gender-related claims is also an analysis of forms of discrimination by the State in failing to extend protection to individuals against certain types of harm. . . "
As Haines states at para 24 of his paper on "Gender-related Persecution" (citing Refugee Appeal No 71427/99[2000] NZAR 545; [2000] INLR 608):
"However discrimination can affect individuals to different degrees and it is necessary to recognise and to give proper weight to the impact of discriminatory measures on women. Various acts of discrimination, in their cumulative effect, can deny human dignity in key ways and should properly be recognised as persecution . . . "
[36]To suffer the insult and indignity of being regarded by one's own community (in Mrs B's words) as 'dirty like contaminated' because one has suffered the gross ill-treatment of a particularly brutal and dehumanising rape directed against that very community is the sort of cumulative denial of human dignity which to my mind is quite capable of amounting to persecution. Of course the treatment feared has to be sufficiently severe, but the severity of its impact upon the individual is increased by the effects of the past persecution. The victim is punished again and again for something which was not only not her fault but was deliberately persecutory of her, her family and her community. Mrs B is fortunate indeed because her husband has stood by her. But Mrs B states that this is seen as a 'big disgrace for a man' and Mr B states that 'according to our culture I should reject her.' The pressure to do so adds to the severity of the ill-treatment they may fear on return.[37]If what they fear is capable of amounting to persecution, is it for a Convention reason? It is certainly capable of being so. In R v Immigration Appeal Tribunal and another, Ex p Shah [1999] 2 AC 629, this House held that women in Pakistan constituted a particular social group, because they shared the common immutable characteristic of gender and were discriminated against as a group in matters of fundamental human rights, from which the State gave them no adequate protection. The fact of current persecution alone is not enough to constitute a social group: a group which is defined by nothing other than that its members are currently being persecuted would not qualify. But women who have been victims of sexual violence in the past are linked by an immutable characteristic which is at once independent of and the cause of their current ill-treatment. They are certainly capable of constituting a particular social group under the Convention.[38]However, it is not suggested that the Kosovan authorities would discriminate against the B family in this way. So the final question is whether the authorities would be able and willing to provide sufficient protection against their ill-treatment at the hands of their own community: see Horvath v Secretary of State for the Home Department [2001] 1 AC 489. This has not been explored in evidence or argument. The most one can say is that it is not easy to protect against this sort of deep-seated prejudice but that in international law there is a clear duty to do so. The Convention on the Elimination of All Forms of Discrimination against Women, article 5(a), requires States parties to take all appropriate measures "to modify the social and cultural patterns of conduct of men and women, with a view to achieving the elimination of prejudices and customary and all other practices which are based on the idea of the inferiority or the superiority of either of the sexes or on stereotyped roles for men and women." The Home Office Immigration and Nationality Directorate API on Gender issues in the asylum claim , para 5, points out that:
"The existence of particular laws or social policies/practices (including traditions and cultural practices) or the manner in which they are implemented may themselves constitute or involve a failure of protection. . . . Women may be subject to gender-related abuse resulting from social customs or conventions because there is no effective means of legal recourse to prevent, investigate or punish such acts."
[39]Understandable concentration on the totality of the abuse which this family had suffered may have obscured the importance of these gender-related issues. The evidence may not have been sufficient to support a claim of the sort I have been discussing, but I regret that the issues were not fully explored at an earlier stage in the proceedings. As, happily, the family have now been given indefinite leave to remain in this country, it is unnecessary to do so now.[40]For these, together with the reasons given by my noble and learned friends, Lord Hope of Craighead and Lord Brown of Eaton-under-Heywood, on the other part of the case, I too would dismiss this appeal. LORD BROWN OF EATON-UNDER-HEYWOOD My Lords,[41]These two appeals raise an important question upon the proper construction and application of the 1951 Convention Relating to the Status of Refugees (Cmd 9171) ("the Convention") as amended by the 1967 Protocol (Cmnd 3906). Assume that a person flees his home country at a time when he has a well-founded fear of being persecuted (and has already been persecuted) for a Convention reason, later arriving in another country where he claims refugee status under the Convention. Assume next that, before his claim has been finally determined, the circumstances in his home country change so that he no longer has a current well-founded fear of persecution; he could, indeed, safely return home. Assume finally, however, that compelling reasons exist arising out of his previous persecution for him not to return home. Is he, in those circumstances, entitled to protection under the Convention? Is he, in other words, when finally his asylum claim comes to be determined, entitled to be treated for all purposes as a refugee—entitled, for example, to have travel documents issued to him under article 28 and to have any naturalisation proceedings facilitated and expedited under article 34—or is he dependant for any further protection upon a favourable exercise of discretion by the host country, there being no continuing Convention obligation owed to him? That crucially is the issue before your Lordships on these appeals.[42]It is convenient at once to set out articles 1A and 1C of the Convention, the provisions most directly bearing upon the points arising. Both are to be found under the general heading of article 1: Definition of the Term "Refugee". Article 1A (2) was later amended by the 1967 Protocol so as to delete the words which, for ease of understanding, I have italicised:
