Canada (Attorney General) v. Ward, 1993 CanLII 105 (SCC), [1993] 2 SCR 689

Canada (Attorney General) v. Ward, [1993] 2 S.C.R. 689

Patrick Francis Ward Appellant

v.

The Attorney General of Canada Respondent

and

United Nations High Commissioner for Refugees, Immigration and Refugee Board and Canadian Council for Refugees Interveners

Indexed as: Canada (Attorney General) v. Ward

File No.: 21937.

1992: March 25; 1993: June 30.

Present: La Forest, L'Heureux‑Dubé, Gonthier, Stevenson* and Iacobucci JJ.

on appeal from the federal court of appeal

Immigration ‑‑ Refugee status ‑‑ "Particular social group" ‑‑ Political opinion ‑‑ "Well‑founded fear of persecution" necessary to establishment of claim to Convention refugee status ‑‑ Claimant a former member of Irish terrorist organization sentenced to death by organization for complicity in assisting escape of hostages ‑‑ Claimant citizen of Ireland and of United Kingdom ‑‑ Whether state complicity requirement for persecution ‑‑ Whether terrorist organization a "particular social group" ‑‑ Whether dissention from politico‑military organization basis for persecution for political opinion ‑‑ Whether s. 15 of Charter applicable to definition of Convention refugee ‑‑ Burden of proof of want of protection of each country of nationality ‑‑ Canadian Charter of Rights and Freedoms, s. 15 ‑‑ Immigration Act, 1976, S.C. 1976‑77, c. 52, ss.2(1), 4(2.1), 19(1)(c), (d), (e), (f), (g), (2), 46.04(1)(c).

Appellant was a resident of Northern Ireland. Motivated by a perceived need to "take a stand" in order to protect his family, mainly from the IRA, he voluntarily joined the INLA, a para‑military terrorist group dedicated to the political union of Ulster and the Irish Republic. Appellant, who had been detailed to guard innocent hostages, secured their escape when he learned that they were to be executed. This action was motivated by his conscience.

The police eventually let slip to an INLA member that one of their own had assisted the escape. The INLA, who had suspected appellant, confined and tortured him and sentenced him to death following a court‑martial by a kangaroo court. Appellant escaped from the INLA, sought police protection and was charged for his part in the hostage incident. The INLA, in a pre‑emptive move to prevent appellant's providing evidence to the police about INLA members and their activities, took his wife and children hostage.

Appellant pleaded guilty to the offence of forcible confinement and was sentenced to three years in jail. He did not give evidence against the INLA and never admitted publicly to having released the hostages. Towards the end of his prison sentence, appellant sought the assistance of the prison chaplain for protection from INLA members. The chaplain, with the assistance of police, obtained a Republic of Ireland passport for appellant and airline tickets to Canada.

Appellant arrived in Toronto in December 1985 and sought admission to Canada as a visitor. He became the subject of an inquiry in May, 1986, and claimed Convention refugee status citing a fear of persecution because of his membership in a particular social group (the INLA). The Minister of Employment and Immigration determined that appellant was not a Convention refugee and, as a result, appellant filed an application for redetermination of his claim before the Immigration Appeal Board. The Board allowed the redetermination and found appellant to be a Convention refugee. The Federal Court of Appeal granted the Attorney General of Canada's application under s. 28 of the Federal Court Act to set aside the decision and referred the matter back to the Board for reconsideration.

At issue before this Court were: (1) whether the element of state complicity is required to establish a refugee claim and the nature of the "unwillingness" or "inability" of a claimant to seek the protection of his or her home state; (2) the meaning of "particular social group"; (3) the nature of persecution for political opinion and whether desertion from a politico‑military organization for reasons of conscience may properly ground a claim based on that ground; (4) whether s. 15 of the Charter was applicable; and (5) in cases of multiple nationality, whether the claimant must establish want of protection in all states of citizenship.

Held: The appeal should be allowed.

International refugee law was formulated to serve as a back‑up to the protection owed a national by his or her state. It was meant to come into play only in situations where 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.

"Persecution" includes situations where the state is not in strictness an accomplice to the persecution but is simply unable to protect its citizens. The dichotomy between "unable" and "unwilling" has become somewhat blurred. The inquiry as to whether a claimant meets the "Convention refugee" definition must focus on whether there is a "well‑founded fear", which the claimant must first establish, and all that follows must be "by reason of" that fear. Two categories, both requiring the claimant to be outside his or her state of nationality by reason of that fear, exist. The first requires that the claimant be unable to avail him‑ or herself of that state's protection. It originally related only to stateless persons, but can now include those refused passports or other protections by their state of nationality. The second requires that the claimant be unwilling to avail him‑ or herself of his or her state's protection by reason of that fear. Neither category of the "Convention refugee" definition, however, requires that the state have been involved in the persecution.

The test as to whether a state is unable to protect a national is bipartite: (1) the claimant must subjectively fear persecution; and (2) this fear must be well‑founded in an objective sense. The claimant need not literally approach the state unless it is objectively unreasonable for him or her not to have sought the protection of the home authorities. The Board, if the claimant's fear has been established, is entitled to presume that persecution will be likely and that the fear is well‑founded if there is an absence of state protection. The presumption goes to the heart of the inquiry, which is whether there is a likelihood of persecution. The persecution must be real ‑‑ the presumption cannot be built on fictional events ‑‑ but the well‑foundedness of the fears can be established through the use of such a presumption.

The presumption was of some importance to the Board in this case. It found that the appellant was a credible witness and therefore accepted that he had a legitimate fear of persecution. Since Ireland's inability to protect was established through evidence that state agents had admitted their ineffectiveness, the Board was then able to presume the well‑foundedness of appellant's fears.

The claimant must provide clear and convincing confirmation of a state's inability to protect absent an admission by the national's state of its inability to protect that national. Except in situations of complete breakdown of the state apparatus, it should be assumed that the state is capable of protecting a claimant. This presumption, while it increases the burden on the claimant, does not render illusory Canada's provision of a haven for refugees. It reinforces the underlying rationale of international protection as a surrogate, coming into play where no alternative remains to the claimant.

In distilling the contents of the head of "particular social group", account should be taken of the general underlying themes of the defence of human rights and anti‑discrimination that form the basis for the international refugee protection initiative. A good working rule for the meaning of "particular social group" provides that this basis of persecution consists of three categories: (1) groups defined by an innate, unchangeable characteristic; (2) groups whose members voluntarily associate for reasons so fundamental to their human dignity that they should not be forced to forsake the association; and (3) groups associated by a former voluntary status, unalterable due to its historical permanence.

Exclusions on the basis of criminality have been carefully drafted in the Immigration Act to avoid the admission of claimants who may pose a threat to the Canadian government or to the lives or property of the residents of Canada. These provisions specifically give the Minister of Employment and Immigration enough flexibility to reassess the desirability of permitting entry to a claimant with a past criminal record, where the Minister is convinced that rehabilitation has occurred. This demonstrates that Parliament has not opted to treat a criminal past as a reason to be estopped from obtaining refugee status. The scope of the term "particular social group" accordingly did not need to be interpreted narrowly to accommodate morality and criminality concerns. Such a blanket exclusion is more appropriately to be avoided in the face of an explicit, comprehensive structure for the assessment of these potentially inadmissible claimants.

Appellant did not meet the definition of "Convention refugee" with respect to his fear of persecution at the hands of the INLA upon his return to Northern Ireland. The group of INLA members is not a "particular social group". Its membership is neither characterized by an innate characteristic nor is it an unchangeable historical fact. Its objective of obtaining specific political goals by any means, including violence, cannot be said to be so fundamental to the human dignity of its members that it constitutes a "particular social group". In any event, appellant's fear was not based on his membership. Rather, he felt threatened because of what he did as an individual. His membership in the INLA placed him in the circumstances that led to his fear, but the fear itself was based on his action, not on his affiliation.

A claimant is not required to identify the reasons for the persecution. The examiner must decide whether the Convention definition is met; usually there will be more than one applicable ground.

Political opinion can generally be interpreted to be any opinion on any matter in which the machinery of state, government, and policy may be engaged. The political opinion at issue need not have been expressed outright. Often the claimant is not even given the opportunity to articulate his or her beliefs; often they are imputed to the claimant from his or her actions. The political opinion ascribed to the claimant and for which he or she fears persecution need not necessarily conform to the claimant's true beliefs. The examination of the circumstances should be approached from the perspective of the persecutor, since that is the perspective that is determinative in inciting the persecution. Similar considerations apply to other bases of persecution.

Appellant's fear of being killed by the INLA, should he return to Northern Ireland, stemmed initially from the group's threat of executing the death sentence imposed by its court‑martial. The act for which appellant was so punished was his assistance in the escape of the hostages he was guarding. From this act, a political opinion related to the proper limits to means used for the achievement of political change can be imputed. To appellant, who believed that the killing of innocent people to achieve political change is unacceptable, setting the hostages free was the only option that accorded with his conscience. The persecution appellant fears stemmed from his political opinion as manifested by this act.

Given that the relevant aspects of the majority decision were found to be incorrect for other reasons, recourse to s. 15 of the Charter with respect to "particular social group" and state complicity was unnecessary.

Appellant conceded dual nationality ‑‑ Irish and British. The burden of proof, including a showing of well‑founded fear of persecution in all countries of which the claimant is a national, lies with appellant and not the Minister.

The Board must investigate whether the claimant is unable or unwilling to avail him‑ or herself of the protection of each and every country of nationality. Any home state protection is a claimant's sole option when available since international refugee protection is to serve as "surrogate" shelter coming into play only upon failure of national support. The inability of a state of nationality to protect can be established where the claimant has actually approached the state and been denied protection. Where, as in the case of appellant, the second state has not actually been approached by the claimant, that state should be presumed capable of protecting its nationals. An underlying premise of this presumption is that citizenship carries with it certain basic consequences, such as the right to gain entry to the country at any time. Denial of admittance to the home territory can amount to a refusal of protection. Here, evidence, albeit not expert opinion, was led to establish that British legislation enabled the British Government to prohibit a national from being in, or entering, Great Britain, if the national had been connected with terrorism with regard to Northern Ireland. The applicability of this presumption and its rebuttal depended on the particular circumstances of this case and was to be determined by the Board.

Cases Cited

Considered: Rajudeen v. Minister of Employment and Immigration (1984), 55 N.R. 129; Surujpal v. Minister of Employment and Immigration (1985), 60 N.R. 73; Zalzali v. Canada (Minister of Employment and Immigration), 1991 CanLII 13557 (FCA), [1991] 3 F.C. 605; McMullen v. Immigration and Naturalization Service, 658 F.2d 1312 (1981); Cheung v. Minister of Employment and Immigration, 1993 CanLII 2946 (FCA), [1993] F.C.J. No. 309 (Q.L.), Appeal No. A‑785‑91; Mayers v. Canada (Minister of Employment and Immigration) (1992), 1992 CanLII 2412 (FCA), 97 D.L.R. (4th) 729; Matter of Acosta, Interim Decision 2986, 1985 WL 56042 (B.I.A.); referred to: Artiga Turcios v. I.N.S., 829 F.2d 720 (1987); Arteaga v. I.N.S., 836 F.2d 1227 (1988); Estrada‑Posadas v. I.N.S., 924 F.2d 916 (1991); Minister of Employment and Immigration v. Satiacum (1989), 99 N.R. 171; Astudillo v. Minister of Employment and Immigration (1979), 31 N.R. 121; Arrechea Gonzalez v. Minister of Employment and Immigration (1991), F.C.A. A‑899‑90; Ahmed v. Minister of Employment and Immigration (1990), F.C.A. A‑215‑90; Lai v. Canada (Minister of Employment and Immigration) (1989), 1989 CanLII 10388 (FCA), 8 Imm. L.R. 245; Osorio Cruz v. Minister of Employment and Immigration (1988), I.A.B.D. M88‑20043X; Nalliah v. Minister of Employment and Immigration (1987), I.A.B.D. M84‑1642; Escoto v. Minister of Employment and Immigration (1987), I.A.B.D. T87‑9024X; Incirciyan v. Minister of Employment and Immigration (1987), I.A.B.D. M87‑1541X/M87‑1248; Balareso v. Minister of Employment and Immigration (1985), I.A.B.D. M83‑1542; Andrews v. Law Society of British Columbia, 1989 CanLII 2 (SCC), [1989] 1 S.C.R. 143; I.N.S. v. Elias‑Zacarias, 112 S.Ct. 812 (1992).

Statutes and Regulations Cited

British Nationality Act 1981, 1981 (U.K.), c. 61.

Canadian Charter of Rights and Freedoms, s. 15.

Federal Court Act, R.S.C. 1970 (2nd Supp.), c. 10, s. 28.

Immigration Act, R.S.C., 1985, c. I‑2, ss. 2(1) [as am. by S.C. 1988, c. 35, s. 1 (R.S.C., 1985, c. 28 (4th Supp.), ss. 1(2), 34)], 19.

