Judgments

[1993] 3 F.C. 401

A-1335-91

James Egan and John Norris Nesbit (Appellants) (Plaintiffs)

v.

Her Majesty the Queen in Right of Canada (Respondent) (Defendant)

Indexed as: Egan v. Canada (C.A.)

Court of Appeal, Mahoney, Linden and Robertson JJ.A.—Vancouver, December 9, 1992; Ottawa, April 29, 1993.

Constitutional law — Charter of Rights — Equality rights — Appeal from F.C.T.D. decision denial of spousal benefit under Old Age Security Act not infringing appellant’s rights under Charter, s. 15(1) — Appellants homosexual couple of long-standing — Application for spouse’s allowance denied as appellant not “spouse” under Act, s. 2 — Sexual orientation analogous ground of discrimination under Charter, s. 15(1) — Case law on discrimination reviewed — Similarly situated test rejected by SCC in Andrews case — Discrimination under s. 15(1) requiring distinction based on irrelevant personal differences — Whether distinction based on spousal status as opposed to sexual orientation — Relevance of “other non-spousal relationships” — “Concrete” test more appropriate than “abstract” test to finding of discrimination — Impugned legislation not having disproportionate or more burdensome effect on same-sex couples.

Health and welfare — Application for spouse’s allowance denied as appellant not “spouse” under Old Age Security Act, s. 2 — Whether exclusion of homosexual or same-sex couples from definition of “spouse” discriminatory under Charter, s. 15(1) — Purpose and objective of spouse’s allowance program explained — Finding of discrimination must have regard to purposes and object of impugned legislation, impact on members of disadvantaged group and others excluded — Legislation not having disproportionate or more burdensome effect on same-sex couples.

This was an appeal from a decision of the Trial Division that the denial the of spousal allowance under the Old Age Security Act to one of the appellants did not infringe his equality rights under subsection 15(1) of the Charter. The appellants have resided together, in an homosexual relationship, since 1948. In September of 1986, the appellant Egan, who was then 65 years old, became entitled to receive old age security and guaranteed income supplement benefits. The appellant Nesbit applied to the Department of National Health and Welfare for a spouse’s allowance. His application, in which he described Egan as his spouse, was rejected on the basis that his relationship with the latter did not meet the definition of “spouse” in section 2 of the Act. In determining whether the appellants had been discriminated against on the basis of an analogous ground—sexual orientation—the Trial Judge concluded that there was nothing discriminatory in a law which restricts a benefit to opposite sex couples, and that the benefit which the appellants sought was denied because of their “non-spousal status” and not because of their “sexual orientation”. The issue in this appeal was whether the exclusion of homosexual or same-sex couples from the definition of “spouse” contained in the Act is discriminatory within the meaning of subsection 15(1) of the Charter and if so, whether the impugned legislation can be demonstrably justified as a reasonable limit under section 1 of the Charter.

Held (Linden J.A. dissenting), the appeal should be dismissed.

Per Robertson J.A.: Although sexual orientation is not expressly referred to in subsection 15(1), it is trite law that the list of enumerated grounds is not finite and that the protections available under that section can be extended to those who establish that they are members of a disadvantaged group and can show discrimination on analogous grounds. Sexual orientation can be invoked as an analogous ground of discrimination under subsection 15(1). Andrews v. Law Society of British Columbia is the leading case on discrimination under section 15 of the Charter. This decision expressly rejected the similarly situated test, whereby people who are alike should be treated alike, while those who are unalike should be treated unalike, because that approach excluded any consideration of the nature of the law and could not be applied other than mechanically. Since the appellants’ case was built on the “similarity” between same-sex and opposite-sex couples, the Supreme Court’s decision in Andrews destroys the very premise on which their claim of discrimination rests. It was also held in Andrews that a person is not to be deprived of a benefit because of a distinction based on irrelevant personal differences. The significance of the legal requirement that a distinction be based on an irrelevant personal difference could not be ignored since it is basic to the law of discrimination under section 15 of the Charter. The appellants having identified the distinction in terms of differential treatment being accorded same-sex and opposite-sex couples and then deemed that irrelevant, sexual orientation remains highly relevant. To the extent that the Andrews decision defines discrimination in terms of a distinction based on an irrelevant personal difference, and has ruled out the application of the similarly situated test, the appellants’ claim that the definition of “spouse” in section 2 of the Act is discriminatory must fail.

A finding of discrimination on the basis of sexual orientation is not automatically precluded because a criterion of entitlement is spousal status. Discrimination on the basis of spousal status will be ruled out provided the consequences of the impugned legislation are such that claimants satisfy the Supreme Court’s adverse impact test. The definition of “spouse” found in the Act excludes a broad class of persons cohabiting with another in a non-conjugal relationship such as siblings, friends and relatives. Though same-sex couples are subjected to various forms of “invidious” discrimination, one should ask whether their rights should be determined in isolation of those also affected by the impugned legislation. Where the latter excludes persons other than those who are members of a particular disadvantaged group, proper consideration must be given to the significance of that reality when deciding whether the impact of a distinction is discriminatory. Legislation denying benefits to members of another group may detract from a finding of discrimination contrary to subsection 15(1) of the Charter. There are two different approaches as to what is sufficient to meet the adverse impact test stated in Andrews. The first one, which can be called the “abstract test”, presumes discrimination once it has been found that a distinction is based on a prohibited ground, without regard to the purposes and objects of the impugned legislation, nor the fact that others may also have been denied the benefit. The second approach, which can be labelled the “concrete test” and is mandated by Andrews, is a contextual one and requires a full analysis of the immediate impact of the legislation on members of the disadvantaged group, as well as others who are excluded, as measured against the purposes and objects of the legislation. In view of the objective of the spouse’s allowance program elaborated on by the then Minister of National Health and Welfare, it cannot be said that the legislative scheme has a disproportionate or more burdensome effect on same-sex couples. The denial of the spouse’s allowance does not cross the line from “distinction” to “discrimination”. The Trial Judge was right in holding that the appellants did not benefit because of their non-spousal status rather than because of their sexual orientation.

Per Mahoney J.A. (concurring in the result): The discrimination perceived in the definition of “spouse” is not discrimination directed to homosexuals generally nor to all homosexual couples but only those who have both established a menage and proclaimed their relationship. The attack is, in substance, on the failure of the definition to comprehend the concept of common law marriage between persons of the same sex. As the definition pertains to the appellants and other homosexual couples who hold themselves out as spouses, the conclusion that the distinction made by the definition is discrimination can only be reached by mechanical application of the similarly situated test.

Per Linden J.A. (dissenting): There are two different levels of analysis in a subsection 15(1) inquiry, depending on whether a party seeks an individual remedy under subsection 24(1) of the Charter or a general remedy under subsection 52(1). A party seeking an individualized remedy under subsection 24(1) must prove that the impugned distinction has, directly or indirectly, compromised his equality rights personally, whereas in a remedy under subsection 52(1), it must be established that the relevant legislation violates the rights of the group. The appellants are seeking a subsection 52(1) remedy as a group—gay men and lesbians—rather than individually. In order to make out a claim under subsection 15(1) of the Charter, a party must establish that a law draws a distinction or has a separate impact which violates one of the four equalities (equality before the law, equality under the law, equal protection of the law and equal benefit of the law) and which does so with discrimination. By virtue of the words “of the opposite sex” found in the definition of “spouse” in section 2 of the Old Age Security Act, gay and lesbian partners are specifically excluded from eligibility for spouse’s allowance benefits independent of the exclusion of other groups or individuals. The effect of that distinction is to deny otherwise qualified gay men and lesbians the equal benefit of the law. Subsection 15(1) prohibits discrimination not only on the listed grounds, but also on analogous grounds. The judicial recognition that sexual orientation is an analogous ground of discrimination prohibited under subsection 15(1) is buttressed by a similar legislative recognition that gay men and lesbians have historically suffered disadvantage and prejudice and should be protected against discrimination. Gay men and lesbians are legally, economically, socially and politically disadvantaged. The exclusion of other groups from benefits under the spouse’s allowance program cannot change the fact that lesbian and gay partners are excluded on the basis of their sexual orientation. Evidence that a group is independently disadvantaged is clearly important for determining whether this group is disadvantaged by a distinction and thereby discriminated against. Under the definition of “spouse” in section 2 of the Old Age Security Act, a distinction is drawn on the basis of sexual orientation with the effect that gay and lesbian partners are denied benefits in the form of spouse’s allowance. Since gay men and lesbians are generally disadvantaged as a group through stereotyping and prejudice, there was no requirement that the appellant, Nesbit, prove that he was personally discriminated against. And the fact that he received benefits under a different plan cannot cure the discriminatory impact of a denial of benefits under the spouse’s allowance program based on a personal characteristic related to an enumerated or analogous ground of discrimination. By excluding lesbian and gay partners from the ambit of that program, Parliament has acted in violation of subsection 15(1) of the Charter.

In order to demonstrate that a limit on a Charter right is reasonable and demonstrably justified in a free and democratic society, it must first be established that the objective which the limitation is designed to promote is “of sufficient importance to warrant overriding a constitutionally protected right or freedom”. The appellants conceded that the objective of the spouse’s allowance program under the Old Age Security Act is pressing and substantial so as to satisfy the first requirement under the Oakes test. The second requirement involves a proportionality test which includes three components. Only the first element of this test is met. It was agreed that the means employed under the spouse’s allowance program are rationally connected to achieving the objective of that program. However, the latter does not impair the right guaranteed in subsection 15(1) as little as possible. Extending benefits to gay and lesbian partners in addition to heterosexual partners is a suitable alternative means for achieving the same objective just as effectively. Therefore, the Charter violation occasioned by section 2 of the Old Age Security Act cannot be said to be reasonable and the minimal impairment condition of the proportionality test is not satisfied. Moreover, this is not an instance in which a Charter right is marginally affected, the violation of that right being clear and direct. Since the effects of the measure on the right to receive benefits are not proportional to the objective of the legislation, the impugned provision cannot be saved under section 1 of the Charter.

Where a benefit conferring scheme is involved, the preferred course of action is to read down (that is to sever or strike out) or read in an inconsistent provision rather than striking down the program. The Supreme Court of Canada has identified a number of factors which support the assumption of legislative approval for both reading in and reading down rather than striking down the impugned legislation. The appropriate remedy to bring section 2 of the Old Age Security Act into compliance with subsection 15(1) of the Charter would be to read down (that is to sever or strike out) the words “of the opposite sex” of the definition of “spouse” and to read into the phrase “if the two persons publicly represent themselves as husband and wife” the words “or as in an analogous relationship”.

STATUTES AND REGULATIONS JUDICIALLY CONSIDERED

An Act to Amend the Human Rights Act, S.N.B. 1992, c. 30.

An Act to amend the Old Age Security Act, to repeal the Old Age Assistance Act and to amend other Acts in consequence thereof, S.C. 1974-75-76, c. 58.

Bill C-108, An Act to amend the Canadian Human Rights Act and other Acts in consequence thereof, 3rd Sess., 34th Parl. 1992.

Bill C-6, An Act to amend the Old Age Security Act, S.C. 1979, c. 4.

Canadian Charter of Rights and Freedoms, being Part I of the Constitution Act, 1982, Schedule B, Canada Act 1982, 1982, c. 11 (U.K.) [R.S.C., 1985, Appendix II, No. 44], ss. 1, 15, 24.

Canadian Human Rights Act, R.S.C., 1985, c. H-6, s. 3.