"A For the purposes of the present Convention, the term 'refugee' shall apply to any person who: (1) Has been considered a refugee under the Arrangements of 12 May 1926 and 30 June 1928 or under the Conventions of 28 October 1933 and 10 February 1938, the Protocol of 14 September 1939 or the Constitution of the International Refugee Organization; Decisions of non-eligibility taken by the International Refugee Organization during the period of its activities shall not prevent the status of refugee being accorded to persons who fulfil the conditions of paragraph 2 of this section; (2) As a result of events occurring before 1 January 1951 and 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; 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. . . . C This Convention shall cease to apply to any person falling under the terms of section A if: (1) He has voluntarily re-availed himself of the protection of the country of his nationality; or (2) Having lost his nationality, he has voluntarily re-acquired it; or (3) He has acquired a new nationality, and enjoys the protection of the country of his new nationality; or (4) He has voluntarily re-established himself in the country which he left or outside which he remained owing to fear of persecution; or (5) He can no longer, because the circumstances in connection with which he has been recognized as a refugee have ceased to exist, continue to refuse to avail himself of the protection of the country of his nationality; Provided that this paragraph shall not apply to a refugee falling under section A(1) of this article who is able to invoke compelling reasons arising out of previous persecution for refusing to avail himself of the protection of the country of nationality; (6) Being a person who has no nationality he is, because the circumstances in connection with which he has been recognized, as a refugee have ceased to exist, able to return to the country of his former habitual residence; Provided that this paragraph shall not apply to a refugee falling under section A(1) of this article who is able to invoke compelling reasons arising out of previous persecution for refusing to return to the country of his formed habitual residence."
[43]It is article 1 C (5) which lies at the heart of this case, the appellants' principal argument being that the decision-maker (be it the Secretary of State or, on appeal, the adjudicator or Immigration Appeal Tribunal) must, in determining their asylum claims, give effect to the compelling reasons (for present purposes to be assumed) for not returning them to their home country by granting them refugee status. Although their claimed entitlement to such status arises under section A (2) of article 1—1A (2) as I shall henceforth call it, similarly abbreviating the other provisions—and not, therefore, under 1A (1) to which the proviso expressly refers, and although there has been no prior determination of their asylum claims and so no question of their having "been recognised as" refugees, at least in any formal sense, it is their central contention that they are nevertheless entitled to benefit from the 1C (5) proviso.[44]To succeed on their appeals, the appellants must accordingly make good each of two arguments: first, that they are to be regarded as having been "recognised" as refugees within the meaning of 1C (5) simply by virtue of having at some time past fulfilled the criteria for refugee status under 1A (2); secondly, that on its true interpretation the proviso to 1C (5) applies no less to 1A (2) refugees than to 1A (1) refugees ("statutory refugees", as 1A (1) refugees are generally known).[45]The appellants failed in both arguments before the Court of Appeal (Lord Phillips of Worth Matravers MR, Chadwick and Keene LLJ, the judgment of the court being given by Keene LJ on 14 October 2002: [2003] 1 WLR 241), as they had in the courts below—in Mr Hoxha's case before Jackson J on 24 July 2001, in B's case before Turner J on 15 January 2002.[46]Having now sought to identify the core issue arising, let me turn briefly to the facts of these two cases—very briefly since, for reasons which will be apparent, the outcome of the appeal cannot depend on the facts of any individual case.[47]Both appellants are ethnic Albanians from Kosovo, citizens of the Federal Republic of Yugoslavia. Both suffered gross ill-treatment by the Serbian authorities in the period prior to June 1999 when NATO succeeded in driving the Serb army out of Kosovo and replacing it with international peace-keeping forces from UNMIK and KFOR. Both in fact fled Kosovo before June 1999, Mr Hoxha in November 1998 when he went to Albania (where he had already been from September 1997 to October 1998); B and his wife and two sons in early 1999 when they went to Macedonia. Both later travelled to the United Kingdom, entering this country clandestinely in the backs of lorries, Mr Hoxha in or shortly before June 2000, B and his family on 26 July 1999. Both claimed asylum on arrival, their claims being refused by the Secretary of State respectively on 10 June 2000, and on 12 June 2000. As already noted, their subsequent appeals have consistently failed.[48]The appellants' ill-treatment whilst in Kosovo was appalling. In September 1997 Mr Hoxha was shot three times in the leg whilst trying to protect his father from Serb soldiers and paramilitaries who had forced their way into his house. After treatment in an Albanian hospital and a year's stay in that country he returned to Kosovo where again, in October-November 1998, he was attacked by Serb soldiers who broke his leg with a metal bar. B and his family's treatment was yet worse. In October 1998 Serb police ransacked his house, beating him and then stabbing him with a knife. They then slashed his eight year old son across the stomach and, when his wife intervened, raped her in front of B and their ten year old son, the whole incident being witnessed by 20 or 30 ethnic Albanian neighbours. There is, as a result, a serious risk that if the family returned to Kosovo they would be ostracised by the rest of their community.