Immigration Act, 1976, S.C. 1976‑77, c. 52, s. 2(1) [as am. by S.C. 1988, c. 35, s. 1], 4(2.1) [as ad. by Miscellaneous Statute Law Amendment Act, 1981, S.C. 1980‑81‑82‑83, c. 47, s. 3], 19(1)(c), (d), (e), (f), (g), (2) [as am. by S.C. ibid., ss. 23, 53], 46.04(1)(c) [as ad. by S.C. 1988, c. 35, s. 14].

Immigration Regulations, 1978, SOR/78‑172, s. 19(4)(j).

Interpretation Act, R.S.C., 1985, c. I‑21, s. 33(2).

Prevention of Terrorism (Temporary Provisions) Act 1984, 1984 (U.K.), c. 8, later replaced by Prevention of Terrorism (Temporary Provisions) Act 1989, 1989 (U.K.), c. 4, ss. 4, 5.

Authors Cited

Compton, Daniel. "Asylum for Persecuted Social Groups: A Closed Door Left Slightly Ajar ‑‑ Sanchez‑Trujillo v. INS, 801 F.2d 1571 (9th Cir. 1986)" (1987), 62 Wash. L. Rev. 913.

Convention Relating to the Status of Refugees (Geneva, July 28, 1951), Can. T.S. 1969, No. 6.

Foighel, Isi. "The Legal Status of the Boat‑People", 48 Nordisk Tidsskrift for International Relations 217.

Goodwin‑Gill, Guy S. The Refugee in International Law. Oxford: Clarendon Press, 1983.

Grahl‑Madsen, Atle. The Status of Refugees in International Law. (n.p.) Netherlands: A. W. Sijthoff‑Leyden, 1966.

Graves, Maureen. "From Definition to Exploration: Social Groups and Political Asylum Eligibility" (1989), 26 San Diego L. Rev. 739.

Gross, Douglas. "The Right of Asylum Under United States Law" (1980), 80 Colum. L. Rev. 1125.

Hathaway, James C. The Law of Refugee Status. Toronto: Butterworths, 1991.

Helton, Arthur C. "Persecution on Account of Membership in a Social Group As a Basis for Refugee Status" (1983), 15 Colum. Hum. Rts. L. Rev. 39.

Hyndman, Patricia. "The 1951 Convention Definition of Refugee: An Appraisal with Particular Reference to the Case of Sri Lankan Tamil Applicants" (1987), 9 Hum. Rts. Q. 49.

Plender, Richard. "Admission of Refugees: Draft Convention on Territorial Asylum" (1977-78), 15 San Diego L. Rev. 45.

Takkenberg, Alex and Christopher C. Tahbaz. The Collected Travaux Préparatoires of the 1951 Geneva Convention relating to the Status of Refugees. Vol. 1, Early History and the Ad Hoc Committee on Statelessness and Related Problems 16 January ‑ 16 February 1950 Lake Success, New York and Vol. 3, The Conference of Plenipotentiaries on the Status of Refugees and Stateless Persons 2 ‑ 25 July 1951 Geneva, Switserland [sic]. Amsterdam: Dutch Refugee Council, under the auspices of the European Legal Network on Asylum, 1990.

United Nations. Economic and Social Council. Ad Hoc Committee on Statelessness and Related Problems. First Session. Summary record of the Fifth Meeting, Lake Success, New York, 18 January 1950. Mr. Henkin (United States Delegate). UN Doc. E/AC.32/SR.5.

United Nations. Economic and Social Council. Ad Hoc Committee on Statelessness and Related Problems. Report of the Ad Hoc Committee on Statelessness and Related Problems. Lake Success, New York, 16 January to 16 February 1950. UN Doc. E/1618 and Corr. 1 ‑ E/AC.32/5.

United Nations. Economic and Social Council. Ad Hoc Committee on Statelessness and Related Problems. United Kingdom. Revised draft proposal for Article 1. UN Doc. E/AC.32/L.2/Rev. 1.

United Nations. Economic and Social Council. Ad Hoc Committee on Statelessness and Related Problems. United States of America: Memorandum on the Definition Article of the Preliminary Draft Convention Relating to the Status of Refugees (and Stateless Persons). UN Doc. E/AC.32/L.4.

United Nations. General Assembly. Conference of Plenipotentiaries on the Status of Refugees and Stateless Persons. Summary Record of the Third Meeting held at the Palais des Nations, Geneva, on Tuesday, 3 July 1951. Mr. Petren (Swedish Delegate). UN Doc A/CONF.2/SR.3.

United Nations. Office of the United Nations High Commissioner for Refugees. Handbook on Procedures and Criteria for Determining Refugee Status. Geneva: 1988.

van der Veen, Job. "Does Persecution by Fellow‑Citizens in Certain Regions of a State Fall Within the Definition of `Persecution' in the Convention Relating to the Status of Refugees of 1951? Some Comments Based on Dutch Judicial Decisions" (1980), 11 Netherlands Y.B. Intl. L. 167.

APPEAL from a judgment of the Federal Court of Appeal, 1990 CanLII 13041 (FCA), [1990] 2 F.C. 667, 67 D.L.R. (4th) 1, 10 Imm. L.R. (2d) 189, 108 N.R. 60, allowing an application to review and set aside a judgment of the Immigration Appeal Board (1988), 1988 CanLII 10335 (CA IRB), 9 Imm. L.R. (2d) 48, finding appellant to be a convention refugee. Appeal allowed.

Peter A. Rekai, M. Christina F. Kurata, LeVern L. Robertson and Constance Nakatsu, for the appellant.

Roslyn J. Levine and Nanette Rosen, for the respondent.

Ronald B. Shacter and Phyllis Gordon, for the intervener Canadian Council for Refugees.

Brian A. Crane, Q.C., and Gerald Stobo, for the intervener Immigration and Refugee Board.

Written submission only for the intervener United Nations High Commissioner for Refugees.

//La Forest J.//

The judgment of the Court was delivered by

La Forest J. -- This case raises, for the first time in this Court, several fundamental issues respecting the definition of a "Convention refugee" in s. 2(1) of the Immigration Act, 1976, S.C. 1976-77, c. 52, which reads:

2. . . .

"Convention refugee" means any person who, by reason of a well-founded fear of persecution for reasons of race, religion, nationality, membership in a particular social group or political opinion,

(a) is outside the country of his nationality and is unable or, by reason of such fear, is unwilling to avail himself of the protection of that country, or

(b) not having a country of nationality, is outside the country of his former habitual residence and is unable or, by reason of such fear, is unwilling to return to that country . . .

This definition was revised somewhat by S.C. 1988, c. 35, s. 1 (R.S.C., 1985, c. 28 (4th Supp.), s. 1(2)), to its current version in the Immigration Act, R.S.C., 1985, c. I-2:

2. (1) . . .

"Convention refugee" means any person who

(a) by reason of a well-founded fear of persecution for reasons of race, religion, nationality, membership in a particular social group or political opinion,

(i) is outside the country of the person's nationality and is unable or, by reason of that fear, is unwilling to avail himself of the protection of that country, or

(ii) not having a country of nationality, is outside the country of the person's former habitual residence and is unable or, by reason of that fear, is unwilling to return to that country, and

(b) has not ceased to be a Convention refugee by virtue of subsection (2),

but does not include any person to whom the Convention does not apply pursuant to section E or F of Article 1 thereof, which sections are set out in the schedule to this Act;

The questions raised are the extent to which a claimant's "well-founded fear of persecution" must emanate from the state from which the claimant flees, as well as the scope of the enumerated grounds of persecution, particularly "membership in a particular social group" and "political opinion".

Facts

The appellant, Patrick Francis Ward, was born in Northern Ireland in 1955. He joined the Irish National Liberation Army (INLA) in 1983 as a volunteer. Ward described the INLA as a ruthless para-military organization more violent than the Irish Republican Army (IRA), with a military-like hierarchy and strict discipline. Before joining as a volunteer, he had loose connections with the INLA in that he had sympathies for their cause. Indeed, Ward had been convicted of the offences of possession of firearms, conspiracy to convey things unlawfully into Northern Ireland, and contributing to acts of terrorism. He testified that with the constant turmoil in Northern Ireland, people were forced to "take a stand" to protect their loved ones and that his joining the INLA stemmed in part from a desire to protect himself and his family, mainly from the IRA.

Ward's first task as a member of the INLA was to assist in guarding two of the organization's hostages at a farm house in the Republic of Ireland. One day after Ward's guard duties commenced, the INLA ordered the hostages executed. He wanted no part in the execution of these innocent hostages, and underwent what he described as a "predicament of moral conscience". As a result, he resolved to release the hostages and succeeded in doing so without revealing himself to the INLA.

Some time later, the police let slip to an INLA member that one of their own had assisted the hostages in their escape. The INLA suspected Ward, and he was confined and tortured. Although he never admitted his role in the escape, Ward was court-martialled by a kangaroo court and sentenced to death. However, he managed to escape and sought police protection. The police in turn charged him for his part in the hostage incident, based on finding his fingerprints at the farm where the hostages had been held.

Ward expressed concern to the police about his wife and children. The police checked on them, only to discover that they had been taken hostage by the INLA in a pre-emptive move to prevent the claimant from "turning supergrass", the colloquial term for providing evidence to the police about INLA members and their activities.

Ward pleaded guilty to the offence of forcible confinement and was sentenced to three years in jail. He did not "turn supergrass"; nor did he ever admit publicly to having released the hostages. Towards the end of his prison sentence, Ward sought the assistance of the prison chaplain for protection from INLA members. The chaplain, with the assistance of police, obtained a Republic of Ireland passport for Ward and airline tickets to Canada. Ward arrived in Toronto in December 1985 and sought admission to Canada as a visitor. He became the subject of an inquiry in May 1986 and claimed Convention refugee status. His claim was based on a fear of persecution because of his membership in a particular social group, namely the INLA. The Minister of Employment and Immigration determined that Ward was not a Convention refugee and, as a result, he filed an application for redetermination of his claim before the Immigration Appeal Board. The Board allowed the redetermination and found Ward to be a Convention refugee.

The respondent, the Attorney General of Canada, brought an application under s. 28 of the Federal Court Act, R.S.C. 1970 (2nd Supp.), c. 10, to review and set aside the decision of the Board. This application was granted by the Federal Court of Appeal, which set aside the decision and referred the matter back to the Board for reconsideration.

Judgments

Immigration Appeal Board (1988), 1988 CanLII 10335 (CA IRB), 9 Imm. L.R. (2d) 48 (K. J. Arkin for the Board)

The Board approached the case on the basis that two issues fell to be decided pertaining to the definition of a "Convention refugee": whether the definition contemplates a claimant whose country of nationality is unable to protect him adequately, and whether the definition requires state complicity in the persecution of the claimant. On the latter question, the Board found the authorities inconclusive but ruled that the definition does not necessarily contemplate state complicity in the persecution of a claimant, and, at p. 59, that "the state's being unable to offer effective protection is sufficient".

Turning to the first issue, the Board, at p. 59, found the requirement that the claimant be unable or unwilling to avail himself of the protection of his home state was "inextricably intertwined" with the state's inability to offer effective protection. Moreover, the Board reasoned as follows, at p. 59, on the link between persecution and protection:

Fear of persecution and lack of protection are also interrelated elements. Persecuted persons clearly do not enjoy the protection of their country of origin and evidence of the lack of protection may create a presumption as to the likelihood of persecution and to the well-foundedness of any fear.

As such, the Board concluded as follows, at p. 60:

In view of the basic nature of the test imposed by the definition of Convention refugee, i.e., whether or not the applicant has a well-founded fear of persecution for one of the enumerated reasons, it is reasonable, even necessary, to consider the state's ability to provide adequate protection to the applicant: to the extent that the state is unable to protect the individual, the applicant will have good reason to fear persecution. The reason for the state's inability to provide adequate protection from persecution seems irrelevant. The question in any such case then becomes whether or not there exists "adequate" protection. [Emphasis in original.]

On the key question of the state's ability to protect Ward, the Board, at p. 54, found Ward to be a "completely credible witness". It accepted that his life would be in danger if he were required to return to Northern Ireland because of the death sentence passed by the INLA and the threat he posed to that organization's security. Although Irish police had offered Ward protection, such protection would not be effective.

The Board turned its mind to the question of Ward's nationality, a question of immediate relevance given the proviso in the statutory definition that a refugee claimant be unable or unwilling to avail himself of the protection of "the country of [his] nationality". On this point, the Board found as follows, at p. 54:

Clearly, the evidence established that the claimant is a citizen of Ireland, both Northern Ireland and the Republic of Ireland. However, no evidence was presented to the Board to establish that the claimant is also a citizen of the United Kingdom. In response to questions put to him in cross-examination, the claimant testified that as a citizen of Northern Ireland, he is entitled to live in Britain unless he is excluded under the Protection of Terrorism Actof the United Kingdom, whereunder anyone with terrorist connections can be refused entry to British mainland. While the respondent questioned the reasonableness of the claimant's fear of the INLA were he to return to Britain, the respondent did not establish either the claimant's right to live in Britain or the claimant's right to citizenship in the United Kingdom. Accordingly, the Board finds the claimant's country of nationality to be Northern Ireland and the Republic of Ireland.