Charter of Human Rights and Freedoms, R.S.Q., c. C-12, s. 10 (as am. by S.Q. 1992, c. 61, s. 3).

Constitution Act, 1867, 30 & 31 Vict., c. 3 (U.K.) (as am. by Canada Act 1982, 1982, c. 11 (U.K.), Schedule to the Constitution Act, 1982, Item 1) [R.S.C., 1985, Appendix II, No. 5], s. 24.

Constitution Act, 1982, Schedule B, Canada Act 1982, 1982, c. 11 (U.K.) [R.S.C., 1985, Appendix II, No. 44], s. 52(1).

Guaranteed Available Income for Need Act, R.S.B.C. 1979, c. 158.

Human Rights Act, R.S.N.S. 1989, c. 214, s. 12 (as am. by S.N.S. 1991, c. 12, s. 1).

Human Rights Act, R.S.Y. 1986 (Suppl.), c. 11.

Human Rights Code, R.S.O. 1990, c. H.19, ss. 1, 2, 3, 5, 6.

Medical Services Act, R.S.B.C., 1979, c. 255.

Medical Services Act Regulations, B.C. Reg. 144/68, s. 2.01 (as am. by B.C. Reg. 5/77).

Old Age Pensions Act, R.S.C. 1927, c. 156.

Old Age Security Act, R.S.C., 1985, c. O-9, ss. 2 (as am. by R.S.C., 1985 (1st Supp.), c. 34, s. 1), 19(1),(2) (as am. idem s. 2).

Old Age Security Act, S.C. 1951, c. 18.

The Government Annuities Act, 1908, S.C. 1908, c. 5.

The Human Rights Code, S.M. 1987-88, c. 45, s. 9(2).

CASES JUDICIALLY CONSIDERED

FOLLOWED:

Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; (1989), 56 D.L.R. (4th) 1; [1989] 2 W.W.R. 289; 34 B.C.L.R. (2d) 273; 36 C.R.R. 193; 91 N.R. 255; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; (1990), 76 D.L.R. (4th) 545; 91 CLLC 17,004; 2 C.R.R. (2d) 1; 118 N.R. 1; 45 O.A.C. 1; R. v. Turpin, [1989] 1 S.C.R. 1296; (1989), 48 C.C.C. (3d) 8; 69 C.R. (3d) 97; 96 N.R. 115; 34 O.A.C. 115; R. v. Oakes, [1986] 1 S.C.R. 103; (1986), 26 D.L.R. (4th) 200; 24 C.C.C. (3d) 321; 50 C.R. (3d) 1; 19 C.R.R. 308; 14 O.A.C. 335.

APPLIED:

Haig v. Canada (1992), 9 O.R. (3d) 495 (C.A.); affg Haig v. Canada (1991), 5 O.R. (3d) 245; 86 D.L.R. (4th) 617 (Gen. Div.).

CONSIDERED:

Schachter v. Canada, [1992] 2 S.C.R. 679; (1992), 93 D.L.R. (4th) 1; 10 C.R.R. (2d) 1; 139 N.R. 1; Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554; Layland v. Ontario (Minister of Consumer and Commercial Relations), [1993] O.J. 575 (QL); R. v. Swain, [1991] 1 S.C.R. 933; (1991), 75 O.R. (2d) 388; 71 D.L.R. (4th) 551; 63 C.C.C. (3d) 481; 5 C.R. (4th) 253; 3 C.R.R. (2d) 1; 125 N.R. 1; 47 O.A.C. 81; Veysey v. Canada (Commissioner of the Correctional Service), [1990] 1 F.C. 321; (1989), 39 Admin. L.R. 161; 44 C.R.R. 364; 29 F.T.R. 74 (T.D.); Brown v. British Columbia (Minister of Health) (1990), 66 D.L.R. (4th) 444; 42 B.C.L.R. (2d) 294 (S.C.); Knodel v. British Columbia (Medical Services Commission), [1991] 6 W.W.R. 728; 58 B.C.L.R. (2d) 356; 91 CLLC 17,023 (B.C.S.C.); Andrews v. Ontario (Minister of Health) (1988), 64 O.R. (2d) 258; 49 D.L.R. (4th) 584; 9 C.H.R.R. D/5089; 88 CLLC 17,023 (H.C.); Dred Scott v. Sandford, 19 How. 393 (1856 U.S. S.Ct.); Janzen v. Platy Enterprises Ltd., [1989] 1 S.C.R. 1252; (1989), 59 D.L.R. (4th) 352; [1989] 4 W.W.R. 39; 58 Man.R. (2d) 1; 25 C.C.E.L. 1; 10 C.H.R.R. D/6205; 89 CLLC 17,011; 47 C.R.R. 274; Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219; (1989), 59 D.L.R. (4th) 321; [1989] 4 W.W.R. 193; 58 Man. R. (2d) 161; 26 C.C.E.L. 1; 10 C.H.R.R. D/6183; 89 CLLC 17,012; 94 N.R. 373; Tétreault-Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22; (1991), 81 D.L.R. (4th) 358; 50 Admin. L.R. 1; 36 C.C.E.L. 117; 91 CLLC 14,023; 4 C.R.R. (2d) 12; 126 N.R. 1.

REFERRED TO:

United States v. Carolene Products Co., 304 U.S. 144 (1938); R. v. Hess; R. v. Nguyen, [1990] 2 S.C.R. 906; [1990] 6 W.W.R. 289; 59 C.C.C. (3d) 161; 79 C.R. (3d) 332; 50 C.R.R. 71; 119 N.R. 353; 46 O.A.C. 13; Veysey v. Canada (Correctional Service) (1990), 43 Admin. L.R. 316; 34 F.T.R. 240; 109 N.R. 300 (F.C.A.); Leshner v. Ontario (No. 2) (1992), 16 C.H.R.R. D/184 (Ont. Bd. Inq.); Douglas v. Canada, [1993] 1 F.C. 264 (T.D.); Reference re Meaning of the word “Persons” in s. 24 of the B.N.A. Act, 1867, [1928] S.C.R. 276; [1928] 4 D.L.R. 98; Re Section 24 of B.N.A. Act, [1930] 1 D.L.R. 98; [1930] A.C. 124 (P.C.); Hunter et al. v. Southam Inc., [1984] 2 S.C.R. 145; (1984), 55 A.R. 291; 11 D.L.R. (4th) 641; [1984] 6 W.W.R. 577; 33 Alta. L.R. (2d) 193; 27 B.L.R. 297; 14 C.C.C. (3d) 97; 2 C.P.R. (3d) 1; 41 C.R. (3d) 97; 9 C.R.R. 355; 84 DTC 6467; 55 N.R. 241; Ontario Human Rights Commission and O’Malley v. Simpsons-Sears Ltd. et al., [1985] 2 S.C.R. 536; (1985), 52 O.R. (2d) 799; 23 D.L.R. (4th) 321; 17 Admin. L.R. 89; 9 C.C.E.L. 185; 86 CLLC 17,002; 64 N.R. 161; 12 O.A.C. 241; Canadian National Railway Co. v. Canada (Canadian Human Rights Commission), [1987] 1 S.C.R. 1114; (1987), 40 D.L.R. (4th) 193; 27 Admin. L.R. 172; 8 C.H.R.R. D/4210; 87 CLLC 17,022; 76 N.R. 161; Rudolf Wolff & Co. v. Canada, [1990] 1 S.C.R. 695; (1990), 69 D.L.R. (4th) 329; 43 Admin. L.R. 1; 41 C.P.C. (2d) 1; 46 C.R.R. 263; 106 N.R. 1; 39 O.A.C. 1; Plessy v. Ferguson, 163 U.S. 537 (1896); Brown v. Board of Education of Topeka, 347 U.S. 483 (1954); R. v. Big M Drug Mart Ltd. et al., [1985] 1 S.C.R. 295; (1985), 60 A.R. 161; 18 D.L.R. (4th) 321; [1985] 3 W.W.R. 481; 37 Alta. L.R. (2d) 97; 18 C.C.C. (3d) 385; 85 CLLC 14,023; 13 C.R.R. 64; 85 CLLC 14,023; 58 N.R. 81; R. v. Edwards Books and Art Ltd., [1986] 2 S.C.R. 713; (1986), 35 D.L.R. (4th) 1; 30 C.C.C. (3d) 385; 87 CLLC 14,001; 55 C.R. (3d) 193; 28 C.R.R. 1; 71 N.R. 161; 19 O.A.C. 239; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; (1989), 58 D.L.R. (4th) 577; 25 C.P.R. (3d) 417; 94 N.R. 167; R. v. Chaulk, [1990] 3 S.C.R. 1303; [1991] 2 W.W.R. 385; (1990), 69 Man. R. (2d) 161; 62 C.C.C. (3d) 193; 2 C.R. (4th) 1; 1 C.R.R. (2d) 1; 119 N.R. 161; Committee for the Commonwealth of Canada v. Canada, [1991] 1 S.C.R. 139; (1991), 77 D.L.R. (4th) 385; 4 C.R.R. (2d) 60; 120 N.R. 241; R. v. Wholesale Travel Group Inc., [1991] 3 S.C.R. 154; (1991), 84 D.L.R. (4th) 161; 67 C.C.C. (3d) 193; 38 C.P.R. (3d) 451; 8 C.R. (4th) 145; 7 C.R.R. (2d) 36; 130 N.R. 1; 49 O.A.C. 161; Canada (Attorney General) v. Mossop, [1991] 1 F.C. 18; (1990), 71 D.L.R. (4th) 661; 32 C.C.E.L. 276; 12 C.H.R.R. D/355; 90 CLLC 17,021 (C.A.); Vogel v. Manitoba (1992), 90 D.L.R. (4th) 84; [1992] 3 W.W.R. 131 (Man. Q.B.); Bliss v. Attorney General (Can.), [1979] 1 S.C.R. 183; (1978), 92 D.L.R. (3d) 417; [1978] 6 W.W.R. 711; 78 CLLC 14,175; 23 N.R. 527; Meinhold v. U.S. (Department of Defense), [1993] W.L. 15899 (C.D. Cal.) (Westlaw).

AUTHORS CITED

Bayefsky, Anne F. “A Case Comment on the First Three Equality Rights Cases Under the Canadian Charter of Rights and Freedoms: Andrews, Workers’ Compensation Reference, Turpin” (1990), 1 S.C.L.R. (2d) 503.

Beaudoin, Gérald-A. and Edward Ratushny, The Canadian Charter of Rights and Freedoms, 2nd ed., Toronto: Carswell Co., 1989. (c. 14 “The Equality Rights” by William Black and Lynn Smith, at p. 563.).

Canada. House of Commons Debates, vol. VI, 1st Sess., 30th Parl., 24 Eliz. II, 1975.

Canada. House of Commons Debates, vol. I, 1st Sess., 31st Parl., 29 Eliz. II, 1979.

Canada. House of Commons. Report of the Parliamentary Committee on Equality Rights: Equality for All, Ottawa, Queen’s Printer, 1985.

Canada, House of Commons, Standing Committee on Health, Welfare and Social Affairs, Minutes of Proceedings and Evidence, Issue No. 25 (June 12, 1975).

“Developments in the Law—Sexual Orientation and the Law” (1989), 102 Harv. L. Rev. 1509.

Elliott, David W. “Comment on Andrews v. Law Society of British Columbia and section 15(1) of the Charter: the Emperor’s New Clothes?” (1990), 35 McGill L.J. 235.