[49]Besides the central argument for refugee status based on 1C (5), both appellants (although more particularly B) advance a subsidiary argument that in any event, even without resort to 1C (5), they should be recognised as 1A (2) refugees because of their fear, not of further or future persecution, but rather of the continuing effects of past persecution, most notably B and his family's fear of likely ostracism by their community. On this subsidiary point I need say no more than that it cannot succeed for the reasons given by my noble and learned friend, Baroness Hale of Richmond, with whose judgment I entirely agree.[50]I should next briefly note the appellants' present position. Were Mr Hoxha's present appeal to fail he could still apply for discretionary leave to remain and advance any Human Rights Act claim available to him. Subject to that, however, he could be returned to Kosovo. B and his family, by contrast, were granted indefinite leave to remain on 9 March 2004 (well after the Court of Appeal's judgment). This was granted pursuant to a Home Office concession extended generally in late 2003 to families who had applied for asylum before 2 October 2000 and who had remained here since that date with at least one dependent child under the age of 18.[51]Finally before coming to the detailed arguments on 1C (5) it is, I think, helpful to take note of the one previous decision of this House which on any view bears closely on the present question and which could, indeed, be thought to stand squarely in the appellants' path: Adan v Secretary of State for the Home Department [1999] 1 AC 293. Mr Adan had fled Somalia in 1988 owing to a well-founded fear of persecution at the hands of the then government. Having made his way to the United Kingdom he was refused asylum but granted exceptional leave to remain—a leave which safeguarded him against return to Somalia but denied him several benefits attaching to refugee status. By this time Somalia was in the grip of civil war between opposing clans, a conflict which put all sections of society (and would have put Mr Adan on return) at grave but equal risk of death or torture against which no protection was afforded. The risk being common to all, the House held that it did not give rise to a well-founded fear of being "persecuted" within the meaning of 1A (2). That defeated one of Mr Adan's arguments. But it left open another, an argument of obvious present relevance. This was that Mr Adan had no need to show a present fear of persecution; it was enough to show a fear of persecution when he left Somalia—a "historic fear" as it was called—coupled with a present inability to avail himself of his country's protection. This argument too, however, failed.[52]Lord Lloyd of Berwick (with whose speech all the other members of the Appeal Committee agreed) found compelling reasons in the language of 1A (2)—with its constant emphasis on the present tense—for holding that it could only be satisfied by proof of a current well-founded fear of persecution. He then turned (p306) to 1C (5):
"I had at first thought that article 1C (5) provided a complete answer to [Mr Adan's] argument. If a present fear of persecution is an essential condition of remaining a refugee, it must also be an essential condition for becoming a refugee. But it was pointed out in the course of argument that article 1C (5) only applies to [nationals who are outside their country owing to a well-founded fear of persecution for a Convention reason, and, owing to such fear, are unwilling to avail themselves of the protection of their country]. It does not help directly as to [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]. This is true. But the proviso does shed at least some light on the intended contrast between article 1A (1) and 1A (2). Article 1A (1) is concerned with historic persecutions. It covers those who qualified as refugees under previous Conventions. They are not affected by article 1C (5) if they can show compelling reasons arising out of previous persecution for refusing to avail themselves of the protection of their country. It would point the contrast with article 1A (1), and make good sense, to hold that article 1A (2) is concerned, not with previous persecution at all, but with current persecution, in which case article 1C (5) would take effect naturally when, owing to a change of circumstance, the refugee ceases to have a fear of current persecution."
[53]Finally, with regard to this argument Lord Lloyd (p308) said this:
"I am glad to have reached that conclusion. For a test which required one to look at historic fear, and then ask whether that historic fear which, ex hypothesi, no longer exists is nevertheless the cause of the asylum-seeker being presently outside his country is a test which would not be easy to apply in practice. This is not to say that historic fear may not be relevant. It may well provide evidence to establish present fear. But it is the existence, or otherwise, of present fear which is determinative."
[54]Lord Slynn of Hadley said (p301) that there seemed to him:
"force in [Mr Adan's] argument that on humanitarian grounds a person who leaves his own country because of a well-founded fear of being persecuted for a Convention reason and later is unable, or, owing to that fear, is unwilling, to avail himself of that country's protection even when the grounds for his fear have gone, should be able to claim the status of a refugee."
He concluded however:
"[T]he coherence of the scheme requires that the well-founded fear, the first matter to be established, is also a current fear. The existence of what has been called a historic fear is not sufficient in itself, though it may constitute important evidence to justify a claim of a current well-founded fear."