However, in a footnote to its reasons, at p. 55, the Board went on to note the following:

Had the Board concluded that the claimant was also a national of the United Kingdom, the Board would have made a finding that the claimant's life would be in danger from the INLA if he was returned to the United Kingdom.

In the result, the Board determined that Ward was a Convention refugee.

Federal Court of Appeal, 1990 CanLII 13041 (FCA), [1990] 2 F.C. 667

On appeal to the Federal Court of Appeal, the Attorney General advanced three basic arguments: the Board failed to consider whether the INLA was a "particular social group" within the terms of s. 2(1) of the Act; the Board erred in finding that there was no requirement of state complicity in "persecution"; and it erred in finding that Ward's only countries of nationality were Northern Ireland and the Republic of Ireland. Urie J.A., writing for himself and Marceau J.A., found that the Board had erred on the first and third of these points. MacGuigan J.A. held that the Board had erred only with respect to the third issue.

On the first question, Urie J.A. reasoned that persecution for reasons of social group membership can occur only when the group's activities are perceived to be a possible danger to the government. He stated, at p. 677:

The INLA activities are clearly contrary to the interests of the government of Northern Ireland and of the United Kingdom. But mere membership does not, of itself, substantiate a claim to refugee status. A fortiori, membership does not substantiate a claim to refugee status based upon a fear arising from acts committed by a member of the group contrary to the interests of the group, which group interests are themselves contrary to the well-being of the state. [Emphasis in original.]

In other words, if the claimant's fear arose from within the group itself and not the state, it cannot provide the basis of a claim of persecution. Urie J.A. was not persuaded that Ward, who feared persecution from the organization to which he belonged, was entitled to the protection afforded bona fide refugees who meet all the elements in the definition of Convention refugee. The fact that he was a member who had acted contrary to the interests of the INLA did not bring him within the definition. Urie J.A. remarked, at p. 678, that "[i]f such a view were to be taken anyone who dissents on anything could be said to be a member of a particular social group", a proposition he considered absurd. He rejected the argument that any group engaged in political activity would fall under the definition of a social group. Such an approach, he reasoned, would render the "political opinion" segment of the "Convention refugee" definition redundant.

In dissent, MacGuigan J.A. opined that there could be no serious argument that the INLA is not literally a particular social group since its members (at p. 689) "are united in a stable association with common purposes". He did not agree that "social group" must be deemed to exclude terrorists. However, even conceding this point, he noted that Ward had abandoned the group because of its terrorism and that the social group here at issue included members and former members of the INLA. The group's general commitment to terrorism did not, in his view, mean that Ward, as an individual, was unable to terminate his adherence to it. For MacGuigan J.A., the "true gravamen" of Ward's fear of persecution sprang from his membership in the organization, rather than from his misbehaviour as a member, since the INLA's motivation in sentencing him to death was, at least in part, to prevent future disclosures about the activities of the group. He further noted that a determination that Ward was a Convention refugee would not automatically entitle him to remain in Canada, as he would still be subject to the exceptions in s. 19 of the Act relating to previous convictions, espionage or subversion.

On the second issue, the need for state complicity in persecution, Urie J.A. appears to have decided that such state complicity is a prerequisite for "persecution" under the Act. In support of this, he turned to the requirements of the definition that a claimant be "unable" or "unwilling" to seek the assistance of his home state. Urie J.A. found that being "unable" to avail oneself of the protection of his national state meant, at p. 680, "quite literally that the claimant cannot, because of his physical inability to do so, even seek out the protection of his state. These imply circumstances over which he has no control and is not a concept applicable in facts of this case." On the "unwillingness" branch of the test, Urie J.A. made the following remarks, at p. 680:

If a claimant is "unwilling" to avail himself of the protection of his country of nationality, it is implicit from that fact that his unwillingness stems from his belief that the State and its authorities, cannot protect him from those he fears will persecute him. That inability may arise because the State and its authorities are either themselves the direct perpetrators of the feared acts of persecution, assist actively those who do them or simply turn a blind eye to the activities which the claimant fears. While there may well be other manifestations of it, these possibilities clearly demonstrate that for the claimant to be unwilling to avail himself of the protection of his country of nationality, to provide the foundation for a claim to be a refugee he must establish that the State cannot protect him from the persecution he fears arising, in this case, from his former membership in the INLA, i.e., he must establish that what he fears is in fact persecution as that term is statutorily and jurisprudentially understood. On that basis the involvement of the State is sine qua non where unwillingness to avail himself of the protection is the fact. [Emphasis in original.]

Urie J.A. found that the Board had confused the determination of persecution and ineffective protection. He also rejected the Board's finding that evidence of the lack of protection may create a presumption regarding the likelihood of persecution and the well-foundedness of any fear.

MacGuigan J.A. rejected the contention that the Board erred in its definition of persecution. In his view, the wording of s. 2(1)(a) of the Act does not necessarily import state complicity. While agreeing that "is unable" probably means literally unable, he found no reason to limit the sense of "is unwilling" to a single meaning. He stated, at pp. 697-98:

In sum, I believe that taking into account (1) the literal text of the statute, (2) the absence of any decisive Canadian precedents, and (3) the weight of international authority, the Board's interpretation of the statutory definition is the preferable one. No doubt this construction will make eligible for admission to Canada claimants from strife-torn countries whose problems arise, not from their nominal governments, but from various warring factions, but I cannot think that this is contrary to "Canada's international legal obligations with respect to refugees and . . . its humanitarian tradition with respect to the displaced and the persecuted".

In his view, then, persecution need not emanate from the state.

The third argument of the Attorney General, we saw, was that the Board erred in holding that no evidence had been presented to establish that Ward was a citizen of the United Kingdom, as well as of Northern Ireland and the Republic of Ireland. Ward replied that while Northern Ireland was part of the United Kingdom, he did not have an unrestricted right to live anywhere in the United Kingdom as a result of the Prevention of Terrorism (Temporary Provisions) Act 1984, 1984 (U.K.), c. 8, under which he could be refused admission because of his terrorist activities. On this question Urie J.A. cited, at p. 685, the second paragraph of Art. 1(A)(2) of the Convention, which, while "not binding upon us since it has not been incorporated into Canadian law, . . . persuasive as forming a logical construction of the Convention refugee definition". Urie J.A. held, at p. 683, that "if it is found that he has more than one country of nationality the claimant is obliged to establish his unwillingness to avail himself of the protection of each of his countries of nationality before he can be considered to be a Convention refugee" (emphasis in original). In this respect, Urie J.A. remarked, at p. 685:

. . . I am of the opinion that a refugee claimant must establish that he is unable or unwilling to avail himself of all of his countries of nationality. It is the nationality of the claimant which is of prime importance. The right to live in his country of nationality becomes relevant only in the discharge of the onus on him of proving that he is unable to avail himself of the country of which he has established he is a national. [Emphasis in original.]

Not only did the Board fail to address the issue, he stated, at p. 685: "it compounded the error because it perceived that it was the Crown which had the onus of establishing `either the claimant's right to live in Britain or the claimant's right to citizenship in the United Kingdom'." Urie J.A. noted that s. 8(1) of the Act states that the burden of proof for a person seeking to enter Canada rests on that person. On this point, MacGuigan J.A. was largely in agreement with the majority. All three judges were of the view that the issue of whether Ward could avail himself of the protection of the United Kingdom should be returned to the Board for determination.

Issues

I propose to approach the issues raised by the parties in the following order:

A. Persecution and State Complicity

(a) Is the element of state complicity, either through direct persecution, collusion with the persecuting agents, or wilful blindness to the actions of the persecuting agents, a requisite element in establishing a refugee claimant's "unwillingness" to avail him- or herself of the protection of his or her country of nationality?

(b) Is a claimant considered "unable" to avail him- or herself of the protection of the state only in those circumstances where he or she is physically unable to seek out this protection?

B. Membership in a Particular Social Group

(a) What is the meaning of the phrase, "particular social group", as used in the definition of Convention refugee in s. 2(1) of the Immigration Act, 1976.

(b) Is there any basis for the exclusion of some kinds of social groups as a result of their objectives or the unlawful methods employed by their members?

C. Political Opinion

Whether desertion or dissension from a politico-military organization for reasons of conscience may properly ground a claim to be a Convention refugee on the basis of a well-founded fear of persecution for reasons of political opinion.

D. Section 15 of the Canadian Charter of Rights and Freedoms

Whether the interpretation of "Convention refugee" by the majority of the Federal Court of Appeal is consistent with s. 15 of the Charter.

E. Double Nationality

Where evidence establishes that a refugee claimant has more than one country of nationality, does the claimant have the burden of establishing that he or she is unwilling or unable to avail him- or herself of the protection of each country of nationality, pursuant to the definition of "Convention refugee"?

Analysis

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. The international community intended that persecuted individuals be required to approach their home state for protection before the responsibility of other states becomes engaged. For this reason, James Hathaway refers to the refugee scheme as "surrogate or substitute protection", activated only upon failure of national protection; see The Law of Refugee Status (1991), at p. 135. With this in mind, I shall now turn to the particular elements of the definition of "Convention refugee" that we are called upon to interpret.

A. Persecution and State Complicity

The persecution alleged by the appellant emanates from non-state actors, the INLA; the Government of Ireland is in no way involved in it. This case, then, raises the question whether state involvement is a prerequisite to "persecution" under the definition of "Convention refugee" in the Act. The precise issues are phrased differently by the parties, but can be summarized in the following fashion. First, is there a requirement that "persecution" emanate from the state? Second, does it matter whether the claim is based on the "unable" or "unwilling" branch of the definition? In my view, the answer to both these questions is no. A third issue is the test for establishing a "well-founded fear of persecution" under the Act.

The respondent Attorney General, while arguing that state complicity is a prerequisite to persecution, conceded that a state's inability to protect its citizens from persecution is sufficient state complicity to satisfy the Convention definition. She also conceded that the Government of Ireland was unable to protect the appellant. As such, the respondent confined her argument to the fact that the appellant did not establish before the tribunal that the United Kingdom was similarly unable to protect him. On the second issue, she maintained that when a claimant asserts that he or she is "unwilling" to seek the protection of his or her home state, he or she must also establish that the reason for such unwillingness is state complicity (which, it is conceded, can be extended to the state's inability to protect). The respondent also contended that there is no such prerequisite for state complicity when the refugee asserts that he or she is "unable" to seek the protection of his or her home state. The appellant argued that the definition of persecution must be "neutral", with no requirement for state complicity. Further, he also accepted that there is a distinction between "unable" and "unwilling", but that a claimant's unwillingness can relate back to persecution neutrally defined. The unwillingness, when combined with the inability of the claimant's state to protect him or her from the persecution, will ground a refugee claim.

When one considers the arguments of the appellant and respondent, it becomes apparent that their positions are in reality almost congruent, differing only as to the point at which the inability of the state to protect becomes a necessary ingredient of the definition. The real difference between the parties is on the question of the appellant's unwillingness to return to Great Britain as well as Ireland, a matter that is discussed later as a separate issue.

The intervener Council for Refugees agrees that the Convention definition does require a claimant to demonstrate an inability by his or her state to protect from non-governmental acts of persecution. It contends that this is inherent in the definition rather than a question arising from the term "unwilling". It argues that "unable" and "unwilling" refer only to the refugee claimant's situation outside the country, vis-à-vis the consular officials of his or her home country.

The United Nations High Commissioner for Refugees intervened to argue that the distinction between "unable" and "unwilling" is irrelevant to this appeal, that there is no requirement for state complicity in the definition, and that the proper focus should be on whether the claimant, because of the state's inability to protect, is "unable" or "unwilling" to seek the protection of the authorities in his or her home state. The High Commissioner also endorses the position of the Board that the absence of protection may create a sufficient evidentiary basis for a presumption of a well-founded fear by the claimant. For its part, the Board intervened to argue against any state complicity requirement, maintaining instead that the interpretation of the "Convention refugee" definition should be flexible enough to allow the Board to respond on a case by case basis, given the variety of conditions in the contemporary world that give rise to refugee movements.

In sum, the parties, including the respondent, appear to be unanimous in concluding that the court below was wrong to suggest that the claimant's fear must emanate from the state. As well, there is substantial agreement that a state's inability to protect is an integral component of the notion of a Convention refugee, although the parties differ as to the point in the analysis at which such component is injected into the definition. I find that the consensus reached by the parties is substantially correct. As will be apparent, the majority of the court below would appear to be isolated in its views on state complicity. The majority placed undue emphasis on the distinction between "unwilling" and "unable" in this case.

It is perhaps useful to begin by returning to the text in question:

2. (1) . . .