Gibson, Dale “Analogous Grounds of Discrimination Under the Canadian Charter: Too Much Ado About Next to Nothing” (1991), 29 Alta. L. Rev. 772.

Gibson, Dale “Equality for Some” (1991), 40 UNBLJ 2.

Wolfson, Evan “Civil Rights, Human Rights, Gay Rights: Minorities and the Humanity of Different” (1991), 14 Harv. Jl.L. & Pub. Pol’y 21.

APPEAL from a Trial Division decision ([1992] 1 F.C. 687; (1991), 87 D.L.R. (4th) 320; 1 C.E.B. & P.G.R. 8110; 47 F.T.R. 305; 38 R.F.L. (3d) 294 (T.D.)) holding that the denial to a partner in a homosexual relationship of an Old Age Security Act spousal allowance equal to that payable to an opposite sex common law spouse did not infringe equality rights under subsection 15(1) of the Charter. Appeal dismissed.

COUNSEL:

Joseph J. Arvay, Q.C. for appellants (plaintiffs).

Harry J. Wruck, Q.C., Faye E. Campbell, Q.C. and Lisa Hitch for respondent (defendant).

SOLICITORS:

Arvay, Finlay, Vancouver, for appellants (plaintiffs).

Deputy Attorney General of Canada for respondent (defendant).

The following are the reasons for judgment rendered in English by

Mahoney J.A.: I have had the opportunity to read in draft the comprehensive reasons for judgment proposed by my colleagues. They have canvassed the authorities and pertinent literature exhaustively thus sparing me extensive recitation.

This is an appeal from a reported decision of the Trial Division,[1] which concluded that rights guaranteed the appellants by subsection 15(1) of the Canadian Charter of Rights and Freedoms [being Part I of the Constitution Act, 1982, Schedule B, Canada Act 1982, 1982, c. 11 (U.K.) [R.S.C., 1985, Appendix II, No. 44]][2] were not infringed by the denial to Nesbit of spousal benefit under the Old Age Security Act[3] consequent upon the definition of “spouse” by that Act.

2. …

“spouse”, in relation to any person, includes a person of the opposite sex who is living with that person, having lived with that person for at least one year, if the two persons have publicly represented themselves as husband and wife. [Underlining added.]

The appellants seek, as a remedy under subsection 52(1) of the Constitution Act, 1982 [Schedule B, Canada Act 1982, 1982, c. 11 (U.K.) [R.S.C., 1985, Appendix II, No. 44]], a declaration that, in the Act,

… the definition of “spouse”, in relation to any person, includes a person of the same sex who is living with that person, having lived with that person for at least one year, if the two persons have publicly represented themselves as spouses.

Mr. Justice Linden has crafted a declaration that better meets the demands of Schachter v. Canada.[4] The learned Trial Judge made the following finding:[5]

I think it is fair to say that had Nesbit been a woman cohabiting with Egan substantially on the same terms as he in fact cohabited with Egan he would have been eligible for the spouse’s allowance.

That is an unassailable conclusion of fact based on the evidence.

The infringement alleged is discrimination based on sexual orientation. That was conceded to be a ground analogous to discrimination based on “sex”. The weight of authority supports that concession as a matter of law. Nesbit has, by reason of the definition, been denied a benefit under the law equal to that to which an opposite sex common law spouse is entitled. The appellants, as a couple, have likewise been denied that benefit.

As a criterion for determining whether different treatment of different persons prescribed by law is discrimination, the similarly situated test was mortally wounded by the Supreme Court of Canada in Andrews v. Law Society of British Columbia[6] and laid to rest in McKinney v. University of Guelph.[7] It was utterly rejected because, in the words of La Forest J.:

Simply put, I do not believe that the similarly situated test can be applied other than mechanically….

In rejecting the test, McIntyre J., in Andrews, had said:[8]

Recognizing that there will always be an infinite variety of personal characteristics, capacities, entitlements and merits among those subject to a law, there must be accorded, as nearly as may be possible, an equality of benefit and protection and no more of the restrictions, penalties or burdens imposed upon one than another. In other words, the admittedly unattainable ideal should be that a law expressed to bind all should not because of irrelevant personal differences have a more burdensome or less beneficial impact on one than another.

…

The test as stated, however, is seriously deficient in that it excludes any consideration of the nature of the law.

…

Consideration must be given to the content of the law, to its purpose, and its impact upon those to whom it applies, and also upon those whom it excludes from its application.

In the present case, the content of the definition, its impact on those to whom it extends benefit and its impact on those excluded are all clear. What remains is the requirement that it be considered in its relation to its purpose in the context of the Old Age Security Act.

Part I of the Act provides for the payment of a monthly pension at age 65 to Canadian citizens and permanent residents. That pension is payable on application regardless of need. Part II provides for an additional payment to a pensioner, called a guaranteed income supplement, on the basis of need. The income of the pensioner’s spouse, if any, is taken into account in determining need. Part III provides for the payment of a monthly spouse’s allowance, at age 60, to the spouse of a pensioner who is in receipt of a guaranteed income supplement.

The spouse’s allowance was enacted in 1975.[9] In testifying before the Standing Committee on Health, Welfare and Social Affairs, the then Minister of National Health and Welfare, Hon. Marc Lalonde, stated:[10]

Its objective is clear and singular in purpose. It is to ensure that when a couple is in a situation where one of the spouses has been forced to retire, and that couple has to live on the pension of a single person, that there should be a special provision, when the breadwinner has been forced to retire at or after 65, to make sure that particular couple will be able to rely upon an income which would be equivalent to both members of the couple being retired or 60 [sic] years of age and over. That is the purpose of this Bill, no more than that, no less than that.

The legislation is consistent with that. It recognizes the obligation of a conjugal spouse to support his or her partner financially and treats pensioner and spouse as a couple entitled to benefit on the basis of need. The impact of the definition is to deny that benefit to otherwise qualified couples who are not married either pursuant to statute or at common law.

Many couples live together in relationships excluded from the definition. Cohabitation by siblings is a commonplace example; persons otherwise related by blood or marriage do so as well and so do persons not related. They do so for countless personal reasons and combinations thereof, for example: mere convenience, the advantage of pooling resources, shared interests, congeniality, friendship and affection not involving sexual attraction, to have someone with them in disability, failing health or in fear of it, or simply to avoid loneliness and seclusion. Unless subjective pressures are in play, sex, whether same or opposite, need not be a consideration in the choice of a live-in companion. There are those, like the appellants, whose sexual orientation is a determining factor in their choice of partner. Many, possibly most, of those couples do not publicly represent themselves as spouses so that they would benefit from the remedy the appellants seek.

The discrimination perceived in the definition is not discrimination directed to homosexuals generally nor to all homosexual couples but only those who have both established a menage and proclaimed their relationship. The attack is, in substance, on the failure of the definition to comprehend the concept of common law marriage between persons of the same sex. It is precisely and only because Nesbit is similarly situated to the common law spouse of the opposite sex included in the definition that it can be rationally argued that his exclusion by the definition is discriminatory while leaving to another day the exclusion of other non-conjugal couples.

As I appreciate it, to reach the conclusion urged by the appellants, one must distinguish homosexual couples from the general class of non-conjugal couples denied spouse’s allowance. The distinction is necessarily made on the basis of an irrelevant personal difference, sexual orientation; there is no other identifiable difference. One then accepts that, because they have maintained their relationship for at least a year and publicly represented themselves as spouses, they are not legitimately to be distinguished from common law spouses of the opposite sex for pension purposes. That seems to me to invert the teaching of Andrews. As the definition pertains to the appellants and other homosexual couples who hold themselves out as spouses, the conclusion that the distinction made by the definition is discrimination can, in my respectful opinion, only be reached by mechanical application of the similarly situated test.

The learned Trial Judge concluded:[11]

When compared to the unit or group which benefits by the challenged law the plaintiffs fall into the general group of non-spouses and do not benefit because of their non-spousal status rather than because of their sexual orientation.

I agree with that conclusion and, with respect, do not find the reasoning leading to it circular. In my view, the conclusion follows from a consideration of the nature of the Old Age Security Act and the purpose of the spouse’s allowance, not mere acceptance of the definition.

I agree in the result with Mr. Justice Robertson and share his hope that the respondent will not pursue the costs to which she is entitled. I would dismiss the appeal with costs.

* * *

The following are the reasons for judgment rendered in English by

Linden J.A. (dissenting): I am unable to agree.

The issue on this appeal is easy to state but difficult to resolve: does the definition of “spouse” in section 2 of the Old Age Security Act, R.S.C., 1985, c. O-9 violate subsection 15(1) of the Canadian Charter of Rights and Freedoms? Section 2 of that Act provides the following definition:

2. In this Act,

…

“spouse”, in relation to any person, includes a person of the opposite sex who is living with that person, having lived with that person for at least one year, if the two persons have publicly represented themselves as husband and wife;

This is a revised version of an earlier definition added to the Old Age Security Act in 1975 [S.C. 1974-75-76, c. 58, s. 1], which was meant to bring “common law spouses”, who had not been included prior to that time, under the newly enacted spouse’s allowance program. This definition sets out who may be eligible to receive spouse’s allowance benefits paid under subsection 19(1) of the Old Age Security Act. The use of the words “a person of the opposite sex” in the definition clearly restricts eligibility to partners in heterosexual relationships and excludes gay men and lesbians from eligibility.

James Egan and John Nesbit challenge this definition, arguing that it discriminates against them and other gay men on the ground of sexual orientation contrary to subsection 15(1) of the Charter, which reads as follows:

15. (1) Every individual is equal before and under the law and has the right to the equal protection and equal benefit of the law without discrimination and, in particular, without discrimination based on race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.

As a Charter challenge, and one based on sexual orientation, this case is very different from the judgment of the Supreme Court in Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554, because it was argued on the ground of discrimination based on family status, not on sexual orientation, as this case is. In fact, the Supreme Court rejected Brian Mossop’s claim partly because it relied on family status rather than sexual orientation as the ground of discrimination. More importantly, Mossop was concerned basically with a narrow issue of statutory interpretation involving the Canadian Human Rights Act [R.S.C., 1985, c. H-6], not a constitutional challenge; this case, on the other hand, involves a direct challenge to the constitutionality of a legislative provision (Mossop, supra, per Lamer C.J., at page 578). The Supreme Court in Mossop mentioned this point repeatedly, indicating that the analysis might well have been different had the Charter been relied upon. Chief Justice Lamer, writing for the majority, referred to the narrow reach of the Mossop decision, explaining that “this decision [should not] be interpreted as meaning that homosexual couples cannot constitute a `family’ for the purposes of legislation other than the CHRA” (Mossop, supra, per Lamer C.J., at page 582.) The case before us does not raise the issue of the definition of family.

Similarly, this case is also very different from Layland v. Ontario (Minister of Consumer and Commercial Relations), [1993] O.J. 575 (QL) in which the issue of marriage between gay and lesbian partners was considered. Indeed, in Layland, Justice Southey expressly stated (at page 23 (QL)):

Whether parties to homosexual unions should receive the same benefits as parties to a marriage, without discrimination because of the nature of their unions, is another question.

Thus, these two recent decisions are of little help in resolving this case which focusses on a rather different issue: Does the Charter permit the exclusion of gay and lesbian partners from the spouse’s allowance program or does that amount to discrimination founded on sexual orientation? The majority, affirming the Trial Judge, holds that they may be excluded; I dissent from that view.