[55]In the light of that authority (from which your Lordships are not in any way invited to depart), sympathetic though inevitably one is to these appellants and persuasive though for my part I acknowledge their humanitarian case to be, their legal arguments appear distinctly unpromising.[56]Their case comes to this. To qualify for refugee status they have to satisfy the requirements of 1A (2). This they seek to do—in the face of Adan's requirement that they demonstrate a current well-founded fear—by resort to a cessation provision, 1C (5). 1C (5), quite apart from appearing to apply not when first an asylum seeker's refugee status is determined but only in connection with its possible later loss, in any event appears not to solve but to compound the appellants' difficulties, expressly postulating as it does that the circumstances earlier giving rise to refugee status "have ceased to exist" i.e. that by now they no longer have a well-founded fear. To escape this further difficulty, however, the appellants seek to invoke the "compelling reasons" proviso notwithstanding its apparent limitation to 1A(1) refugees. Putting it another way, the appellants seek by way of the proviso to disapply a cessation provision which, were it to apply, would itself take effect not to confer on them but rather to deny them refugee protection ("This Convention shall cease to apply"). Quite how the disapplication of a provision itself otherwise disapplying the Convention can assist an asylum seeker to qualify for Convention protection in the first place is not altogether easy to understand. Plainly, moreover, the argument is irreconcilable with the passage already cited from Lord Lloyd's speech in Adan [1999] 1 AC 293, 306, where he points to the contrast logically and intentionally struck in 1C (5) between on the one hand 1A (1) refugees, who have already been "considered" refugees (and thus recognised as such) and who, although potentially amenable to the loss of that status under 1C (5), will not in fact lose it if they can show "compelling reasons", and on the other hand 1A(2) refugees who must demonstrate a current well-founded fear of persecution not only when first seeking recognition of their status but also thereafter in order not to lose it.[57]Before turning to some of the many texts put before the House in connection with the true construction of 1C (5), it is convenient next, whilst still laying the ground for your Lordships' consideration of both the questions arising—what is meant by "recognized" and the reach of the proviso given its reference to 1A (1)—to set out the more directly relevant paragraphs of the UNHCR 1979 Handbook, issued, as its foreword (para iv) states, pursuant to a request from the Executive Committee in 1977:
"for the guidance of Governments a handbook relating to procedures and criteria for determining refugee status"
. Noteworthy amongst the Committee's express recommendations was:
"v. If the applicant is recognised as a refugee, he should be informed accordingly and issued with documentation certifying his refugee status."
(emphasis added)[58]The foreword to the Handbook states (para ii) that the Convention and the 1967 Protocol apply "to persons who are refugees as therein defined" and continues:
"The assessment as to who is a refugee, i.e. the determination of refugee status under the 1951 Convention and the 1967 Protocol, is incumbent upon the Contracting State in whose territory the refugee applies for recognition of refugee status."
(emphasis added)[59]The foreword was revised in 1992; the body of the Handbook, however, remained as initially published in 1979. It includes the following paragraphs:"26. With respect to the treatment within the territory of States, this is regulated as regards refugees by the main provisions of the 1951 Convention and 1967 Protocol … Furthermore, attention should be drawn to Recommendation E contained in the Final Act of the conference of Plenipotentiaries which adopted the 1951 Convention: 'The Conference expresses the hope that the Convention relating to the Status of Refugees will have value as an example exceeding its contractual scope and that all nations will be guided by it in granting so far as possible to persons in their territory as refugees and who would not be covered by the terms of the Convention, the treatment for which it provides.'27. This recommendation enables States to solve such problems as may arise with regard to persons who are not regarded as fully satisfying the criteria of the definition of the term 'refugee'.28. A person is a refugee within the meaning of the 1951 Convention as soon as he fulfils the criteria contained in the definition. This would necessarily occur prior to the time at which his refugee status is formally determined. Recognition of his refugee status does not therefore make him a refugee but declares him to be one. He does not become a refugee because of recognition, but is recognized because he is a refugee. 30. The provisions of the 1951 Convention defining who is a refugee consist of three parts, which have been termed respectively 'inclusion,' 'cessation' and 'exclusion' clauses. 31. The inclusion clauses define the criteria that a person must satisfy in order to be a refugee. They form the positive basis upon which the determination of refugee status is made. 112. [Under the heading: Cessation clauses] Once a person's status as a refugee has been determined, it is maintained unless he comes within the terms of one of the cessation clauses. This strict approach towards the determination of refugee status results from the need to provide refugees with the assurance that their status will not be subject to constant review in the light of temporary changes - not of a fundamental character - in the situation prevailing in their country of origin. 135. Circumstances [in 1C (5)] refer to fundamental changes in the country, which can be assumed to remove the basis of the fear of persecution. A mere - possibly transitory - change in the facts surrounding the individual refugee's fear, which does not entail such major changes of circumstances, is not sufficient to make this clause applicable. A refugee's status should not in principle be subject to frequent review to the detriment of his sense of security, which international protection is intended to provide. 136. The second paragraph of this clause contains an exception to the cessation provision contained in the first paragraph. It deals with the special situation wh[60]True it is that 1C (5), no less than 1A (2), appears in the Convention under the heading "Definition of the Term 'Refugee'". True it is, too, as para 28 of the Handbook neatly points out, that someone recognised to be a refugee must by definition have been one before his refugee status has been determined. But it by no means follows that, because someone has been a refugee before his status comes to be determined, any change in circumstances in his home country falls to be considered under 1C (5) rather than under 1A (2). Quite the contrary. As has been seen, the Handbook is replete with references to the "determination" of a person's refugee status and his "recognition" as such. Article 9 of the Convention itself, indeed, allows certain provisional measures to be taken "pending a determination by the Contracting State that that person is in fact a refugee". The whole scheme of the Convention points irresistibly towards a two-stage rather than composite approach to 1A (2) and 1C (5). Stage 1, the formal determination of an asylum-seeker's refugee status, dictates whether a 1A (2) applicant (who may, indeed, be someone previously held not to qualify as a statutory refugee by the International Refugee Organisation—see the second paragraph of 1A (1)), is to be recognised as a refugee. 1C (5), a cessation clause, simply has no application at that stage, indeed no application at any stage unless and until it is invoked by the State against the refugee in order to deprive him of the refugee status previously accorded to him.[61]Para 112 of the Handbook makes all this perfectly plain. So too, more recently, did the UNHCR Lisbon Roundtable Meeting of Experts held in May 2001 in their Summary Conclusions:
"26. In principle, refugee status determination and cessation procedures should be seen as separate and distinct processes, and which should not be confused."