"Convention refugee" means any person who, by reason of a well-founded fear of persecution for reasons of race, religion, nationality, membership in a particular social group or political opinion,

(a) is outside the country of his nationality and is unable or, by reason of such fear, is unwilling to avail himself of the protection of that country, or

(b) not having a country of nationality, is outside the country of his former habitual residence and is unable or, by reason of such fear, is unwilling to return to that country . . . .

The section appears to focus the inquiry on whether there is a "well-founded fear". This is the first point the claimant must establish. All that follows must be "by reason of" that fear. The first category requires the claimant to be outside the country of nationality by reason of that fear and unable to avail him- or herself of its protection. The second requires that the claimant be both outside the country of nationality and unwilling to avail him- or herself of its protection, by reason of that fear. Thus, regardless of the category under which the claimant falls, the focus is on establishing whether the fear is "well-founded". It is at this stage that the state's inability to protect should be considered. The test is in part objective; if a state is able to protect the claimant, then his or her fear is not, objectively speaking, well-founded. Beyond this point, I see nothing in the text that requires the state to be complicit in, or be the source of, the persecution in question.

State Complicity

My conclusion that state complicity in persecution is not a pre-requisite to a valid refugee claim is reinforced by an examination of the history of the provision, the prevailing authorities, and academic commentary. On the first point, the parties argue that there is no evidence in the drafting history, the Travaux préparatoires, suggesting that persecution was linked to state action. The draft proposed by the United States delegate mentions only the omission of "person[s] who leave. . . or ha[ve] left [their] country of nationality or of former habitual residence for reasons of purely personal convenience" from the definition of "Convention refugee"; see UN doc. E/AC.32/L.4 (January 18, 1950), paragraph B., at p. 3. The revised draft proposed by the United Kingdom did not qualify the word "persecution" in any way, though it did make reference to state authorities in requiring that the claimant "does not wish to return to that country for good and sufficient reason or is not allowed by the authorities of that country to return there"; see UN doc. E/AC.32/L.2/Rev. 1 (January 19, 1950). The omission of a reference to state action does not tell us much, however. The question was apparently never discussed, and the text does not reveal that any link to state action is required.

While the drafting history of the Convention may not go far in justifying the exclusion of state complicity from the interpretation of "Convention refugee", other sources provide more convincing support. A much-cited guide on this question is paragraph 65 of the UNHCR Handbook on Procedures and Criteria for Determining Refugee Status ("UNHCR Handbook"). While not formally binding on signatory states, the Handbook has been endorsed by the states which are members of the Executive Committee of the UNHCR, including Canada, and has been relied upon by the courts of signatory states. Paragraph 65 of the UNHCR Handbook reads:

65. 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. [Emphasis added.]

The position reflected in the UNHCR Handbook, therefore, is that acts by private citizens, when combined with state inability to protect, constitute "persecution".

The absence of a state complicity requirement has also been endorsed by academics; see Job van der Veen, "Does Persecution by Fellow-Citizens in Certain Regions of a State Fall Within the Definition of `Persecution' in the Convention Relating to the Status of Refugees of 1951? Some Comments Based on Dutch Judicial Decisions" (1980), 11 Netherlands Y.B. Intl. L. 167, at p. 172; J. Hathaway, supra, at p. 127; Guy S. Goodwin-Gill, The Refugee in International Law (1983), at p. 42; Patricia Hyndman, "The 1951 Convention Definition of Refugee: An Appraisal with Particular Reference to the Case of Sri Lankan Tamil Applicants" (1987), 9 Hum. Rts. Q. 49, at p. 67; Douglas Gross, "The Right of Asylum Under United States Law" (1980), 80 Colum. L. Rev. 1125, at p. 1139; Atle Grahl-Madsen, The Status of Refugees in International Law (1966), at p. 191.

Canadian decisions reflect the growing consensus that state complicity is not necessary. Two recent cases in the Federal Court of Appeal should be noted. First, in Rajudeen v. Minister of Employment and Immigration (1984), 55 N.R. 129, the court seems to suggest that a state's inability to protect is a sub-set of state complicity. The case involved a refugee claimant from Sri Lanka, who was persecuted by other citizens because of his religious convictions. The police were largely indifferent to this persecution. Heald J.A., writing for the majority, found that persecution need not be at the hands of state agents. As for "unwillingness", he found that the police indifference justified the claimant's reluctance to seek their protection. Stone J.A. concurred, stating at p. 135:

Obviously, an individual cannot be considered a "Convention refugee" only because he has suffered in his homeland from the outrageous behaviour of his fellow citizens. To my mind, in order to satisfy the definition the persecution complained of must have been committed or been condoned by the state itself and consist either of conduct directed by the state toward the individual or in it knowingly tolerating the behaviour of private citizens, or refusing or being unable to protect the individual from such behaviour.

As I understand him, Stone J.A. argues that there must be state complicity, but that concept is broadly defined to include a state's inability to protect its citizen from private persecution.

The facts of the second case, Surujpal v. Minister of Employment and Immigration (1985), 60 N.R. 73, are somewhat similar. There the claimants claimed to have been persecuted by non-state agents because they were members of the opposition. They sought assistance from the police, and were refused. In an oral judgment, MacGuigan J.A. stressed the "police complicity" in the persecution. The majority in the present case seized upon this phrase as evidence that the proper test is state involvement in the persecution. However, MacGuigan J.A. observed that his statement in Surujpal was made in the context of the facts before him, and he appears to suggest that he was not attempting to elucidate a test, but was simply describing the conduct in that case. For him "state complicity" also appears to be sufficiently broad to encompass the state's inability to protect.

This approach is confirmed by the court's recent judgment in Zalzali v. Canada (Minister of Employment and Immigration), 1991 CanLII 13557 (FCA), [1991] 3 F.C. 605, where Décary J.A. bases his reasons on the inability of the state to protect. In doing so, Décary J.A. endorsed the position articulated in paragraph 65 of the UNHCR Handbook. I shall canvass this decision in more detail later.

The jurisprudence in the United States, which is also a party to the Convention, also supports the interpretation that "a well-founded fear of persecution" includes the actions of non-governmental persecutors where the state cannot or will not protect the claimant from those actions. In McMullen v. Immigration and Naturalization Service, 658 F.2d 1312 (9th Cir. 1981), at p. 1315, the Court of Appeal interpreted "likelihood of persecution" in the context of deciding whether a deserter from the Provisional IRA was deportable. The court found the concept to include "[p]ersecution by the government or by a group which the government is unable to control". This principle was reiterated in Artiga Turcios v. I.N.S., 829 F.2d 720 (9th Cir. 1987), at p. 723; Arteaga v. I.N.S., 836 F.2d 1227 (9th Cir. 1988), at p. 1231; and Estrada-Posadas v. I.N.S., 924 F.2d 916 (9th Cir. 1991), at p. 919.

The international community was meant to be a forum of second resort for the persecuted, a "surrogate", approachable upon failure of local protection. The rationale upon which international refugee law rests is not simply the need to give shelter to those persecuted by the state, but, more widely, to provide refuge to those whose home state cannot or does not afford them protection from persecution. The former is, of course, comprised in the latter, but the drafters of the Convention had the latter, wider purpose in mind. The state's inability to protect the individual from persecution founded on one of the enumerated grounds constitutes failure of local protection.

I, therefore, conclude that persecution under the Convention includes situations where the state is not in strictness an accomplice to the persecution, but is simply unable to protect its citizens.

Unable/Unwilling

I now turn to the second question. I would agree with the court below that "unable" and "unwilling" have different meanings, which are fairly apparent on their face. One can say that "unable" means physically or literally unable, and that "unwilling" simply means that protection from the state is not wanted for some reason, though not impossible. This would, at first sight, seem to be a clear distinction, but as we shall see it has become somewhat blurred.

There is some evidence from the Travaux préparatoires on the development of the distinction between the two concepts. The Report of the First Ad Hoc Committee on Statelessness and Related Problems, February 17, 1950 (U.N. Doc. E/1618 and Corr. 1), contained a draft Convention which included a definition of "refugee" that was conceptually similar to the current definition. However, the draft version linked "unwilling" with claimants who were entitled to seek the protection of their state, whereas "unable" was used in connection with stateless individuals. The Committee commented as follows (at p. 415 of the Travaux préparatoires):

The Committee agreed that for the purposes of this sub-paragraph . . . and therefore for the draft convention as a whole, "unable" refers primarily to stateless refugees, but includes also refugees possessing a nationality who are refused passports or other protection by their own government. "Unwilling" refers to refugees who refuse to accept the protection of the government of their nationality.

This has generally been taken as creating a distinction between refugees with a nationality and those who are stateless; see Goodwin-Gill, supra, at p. 25, n. 23. But when the definition was revised to its current form, "unable" was used in connection with both nationals and stateless persons. The Board argues that this revision demonstrates that the term "unable" can apply to those with a nationality, and that the distinction between "unable" and "unwilling" has become blurred. Indeed, this argument is supported by the commentary in the UNHCR Handbook, paragraphs 98-100:

98. Being unable to avail himself of such protection implies circumstances that are beyond the will of the person concerned. There may, for example, be a state of war, civil war or other grave disturbance, which prevents the country of nationality from extending protection or makes such protection ineffective. Protection by the country of nationality may also have been denied to the applicant. Such denial of protection may confirm or strengthen the applicant's fear of persecution, and may indeed be an element of persecution.

99. What constitutes a refusal of protection must be determined according to the circumstances of the case. If it appears that the applicant has been denied services (e.g., refusal of a national passport or extension of its validity, or denial of admittance to the home territory) normally accorded to his co-nationals, this may constitute a refusal of protection within the definition.

100. The term unwilling refers to refugees who refuse to accept the protection of the Government of the country of their nationality. It is qualified by the phrase "owing to such fear". Where a person is willing to avail himself of the protection of his home country, such willingness would normally be incompatible with a claim that he is outside that country "owing to well-founded fear of persecution". Whenever the protection of the country of nationality is available, and there is no ground based on well-founded fear for refusing it, the person concerned is not in need of international protection and is not a refugee. [Emphasis in original.]

This would appear to be an entirely reasonable reading of the current definition. With respect to "unable", it would appear that physical or literal impossibility is one means of triggering the definition, but it is not the only way. Thus ineffective state protection is encompassed within the concept of "unable" and "unwilling", and I am left with the conclusion that the appellant here could have pursued his claim under either category. As such, the distinctions made in the court below were really of no great importance for the purposes of this case.

The majority in the court below, although somewhat unclear on the point, appeared to suggest that "unable" requires no state complicity, but that "unwilling" does. This dichotomy is not, in my view, supported by text of the section or the relevant authorities. As MacGuigan J.A. noted in dissent, the distinction begs the real question of what state complicity means. As we have seen, all parties agree at a minimum that state complicity encompasses an inability to protect. Thus, even if the Court of Appeal's dichotomy were supportable, it would not preclude the appellant's refugee claim.

The Court of Appeal again considered the dichotomy in Zalzali v. Canada (Minister of Employment and Immigration), supra. That case involved a Lebanese national claiming fear of persecution from one of the various warring militias in that country. His persecutors were thus not agents of the state. Nonetheless, the Court of Appeal ruled that he fell within the definition of a "refugee". The court there accepted the dichotomy between "unable" and "unwilling" as used in its judgment in the present case and concluded that state complicity was a sine qua non of persecution only under the latter term. The court further found that the claimant was "unable" to seek the protection of the Lebanese government, as that government had quite literally ceased to exist during Lebanon's civil war.

Décary J.A., writing for the court, concluded at p. 611 that there can be persecution within the meaning of the Act where there is no form of guilt, complicity or participation by the state. His conclusions are largely stated in the context of the "unable" branch of the definition, in deference to its judgment in the present case. However, there are hints in his reasons that he would be willing to apply the same analysis to the "unwilling" branch of the section. He notes that Court of Appeal's reasons in the present case should be applied "with the utmost caution", and his conclusions on state complicity are stated initially without reference to the dichotomy between "unwilling" and "unable". Indeed, much of his reasoning is not grounded in the dichotomy.

Décary J.A. draws on a variety of sources to conclude that persecution can arise from one's fellow nationals, when the government is unable to protect the victim against what they are doing. I am persuaded by the reasoning of these authorities that there is no requirement for state complicity in the Act.

The Council for Refugees and the Board argued, convincingly in my view, that there is simply no need for a judicial gloss of the meaning of "unwilling" and "unable". As the Council argued, there is a clear distinction between the state's being unable to protect its citizens while they are situated in that state (which is considered in the "fear of persecution" analysis) and the individual's being "unable" to avail him- or herself of that protection, which refers to the relationship between the individual and the state outside the country.

Whether the claimant is "unwilling" or "unable" to avail him- or herself of the protection of a country of nationality, state complicity in the persecution is irrelevant. The distinction between these two branches of the "Convention refugee" definition resides in the party's precluding resort to state protection: in the case of "inability", protection is denied to the claimant, whereas when the claimant is "unwilling", he or she opts not to approach the state by reason of his or her fear on an enumerated basis. In either case, the state's involvement in the persecution is not a necessary consideration. This factor is relevant, rather, in the determination of whether a fear of persecution exists.