1.         FACTS

James Egan and John Nesbit have lived together since 1948 in an openly gay relationship. Over the past 45 years, they have shared a long term commitment to each other, thoroughly intermingling their daily lives. They share joint bank accounts, credit cards and property ownership. By their wills, Egan and Nesbit appointed each other as their respective executors and beneficiaries. They also hold absolute power of attorney for each other. Egan and Nesbit publicly exchanged rings and, to their friends and family, they refer to themselves as partners.

On September 14, 1986, Egan reached 65 years of age. On October 1, 1986, Egan became eligible and began receiving old age security and guaranteed income supplement. By letter dated February 25, 1987, Egan applied on behalf of John Nesbit for a spouse’s allowance, pursuant to section 19 [as am. by R.S.C., 1985 (1st Supp.), c. 34, s. 2] of the Old Age Security Act. The appellants were advised by letter from the Department of National Health and Welfare that Nesbit was ineligible for spouse’s allowance.

On July 24, 1989, Nesbit applied for spouse’s allowance pursuant to the provisions of Part III of the Act. In the application, Egan was described as Nesbit’s spouse. On September 8, 1989, Nesbit’s application was also denied by the Department of National Health and Welfare on the basis that he was not the spouse of Egan, as defined in section 2 of the Act. Egan and Nesbit appealed this determination to the Trial Division of this Court, which dismissed their appeal (Egan v. Canada, [1992] 1 F.C. 687 (T.D.)).

The Trial Judge concluded that the spouse’s allowance program is not discriminatory. The Trial Judge acknowledged that the definition of spouse in section 2 of the Act creates a distinction, but determined that that distinction was not discriminatory. He explained at page 703:

The legislation denies the financial benefits, the spouse’s allowance, to homosexual couples which benefits are accorded to heterosexual couples where one spouse has reached the age of 65 and the other is between the age of 60 and 65 but, in my view, that distinction is not made upon the basis of the sexual orientation of the plaintiffs and thus does not discriminate against them on that basis.

The Trial Judge then considered on what basis the distinction was drawn. After reviewing the submissions of counsel, the Trial Judge decided that the distinction created under the spouse’s allowance program is between spouses and non-spouses. He concluded at pages 703-704:

The plaintiffs in this matter do not come within the definition assigned by Parliament to the group which it intended to benefit ….

When compared to the unit or group which benefits by the challenged law the plaintiffs fall into the general group of non-spouses and do not benefit because of their non-spousal status rather than because of their sexual orientation.

In my view, however, this analysis, which is largely adopted by the majority, is circular, returning us to the original issue without satisfactorily resolving it. The issue before this Court is whether the definition of spouse in the Old Age Security Act creates a distinction in the distribution of benefits which is discriminatory on the basis of sexual orientation. That issue cannot be resolved simply by stating that the distinction drawn by the Act is based on the definition of spouse rather than on sexual orientation. It is, after all, the definition of spouse that is being attacked as discriminatory. The question before us is whether the exclusion of gay and lesbian partners from the definition of spouse, and therefore from eligibility for spouse’s allowance, contravenes subsection 15(1) of the Charter. Thus, in my view, this Court must subject the definition of spouse in the Old Age Security Act to a thorough subsection 15(1) analysis.

2.         REMEDIES GENERALLY UNDER SUBSECTIONS 24(1) AND 52(1)

There are two different levels of analysis that may be employed in a subsection 15(1) inquiry. Which level should be used depends on whether a party is seeking an individual remedy under subsection 24(1) or whether the party is seeking a general remedy under subsection 52(1). Following the decision of the Supreme Court in Schachter v. Canada, [1992] 2 S.C.R. 679, it is now clear that subsections 24(1) and 52(1) serve distinct remedial functions. Subsection 24(1) empowers courts to devise individualized remedies. That subsection operates where a law is not by itself unconstitutional but where action taken under that law violates a person’s Charter rights. In that situation, it is the impact or effect of the action on the individual which is relevant. Accordingly, when a party bringing a subsection 15(1) claim seeks an individualized remedy under subsection 24(1), they will have to show that the impugned distinction has, directly or indirectly, compromised their equality rights personally. Evidence that action undertaken pursuant to a particular law violates the equality rights of a group generally will be probative, but by itself it will not be determinative of whether the subsection 15(1) rights of an individual have been violated. Since subsection 24(1) provides individualized remedies, the ultimate focus of a Charter analysis leading to a subsection 24(1) remedy must be on the party personally.

Unlike subsection 24(1), subsection 52(1) applies to invalidate all or part of a law, when the law itself, not an act done under the law, is held to be unconstitutional. To establish that a law violates the Charter and is therefore unconstitutional, reference must be had to the purpose or effect of the law generally. The constitutionality of a law cannot be confirmed by establishing that it does not violate a particular individual’s rights. You cannot prove that something is constitutional generally by showing that it is constitutional in one instance. Thus, where a party raising a subsection 15(1) claim contests the validity of a law under subsection 52(1), it must be established that the relevant legislation violates the rights of the group (i.e., gay men, lesbians, women, African-Canadians, etc.). Of course, where relief is sought under subsection 52(1), the fact that the equality rights of the party raising the claim are violated personally may serve as evidence that the subsection 15(1) equality rights of the group have been violated. Where a provision contravenes the equality rights of an individual as a member of a group, it follows that the provision violates the equality rights of the group. This is not to suggest, however, that a member of a particular group cannot succeed in challenging the validity of a law where there is no indication that their equality rights have been personally infringed. A remedy under subsection 52(1) does not in any way depend upon proof that the subsection 15(1) rights of an individual bringing a claim have been violated.

The need to establish a violation of the equality rights of both the group and the individual will only arise where subsections 24(1) and 52(1) are both relied upon. However, as Chief Justice Lamer observed in Schachter v. Canada, [1992] 2 S.C.R. 679, a subsection 24(1) remedy will rarely be available in conjunction with the operation of section 52. In those exceptional circumstances in which both sections apply, both levels of analysis—individual and general—will be required. Where, as here, that is not the case, the appropriate level of analysis to be used throughout the subsection 15(1) inquiry will correspond to the remedy pursued.

In the case before us, the parties are essentially seeking a subsection 52(1) remedy. Accordingly, my subsection 15(1) analysis will focus on the group—gay men and lesbians—rather than on the parties personally. My concern will mainly be the effect of the spouse’s allowance program on that group generally, rather than on the appellants individually.

3.         SUBSECTION 15(1) OF THE CHARTER

Subsection 15(1) falls under the heading “Equality Rights” in the Charter. In Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, at page 164, the first Supreme Court decision addressing subsection 15(1), Mr. Justice McIntyre observed:

The concept of equality has long been a feature of Western thought. As embodied in s. 15(1) of the Charter, it is an elusive concept and, more than any of the other rights and freedoms guaranteed in the Charter, it lacks precise definition.

Without attempting a precise definition, the Supreme Court has clearly outlined the purpose of subsection 15(1). In R. v. Swain, [1991] 1 S.C.R. 933, at page 994, Chief Justice Lamer proclaims that subsection 15(1) is aimed at “remedying or preventing discrimination against groups suffering social, political and legal disadvantage in Canadian society.” Similarly, in R. v. Turpin, [1989] 1 S.C.R. 1296, at page 1333, Madam Justice Wilson, writing for the Court, declared again that the purposes of subsection 15(1) lie “in remedying or preventing discrimination against groups suffering social, political and legal disadvantage in our society.” It is with this laudable goal in mind that subsection 15(1) must be applied.

The purpose of section 15 enunciated by the Supreme Court is reflected in the “enumerated and analogous grounds” approach adopted by that Court. As Mr. Justice McIntyre explained in Andrews v. Law Society of British Columbia, supra, at page 182:

The … “enumerated and analogous grounds” approach most closely accords with the purposes of s. 15 and the definition of discrimination outlined above and leaves questions of justification to s. 1. However, in assessing whether a complainant’s rights have been infringed under s. 15(1), it is not enough to focus only on the alleged ground of discrimination and decide whether or not it is an enumerated or analogous ground. The effect of the impugned distinction or classification on the complainant must be considered.

The point of this approach is to view the issues raised by a section 15 claim in light of the social, political and historical circumstances of the group raising the challenge. With an understanding of the circumstances of the group coming within the enumerated or analogous ground, a court is able to better appreciate the implications of a distinction that denies a benefit or imposes a burden on the group. This is captured in the specific steps to be followed in an analysis under subsection 15(1).

An inquiry under subsection 15(1) of the Charter involves two stages of analysis: the first refers to equality rights and the second refers to freedom from discrimination. In order to make out a claim under subsection 15(1), a party must establish that a law draws a distinction or has a disparate impact which violates one of the four equalities and which does so with discrimination. Chief Justice Lamer recently summarized the basic framework for an inquiry under subsection 15(1) in R. v. Swain, [1991] 1 S.C.R. 933, at page 992:

The court must first determine whether the claimant has shown that one of the four basic equality rights has been denied (i.e., equality before the law, equality under the law, equal protection of the law and equal benefit of the law). This inquiry will focus largely on whether the law has drawn a distinction (intentionally or otherwise) between the claimant and others, based on personal characteristics. Next, the court must determine whether the denial can be said to result in “discrimination”. This second inquiry will focus largely on whether the differential treatment has the effect of imposing a burden, obligation or disadvantage not imposed upon others or of withholding or limiting access to opportunities, benefits and advantages available to others. Furthermore, in determining whether the claimant’s s. 15(1) rights have been infringed, the court must consider whether the personal characteristic in question falls within the grounds enumerated in the section or within an analogous ground, so as to ensure that the claim fits within the overall purpose of s. 15—namely, to remedy or prevent discrimination against groups subject to stereotyping, historical disadvantage and political and social prejudice in Canadian society.

This summary charts the direction of a subsection 15(1) inquiry. It requires some considerable elaboration.

(a)       The four equalities

At the outset, as the Chief Justice explained, the Court must first determine whether a classification exists or a distinction has been drawn—intentionally or otherwise—which violates one of the four basic equality rights:

1. equality before the law;

2. equality under the law;

3. equal protection of the law; or,

4. equal benefit of the law.

At this stage, explanations, justifications and even questions of discrimination need not be addressed. Thus, a test which asks “whether a distinction is ‘unreasonable’, -‘invidious’, ‘unfair’ or ‘irrational’ imports limitations into s. 15 which are not there…. The equality rights must be given their full content divorced from justificatory factors properly considered under s. 1.” (See R. v. Turpin, [1989] 1 S.C.R. 1296, at page 1328, per Madam Justice Wilson). The inquiry at this stage, therefore, is intended to be straightforward, but, we must still determine which level of analysis is appropriate in the circumstances.

As explained above, the correct level of analysis depends on whether a party is pursuing an individual remedy under subsection 24(1), or whether the party is seeking a general remedy under subsection 52(1). Since the appellants in this case are relying mainly on subsection 52(1), we must determine whether, as a group, gay men and lesbians are denied one of the four equalities. Our inquiry is not limited to the individual circumstances of the appellants. The issue is whether the Old Age Security Act, and the definition of spouse in particular, violates one of the four equality rights of gay men and lesbians generally.