[62]Many other of the documents and writings put before your Lordships point the same way. And so, of course, does the language of 1C (5) itself. The words "the circumstances in connection with which he has been recognised as a refugee" could hardly be clearer. They expressly postulate that the person concerned "has been recognised as a refugee", not that he "became" or "was" a refugee.[63]This provision, it shall be borne in mind, is one calculated, if invoked, to redound to the refugee's disadvantage, not his benefit. Small wonder, therefore, that all the emphasis in paras 112 and 135 of the Handbook is upon the importance of ensuring that his recognised refugee status will not be taken from him save upon a fundamental change of circumstances in his home country. As the Lisbon Conference put it in para 27 of their conclusions: " … the asylum authorities should bear the burden of proof that such changes are indeed fundamental and durable".[64]Many other UNHCR publications are to similar effect. A single further instance will suffice, taken from the April 1999 Guidelines on the application of the cessation clauses:
"2. The cessation clauses set out the only situations in which refugee status properly and legitimately granted comes to an end. This means that once an individual is determined to be a refugee, his/her status is maintained until he/she falls within the terms of one of the cessation clauses. This strict approach is important since refugees should not be subjected to constant review of their refugee status. In addition, since the application of the cessation clauses in effect operates as a formal loss of refugee status, a restrictive and well-balanced approach should be adopted in their interpretation."
[65]The reason for applying a "strict" and "restrictive" approach to the cessation clauses in general and 1C (5) in particular is surely plain. Once an asylum application has been formally determined and refugee status officially granted, with all the benefits both under the Convention and under national law which that carries with it, the refugee has the assurance of a secure future in the host country and a legitimate expectation that he will not henceforth be stripped of this save for demonstrably good and sufficient reason. That assurance and expectation simply does not arise in the earlier period whilst the refugee's claim for asylum is under consideration and before it is granted. Logically, therefore, the approach to the grant of refugee status under 1A (2) does not precisely mirror the approach to its prospective subsequent withdrawal under 1C (5).[66]That said, however, it would seem to me appropriate that in the initial determination of an asylum claim under 1A (2) the decision-maker, in a case where plainly the applicant fled his home country as a genuine refugee from Convention persecution, should not too readily reach the view that he could now safely be returned to it. Not only, as both Lord Slynn and Lord Lloyd observed in Adan [1999] 1 AC 293, may historic fear constitute important evidence tending to establish a current fear; so too it justifies some scepticism on the part of the decision-maker as to whether in truth the change in home circumstances is sufficiently clear and firm as to warrant the refusal of refugee status. That essentially is the point I was trying to make in the Court of Appeal in Mohammed Arif v Secretary of State for the Home Department [1999] Imm AR 271 where, at p 276, I suggested that, depending always on the particular facts of the case, there might well be "an evidential burden on the Secretary of State to establish that [the asylum seeker] could safely be returned home." Although "some reservations as to the utility of the language of burden of proof" were expressed in the later Court of Appeal decision in S v Secretary of State for the Home Department [2002] INLR 416,431, I remain unrepentant. It seems to me only right that in a case where the Secretary of State is contending that a country once plainly unsafe (like, say, Sri Lanka or Kosovo) has now become safe, he should place before the appellate authority sufficient material to satisfy them of that critical fact. There can, of course, be no doubt that the Arif approach was satisfied here. As Jackson J observed in Mr Hoxha's case: "In cases arising from Kosovo the Secretary of State can discharge that evidential burden by pointing to the presence of UNMIK and KFOR since June 1999". (June 1999, your Lordships may note, was before either appellant in fact arrived in this country.)[67]From all this it follows that, even were the proviso to 1C (5) capable of availing the appellants, their appeals must fail since clearly 1C (5) has no application. Recognising, however, that were the proviso indeed to encompass them as well as statutory refugees this would involve a substantially greater (and in this event rather disturbing) mismatch between the approach respectively to recognition under 1A (2) and cessation under 1C (5), it is right for your Lordships to address the point.[68]The appellants advance two principal arguments with regard to the proviso, pointing first to its drafting history and secondly to current state practice. As to the drafting history there can be no doubt whatever that, when first the Convention took effect, the proviso was intended to apply, as indeed it expressly states, to statutory refugees only. As Grahl-Madsen said in The Status of Refugees in International Law , vol 1 (1966) p410:
"What the drafters of the Convention had in mind was the situation of refugees from Germany and Austria who were unwilling to return to the scene of the atrocities which they and their kin had experienced or to avail themselves of the protection of a country which had treated them so badly."