Test for Determining Fear of Persecution

In the court below, Urie J.A. appears to have taken greatest exception to the linkages made by the Board between various concepts inherent in the definition. Specifically, the Board linked the claimant's unwillingness to the state's inability to protect, and tied the former concept to the well-foundedness of the fear of persecution. These appear to be unobjectionable propositions. The problem for Urie J.A. arose from the Board's conclusion that a lack of state protection creates a presumption of persecution and well-foundedness of the claimant's fears. Although not cited, the Board's expression of this presumption is taken almost verbatim from Goodwin-Gill, supra, at p. 38.

Urie J.A. appears to have concluded that the Board erred by making some sort of algebraic link between the various propositions advanced, thereby engaging in a process of circular reasoning. In his view, the Board developed a causative relationship between the claimant's unwillingness and persecution, by linking both concepts to the issue of whether the claimant's fear is well-founded. However, in Zalzali Décary J.A. had occasion to comment on his colleague's concerns, at p. 610, as follows:

In Ward, at 680, Urie, J.A., said it was important to avoid confusing "the determination of persecution and ineffective protection" and that "the two concepts must be addressed and satisfied independently" but, if I understand his conclusion correctly, as indicated at p. 681, he was anxious to avoid as a matter of fact having one (ineffective protection) serve as a presumption in favour of the other (persecution). I do not think he meant to say that these two concepts could not be interconnected for the purposes of interpreting the definition of a refugee in law. In my view, to accurately define what a refugee is it is important to examine the wording as a whole and interpret the whole in light of its component parts.

With respect to both Décary and Urie JJ.A., it is not clear to me that the Board purported to make an algebraic link, at least in the causative sense that Urie J.A. perceives.

It is clear that the lynch-pin of the analysis is the state's inability to protect: it is a crucial element in determining whether the claimant's fear is well-founded, and thereby the objective reasonableness of his or her unwillingness to seek the protection of his or her state of nationality. Goodwin-Gill's statement, the apparent source of the Board's proposition, reads as follows, at p. 38:

Fear of persecution and lack of protection are themselves interrelated elements. The persecuted clearly do not enjoy the protection of their country of origin, while 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. [Emphasis added.]

Having established that the claimant has a fear, the Board is, in my view, entitled to presume that persecution will be likely, and the fear well-founded, if there is an absence of state protection. The presumption goes to the heart of the inquiry, which is whether there is a likelihood of persecution. But I see nothing wrong with this, if the Board is satisfied that there is a legitimate fear, and an established inability of the state to assuage those fears through effective protection. The presumption is not a great leap. Having established the existence of a fear and a state's inability to assuage those fears, it is not assuming too much to say that the fear is well-founded. Of course, the persecution must be real ‑‑ the presumption cannot be built on fictional events ‑‑ but the well-foundedness of the fears can be established through the use of such a presumption.

In this case, the presumption was apparently of some importance to the Board. It found the appellant to be a credible witness, thus accepting that he had a legitimate fear of persecution. Since Ireland's inability to protect was established through evidence that state agents had admitted their ineffectiveness, the Board was then able to presume the well-foundedness of the claimant's fears. In my view, this approach is correct and suffices for a finding of fear of persecution in this case.

More generally, what exactly must a claimant do to establish fear of persecution? As has been alluded to above, the test is bipartite: (1) the claimant must subjectively fear persecution; and (2) this fear must be well-founded in an objective sense. This test was articulated and applied by Heald J.A. in Rajudeen, supra, at p. 134:

The subjective component relates to the existence of the fear of persecution in the mind of the refugee. The objective component requires that the refugee's fear be evaluated objectively to determine if there is a valid basis for that fear.

See also Minister of Employment and Immigration v. Satiacum (1989), 99 N.R. 171 (F.C.A.), at p. 173. In the present case, the only real issue is the objective test. The Board here found Ward to be credible in his testimony, thus establishing the subjective branch. The issue is whether the fear is objectively justifiable.

Does the plaintiff first have to seek the protection of the state, when he is claiming under the "unwilling" branch in cases of state inability to protect? The Immigration Appeal Board has found that, where there is no proof of state complicity, the mere appearance of state ineffectiveness will not suffice to ground a claim. As Professor Hathaway, supra, puts it, at p. 130:

Obviously, there cannot be said to be a failure of state protection where a government has not been given an opportunity to respond to a form of harm in circumstances where protection might reasonably have been forthcoming:

A refugee may establish a well-founded fear of persecution when the official authorities are not persecuting him if they refuse or are unable to offer him adequate protection from his persecutors . . . however, he must show that he sought their protection when he is convinced, as he is in the case at bar, that the official authorities ‑‑ when accessible ‑‑ had no involvement ‑‑ direct or indirect, official or unofficial ‑‑ in the persecution against him. (José Maria da Silva Moreira, Immigration Appeal Board Decision T86-10370, April 8, 1987, at 4, per V. Fatsis.)

This is not true in all cases. Most states would be willing to attempt to protect when an objective assessment established that they are not able to do this effectively. Moreover, it would seem to defeat the purpose of international protection if a claimant would be required to risk his or her life seeking ineffective protection of a state, merely to demonstrate that ineffectiveness.

Like Hathaway, I prefer to formulate this aspect of the test for fear of persecution as follows: only in situations in which state protection "might reasonably have been forthcoming", will the claimant's failure to approach the state for protection defeat his claim. Put another way, the claimant will not meet the definition of "Convention refugee" where it is objectively unreasonable for the claimant not to have sought the protection of his home authorities; otherwise, the claimant need not literally approach the state.

The issue that arises, then, is how, in a practical sense, a claimant makes proof of a state's inability to protect its nationals as well as the reasonable nature of the claimant's refusal actually to seek out this protection. On the facts of this case, proof on this point was unnecessary, as representatives of the state authorities conceded their inability to protect Ward. Where such an admission is not available, however, clear and convincing confirmation of a state's inability to protect must be provided. For example, a claimant might advance testimony of similarly situated individuals let down by the state protection arrangement or the claimant's testimony of past personal incidents in which state protection did not materialize. Absent some evidence, the claim should fail, as nations should be presumed capable of protecting their citizens. Security of nationals is, after all, the essence of sovereignty. Absent a situation of complete breakdown of state apparatus, such as that recognized in Lebanon in Zalzali, it should be assumed that the state is capable of protecting a claimant.

The Federal Court of Appeal's disposition in Satiacum may best be explained as exemplifying such a case of presumption of a state's ability to protect and of objective unreasonability in the claimant's failure to avail himself of this protection. In that case, an American Indian chief who was convicted of federal criminal charges fled to Canada before sentencing. Arrested in Canada a year later, he claimed refugee status. The persecution he alleged to have feared was a risk to his life if incarcerated in a federal prison. The Federal Court of Appeal found that Satiacum's fear did not meet the objective component of the test for fear of persecution, as it must be presumed that the United States judicial system is effective in affording a citizen just treatment. The court stated, at p. 176:

In the absence of exceptional circumstances established by the claimant, it seems to me that in a Convention refugee hearing, as in an extradition hearing, Canadian tribunals have to assume a fair and independent judicial process in the foreign country. In the case of a nondemocratic State, contrary evidence might be readily forthcoming, but in relation to a democracy like the United States contrary evidence might have to go to the extent of substantially impeaching, for example, the jury selection process in the relevant part of the country, or the independence or fair-mindedness of the judiciary itself.

Although this presumption increases the burden on the claimant, it does not render illusory Canada's provision of a haven for refugees. The presumption serves to reinforce the underlying rationale of international protection as a surrogate, coming into play where no alternative remains to the claimant. Refugee claims were never meant to allow a claimant to seek out better protection than that from which he or she benefits already.

In summary, I find that state complicity is not a necessary component of persecution, either under the "unwilling" or under the "unable" branch of the definition. A subjective fear of persecution combined with state inability to protect the claimant creates a presumption that the fear is well-founded. The danger that this presumption will operate too broadly is tempered by a requirement that clear and convincing proof of a state's inability to protect must be advanced. I recognize that these conclusions broaden the range of potentially successful refugee claims beyond those involving feared persecution at the hands of the claimant's nominal government. As long as this persecution is directed at the claimant on the basis of one of the enumerated grounds, I do not think the identity of the feared perpetrator of the persecution removes these cases from the scope of Canada's international obligations in this area. On this note, I now turn to a consideration of these enumerated grounds.

B. Membership in a Particular Social Group

Section 2(1) of the Act limits the grounds for a Convention refugee's well-founded fear of persecution to five possibilities: "race, religion, nationality, membership in a particular social group or political opinion". The appellant justifies his claim to Convention refugee status on the basis of his well-founded fear of persecution at the hands of the INLA, should he return to Northern Ireland, by reason of his membership in a particular social group, i.e., the INLA. The first issue to be addressed, therefore, is the scope of "particular social group" and whether this enumerated basis of persecution embraces INLA members.

Attempts at defining the range of the category of "particular social group" in this case were not made until reaching the Federal Court of Appeal. The Immigration Appeal Board did not broach the issue, seemingly assuming that the INLA did indeed constitute a particular social group. In the Court of Appeal, the majority adopted a very narrow definition, at p. 674, excluding "groups who by acts of terrorism seek to promote their aims, in this case the overthrow of the duly constituted authority". MacGuigan J.A., on the other hand, delineated the reach of this category loosely, at p. 689, including within it any "stable association with common purposes". In my opinion, the proper scope of "particular social group" lies in between these two extremes, but would still exclude from its ambit Ward's membership in the INLA.

Canadian jurisprudence in which "particular social group" is interpreted has, until very recently, been quite sparse; the cases that did deal with this notion were usually handled on their own particular facts and lacked guidance with respect to a more general interpretation of the category: see Astudillo v. Minister of Employment and Immigration (1979), 31 N.R. 121 (F.C.A.), Arrechea Gonzalez v. Minister of Employment and Immigration (1991), F.C.A. A-899-90, Ahmed v. Minister of Employment and Immigration (1990), F.C.A. A-215-90, Lai v. Canada (Minister of Employment and Immigration) (1989), 1989 CanLII 10388 (FCA), 8 Imm. L.R. 245, Osorio Cruz v. Minister of Employment and Immigration (1988), I.A.B.D. M88-20043X, Nalliah v. Minister of Employment and Immigration (1987), I.A.B.D. M84-1642, Escoto v. Minister of Employment and Immigration (1987), I.A.B.D. T87-9024X, Incirciyan v. Minister of Employment and Immigration (1987), I.A.B.D. M87-1541X/M87-1248 and Balareso v. Minister of Employment and Immigration (1985), I.A.B.D. M 83-1542. Recently, the Federal Court of Appeal has begun to articulate a test which attempts to achieve a middle ground between the two positions advanced by the majority and the minority in the Court of Appeal in the present case: see Cheung v. Minister of Employment and Immigration, 1993 CanLII 2946 (FCA), [1993] F.C.J. No. 309 (Q.L.), Appeal No. A-785-91 (F.C.A.) and Mayers v. Canada (Minister of Employment and Immigration) (1992), 1992 CanLII 2412 (FCA), 97 D.L.R. (4th) 729. I shall address these two decisions in some detail below. International and foreign sources are also of considerable significance in the study of the meaning of "particular social group" and specifically in evaluating the test proposed recently by the Federal Court of Appeal. An examination of the Canadian and foreign doctrine and jurisprudence reveals three advocated approaches:

(1) A very wide definition, similar to that adopted by MacGuigan J.A., pursuant to which the class serves as a safety net to prevent any possible gap in the other four categories;

(2) A narrower definition that confines its scope by means of some appropriate limiting mechanism, recognizing that this class is not meant to encompass all groups; and

(3) An even narrower definition, paralleling that formulated by the majority of the Federal Court of Appeal, that responds to concerns about morality and criminality by excluding terrorists, criminals and the like.

I shall consider each of these suggested definitions in turn.

"Particular Social Group" as Safety Net

The broad definition of "particular social group", comprising basically any alliance of individuals with a common objective, is most forcefully advocated by Arthur C. Helton, Director of the Political Asylum Project of the Lawyers Committee for International Human Rights. In his article, "Persecution on Account of Membership in a Social Group As a Basis for Refugee Status" (1983), 15 Colum. Hum. Rts. L. Rev. 39, at p. 45, Helton sets out his view of the scope of this category in these terms:

The intent of the framers of the Refugee Convention was not to redress prior persecution of social groups, but rather to save individuals from future injustice. The "social group" category was meant to be a catch-all which could include all the bases for and types of persecution which an imaginative despot might conjure up.

Isi Foighel refers to this category as a "safety net". "[T]his category", he states, "was to include also race and ethnicity and, furthermore, was to operate as a kind of comprehensive provision for the categories of persons who had a legitimate claim upon being considered refugees in the international sense, although they were not clearly included in the categories specifically mentioned". See Isi Foighel, "The Legal Status of the Boat-People", 48 Nordisk Tidsskrift for International Relations 217, at pp. 222-23. This interpretation essentially characterizes an association of people as a "particular social group" merely by virtue of their common victimization as the objects of persecution.