Subject to certain restrictions, a person between the ages of sixty and sixty-five who is the spouse of a pensioner is eligible to receive monthly spouse’s allowance payments under section 19 of the Old Age Security Act. As indicated above, the term “spouse” is defined in section 2 of the Act as follows:

2. In this Act

…

“spouse”, in relation to any person, includes a person of the opposite sex who is living with that person, having lived with that person for at least one year, if the two persons have publicly represented themselves as husband and wife;

The word “includes” signifies that, in addition to the conventional definition of a spouse as a person legally married to another person, the Act also “includes” what are generally referred to as “common law spouses” (although the definition may not correspond precisely with the legal description of a common law spouse.). In contrast, the words “of the opposite sex” are used to create an exclusion. By virtue of those words, persons in same sex relationships are excluded from receiving spouse’s allowance. This definition specifically distinguishes between couples of the opposite sex and same sex couples, excluding the latter. Other types of relationships are also excluded by the definition of spouse under the Act, but that is not determinative of the question of whether a distinction has been drawn which violates one of the four equalities. In any event, it is obvious that gay and lesbian partners are specifically excluded from eligibility for spouse’s allowance benefits independent of the exclusion of other groups or individuals.

On this reading of the definition of the word spouse in section 2 of the Act, heterosexual partners are eligible to apply for and receive spouse’s allowance, subject to certain qualifications, while gay and lesbian partners are not eligible to receive such an allowance. This is an obvious example in which a distinction is drawn between partners in gay and lesbian relationships and partners in heterosexual relationships. The effect of that distinction is to deny spouse’s allowance to men and women in gay and lesbian relationships. Monthly spouse’s allowance payments are clearly a benefit of the law as that phrase is used in subsection 15(1) of the Charter. In fact, spouse’s allowance is specifically included in the definition of “benefit” in the Old Age Security Act (see section 2). One cannot escape the conclusion, then, that the effect of the distinction drawn in the definition of spouse in section 2 of the Old Age Security Act is to deny otherwise qualified gay men and lesbians the equal benefit of the law.

(b)       Discrimination under subsection 15(1)

After determining whether the impugned distinction violates one of the four equality rights, the Court must next consider whether the distinction in question constitutes discrimination for the purposes of subsection 15(1). This is the key component of the subsection 15(1) analysis. As Madam Justice Wilson explained in Turpin, supra, at page 1331:

The internal qualification in s. 15 that the differential treatment be “without discrimination” is determinative of whether or not there has been a violation of the section. It is only when one of the four equality rights has been denied with discrimination that the values protected by s. 15 are threatened and the court’s legitimate role as the protector of such values comes into play.

There are two elements necessary to a finding of discrimination under subsection 15(1). At the outset, a claim of discrimination must be brought within the ambit of subsection 15(1) by establishing that the relevant distinction is drawn on an “enumerated or analogous” ground of discrimination. If it can be established that a claim falls within an enumerated or analogous ground, the final issue is whether the impugned distinction leading to the claim has a discriminatory impact or effect. As the Supreme Court jurisprudence under section 15 makes clear, the approach to this last issue remains the same regardless of whether the claim falls within an enumerated ground or whether it falls within an analogous ground. The Court has always applied the same test whether the group raising the challenge relied on an enumerated or an analogous ground.

(i)         Enumerated or analogous ground

The prohibited grounds listed in subsection 15(1) are representative rather than exhaustive. Accordingly, subsection 15(1) prohibits discrimination not only on the listed grounds, but also on grounds analogous to those expressly set out in the subsection (Andrews v. Law Society of British Columbia, supra, at page 182; Turpin, supra, at page 1332; Swain, supra). However, what counts as analogous cannot be decided simply by searching for similarities between a particular case and the listed grounds. To determine whether a group falls within a category analogous to those specifically enumerated in subsection 15(1), it is necessary to look beyond the circumstances of the particular case. As Madam Justice Wilson explained in Andrews, supra at page 152:

I emphasize, moreover, that this is a determination which is not to be made only in the context of the law which is subject to challenge but rather in the context of the place of the group in the entire social, political and legal fabric of our society.

In other words, the issue at this stage is not whether the group is disadvantaged by the distinction drawn in this particular case, but rather whether the group characteristic is one on which political and social prejudice, stereotyping or historical disadvantage has been, is, or may be based. (See Turpin, supra, at page 1333; Swain, supra, at page 992).

The Supreme Court of Canada has, in the past, on a number of occasions, referred to the “discrete and insular minority” approach embraced by the Supreme Court of the United States in the now famous footnote 4 in United States v. Carolene Products Co., 304 U.S. 144 (1938), at pages 152-153. Under that approach, a court inquires into whether a group is a discrete and insular minority by considering whether they are a disadvantaged group within society in the sense that they are politically disempowered. Such a group will undoubtedly fall within an enumerated or analogous ground of discrimination for the purposes of subsection 15(1). Nevertheless, the “discrete and insular minority” test is but one criterion to be considered in assessing whether a distinction is discriminatory. Certainly a discrete and insular minority will be defined by a characteristic associated with an enumerated or analogous ground of discrimination, but there may be other groups that fall within the enumerated or analogous grounds that do not constitute a discrete and insular minority. A group may be socially, legally, or historically disadvantaged without necessarily being politically disempowered. Indeed, recognizing the limitations of the “discrete and insular minority” approach, Madam Justice Wilson explained that finding a “‘discrete and insular minority’ … is not an end in itself but merely one of the analytical tools which are of assistance in determining whether the interest advanced by a particular claimant is the kind of interest s. 15 of the Charter is designed to protect.” (Turpin, supra, at page 1333; R. v. Hess; R. v. Nguyen, [1990] 2 S.C.R. 906, at page 944.)

Thus, inquiring into whether a group is a discrete and insular minority may play a role in bringing a group within an enumerated or analogous ground of discrimination. Such an inquiry will not, however, preclude a group from coming within the ambit of the protection of subsection 15(1) since other criteria may also be taken into consideration. As explained below, whether a group constitutes a discrete and insular minority or is otherwise independently disadvantaged will also play an important role in the final stage of the analysis, informing the assessment of whether the impugned distinction is discriminatory.

Turning to the facts of this case, Egan and Nesbit claim that section 2 of the Old Age Security Act discriminates on the basis of sexual orientation; it is immediately evident, however, that sexual orientation is absent from the list of prohibited grounds of discrimination in subsection 15(1). This raises the question whether sexual orientation is an analogous ground for the purposes of subsection 15(1). Although not among the enumerated grounds of discrimination, a number of recent court decisions have treated sexual orientation as an analogous ground for the purposes of subsection 15(1) of the Charter. In Veysey v. Canada (Commissioner of the Correctional Service), [1990] 1 F.C. 321 (T.D.), at page 329 (appeal to the F.C.A. dismissed on other grounds [(1990), 43 Admin L.R. 316]), Dubé J. recognized sexual orientation as an analogous ground. Shortly afterwards, Mr. Justice Coultas ruled that discrimination based on sexual orientation contravenes the equality provisions of the Charter (Brown v. British Columbia (Mininister of Health) (1990), 66 D.L.R. (4th) 444 (B.C.S.C.)). In another British Columbia case, Madam Justice Rowles premised her decision on the view that sexual orientation is an analogous ground, a point that had been conceded by the parties (Knodel v. British Columbia (Medical Services Commission), [1991] 6 W.W.R. 728 (B.C.S.C.), at page 743). Earlier this year, in Haig v. Canada (1992), 9 O.R. (3d) 495 (C.A.), the Ontario Court of Appeal noted that counsel for the Attorney General of Canada conceded that sexual orientation is an analogous ground. Mr. Justice Krever remarked (at page 501) “I agree and add that, as a matter of law, the concession is right.” Similar findings have been made by human rights tribunals (Leshner v. Ontario (No. 2) (1992), 16 C.H.R.R. D/184 (Ont. Bd. Inq.), at page D/200) and have been authorized by a consent judgment of the Federal Court (Douglas v. Canada, [1993] 1 F.C. 264 (T.D.)). In fact, the Trial Judge in the case at bar accepted the defendant’s concession that sexual orientation is an analogous ground of discrimination for the purposes of subsection 15(1). (Egan v. Canada, supra, at page 700). The only case in which a Court has ruled that sexual orientation is not covered by subsection 15(1) is Andrews v. Ontario (Minister of Health) (1988), 64 O.R. (2d) 258 (H.C.). However, that case relied on the similarly situated approach to subsection 15(1), which has since been rejected by the Supreme Court in Andrews v. Law Society of British Columbia, supra.

This judicial recognition that sexual orientation is an analogous ground of discrimination prohibited under subsection 15(1) is buttressed by a similar legislative recognition that gay men and lesbians have historically suffered disadvantage and prejudice and should be protected against discrimination. Discrimination on the basis of sexual orientation is expressly prohibited under the human rights statutes in Manitoba, New Brunswick, Nova Scotia, Ontario, Quebec, and the Yukon Territory as well as in eight states of the United States of America: California, Connecticut, Hawaii, Massachusetts, Minnesota, New Jersey, Vermont and Wisconsin. In 1985, the House of Commons Parliamentary Committee on Equality Rights released a report entitled Equality for All (Ottawa: Queen’s Printer) recommending the inclusion of sexual orientation among the prohibited grounds of discrimination in the Canadian Human Rights Act, R.S.C., 1985, c. H-6. Following this report, a proposed amendment to the Act which reflects the Committee’s recommendation (Bill C-108, 1992) has been introduced into Parliament.

Consequently, as two historically disadvantaged groups, disadvantaged on the common ground of sexual orientation, lesbians and gay men are now entitled to the protection of subsection 15(1) of the Charter. In Andrews v. Law Society of British Columbia, supra, at page 153, Madam Justice Wilson explained that “It is consistent with the constitutional status of s. 15 that it be interpreted with sufficient flexibility to ensure the [cad96]unremitting protection[cad39] of equality rights in the years to come.” Wilson J. reasoned that the range of disadvantaged groups in society has changed and will continue to change with changing political and social circumstances. A brief look back into legal history reveals shameful instances in which groups we now consider to be entitled to protection from discrimination have been excluded from constitutional recognition. For instance, in 1856 the United States Supreme Court ruled in Dred Scott v. Sandford, 19 How. 393 (1856 U.S. S.Ct.), at pages 404-405, that individuals whose ancestors were brought to the United States from Africa and sold as slaves were not citizens under the Constitution. Chief Justice Taney stated:

The words “people of the United States” and “citizens” are synonymous terms, and mean the same thing…. The question before us is, whether the class of persons described in the plea in abatement compose a portion of this people, and are constituent members of this sovereignty? We think they are not, and they are not included, and were not intended to be included, under the word “citizens” in the Constitution, and can therefore claim none of the rights and privileges which that instrument provides for and secures to citizens of the United States. On the contrary, they were at that time considered as a subordinate and inferior class of beings, who had been subjugated by the dominant race, and, whether emancipated or not, yet remained subject to their authority, and had no rights or privileges but such as those who held the power and the Government might choose to grant them.

Similarly, in 1928 the Supreme Court of Canada compared women to “criminals and lunatics and minors” when it ruled that women were not “qualified persons” within the meaning of section 24 of the Constitution Act, 1867 [30 & 31 Vict., c. 3 (U.K.) (as am. by Canada Act 1982, 1982, c. 11 (U.K.), Schedule to the Constitution Act, 1982, Item 1) [R.S.C., 1985, Appendix II, No. 5]] and therefore could not be appointed to the Senate (Reference re Meaning of the word “Persons” in s. 24 of the B.N.A. Act, 1867, [1928] S.C.R. 276, revd by Re Section 24 of B.N.A. Act, [1930] 1 D.L.R. 98 (P.C.)). Similar shameful examples of indefensible discrimination in the past can be found here and elsewhere in relation to race, religion, nationality, disability, gender, age, common law unions, divorced people, children born outside of marriage and many others.