[69]It is the appellants' submission, however, that the 1967 Protocol, recognising as it did the need thenceforth to deal also with new refugee situations and not just those which had arisen before 1 January 1951, must be recognised as having changed the scope of the proviso notwithstanding the Protocol's failure actually to amend it. The appellants point to the Preamble to the Protocol—"it is desirable that equal status should be enjoyed by all refugees covered by the definition in the Convention irrespective of the dateline, 1 January 1951"—and urge too the great importance accorded generally in international law to the principle of non-discrimination. To my mind, however, the argument is an impossible one. The language of the proviso is clear and unambiguous. The failure in 1967 to amend it, as it could so easily have been amended simply by deleting the "(1)" from the reference to "section A (1)", seems to me eloquent of the continuing intention of the contracting parties to confine the benefit of the proviso to the diminishing number of statutory refugees who would otherwise have remained vulnerable to the loss of their Convention rights under the cessation clause. (It may be noted for good measure that the Protocol, by article 1 (3), expressly preserved certain geographical limitations on the Convention's scope, namely any declarations previously made by Contracting States under article 1 B limiting their Convention obligations to those claiming refugee status under 1A (2) from "events occurring in Europe".)[70]Professor Guy Goodwin-Gill, having pointed out that the UNHCR statute on the Convention itself dealt differently with the position, describes the Convention's limitation of the right to invoke compelling reasons for non-return to statutory refugees as "perverse" ( The Refugee in International Law , 2nd ed (1996), p 87). To my mind, however, that puts it too high: one can surely understand why some Contracting States at least would wish to limit the proviso's scope. There is, indeed, a striking modern example of precisely this evident reluctance to extend the range of Convention protection. Despite the EU Commission's 2001 proposal in connection with the then proposed Council Directive on the Minimum Standards for Refugee Protection that any "Member State invoking this cessation clause should ensure that an appropriate status, preserving previously acquired rights, is granted to persons who are unwilling to leave the country for compelling reasons arising out of previous persecution or experiences of serious and unjustified harm", the Directive as enacted—Council Directive 2004/83/EC—conspicuously omitted any such provision. Instead, all that article 11.2 requires with regard to the relevant cessation clauses is that "Member States shall have regard to whether the change of circumstances is of such a significant and non-temporary nature that the refugee's fear of persecution can no longer be regarded as well-founded."[71]I conclude, therefore, that the appellants' argument derives no support whatever from the drafting history of the proviso. The 1967 Protocol, indeed, so far from advancing their case, constitutes yet another obstacle in their path.[72]I turn therefore to the second limb of the appellant's argument, Mr Manjit Gill QC's reliance on state practice. The argument arises under article 31 of the Vienna Convention on the Law of Treaties, 1969:
"1. A treaty shall be interpreted in good faith in accordance with the ordinary meaning to be given to the terms of the treaty in their context and in the light of its object and purpose. 3. There shall be taken into account, together with the context: (a) any subsequent agreement between the parties regarding the interpretation of the treaty or the application of its provisions; (b) any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation."
[73]Sir Ian Sinclair QC, in The Vienna Convention on the Law of Treaties , 2nd ed, (1984) says at p138: "It should of course be stressed that para 3 (b) of article 31 of the Convention does not cover subsequent practice in general, but only a specific form of subsequent practice—that is to say, concordant subsequent practice common to all the parties".[74]Anthony Aust in Modern Treaty Law and Practice (2000) states, at p 195:
"It is not necessary to show that each party has engaged in a practice, only that all have accepted it, albeit tacitly. But, if a clear difference of opinion between the parties exists, the practice may not be relied upon as a supplementary means of interpretation."
[75]That, submit the appellants, with some support from Professor Goodwin Gill, puts it too high: universal agreement by all the treaty parties is not necessary; a state can be bound by the general practice of other states even against its wishes.[76]For the purposes of the present appeals it seems to me unnecessary to resolve the apparent differences between the several commentators upon the degree of uniformity of approach necessary in the case of multi-lateral treaties to support an interpretation based on state practice. Having regard to the clarity of the "ordinary meaning" born by the proviso to 1C (5), only the most compelling case founded on "subsequent practice" could properly give rise to a different and apparently contradictory interpretation from that obviously first intended. Until very recently, as we shall see, none of the many UNHCR pronouncements down the years appeared to support such a case. So far from suggesting that, as a result of state practice, contracting states had become obliged to treat the proviso as extending to all refugees, one repeatedly finds instead the language of aspiration and exhortation. Paragraph 26 of the Handbook expresses "the hope" that nations will grant refugee protection beyond the Convention's "contractual scope". Paragraph 27 characterises this as a "recommendation". Paragraph 136 points out encouragingly that the proviso, reflecting as it does a general humanitarian principle, "could" also be applied to A1 (2) refugees. Executive Committee Conclusion 69 of 1992:
"(e) recommends, so as to avoid hardship cases, that states seriously consider an appropriate status, preserving previously acquired rights, for persons who have compelling reasons arising out of previous persecution for refusing to re-avail themselves of the protection of their country …"
[77]Paragraph 31 of the UNHCR guidelines of 1999 states:
"Formally speaking, this provision [the proviso] applies only to a very small group of refugees in the present day context. However, there is nothing to prevent it being applied on humanitarian grounds to other than statutory refugees."