This wide approach has been promoted by several other writers in the field. Guy Goodwin-Gill, in The Refugee in International Law, supra, at p. 30, describes as essential to the definition "the factor of shared interests, values, or background ‑‑ a combination of matters of choice with other matters over which members of the group have no control". Goodwin-Gill goes so far so as to enumerate as relevant uniting characteristics, in addition to ethnic, cultural and linguistic origin, education and family background, the factors of economic activity, shared values, outlook and aspirations. Daniel Compton, in "Asylum for Persecuted Social Groups: A Closed Door Left Slightly Ajar ‑‑ Sanchez-Trujillo v. INS, 801 F.2d 1571 (9th Cir. 1986)" (1987), 62 Wash. L. Rev. 913, at p. 923, delimits the broad range of "particular social group" as "a recognized grouping within a society, a group that shares some common experience". Occasionally, it is true, these writers appear to qualify their approach somewhat by referring to "legitimate" groups or "invidious" persecution. But their essential theme remains that as long as some common thread binds the set of individuals together, whether on the basis of background, habits or status, the requirement that the feared persecution be based on membership in a particular social group is met.

The proponents of this expansive view rely on the genesis of the category of "particular social group". It was suggested as a last-minute expansion of the Convention's definition of "refugee" by the Swedish delegate (A/CONF.2/SR.3, at p. 14):

Mr. PETREN (Sweden) . . .

In the first place, experience had shown that certain refugees had been persecuted because they belonged to particular social groups. The draft Convention made no provision for such cases, and one designed to cover them should accordingly be included.

The proponents of the liberal approach justify their position by seizing upon this limited discussion of the addition's background. Grahl-Madsen, supra, at p. 219, justifies his wide definition, for example, on the basis of the intent of the framers. He asserts:

The reason membership of a particular social group' was added by the Conference of Plenipotentiaries as [sic] an afterthought. Many cases falling under this term are also covered by the terms discussed above, but the notion of social group' is of broader application than the combined notions of racial, ethnic, and religious groups, and in order to stop a possible gap, the Conference felt that it would be as well to mention this reason for persecution explicitly.

Others make the same point; see Maureen Graves, "From Definition to Exploration: Social Groups and Political Asylum Eligibility" (1989), 26 San Diego L. Rev. 739, at pp. 747-49; Compton, supra, at pp. 925-26.

In my view, the supporters of the wide definition exaggerate the implications of the intention of the framers. The fact that this class was added to enlarge the range of cases falling within the definition of "refugee" therein was initially a Cold War reaction aimed at ensuring a haven for capitalists fleeing the persecution they encountered in Eastern Bloc regimes after the World War II. Daniel Compton, supra, made this historical observation at pp. 925-26:

The most well-known examples of social group-based persecution at [the time of drafting the Convention] occurred in Eastern Europe following the rise of the Communist regimes. Subsequent cases from European courts of nations party to the Convention have recognized, for example, the "capitalist class" and "independent businessmen" and their families as valid social groups in granting refugee status to persons fleeing Eastern Europe. Examples such as these are probably what the Swedes had in mind.

See also R. Plender, "Admission of Refugees: Draft Convention on Territorial Asylum" (1977-78), 15 San Diego L. Rev. 45, at p. 52; and Grahl-Madsen, supra, at pp. 219-20, who reviews the foreign jurisprudence on these Cold War cases. The persecution in the "Cold War cases" was imposed upon the capitalists not because of their contemporaneous activities but because of their past status as ascribed to them by the Communist leaders. Given this historical origin, the definition of "particular social group" must, at the very least, embrace these types of situations. The scope of "particular social group", however, was not meant to be limited to that specific historical circumstance and no one has ever so contended. The ambit of this portion of the definition of "Convention refugee" must be evaluated on the basis of the basic principles underlying the treaty.

As explained earlier, international refugee law was meant to serve as a "substitute" for national protection where the latter was not provided. For this reason, the international role was qualified by built-in limitations. These restricting mechanisms reflect the fact that the international community did not intend to offer a haven for all suffering individuals. The need for "persecution" in order to warrant international protection, for example, results in the exclusion of such pleas as those of economic migrants, i.e., individuals in search of better living conditions, and those of victims of natural disasters, even when the home state is unable to provide assistance, although both of these cases might seem deserving of international sanctuary.

Similarly, the drafters of the Convention limited the included bases for a well-founded fear of persecution to "race, religion, nationality, membership in a particular social group or political opinion". Although the delegates inserted the social group category in order to cover any possible lacuna left by the other four groups, this does not necessarily lead to the conclusion that any association bound by some common thread is included. If this were the case, the enumeration of these bases would have been superfluous; the definition of "refugee" could have been limited to individuals who have a well-founded fear of persecution without more. The drafters' decision to list these bases was intended to function as another built-in limitation to the obligations of signatory states. The issue that arises, therefore, is the demarcation of this limit.

The UNHCR Handbook does not appear to address this issue specifically. Paragraphs 77-79 deal with the meaning of "membership of a particular social group":

77. A "particular social group" normally comprises persons of similar background, habits or social status. A claim to fear of persecution under this heading may frequently overlap with a claim to fear of persecution on other grounds, i.e. race, religion or nationality.

78. Membership of such a particular social group may be at the root of persecution because there is no confidence in the group's loyalty to the Government or because the political outlook, antecedents or economic activity of its members, or the very existence of the social group as such, is held to be an obstacle to the Government's policies.

79. Mere membership of a particular social group will not normally be enough to substantiate a claim to refugee status. There may, however, be special circumstances where mere membership can be a sufficient ground to fear persecution.

The language is sufficiently general that it may, on one view of the matter, be interpreted as accepting the expansive approach just discussed. But that is far from certain. The handbook may, I think, with equal consistency, be read more narrowly. That, having regard to the context and purpose of the treaty, appears to me to be the better approach.

"Particular Social Group" Limited by Anti-Discrimination Notions

Underlying the Convention is the international community's commitment to the assurance of basic human rights without discrimination. This is indicated in the preamble to the treaty as follows:

CONSIDERING that the Charter of the United Nations and the Universal Declaration of Human Rights approved on 10 December 1948 by the General Assembly have affirmed the principle that human beings shall enjoy fundamental rights and freedoms without discrimination.

This theme outlines the boundaries of the objectives sought to be achieved and consented to by the delegates. It sets out, in a general fashion, the intention of the drafters and thereby provides an inherent limit to the cases embraced by the Convention. Hathaway, supra, at p. 108, thus explains the impact of this general tone of the treaty on refugee law:

The dominant view, however, is that refugee law ought to concern itself with actions which deny human dignity in any key way, and that the sustained or systemic denial of core human rights is the appropriate standard.

This theme sets the boundaries for many of the elements of the definition of "Convention refugee". "Persecution", for example, undefined in the Convention, has been ascribed the meaning of "sustained or systemic violation of basic human rights demonstrative of a failure of state protection"; see Hathaway, supra, at pp. 104-105. So too Goodwin-Gill, supra, at p. 38, observes that "comprehensive analysis requires the general notion [of persecution] to be related to developments within the broad field of human rights". This has recently been recognized by the Federal Court of Appeal in the Cheung case.

In similar fashion, the enumeration of specific foundations upon which the fear of persecution may be based to qualify for international protection parallels the approach adopted in international anti-discrimination law. Thus Goodwin-Gill, supra, at p. 39, notes:

The references to `race, religion, nationality, membership of [sic] a particular social group, or political opinion' illustrate briefly the characteristics of individuals and groups which are considered worthy of special protection. These same factors have figured in the development of the fundamental principle of non-discrimination in general international law, and have contributed to the formulation of other fundamental human rights.

In distilling the contents of the head of "particular social group", therefore, it is appropriate to find inspiration in discrimination concepts. Hathaway, supra, at pp. 135-36, explains that the anti-discrimination influence in refugee law is justified on the basis of those sought to be protected thereby:

The early refugee accords did not articulate this notion of disfranchisement or breakdown of basic membership rights, since refugees were defined simply by specific national, political, and religious categories, including anti-Communist Russians, Turkish Armenians, Jews from Germany, and others. The de facto uniting criterion, however, was the shared marginalization of the groups in their states of origin, with consequent inability to vindicate their basic human rights at home. These early refugees were not merely suffering persons, but were moreover persons whose position was fundamentally at odds with the power structure in their home state. It was the lack of a meaningful stake in the governance of their own society which distinguished them from others, and which gave legitimacy to their desire to seek protection abroad.

The manner in which groups are distinguished for the purposes of discrimination law can thus appropriately be imported into this area of refugee law.

This theme of international concern for discrimination and human rights seems to underlie the recent trend in the jurisprudence of the Federal Court of Appeal. In Mayers v. Canada (Minister of Employment and Immigration), supra, the court reviewed the decision of a credible basis panel. Pursuant to this decision, it was found that there was some evidence upon which the Refugee Division might determine the applicant to be a Convention refugee in her claim to fear persecution on the basis of membership in the particular social group of "Trinidadian women subject to wife abuse". Although not strictly necessary to this review, Mahoney J.A. addressed the question of whether this group could meet the definition of Convention refugee. In doing so, he articulated the following test, at p. 737, proposed by counsel for the applicant:

. . . a particular social group means: (1) a natural or non-natural group of persons with (2) similar shared background, habits, social status, political outlook, education, values, aspirations, history, economic activity or interests, often interests contrary to those of the prevailing government, and (3) sharing basic, innate, unalterable characteristics, consciousness and solidarity, or (4) sharing a temporary but voluntary status, with the purpose of their association being so fundamental to their human dignity that they should not be required to alter it.

In Cheung v. Minister of Employment and Immigration, supra, the court was more directly confronted with the question of the test for "particular social group", in deciding whether women in China who have more than one child and are faced with forced sterilization constitute such a group. In order to make this evaluation, Linden J.A. adopted the test proposed in Mayers v. Canada (Minister of Employment and Immigration), supra. In applying the test to the facts before him, Linden J.A. held:

It is clear that women in China who have one child and are faced with forced sterilization satisfy enough of the above criteria to be considered a particular social group. These people comprise a group sharing similar social status and hold a similar interest which is not held by their government. They have certain basic characteristics in common. All of the people coming within this group are united or identified by a purpose which is so fundamental to their human dignity that they should not be required to alter it on the basis that interference with a woman's reproductive liberty is a basic right "ranking high in our scale of values" (E. (Mrs.) v. Eve, 1986 CanLII 36 (SCC), [1986] 2 S.C.R. 388).

In this way, the focus of the inquiry was on the basic right of reproductive control.

This approach to delineating the scope of "particular social group" is developed further in American quasi-judicial authority. In Matter of Acosta, Interim Decision 2986, 1985 WL 56042 (B.I.A.) (Database FIM-81A), the United States Board of Immigration Appeals was confronted with the claim for asylum of an El Salvador citizen. The claimant based his fear of persecution on his membership in a cooperative organization of taxi drivers. According to the claimant, members of the cooperative had been targeted by anti-government guerrillas for having refused to comply with the latter's requests to engage in work stoppages. In finding that the cooperative did not constitute a "particular social group", the Board defined this term in a manner that reflects classic discrimination analysis. It stated, at pp. 37-39:

We find the well-established doctrine of ejusdem generis, meaning literally, of the same kind,' to be most helpful in construing the phrase membership in a particular social group.' That doctrine holds that general words used in an enumeration with specific words should be construed in a manner consistent with the specific words. See, e.g., Cleveland v. United States, 329 U.S. 14 (1946); 2A C. Sands, supra, s 47.17. The other grounds of persecution in the Act and the Protocol listed in association with membership in a particular social group' are persecution on account of race,' religion,' nationality,' and `political opinion.' Each of these grounds describes persecution aimed at an immutable characteristic: a characteristic that either is beyond the power of an individual to change or is so fundamental to individual identity or conscience that it ought not be required to be changed. See A. Grahl-Madsen, supra, at 217; G. Goodwin-Gill, supra, at 31. Thus, the other four grounds of persecution enumerated in the Act and the Protocol restrict refugee status to individuals who are either unable by their own actions, or as a matter of conscience should not be required, to avoid persecution.

Applying the doctrine of ejusdem generis, we interpret the phrase persecution on account of membership in a particular social group' to mean persecution that is directed toward an individual who is a member of a group of persons all of whom share a common, immutable characteristic. The shared characteristic might be an innate one such as sex, color, or kinship ties, or in some circumstances it might be a shared past experience such as former military leadership or land ownership. The particular kind of group characteristic that will qualify under this construction remains to be determined on a case-by-case basis. However, whatever the common characteristic that defines the group, it must be one that the members of the group either cannot change, or should not be required to change because it is fundamental to their individual identities or consciences. Only when this is the case does the mere fact of group membership become something comparable to the other four grounds of persecution under the Act, namely, something that either is beyond the power of an individual to change or that is so fundamental to his identity or conscience that it ought not be required to be changed. By construing persecution on account of membership in a particular social group' in this manner, we preserve the concept that refuge is restricted to individuals who are either unable by their own actions, or as a matter of conscience should not be required, to avoid persecution.