One lesson to be learned from these cases is that a decision excluding a group from Charter protection cannot be justified by reference to the discriminatory treatment accorded that group in the past. We must be prepared to admit new grounds of discrimination; if we fail to do so, we perpetuate that discrimination. Certainly that was the view of the Privy Council when it overturned the Supreme Court’s infamous ruling in the Persons case. In a passage which continues to guide the interpretation of our Constitution, Lord Sankey L.C. in Re Section 24 of B.N.A. Act, supra, at pages 106-107, issued the following statement:

The B.N.A. Act planted in Canada a living tree capable of growth and expansion within its natural limits. The object of the Act was to grant a Constitution to Canada.

…

Their Lordships do not conceive it to be the duty of this Board—it is certainly not their desire—to cut down the provisions of the Act by a narrow and technical construction, but rather to give it a large and liberal interpretation ….

The living tree doctrine applies equally to the Canadian Charter of Rights and Freedoms. (See Hunter et al. v. Southam Inc., [1984] 2 S.C.R. 145, at page 155.) We should, therefore, keep this metaphor clearly in mind when determining whether a ground of discrimination should be treated as an analogous ground for the purposes of subsection 15(1).

As with enumerated grounds such as race and sex, a person’s sexual orientation has been a basis for discrimination and persecution throughout history. Gay men and lesbians are legally, economically, socially and politically disadvantaged. Lesbians and gay men endure the constant threat of verbal, physical, and sexual abuse. Harassment is not uncommon. For most of this century, certain forms of gay male sexual expression have been criminalized. Indeed, lesbians and gay men have often felt that they must conceal their lifestyles because of these difficulties. Certainly, gay men and lesbians have been excluded from aspects of public life in the past, including participation in the Armed Forces (Douglas, supra; Meinhold v. U.S. (Department of Defense), [1993] W.L. 15899 (C.D. Cal.) (Westlaw)). In short, lesbians and gay men suffer widespread stereotyping and prejudice in our society, as they have throughout history. Thus, both the weight of recent authority and simple justice leads me to the conclusion that sexual orientation is an analogous ground of discrimination for the purposes of subsection 15(1) of the Charter.

Accepting that sexual orientation is an analogous ground of discrimination, it remains to be determined whether the spouse’s allowance program discriminates on that ground.

(ii)        Discrimination

If, as here, the ground on which the impugned distinction is drawn falls within an enumerated or analogous ground of discrimination for the purposes of subsection 15(1), the final consideration is whether that distinction has a discriminatory impact or effect. The definition of discrimination for the purposes of subsection 15(1) was set out by Mr. Justice McIntyre in Andrews, supra, at page 174:

I would say then that discrimination may be described as a distinction, whether intentional or not but based on grounds relating to personal characteristics of the individual or group, which has the effect of imposing burdens, obligations, or disadvantages on such individual or group not imposed upon others, or which withholds or limits access to opportunities, benefits, and advantages available to other members of society.

According to this definition, a distinction may be discriminatory regardless of whether that distinction was drawn intentionally or unintentionally (Ontario Human Rights Commission and O’Malley v. Simpsons-Sears Ltd. et al., [1985] 2 S.C.R. 536; Canadian National Railway Co. v. Canada (Canadian Human Rights Commission), [1987] 1 S.C.R. 1114). A distinction that is facially neutral will not survive subsection 15(1) scrutiny if it has a disparate impact on a particular group or an individual member of that group. Unintentional systemic discrimination will no more be tolerated under subsection 15(1) of the Charter than will direct intentional discrimination.

While a distinction must be based on grounds relating to personal characteristics of the individual or group in order to be discriminatory, the words “based on” do not mean that the distinction must be designed with reference to those grounds. Rather, the relevant consideration is whether the distinction affects the individual or group in a manner related to their personal characteristics (Janzen v. Platy Enterprises Ltd., [1989] 1 S.C.R. 1252, at page 1279; Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219). For example, in the Brooks decision, the Supreme Court ruled that a distinction based on pregnancy but not based directly on gender constitutes sex discrimination. Chief Justice Dickson reasoned that distinctions based on pregnancy are, “at least, strongly, ‘sex related’.” (Brooks, supra, at page 1244.) Further, in Janzen, the Supreme Court held that sexual harassment is a form of sex discrimination even though gender may not be the only characteristic considered by a harasser. Dickson C.J. explained at pages 1288-1289 that:

The fallacy … is the belief that sex discrimination only exists where gender is the sole ingredient in the discriminatory action and where, therefore, all members of the affected gender are mistreated identically. While the concept of discrimination is rooted in the notion of treating an individual as part of a group rather than on the basis of the individual’s personal characteristics, discrimination does not require uniform treatment of all members of a particular group. It is sufficient that ascribing to an individual a group characteristic is one factor in the treatment of that individual. If a finding of discrimination required that every individual in the affected group be treated identically, legislative protection against discrimination would be of little or no value.

In addition, the Court recognized that, while theoretically both men and women may be subject to sexual harassment, “in the present sex stratified labour market, those with the power to harass sexually will predominantly be male and those facing the greatest risk of harassment will tend to be female”. (Janzen, supra, at page 1284). Thus, where a distinction impacts on a group because of a characteristic related to the group, but affects only a portion of that group, it may be found to be discriminatory even where the exclusion is not entirely restricted to that group. Where a distinction specifically or disproportionately impacts on a group, it may be found to be discriminatory even though the distinction impacts on other groups. For example, it is no answer to a charge of sex discrimination to say that people of colour are also excluded by a distinction. What matters is only that women are excluded as women (or as women of colour). This may be true regardless of whether the exclusion is direct or indirect, particularly where it is a disadvantaged group that is excluded by the distinction.

In the case before this Court, the impugned distinction is drawn on the basis of whether partners in a relationship are of the same sex or “of the opposite sex”. Strictly speaking, it is true that this is not a distinction based directly on sexual orientation, since being in a same sex relationship is not necessarily the defining characteristic of being gay or lesbian. Obviously, many lesbians and gay men may not be involved in relationships. Nevertheless, the distinction in this case is based on a characteristic or matter related to sexual orientation, since it is lesbians and gay men who may enter into same sex relationships. The words “of the opposite sex” used in the definition of spouse are specifically aimed at excluding lesbian and gay partners from eligibility for spouse’s allowance benefits. Those words are not directed towards excluding same sex relationships such as brother-brother or sister-sister relationships since those relationships are excluded on the same basis that opposite sex relationships such as brother-sister relationships are excluded. The observations of the Trial Judge in this case at page 695, are revealing on this point:

I think it is fair to say that had Nesbit been a woman cohabiting with Egan substantially on the same terms as he in fact cohabited with Egan he would have been eligible for the spouse’s allowance.

Accordingly, it is clear that the only factor excluding Nesbit from eligibility for spouse’s allowance was his being in a gay relationship rather than in a heterosexual relationship. Like Nesbit, gay men and lesbians generally are excluded from eligibility for benefits under the spouse’s allowance program as gay men and lesbians, rather than on any other basis. And since being in a gay or lesbian relationship is a matter related to one’s sexual orientation, it is apparent that gay men and lesbians are, therefore, excluded from the program based on a matter in relation to sexual orientation.

The fact that other groups may also be excluded from benefits under the spouse’s allowance program cannot change the fact that gay men and lesbians are excluded from that program on the basis of a characteristic related to sexual orientation. There are three important reasons why the exclusion of others under the program is irrelevant and is no answer to the finding that lesbian and gay partners are excluded on the basis of their sexual orientation. First, the exclusion of others under the program is irrelevant because the claim here is a case of direct discrimination in which sexual orientation is singled out as a factor excluding gay and lesbian partners from eligibility for spouse’s allowance independent of the factors excluding partners in other types of relationships. In other words, the Act itself specifically brings sexual orientation into issue. Gay men and lesbians do not have to distinguish themselves from other non-spouses because the Act already does that.

Second, the exclusion of lesbian and gay relationships is qualitatively different from the exclusion of other types of relationships. Lesbians and gay men are a disadvantaged group. Under the “enumerated and analogous grounds” approach to subsection 15(1), it is incumbent upon the Court to understand a distinction involving a disadvantaged group in light of their historical, social, or political disadvantage. The disadvantage endured by lesbians and gay men is particularly relevant to their exclusion from spouse’s allowance benefits since the disadvantage they have suffered throughout history and which they continue to suffer in our society currently is premised, in part, on the view that gay and lesbian relationships are not legitimate or worthy. To treat lesbian and gay relationships simply like all other non-spousal relationships is to rely on and perpetuate the prejudiced view of the legitimacy and worth of those relationships. It would be paradoxical indeed if a decision under section 15 were itself to be based on prejudice and stereotyping.

Finally, whether or not other individuals or groups are excluded from benefits is irrelevant to this claim since what really matters is whether there is something about gay and lesbian relationships that justifies their exclusion. That others are also treated as non-spouses does not resolve whether gay and lesbian partners should be treated as spouses for the purposes of the spouse’s allowance program. Referring to the exclusion of others under the definition of spouse results from a refusal to directly address the true reason for the exclusion of gay and lesbian partners. Another way of understanding this point is to recognize that the exclusion of gay and lesbian partners cannot be justified on the basis that they are defined by the Act as being non-spouses since the issue before the Court is whether defining and treating lesbian and gay partners as non-spouses is discriminatory where non-spouses are excluded from benefits. Accordingly, the Court must address the reason why gay and lesbian partners are not treated as spouses under the Act. That reason cannot be simply that others are also excluded by the definition of spouse. The exclusion of gay and lesbian partners cannot pass scrutiny unless there is a valid non-discriminatory reason for their exclusion. Thus, the claim that excluding gay and lesbian partners from eligibility for the spouse’s allowance program is discriminatory cannot be answered by merely indicating that others are also excluded from eligibility. The exclusion of others is not, by itself, a reason for excluding lesbian and gay partners.

In my view, it is no answer to a charge of discrimination to say that admitting one group to the benefits program would leave another unjustly excluded. To use the possibility that other persons may be discriminatorily excluded from benefits under the spouse’s allowance program as a basis for denying the equality claim of gay men and lesbians would be, to quote the Supreme Court, “equality with a vengeance” (Schachter, supra, at page 702). The better solution is to address the claim before the Court, leaving other groups or individuals unjustly excluded under the spouse’s allowance program to advance their own claims, if they wish, and to deal with them on their own merits in due course.

We are left, then, to determine whether the distinction, based on grounds relating to their personal characteristics, excluding gay and lesbian partners from eligibility for spouse’s allowance benefits is discriminatory as that term was defined by Mr. Justice McIntyre in Andrews v. Law Society of British Columbia, supra.

In considering the impact of a distinction based on grounds relating to personal characteristics, Mr. Justice McIntyre in Andrews, supra, at page 174, instructs us to examine whether the distinction has the effect of imposing burdens, obligations, or disadvantages. McIntyre J. also advises that a distinction is discriminatory if it withholds or limits access to opportunities, benefits, and advantages available to other members of society (ibid, at page 174). It is at this stage, when assessing the impact or effect of a distinction, that the complexity of the analysis surfaces. This assessment must consider the setting in which a distinction is drawn, as well as the circumstances of the group or individual affected by the distinction. A formal analysis of the kind associated with the similarly situated test will not be sufficient; rather a contextualized approach is required, one which weighs prejudice and stereotyping.