[78]A change of view, however, appears to have overtaken the UNHCR at around the time of the Lisbon Conference in 2001. Para 18 of the Roundtable's conclusions states:
"Application of the 'compelling reasons' exception to general cessation contained in article 1C (5)-(6) is interpreted to extend beyond the actual words of the provision and is recognised to apply to article 1A (2) refugees. This reflects the general humanitarian principle that is now well-grounded in state practice."
This statement was repeated verbatim in guidelines issued by the UNHCR in February 2003—see para 21. This was supported solely by a footnoted reference to a forthcoming UNHCR publication: Refugee Protection in International Law edited by Feller, Türk and Nicholson . At p 32 of that publication (which came out later in 2003) appears this:
"Another issue of contemporary concern is the question of exceptions to any general declaration of cessation. One exception is that on the basis of 'compelling reasons arising out of previous persecution' as referred to in article 1 C (5) and (6). This is now well established in state practice as extending beyond the actual terms of this provision to apply to refugees under article 1A (2) of the 1951 Convention. In such circumstances, the best state practice in keeping with the spirit of the Convention allows for the continuation of refugee status, although states sometimes accord such individuals subsidiary statuses, which may not necessarily provide a secure legal status or preserve 'previously acquired rights' as stipulated by the Executive Committee [a reference to that Committee's Conclusion No. 69 (e)—see para 76 above]."
[79]Those 2003 publications, it may be noted, post-date the Court of Appeal's judgment in this case. In turn they were followed by a research paper, "The Applicability of the 'Compelling Reasons' exception to cessation for refugees and asylum-seekers" published under the auspices of UNHCR as recently as November 2004. The findings of that paper were summarised as follows:
"Although current international practice regarding article 1C (5) sentence 2 is inconsistent, the research confirms that the application of the 'compelling reasons' exception reflects a general humanitarian principle that is well-grounded in state practice. Some states grant either refugee status or complementary (subsidiary) protection by interpreting article 1C (5) sentence 2 as a humanitarian principle applicable to Convention refugees (article 1A (2)) and asylum-seekers, while others deny the application of article 1C (5) sentence 2 to Convention refugees. According to this research, it appears that a greater number of states apply the former interpretation and grant either refugee status or a subsidiary form of protection to refugees or asylum-seekers invoking the 'compelling reasons' exception. Indeed, out of 20 countries studied, 15 apply article 1C (5) sentence 2 to article 1A (2) refugees, and interpret it as a humanitarian principle or by analogy. It is worth mentioning that amongst these countries, are significant countries of asylum such as Belgium, Canada, France, Germany, South Africa, Switzerland, The Netherlands and The United States. The countries that apply a literal interpretation of article 1C (5) sentence 2, thus limiting its application to statutory refugees, are notably Australia, Poland and the United Kingdom and, in a recent change of position, New Zealand."
[80]In the International Journal of Refugee Law published in January 2004, Vol 16, No 1, p 92 in an article "Exemption from Cessation of Refugee Status in the Second Sentence of article 1C(5)/(6) of the 1951 Refugee Convention", David Milner had expressed the view that "whilst it is clearly the case that extension of the exemption to Convention refugees is indeed well established, it is harder to argue that it has 'generally been accepted', and the practice is certainly far from being universal." The November 2004 subsequent research document refers to Mr Milner's article but suggests that "despite [its] recent date of publication … some of the information regarding national legislation and case law is outdated."[81]It is upon these recent UNHCR publications and above all the 2004 research findings that Mr Manjit Gill founds his main argument on this limb of the case. They invalidate, he submits, an important finding in the judgment below, [2003] 1 WLR 241, 254, para 47:
"A number of states do adopt a more generous approach towards article 1C (5) than is required by the terms of the Convention itself, but they represent on the evidence before us a minority of the signatories to the Convention, who number over 100."
The research is also relied on by Mr Tim Eicke in his written case on behalf of the UNHCR as Intervener. It shows, he submits (at para 10.12):
"that the vast majority of states parties accept this need [to give effect to the general humanitarian principles underlying the proviso in relation to all refugees] and make provision for it in their national law (either through legislation or through developing jurisprudence)."