What is excluded by this definition are "groups defined by a characteristic which is changeable or from which disassociation is possible, so long as neither option requires renunciation of basic human rights"; see Hathaway, supra, at p. 161.

Anti-discrimination law in Canada as embodied by s. 15 of the Charter and the jurisprudence decided thereunder, although still not completely developed, makes reference to very similar criteria. In the seminal equality case of Andrews v. Law Society of British Columbia, 1989 CanLII 2 (SCC), [1989] 1 S.C.R. 143, non-citizenship was held to be an analogous ground of discrimination because it shared the same overarching characteristics of those enumerated in s. 15 of the Charter. In that case, I articulated these common characteristics as follows, at p. 195:

The characteristic of citizenship is one typically not within the control of the individual and, in this sense, is immutable. Citizenship is, at least temporarily, a characteristic of personhood not alterable by conscious action and in some cases not alterable except on the basis of unacceptable costs.

The "analogous grounds" approach to s. 15 of the Charter parallels that of the Federal Court of Appeal in its recent judgments, as well as the United States Immigration Board of Appeals, with respect to the definition of "particular social group" in the distillation of and extrapolation from the common thread running through the enumerated heads.

These types of tests appear to be appropriate to us. Canada's obligation to offer a haven to those fleeing their homelands is not unlimited. Foreign governments should be accorded leeway in their definition of what constitutes anti-social behaviour of their nationals. Canada should not overstep its role in the international sphere by having its responsibility engaged whenever any group is targeted. Surely there are some groups, the affiliation in which is not so important to the individual that it would be more appropriate to have the person dissociate him- or herself from it before Canada's responsibility should be engaged. Perhaps the most simplified way to draw the distinction is by opposing what one is against what one does, at a particular time. For example, one could consider the facts in Matter of Acosta, in which the claimant was targeted because he was a member of a taxi driver cooperative. Assuming no issues of political opinion or the right to earn some basic living are involved, the claimant was targeted for what he was doing and not for what he was in an immutable or fundamental way.

The meaning assigned to "particular social group" in the Act should take into account the general underlying themes of the defence of human rights and anti-discrimination that form the basis for the international refugee protection initiative. The tests proposed in Mayers, supra, Cheung, supra, and Matter of Acosta, supra, provide a good working rule to achieve this result. They identify three possible categories:

(1) groups defined by an innate or unchangeable characteristic;

(2) groups whose members voluntarily associate for reasons so fundamental to their human dignity that they should not be forced to forsake the association; and

(3) groups associated by a former voluntary status, unalterable due to its historical permanence.

The first category would embrace individuals fearing persecution on such bases as gender, linguistic background and sexual orientation, while the second would encompass, for example, human rights activists. The third branch is included more because of historical intentions, although it is also relevant to the anti-discrimination influences, in that one's past is an immutable part of the person.

"Particular Social Group": Exclusion of Criminals and Terrorists

The majority of the Federal Court of Appeal held that international refugee protection should not embrace terrorists, such as members of the INLA. Urie J.A. put it this way, at pp. 674-75:

Counsel pointed out that paragraph 3(g) of the Act recognizes Canada's need to fulfil its "international legal obligations with respect to refugees and to uphold its humanitarian tradition with respect to the displaced and persecuted". To be consistent in the fulfilment of its humanitarian goal, groups who by acts of terrorism seek to promote their aims, in this case the overthrow of the duly constituted authority, should be excluded from those social groups who meet the definition of Convention refugee. To do otherwise, counsel said, would allow Canada to be a haven for persons who admit to sympathizing with or having committed or participated in terrorists acts in other countries, with or without disavowing their support of terrorists.

The mechanism adopted by Urie J.A. to ensure the exclusion of these undesirable claimants, in this way, is a limitation of the scope of the definition of "particular social group". An examination of the Act as a whole, however, reveals that the concerns he articulated are anticipated and provided for elsewhere in the Act. In my view, therefore, such a restriction on the scope of "particular social group" is unnecessary and renders redundant the explicit exclusionary provisions.

The Act lists classes of claimants considered to be inadmissible in s. 19. Several of these relate to concerns about criminality, violence and government subversion. Subsection (1) in relevant part reads:

19. (1) No person shall be granted admission if he is a member of any of the following classes:

(c) persons who have been convicted of an offence that, if committed in Canada, constitutes or, if committed outside Canada, would constitute an offence that may be punishable under any Act of Parliament and for which a maximum term of imprisonment of ten years or more may be imposed, except persons who have satisfied the Governor in Council that they have rehabilitated themselves and that at least five years have elapsed since the termination of the sentence imposed for the offence;

(d) persons who there are reasonable grounds to believe will

(i) commit one or more offences punishable by way of indictment under any Act of Parliament, or

(ii) engage in activity that is part of a pattern of criminal activity planned and organized by a number of persons acting in concert in furtherance of the commission of any offence that may be punishable under any Act of Parliament by way of indictment;

(e) persons who have engaged in or who there are reasonable grounds to believe will engage in acts of espionage or subversion against democratic government, institutions or processes, as they are understood in Canada, except persons who, having engaged in such acts, have satisfied the Minister that their admission would not be detrimental to the national interest;

(f) persons who there are reasonable grounds to believe will, while in Canada, engage in or instigate the subversion by force of any government;

(g) persons who there are reasonable grounds to believe will engage in acts of violence that would or might endanger the lives or safety of persons in Canada or are members of or are likely to participate in the unlawful activities of an organization that is likely to engage in such acts of violence . . . .

Section 19(2) goes on to preclude the granting of admission to persons who have been convicted of offences that would have constituted indictable or summary conviction offences, had they been committed in Canada, unless these persons demonstrate that they have become rehabilitated and certain delineated time periods have elapsed.

A claimant for refugee status in Canada who has established his or her inclusion in the definition of "Convention refugee" must still overcome the hurdle of s. 19 before entry into this country will be permitted. These exclusions on the basis of criminality have been carefully drafted to avoid the admission of claimants who may pose a threat to the Canadian government or to the lives or property of the residents of Canada. The provisions specifically give the Minister of Employment and Immigration enough flexibility, however, to reassess the desirability of permitting entry to a claimant with a past criminal record, where the Minister is convinced that rehabilitation has occurred. In this way, Parliament opted not to treat a criminal past as a reason to be estopped from obtaining refugee status. If the scope of the term "particular social group" were interpreted so as to exclude criminals and terrorists, as the majority of the Court of Appeal did, this legislative decision would be ignored. I think it more appropriate to avoid such a blanket exclusion in the face of an explicit, comprehensive structure for the assessment of these potentially inadmissible claimants.

In the amended Immigration Act, R.S.C., 1985, c. I-2, Parliament has further responded to the concern of keeping out dangerous and criminal claimants by excluding from the definition of "Convention refugee" in s. 2 of the Act any person to whom the Convention does not apply pursuant to s. E or F of Art. 1 thereof, which sections are set out in the schedule to the Act (R.S.C., 1985, c. 28 (4th Supp.), s. 34). The provision of Art. 1 of the Convention relevant for the purposes of this analysis is s. F, which reads:

F. The provisions of this Convention shall not apply to any person with respect to whom there are serious reasons for considering that:

(a) he has committed a crime against peace, a war crime, or a crime against humanity, as defined in the international instruments drawn up to make provision in respect of such crimes;

(b) he has committed a serious non-political crime outside the country of refuge prior to his admission to that country as a refugee;

(c) he has been guilty of acts contrary to the purposes and principles of the United Nations.

The articulation of this exclusion for the "commission" of a crime can be contrasted with those of s. 19 of the Act which refers to "convictions" for crimes. Hathaway, supra, at p. 221, interprets this exclusion to embrace "persons who are liable to sanctions in another state for having committed a genuine, serious crime, and who seek to escape legitimate criminal liability by claiming refugee status". In other words, Hathaway would appear to confine paragraph (b) to accused persons who are fugitives from prosecution. The interpretation of this amendment was not argued before us. I note, however, that Professor Hathaway's interpretation seems to be consistent with the views expressed in the Travaux préparatoires, regarding the need for congruence between the Convention and extradition law; see statement of United States delegate Henkin, U.N. Doc. E/AC.32/SR.5 (January 30, 1950), at p. 5. As such, Ward would still not be excluded on this basis, having already been convicted of his crimes and having already served his sentence. This addition to the Act does answer, however, in a more general fashion, the concerns raised by the majority of the Court of Appeal and renders less forceful the argument that morality and criminality concerns need be accommodated by narrowing the definition of "particular social group".

Is Ward a Member of a Particular Social Group?

Applying the three-pronged interpretation of "particular social group" adopted earlier to the case at bar, Ward does not meet the definition of "Convention refugee" and thus cannot be admitted into Canada on the basis of his fear of persecution at the hands of the INLA upon his return to Northern Ireland.

First, we must define the association at issue. In the Court of Appeal, Ward's affiliation was designated as "member of the INLA" (by Urie J.A., at p. 677) and as "members and former members of the INLA" (by MacGuigan J.A., at p. 691). Ward's claim is that he fears persecution, should he return to Northern Ireland, because the INLA would retaliate to avenge his release of the hostages. This act was effected by Ward qua member of the INLA. Ward also testified that he feared persecution by the INLA because of its concern that he "turn supergrass". This fear is present whether or not Ward renounced his membership in the INLA, as the possibility of revealing organization secrets is present in the case of both present and former members. Thus, no subsequent disassociation from this group by Ward had any impact on his fear. I do not think it appropriate, therefore, to say that Ward's fear was based on his status as a former member of the INLA. The fact that Ward might no longer be a member is merely a result of the persecution feared, not its foundation.

The group of INLA members is not a "particular social group". To review, the test given above includes:

(1) groups defined by an innate or unchangeable characteristic;

(2) groups whose members voluntarily associate for reasons so fundamental to their human dignity that they should not be forced to forsake the association; and

(3) groups associated by a former voluntary status, unalterable due to its historical permanence.

Clearly, the INLA members are not characterized by an innate or unalterable characteristic. The third branch of the definition is not applicable to Ward, since the group is associated in the present and membership is not unchangeable owing to its status as a historical fact. (It seems that this branch of the definition will only come into play when the identity of the persecutor does not coincide with that of the social group as it does in this case. For this prong to be relevant, the social group should no longer be actively affiliated; if the group has disbanded, it cannot possibly persecute.) As for the second branch, the INLA is a voluntary association committed to the attainment of specific political goals by any means, including violence, but I do not believe that this objective can be said to be so fundamental to the human dignity of its members such that it constitutes a "particular social group". The fight for independence from the United Kingdom and unification with the Irish Republic may be very serious political ends for INLA members, but requiring them to abandon their violent means of expressing and achieving these goals does not amount to an abdication of their human dignity.

Moreover, I do not accept that Ward's fear was based on his membership. Rather, in my view, Ward was the target of a highly individualized form of persecution and does not fear persecution because of his group characteristics. Ward feels threatened because of what he did as an individual, and not specifically because of his association. His membership in the INLA placed him in the circumstances that led to his fear, but the fear itself was based on his action, not on his affiliation.

C. Political Opinion

Political opinion was not raised as a ground for fear of persecution either before the Board or the Court of Appeal. It was raised for the first time in this Court by the intervener, the United Nations High Commissioner for Refugees, who, in his factum, expressed the view that the Court of Appeal had "erred in considering that the claimant's fear of persecution was based on membership in an organization". The additional ground was ultimately accepted by the appellant during oral argument. I note that the UNHCR Handbook, at p. 17, paragraph 66, states that it is not the duty of a claimant to identify the reasons for the persecution. It is for the examiner to decide whether the Convention definition is met; usually there will be more than one ground (idem, paragraph 67). While political opinion was raised at a very late stage of the proceedings, the Court has decided to deal with it because this case is one involving human rights and the issue is critical to the case.

Political opinion as a basis for a well-founded fear of persecution has been defined quite simply as persecution of persons on the ground "that they are alleged or known to hold opinions contrary to or critical of the policies of the government or ruling party"; see Grahl-Madsen, supra, at p. 220. The persecution stems from the desire to put down any dissent viewed as a threat to the persecutors. Grahl-Madsen's definition assumes that the persecutor from whom the claimant is fleeing is always the government or ruling party, or at least some party having parallel interests to those of the government. As noted earlier, however, international refugee protection extends to situations where the state is not an accomplice to the persecution, but is unable to protect the claimant. In such cases, it is possible that a claimant may be seen as a threat by a group unrelated, and perhaps even opposed, to the government because of his or her political viewpoint, perceived or real. The more general interpretation of political opinion suggested by Goodwin-Gill, supra, at p. 31, i.e., "any opinion on any matter in which the machinery of state, government, and policy may be engaged", reflects more care in embracing situations of this kind.