The Supreme Court of Canada has unequivocally repudiated the similarly situated test. That test, which continues to flourish in the United States, requires that persons who are similarly situated be similarly treated and that persons who are differently situated be differently treated. However, the Supreme Court in this country has made it clear that, under section 15, a distinction cannot be assessed by means of a simple calculus which demands similar treatment for similarly situated individuals. To be sure, differential treatment is not necessarily discriminatory, while similar treatment may well be discriminatory. As Mr. Justice McIntyre explained in Andrews, supra, at page 164:

It must be recognized at once, however, that every difference in treatment between individuals under the law will not necessarily result in inequality and, as well, that identical treatment may frequently produce serious inequality.

Yet rejecting the similarly situated test does not entail the rejection of all comparative analysis under subsection 15(1). The Supreme Court’s repudiation of the similarly situated test must be understood in light of Mr. Justice McIntyre’s comments in Andrews, supra, at page 164, that equality, as that concept is used in subsection 15(1), is “a comparative concept, the condition of which may only be attained or discerned by comparison with the condition of others in the social and political setting in which the question arises.” The Court rejected the similarly situated test because it was thought to provide too narrow a basis for assessing equality claims. The repudiation of that test should not, therefore, be used to restrict the breadth of the analysis under section 15 so as to impede claims raised by disadvantaged groups. Rather than repudiating comparison under section 15, the Supreme Court was rejecting a limited form of comparison that failed to address the broader social, historical and political circumstances of the group raising the claim independent of the legislation in question. Accordingly, comparison remains an acceptable feature of a subsection 15(1) analysis (e.g., Andrews, supra; R. v. Hess; R. v. Nguyen, supra; Swain, supra; and, Rudolf Wolff & Co. v. Canada, [1990] 1 S.C.R. 695). Indeed, a group raising a section 15 challenge to a benefit conferring scheme can legitimately compare itself to a group receiving benefits under the scheme in order to establish that the exclusion of the first group is discriminatory. That type of analysis is certainly appropriate; however, such comparison cannot take place solely within the four corners of the legislation. The development that has taken place is that the similarly situated test, which operates exclusively within the confines of the impugned legislation, has been replaced by the “enumerated and analogous grounds” approach which requires us to examine the broader social, political and historical circumstances of the group raising the challenge. Frequently those circumstances will only be understood through comparison to the circumstances of other groups in our society.

Having rejected the similarly situated test, Mr. Justice McIntyre endorsed a fuller analysis under subsection 15(1) [at page 168]:

Consideration must be given to the content of the law, to its purpose, and its impact upon those to whom it applies, and also upon those whom it excludes from its application. The issues which will arise from case to case are such that it would be wrong to attempt to confine these considerations within such a fixed and limited formula.

Under this more complex, contextualized analysis, a distinction must be reviewed by taking into consideration the circumstances of the affected group or the individual member of the group. As Mr. Justice McIntyre explained in Andrews v. Law Society of British Columbia, supra, at page 180:

The analysis of discrimination in this approach must take place within the context of the enumerated grounds and those analogous to them.

In applying this “enumerated and analogous” grounds approach, it is important to recognize that an independently disadvantaged group may be more easily disadvantaged by a distinction than an advantaged group in our society (R. v. Turpin, supra, at page 1333, per Madam Justice Wilson). As a result, evidence that a group is independently disadvantaged is clearly important for determining whether a group is disadvantaged by a distinction and thereby discriminated against. In R. v. Turpin, supra, at page 1332, Madam Justice Wilson stated:

If the larger context is not examined, the s. 15 analysis may become a mechanical and sterile categorization process conducted entirely within the four corners of the impugned legislation. A determination as to whether or not discrimination is taking place, if based exclusively on an analysis of the law under challenge is likely, in my view, to result in the same kind of circularity which characterized the similarly situated similarly treated test clearly rejected by this Court in Andrews.

Upon considering the larger context, where there is an indication that a group is independently disadvantaged, the indicia of discrimination—prejudice and stereotyping—will be more readily discoverable and may have a more deleterious effect (R. v. Turpin , supra, at page 1333; Andrews v. Law Society of British Columbia, supra, at page 152, per Madam Justice Wilson). Moreover, while a disadvantaged group is not required to demonstrate that these indicia of discrimination are present, a finding of discrimination will be made more readily where stereotyping or prejudice is shown.

Turning to the case before us, I have already concluded that the definition of spouse in the Old Age Security Act draws a distinction that denies gay and lesbian partners the equal benefit of the law. The final consideration, then, is whether that distinction has a discriminatory impact or effect. Under the definition of discrimination offered by Mr. Justice McIntyre in Andrews, supra, this last consideration is virtually satisfied by definition once we have concluded that a distinction has been drawn on the basis of an analogous ground denying equal benefit of the law to a disadvantaged group or an individual member of the group. The definition of discrimination, set out by Mr. Justice McIntyre in Andrews, supra, at page 174, covers:

… a distinction, whether intentional or not but based on grounds relating to personal characteristics of the individual or group… which withholds or limits access to … benefits … available to other members of society.

Under the definition of spouse in section 2 of the Old Age Security Act, a distinction is drawn on the basis of sexual orientation with the effect that gay and lesbian partners are denied benefits in the form of spouse’s allowance. Since this is primarily a claim for a section 52 remedy, there is no requirement that Nesbit prove that he was personally discriminated against. All of the requisite criteria for a finding of discrimination are, therefore, met.

The respondent in this case argued that the impugned distinction in the spouse’s allowance program is not motivated by stereotypical assumptions but rather is based upon sound economic, administrative and institutional considerations. This type of argument has been unsuccessfully raised in a number of similar cases before the Supreme Court. In Tétreault-Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22, the Court expressly rejected the Government’s contention that its decision was based on economic, administrative, and institutional considerations and was not motivated by stereotypical assumptions. Mr. Justice La Forest declared, “‘[t]his is all irrelevant, since as Andrews v. Law Society of British Columbia made clear … not only does the Charter protect from direct or intentional discrimination; it also protects from adverse impact discrimination, which is what is in issue here’” (supra, at page 41). For this same reason, it is irrelevant whether lesbian and gay partners are specifically excluded under the spouse’s allowance program or whether Parliament has simply failed to include them in the program. Even if gay and lesbian partners were simply left out of the definition of spouse so that the program was merely underinclusive, the program would still be discriminatory. This point was made clear in Brooks, supra, at page 1240, where Chief Justice Dickson stated that “[u]nderinclusion may be simply a backhanded way of permitting discrimination.” Thus, regardless of the form or intent of the exclusion of gay and lesbian partners from the spouse’s allowance program, that exclusion is discriminatory.

The finding that the spouse’s allowance program is discriminatory is confirmed by reference to the supporting considerations which the Supreme Court identified as relevant to a subsection 15(1) analysis. I have already concluded that gay men and lesbians are generally disadvantaged as a group through stereotyping and prejudice. Those indicia of discrimination can be discerned in the circumstances of this case. The exclusion of gay and lesbian partners from eligibility for spouse’s allowance under the Old Age Security Act relies on a stereotyped view of lesbian and gay relationships. The Act, consequently, treats gay and lesbian relationships as less worthy than heterosexual ones. It fails to recognize that long-term affection and mutual support may characterize these relationships, much like heterosexual relationships. The definition of spouse in section 2 of the Act is premised on a negative, stereotyped view of lesbian and gay relationships, which is implicitly contrasted to an idealized view of the heterosexual family. Gay and lesbian relationships share certain attributes of heterosexual ones, but they may also be different, in some respects, from idealized heterosexual families. Nevertheless, they are deserving of respect and dignity. In reality, there are as many diverse gay and lesbian relationships as there are heterosexual relationships. There can be no single, immutable, correct norm for all human relationships.

As Madam Justice L’Heureux-Dubé (dissenting) explained in another context in Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554, at page 634:

It is not anti-family to support protection for non-traditional families. The traditional family is not the only family form, and non-traditional family forms may equally advance true family values.

In order to support stable family relationships, it is not necessary to oppose all non-traditional variations of those relationships. Quite the contrary, loving, continuous and worthy family relationships may be established in a variety of forms, all of which are deserving of respect and support.

This view fits well with the jurisprudence under subsection 15(1) of the Charter. In rejecting the similarly situated test, the Supreme Court has rejected the idea that groups must be the same as or substantially similar to an advantaged group in order to claim equal entitlement to benefits. It follows that, like heterosexual partners, gay and lesbian partners should be equally entitled to spouse’s allowance regardless of whether their relationship corresponds to an idealized heterosexual relationship. Indeed, the altered definition of spouse in the Old Age Security Act was specifically designed to include common law spouses, who had been unfairly excluded before, in an effort to accommodate a new type of heterosexual relationship, that did not correspond to the idealized family at that time. I should emphasize, though, that the fact that gay and lesbian partners are not required to show that their relationship corresponds to an idealized heterosexual relationship does not preclude them from establishing that lesbian and gay relationships share certain attributes with heterosexual relationships in order to show that they come within the purpose of the spouse’s allowance program. Moreoever, as I indicated earlier, the Supreme Court’s rejection of the similarly situated test does not foreclose this type of analysis so long as it is not undertaken in isolation of the broader circumstances of the group in question.

Lesbians and gay men are independently disadvantaged in that they have historically suffered legal, political and social prejudice and stereotyping. As a result, the denial of benefits such as spouse’s allowance may be particularly damaging to lesbians and gay men because this perpetuates those stereotypes and prejudices. This is especially significant in this case since the refusal of spouse’s allowance reflects negatively on the worth of lesbian and gay relationships, contributing to the stereotype that those relationships are without merit. As a result, the fact that similar benefits may be available to gay and lesbian partners under a different scheme—a provincial rather than a federal scheme in this case—cannot remedy the discriminatory impact of denying benefits to gay and lesbian partners on the same footing as heterosexual partners. If at all relevant, the impact of other legislative benefit schemes should only be considered under the section 1 analysis, not under the section 15 equality guarantee. The Supreme Court has made it clear that any limitations on the promise of equality in subsection 15(1) must be addressed under section 1 (Turpin, supra, at page 1328). Restricting a group’s eligibility for equal benefits (merely on the basis of a personal characteristic related to a ground of discrimination) so that those benefits are available only under a separate or different scheme places a limitation of separateness or difference on the promise of equality. Such a compromise is reminiscent of the now-discredited “separate but equal doctrine”, developed by the United States Supreme Court in Plessy v. Ferguson, 163 U.S. 537 (1896), which supported discrimination against African-Americans and other people of colour. That doctrine was widely condemned and was officially rejected in the landmark case of Brown v. Board of Education of Topeka, 347 U.S. 483 (1954). In this country, the separate but equal doctrine was rejected by the Supreme Court in Andrews, supra, at page 166, as a loathsome artifact of the similarly situated approach. One cannot avoid the conclusion that offering benefits to gay and lesbian partners under a different scheme from heterosexual partners is a version of the separate but equal doctrine. That appalling doctrine must not be resuscitated in Canada forty decades after its much-heralded death in the United States.