[82]For my part I would reject these submissions. No doubt the Court of Appeal's judgment can now be seen to have put the appellants' case too low. But the Intervener to my mind puts it too high. I do not regard 15 out of 20 as a "vast majority" and nor do I think it capable of overriding the obvious disagreement of countries as important to the asylum system as Australia, New Zealand and the UK. Nor, indeed, is the appellants' difficulty in establishing the necessary consistency of state practice solely one of numbers. It is noteworthy too that both Feller, Türk and Nicholson and the subsequent summary of research findings describe the grant or continuation of refugee status merely as "best state practice", acknowledging in terms that a number of other states (even amongst the majority who give general effect to the "compelling reasons" exception) do so only by according "subsidiary statuses" to those they let stay. These grants of permission do not necessarily provide a secure legal status or preserve acquired rights. This "subsidiary form of protection," as the research paper calls it, manifestly falls short of continuing refugee status and is inconsistent therefore with a strict Convention obligation to treat A1 (2) refugees in the same way as A1 (1) refugees for the purposes of the 1C (5) proviso. On this ground too, therefore, the state practice argument is in my opinion unsustainable.[83]As Lord Bingham of Cornhill observed in Sepet v Secretary of State for the Home Department [2003] 1 WLR 856, 864, para 11, with regard to a claimed right of conscientious objection to compulsory military service as the basis for the grant of asylum in the light of the many expressions of international opinion supporting that view:
"But resolutions and recommendations of this kind, however sympathetic one may be towards their motivation and purpose, cannot themselves establish a legal rule binding in international law."
[84]From paragraph 72 onwards I have been discussing state practice with regard to the proviso for all the world as if the UK's practice is to invoke 1C (5) irrespective of any "compelling reasons" not to do so. In fact, as Miss Carss-Frisk QC told us, there is only one recorded case of the Secretary of State ever invoking 1C (5) to remove a refugee and the scope of the proviso was not there in point. The reality appears to be that, once the UK grants refugee status, it does not subsequently withdraw it however safe the home country may become. In truth, therefore, for the appellants to succeed in these appeals, they would have to establish that modern state practice requires not merely that refugee status be not withdrawn in "compelling reasons" cases but that it be granted in such cases, a proposition which, as already explained, flies directly in the face of Adan and is in any event irreconcilable with the UNHCR's repeated emphasis on treating the initial determination of refugee status and its subsequent cessation as "separate and distinct processes".[85]It is one thing to invite this House to construe the Convention as a living instrument generously and in the light of its underlying humanitarian purposes; quite another to urge your Lordships effectively to rewrite it so as to create a fresh entitlement to refugee status based upon no more than historic fear and present compelling reasons for non-return, with no need at all for any current fear of persecution. That would be to distort entirely the language and structure of the text and in my judgment do a serious disservice to the cause of human rights generally. As Lord Bingham of Cornhill said, first in Brown v Stott [2003] 1 AC 681, 703, and then again in R (European Roma Rights) v Prague Immigration Officer [2005] 2 WLR 1, para 18:
"'[I]t is generally to be assumed that the parties have included the terms which they wished to include and on which they were able to agree, omitting other terms which they did not wish to include or on which they were not able to agree,' and caution is needed 'if the risk is to be averted that the contracting parties may, by judicial interpretation, become bound by obligations which they did not expressly accept and might not have been willing to accept.'"
Lord Steyn too recognised in the European Roma Rights case (at para 43), the limitations upon the "very important principles of interpretation" set out in the Vienna Convention:
" … they are not capable of filling gaps which were designedly left in the protective scope of the Refugee Convention."
[86]The Convention does not meet all humanitarian needs. It only avails those unable to return to their home country who have a present fear of persecution ( Adan [1999] 1 AC 293). It does not avail conscientious objectors ( Sepet [2003] 1 WLR 856). It cannot be invoked by those who have yet to leave their home country ( European Roma Rights ). It does not protect those whose life is threatened by famine, or civil war, or by persecution for non-Convention reasons. As Lord Hope of Craighead observed in Horvath v Secretary of State for the Home Department [2001] 1 AC 489, 498:
"It is important to note throughout that the humanitarian purposes of the Convention are limited by the tests set out in the article [1A (2)]."
[87]However compelling therefore may be the reasons why these appellants or others like them should not be returned to their home country, there is no basis in the Convention, absent a continuing well-founded fear, for granting them refugee status (or even, indeed, for allowing them to retain such status once granted). That, however, is not to deny such refugees all possibility of relief. Rather, as paragraph 72 of the respondent's written case makes plain:
"The Secretary of State accepts that there will be cases where an individual may be particularly vulnerable by reason of the continuing physical or psychological effects of persecution he has suffered in the past. Such problems may render him less able to cope with difficult conditions in his country of origin. It is open to such an applicant to apply for discretionary leave to remain in the United Kingdom on compassionate grounds."
[88]A refusal of discretionary leave on compassionate grounds might itself, in an extreme case, be judicially reviewable. No such challenge is, however, mounted here. These appellants' claims stand or fall on their entitlement to refugee status. For the reasons given, reasons substantially the same as those appearing in Keene LJ's admirable judgment below, I would hold that their claims fail. I would accordingly dismiss these appeals.

Cited in 38 later judgments

and 29 more of the most senior