Two refinements must be added to the definition of this category. First, the political opinion at issue need not have been expressed outright. In many cases, the claimant is not even given the opportunity to articulate his or her beliefs, but these can be perceived from his or her actions. In such situations, the political opinion that constitutes the basis for the claimant's well-founded fear of persecution is said to be imputed to the claimant. The absence of expression in words may make it more difficult for the claimant to establish the relationship between that opinion and the feared persecution, but it does not preclude protection of the claimant.

Second, the political opinion ascribed to the claimant and for which he or she fears persecution need not necessarily conform to the claimant's true beliefs. The examination of the circumstances should be approached from the perspective of the persecutor, since that is the perspective that is determinative in inciting the persecution. The political opinion that lies at the root of the persecution, therefore, need not necessarily be correctly attributed to the claimant. Similar considerations would seem to apply to other bases of persecution.

Ward's fear of being killed by the INLA, should he return to Northern Ireland, stems initially from the group's threat of executing the death sentence imposed by its court-martial. The act for which Ward was so punished was his assistance in the escape of the hostages he was guarding. From this act, a political opinion related to the proper limits to means used for the achievement of political change can be imputed. Ward had many reasons to go through with the assassination order and only one, that of acting in conformity with his beliefs, for doing what he eventually did. Ward recognized the risk of serious retribution by the INLA upon being caught, as reflected in his testimony before the Immigration Appeal Board:

Q. What type of discipline is it?

A. The discipline is once you are a member you are always a member. And if anybody steps outside those lines of demarcation the only alternative is to assassinate them, do away with them.

Q. So if a person does not tow [sic] the line, once he is a member and he does not tow [sic] the line what happens to him?

A. He will be shot. . . .

Nevertheless, Ward felt that to carry out the INLA's hostage assassination order would have been going too far. He described his reasons for turning the hostages free as follows:

Q. So the order [to shoot the hostages] has come down then, and what happened next?

A. Well I found myself in a predicament.

Q. Yes?

A. Both of conscience and morals, these things all go through your head in a situation like that. Quite frankly, I wanted no part of it at that stage.

Q. Did you express this desire or this feeling to anyone?

A. To one particular person that was involved there. I cannot do this. But rules are rules, if you voice your opinion to the wrong people or too loudly you would be joining the victims.

Q. So what happened next?

A. What happened was, I gave the situation a lot of thought and consideration. I thought of the implications of various actions. Things went through my head and the final conclusion I came to was I could not have any part of it.

Q. The order came down from the Army council. And you knew that it was your responsibility to protect these people so that that order, I guess, could be carried out. Correct?

A. Yes.

Q. And you had a problem with that. You realized that you could not go along with the killing?

A. They were innocent people . . . . I could not live with my own conscience if I permitted this to go on. The decision I came to in my own mind was to try to release him.

To Ward, who believes that the killing of innocent people to achieve political change is unacceptable, setting the hostages free was the only option that accorded with his conscience. The fact that he did or did not renounce his sympathies for the more general goals of the INLA does not affect this. This act, on the other hand, made Ward a political traitor in the eyes of a militant para-military organization, such as the INLA, which supports the use of terrorist tactics to achieve its ends. The act was not merely an isolated incident devoid of greater implications. Whether viewed from Ward's or the INLA's perspective, the act is politically significant. The persecution Ward fears stems from his political opinion as manifested by this act.

The appropriateness of the application of this ground to the facts in this case is confirmed when contrasted with a recent United States Supreme Court disposition of a similar issue. In I.N.S. v. Elias-Zacarias, 112 S.Ct. 812 (1992), a Guatemalan claimant sought asylum because of his fear of persecution at the hands of the anti-government guerrillas owing to his refusal to join them. For the majority, Scalia J. was not convinced that the claimant's motive, nor that perceived by the guerrillas to be his motive, was politically based. He stated, at pp. 815-16:

Even a person who supports a guerrilla movement might resist recruitment for a variety of reasons ‑‑ fear of combat, a desire to remain with one's family and friends, a desire to earn a better living in civilian life, to mention only a few. The record in the present case not only failed to show a political motive on Elias-Zacarias' part; it showed the opposite. He testified that he refused to join the guerrillas because he was afraid that the government would retaliate against him and his family if he did so. Nor is there any indication (assuming, arguendo, it would suffice) that the guerrillas erroneously believed that Elias-Zacarias' refusal was politically based. [Emphasis in original.]

In Ward's case, a contrario, his act was inconsistent with any other possible motive. He was already a member of the INLA; any fear of retaliation could have been dispelled simply by executing the order. The rationale underlying his decision was unequivocal, both in his eyes and in those of the INLA.

A positive labelling of Ward as a "Convention refugee" because of his well-founded fear of persecution for reasons of political opinion meets the concerns of Urie J.A., in the Court of Appeal, who remarked, at p. 678, that it would be absurd to allow Ward into Canada owing to the fact that he had acted contrary to the interests of the INLA, because "[i]f such a view were to be taken anyone who dissents on anything could be said to be a member of a particular social group". Permitting Ward entry on the basis of feared persecution because of political opinion provides the focus needed in this inquiry. Not just any dissent to any organization will unlock the gates to Canadian asylum; the disagreement has to be rooted in a political conviction. This approach to Ward's case would preclude a former Mafia member, for example, from invoking it as precedent.

Section 15 of the Charter

The intervener, Canadian Council for Refugees, has raised the argument that the majority decision of the Federal Court of Appeal imposes two requirements having a discriminatory impact on historically disadvantaged groups such as women and children, by making it more difficult for them to obtain refugee status in Canada. These two requirements are, first, that social group activities be viewed as a possible danger to the state in order to qualify as a social group, and second, that state complicity be present. Essentially, the argument can be reduced to the contention that differential impact will exist since persecution of women and children is less likely to meet these criteria. I do not find this argument convincing, but I need not enter into it further since I have found both these aspects of the majority decision incorrect for other reasons. Recourse to s. 15 of the Charter is, therefore, unnecessary.

Dual Nationality

Ward's citizenship, by virtue of his being a resident of Northern Ireland which forms part of the United Kingdom, effectively endows him with British citizenship; see the British Nationality Act 1981, 1981 (U.K.), c. 61. On January 1, 1983, British citizenship was automatically acquired by all those citizens of the United Kingdom and the Colonies who had the right of abode in the United Kingdom on that date pursuant to the British Nationality Act 1981. During the oral hearing, Ward's counsel effectively admitted the Board's error in this regard and conceded Ward's dual nationality. This makes unnecessary a consideration of burden of proof, but it is right to say that I agree with the Court of Appeal that the Board erred in placing the burden of proof on the Minister. This burden includes a showing of well-founded fear of persecution in all countries of which the claimant is a national.

In considering the claim of a refugee who enjoys nationality in more than one country, the Board must investigate whether the claimant is unable or unwilling to avail him- or herself of the protection of each and every country of nationality. Although never incorporated into the Immigration Act and thus not strictly binding, paragraph 2 of Art. 1(A)(2) of the 1951 Convention infuses suitable content into the meaning of "Convention refugee" on the point. This paragraph of the Convention provides:

Article 1

A. . . .

(2) . . .

In the case of a person who has more than one nationality, the term "the country of his nationality" shall mean each of the countries of which he is a national, and a person shall not be deemed to be lacking the protection of the country of his nationality if, without any valid reason based on a well-founded fear, he has not availed himself of the protection of one of the countries of which he is a national.

As described above, the rationale underlying international refugee protection is to serve as "surrogate" shelter coming into play only upon failure of national support. When available, home state protection is a claimant's sole option. The fact that this Convention provision was not specifically copied into the Act does not render it irrelevant. The assessment of Convention refugee status most consistent with this theme requires consideration of the availability of protection in all countries of citizenship.

This conclusion is bolstered by general rules of statutory interpretation. Section 33(2) of the Interpretation Act, R.S.C., 1985, c. I-21, stipulates that words in the singular include the plural. Consequently, references to "country of nationality" in the definition of "Convention refugee" in s. 2(1) of the Immigration Act should be read as including "countries of nationality".

The appellant argues that the presence of s. 46.04(1)(c) of the Act (resulting from the amendments effected by S.C. 1988, c. 35, s. 14, effective January 1, 1989) is inconsistent with a requirement of demonstrating a lack of protection in all countries of citizenship. Section 46.04(1)(c) precludes eligibility for landed status in Canada for a claimant who has demonstrated his or her status as a Convention refugee, where the claimant is "a national or citizen of a country, other than the country that the person left, or outside of which the person remains, by reason of fear of persecution". The appellant's contention is, essentially, that if the Act's definition of "Convention refugee" were to encompass inability or unwillingness to avail himself of the protection of each country of nationality, then s. 46.04(1)(c) would be redundant.

I am not persuaded by this argument. The right to apply for the status of permanent resident is but one of several consequences flowing from the characterization of a claimant as a Convention refugee. The Convention refugee also benefits from the right to remain in Canada (s. 4(2.1)), the right not to be deported to the country where the refugee has a well-founded fear of persecution (s. 53(1)) and the right to work while in Canada (s. 19(4)(j) of the Immigration Regulations, 1978, SOR/78-172). None of these provisions requires assurance that the claimant has exhausted his or her search for protection in every country of nationality. The exercise of assessing the claimant's fear in each country of citizenship at the stage of determination of "Convention refugee" status, before conferring these rights on the claimant, accords with the principles underlying international refugee protection. Otherwise, the claimant would benefit from rights granted by a foreign state while home state protection had still been available. The reference to other countries of nationality in s. 46.04(1)(c) is probably intended as a double-check on the refugee's lack of access to national protection, in case of changed circumstances or new revelations, before the significant status of permanent resident is bestowed.

As alluded to previously, and as conceded by appellant's counsel to be in error, the Board concluded that it could not make a finding of dual citizenship because there was insufficient evidence to do so. The Board commented, at p. 55, however, that had it

. . . concluded that the claimant was also a national of the United Kingdom, the Board would have made a finding that the claimant's life would be in danger from the INLA if he was returned to the United Kingdom.

This finding, however, is insufficient for the purposes of the determination that must be made by the Board. It does not address the real issue. The fact that Ward's life will be in danger should he be returned either to Ireland or to Great Britain is not disputed by anyone; the question, rather, is whether Ward can be protected from that danger. The Board never made a finding of fact on the real issue ‑‑ the ability of the British to protect Ward.

As explained above, the well-foundedness of a claimant's fear of persecution can be grounded in the concept of "inability to protect", assessed with respect to each and every country of nationality. Since the Board failed to make a finding on this point, as far as Great Britain is concerned, its ultimate finding of fear of persecution there is similarly erroneous. The validity of Ward's claim is dependant upon such a finding. This case must, therefore, be referred back to the Board (now the Immigration and Refugee Board) for a determination as to whether Ward can be afforded protection in Great Britain.

Clearly, the inability of a second state of nationality to protect can be established where the claimant has actually approached the state and been denied protection. Where, as in the case of Ward, the second state has not actually been approached by the claimant, the principles delineated above regarding the home state should apply. In other words, Great Britain should be presumed capable of protecting its nationals.

An underlying premise of this presumption, however, is that citizenship carries with it certain basic consequences. One of these, as noted by MacGuigan J.A., at p. 699, is the right to gain entry to the country at any time. The appellant presented evidence, albeit not by way of expert opinion, of the existence of the Prevention of Terrorism (Temporary Provisions) Act 1984. The current version of this Act (Prevention of Terrorism (Temporary Provisions) Act 1989, 1989 (U.K.), c. 4), which replaced the Prevention of Terrorism (Temporary Provisions) Act 1984, seems to enable the British Government to prohibit a national from being in, or entering, Great Britain, if the national has been "concerned in the commission, preparation or instigation of acts of terrorism" connected with the affairs of Northern Ireland; see ss. 4 and 5 of the Act. Such evidence might serve to rebut the presumption by demonstrating a lack of protection afforded by Great Britain. Denial of admittance to the home territory is offered by the UNHCR in its Handbook, at paragraph 99, as a possible example of what might amount to a refusal of protection. The applicability of the presumption and its rebuttal are matters that depend upon the particular circumstances of this case and which must be determined by the Board.

Conclusion

For these reasons, I would allow the appeal, set aside the order of the Federal Court of Appeal and remit the case back to the Immigration and Refugee Board for an evaluation consistent with these reasons of the appellant's claim with reference to his second state of citizenship, Great Britain.

Appeal allowed.

Solicitors for the appellant: Rekai & Johnson, Toronto.

Solicitor for the respondent: The Deputy Attorney General of Canada, Toronto.

Solicitor for the intervener United Nations High Commissioner for Refugees: United Nations High Commissioner for Refugees, Ottawa.

Solicitor for the intervener Immigration and Refugee Board: Immigration and Refugee Board, Ottawa.

Solicitor for the intervener Canadian Council for Refugees: Parkdale Community Legal Services Inc., Toronto.

• Stevenson J. took no part in the judgment.