The Trial Judge recognized the irrelevance of an alternative benefit program to the issue of discrimination under subsection 15(1). It was acknowledged that, in the exceptional circumstances of this case, Nesbit may actually have received more under the British Columbia G.A.I.N. program as a single individual than he would have under the spouse’s allowance program (Guaranteed Available Income for Need Act, R.S.B.C. 1979, c. 158, as amended). Nesbit, though, only received greater benefits under the Guaranteed Available Income for Need Act because of a disability which prevented him from working. The Trial Judge, at page 698 supra, effectively responded to these facts as follows:

Either the plaintiffs are entitled to claim the spouse’s allowance or they are not. The fact that the plaintiffs have claimed under a provincial social assistance plan and have received payments in excess of those which they would have received under the federal spouse’s allowance had they been treated as spouses under that latter program is not relevant to the question of their entitlement. The question is not whether the plaintiffs are getting equal or higher benefits by claiming as single individuals or as spouses but whether they have been deprived of benefits under the federal program for spouse’s allowances to which they may have been legally entitled.

The discriminatory impact of denying benefits to an individual or group based on their personal characteristics cannot be measured in dollars and cents. Discrimination is as much a psychological or emotional harm as it is a material injustice. In Andrews, supra, at page 171, Mr. Justice McIntyre explained:

It is clear that the purpose of s. 15 is to ensure equality in the formulation and application of the law. The promotion of equality entails the promotion of a society in which all are secure in the knowledge that they are recognized at law as human beings equally deserving of concern, respect and consideration.

Thus, the fact that Nesbit received benefits under a different plan cannot cure the discriminatory impact of a denial of benefits under the spouse’s allowance program based on a personal characteristic related to an enumerated or analogous ground of discrimination. It is no defence to the unconstitutional denial of benefits to say that other potential benefits may be available to those discriminated against under a different scheme.

In addition to the psychological or emotional harm of such a differentiation, there may be another non-economic disadvantage to receiving benefits under the provincial program rather than under the federal spouse’s allowance program. Provincial benefits are not viewed in the same light as payments under the Old Age Security Act (Minutes of Proceedings and Evidence of the Standing Committee on Health, Welfare and Social Affairs, Hearings on Bill C-62 An Act to amend the Old Age Security Act, to repeal the Old Age Assistance Act and to amend other Acts in consequence thereof, S.C. 1974-75-76, c. 58, June 12, 1975, at pages 25:11, 25:18, 25:21 and 25:28). In contrast to spouse’s allowance, receiving “relief payments” under a provincial scheme may be demeaning for some. Whether or not this is the case, it is obvious that the federal Government recognized a need to institute a spouse’s allowance program, notwithstanding the existence of provincial social benefits schemes.

Although the purpose of an Act may be found to be discriminatory, that is not the case here. On the arguments before me, it is evident that partners in gay and lesbian relationships may face situations that fit within the overall purpose of the spouse’s allowance program. The objective of that program was described by the Minister of National Health and Welfare, the Honourable Marc Lalonde as follows (Minutes of Proceedings and Evidence of the Standing Committee on Health, Welfare and Social Affairs, Hearings on Bill C-62, June 12, 1975, at page 25:7):

Its objective is clear and singular in purpose. It is to ensure that when a couple is in a situation where one of the spouses has been forced to retire, and that couple has to live on the pension of a single person, that there should be a special provision, when the breadwinner has been forced to retire at or after 65, to make sure that particular couple will be able to rely upon an income which would be equivalent to both members of the couple being retired or 60 years of age and over. That is the purpose of this Bill, no more than that, no less than that.

The particular circumstances of Nesbit and his partner Egan are evidence that gay men and lesbians may find themselves in precisely the circumstances that the spouse’s allowance program was fashioned to address. Undoubtedly other gay and lesbian partners find themselves in similar circumstances. To engage in a further analysis of the purpose of the spouse’s allowance program at this stage of the analysis would be to pre-empt section 1 of the Charter. However, for clarity, I should point out that the spouse’s allowance program was not one designed to benefit women exclusively, although they may be the main beneficiaries. It is not an affirmative action program, so that others cannot be excluded from benefits on that basis.

Thus, I must conclude that, by excluding lesbian and gay partners from the ambit of the spouse’s allowance program, while obviously considering it necessary to have such a program for heterosexual partners, Parliament has acted in violation of subsection 15(1) of the Charter. In my view, the definition of spouse in section 2 of the Old Age Security Act, therefore, discriminates against gay men and lesbians, contrary to subsection 15(1) of the Charter. It remains to be seen, however, whether that violation may be justified under section 1 of the Charter.

4.         SECTION 1 OF THE CHARTER GENERALLY

The basic framework for handling section 1 of the Charter enunciated in R. v. Oakes, [1986] 1 S.C.R. 103 remains in place, but it has been clarified by subsequent decisions. As described in Oakes, supra, the civil standard of proof—proof by a preponderance of probability—governs the entire section 1 analysis. But, while the civil standard of proof admits of some flexibility, it is clear that the degree of probability must be commensurate with the circumstances.

In order to demonstrate that a limit on a Charter right is reasonable and demonstrably justified in a free and democratic society, the party seeking to uphold the limitation must satisfy two requirements. First, it must be established that the objective which the limitation is designed to promote is “of sufficient importance to warrant overriding a constitutionally protected right or freedom” (R. v. Big M Drug Mart Ltd. et al., [1985] 1 S.C.R. 295, at page 352). At a minimum, an objective must be “pressing and substantial in a free and democratic society” to qualify as sufficiently important (R. v. Oakes, supra, at pages 138-139).

If this requirement is met, the second requirement involves a proportionality test. The proportionality test includes three components. First, the measure limiting the Charter right must be rationally connected to the intended objective. In other words, the measure must be carefully designed to achieve its objective without being arbitrary, unfair, or based on irrational considerations. Second, the limiting measures must impair the Charter right as little as possible. This condition has been modified by decisions subsequent to Oakes, supra (R. v. Edwards Books and Art Ltd., [1986] 2 S.C.R. 713; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; R. v. Chaulk, [1990] 3 S.C.R. 1303; and, Committee for the Commonwealth of Canada v. Canada, [1991] 1 S.C.R. 139). The issue of when the modified version of the minimal impairment branch of the proportionality test is applicable and when, or if, the conventional Oakes version should be relied upon may not yet be settled (see McKinney v. University of Guelph, supra, at pages 398-405; R. v. Chaulk, supra, at pages 1388-1393, per Madam Justice Wilson; and R. v. Wholesale Travel Group Inc., [1991] 3 S.C.R. 154, at page 260, per Madam Justice McLachlin). Under the modified approach, the minimal impairment condition depends on whether Parliament could “reasonably have chosen an alternative means which would have achieved the identified objective as effectively” (Chaulk, supra, at page 1341, per Lamer C.J.). Third, the effects of the measures must be proportional to the significance of the objective which is to be achieved. An objective that is no more than pressing and substantial should not override a Charter right, if the effect of the means used to accomplish that objective severely compromise the rights of an individual or group. A provision limiting a Charter right that fails to satisfy any one of these criteria will not be saved under section 1.

(a)       Applying Section 1 of the Charter

The appellants in this action concede that the objective of the spouse’s allowance program under the Old Age Security Act is pressing and substantial so as to satisfy the first requirement under the Oakes test. At least for the purposes of this appeal, I am prepared to accept the appellants’ concession on this point. In general terms, the objective of the spouse’s allowance program is to ensure that when one partner in a couple retires, that couple will continue to receive income equivalent to the amount that would be earned if both members of the couple were retired. On its face, that objective is sufficiently pressing and substantial to justify the limitation of subsection 15(1) of the Charter. As Chief Justice Lamer remarked, “It will be a rare occasion when a benefit conferring scheme is found to have an unconstitutional purpose” (Schachter, supra, at page 721). As this does not appear to be one of those rare occasions, the first part of the section 1 test is satisfied.

Proceeding to the proportionality branch of the test, counsel for the parties agree that the means employed under the spouse’s allowance program are rationally connected to achieving the objective of that program. No useful purpose would be served by going behind that agreement. In any event, the spouse’s allowance program fails to pass muster under the last two elements of the proportionality test.

As a result of the definition of spouse in the Old Age Security Act, the spouse’s allowance program does not impair the right guaranteed in subsection 15(1) as little as possible. Even applying the modified minimum impairment test, it is clear that Parliament could “reasonably have chosen an alternative means which would have achieved the identified objective as effectively.” In this case, the alternative method is readily apparent. Parliament could have achieved the objective of the spouse’s allowance program equally as effectively without violating subsection 15(1) on the basis of sexual orientation by including lesbian and gay partners in the program. I should point out that any impact on the effectiveness of the spouse’s allowance program of the increased costs resulting from extending benefits to gay and lesbian partners is not relevant at any stage under a section 1 inquiry (see Schachter, supra, at page 709). Therefore, setting aside all budgetary considerations at this stage, providing benefits to lesbian and gay partners under this program would not in any way undermine the effectiveness of the program or compromise the realization of the program’s objective. Extending benefits to gay and lesbian partners in addition to heterosexual partners is a suitable alternative means for achieving the same objective just as effectively.

In their submissions, counsel for the respondent argued that the spouse’s allowance program impairs the rights of gay men and lesbians as little as possible because other programs are intended to address their financial needs. The respondent claimed that the spouse’s allowance program is part of an interconnected social benefits network of both provincial and federal programs. This same argument was earlier rejected as an answer to the charge of discrimination but may be properly considered here as a justification for the limitation of the subsection 15(1) right. However, upon careful scrutiny, it becomes apparent that the benefits available to people in Nesbit’s circumstances under the provincial scheme are not substantially similar to benefits paid under the spouse’s allowance program. The objectives of other social benefits legislation are not co-extensive with the objectives of the spouse’s allowance program. Indeed, if it were true that other programs accomplished precisely the same objectives as the spouse’s allowance program, the latter program would be entirely redundant and therefore unnecessary. That is not the case. In so far as section 19 of the Old Age Security Act responds to the particular circumstances of elderly partners in a dependent relationship, that section is distinct from other social benefits legislation. The provincial program under which Nesbit received benefits provides income for people in need (Guaranteed Available Income for Need Act, supra). The G.A.I.N. program is not integrated with the spouse’s allowance program and does not provide equivalent benefits as compared with individuals between the ages of 60 and 65 under the spouse’s allowance program. It cannot be said that the Guaranteed Available Income for Need Act or any other legislation “attempts to fill the gap” created by the denial of spouse’s allowance benefits to lesbian and gay partners (Tétreault-Gadoury, supra).

I am not persuaded that the Guaranteed Available Income for Need Act was developed with the limitations of the spouse’s allowance program in mind nor even that repairing those limitations was one of the objectives of that Act. As a result, the claim that the needs of gay and lesbian partners were meant to be looked after by provision for subsistence benefits under provincial programs is not convincing. That is not a reasonable alternative to extending eligibility to lesbian and gay partners under the spouse’s allowance program. There is no compelling reason for adopting a discriminatory method of distributing benefits when, as here, a suitable alternative means for achieving the objective of the program is readily available. In order for a discriminatory method of distribution to be reasonable, it must be necessary to use that method rather than a non-discriminatory method. This point was made by Madam Justice McLachlin who explained, “It follows from the fact that the right in question is constitutionally guaranteed as fundamental in our society that it should be trenched on no more than is clearly necessary and justified” (Committee for the Commonwealth of Canada, supra, at page 246).

In this case, the discriminatory method of distribution is not necessary since extending eligibility for spouse’s allowance to gay and lesbian partners is a reasonable alternative. In light of the existence of a suitable alternative means for achieving the same objective equally effectively, the Charter violation occasioned by section 2 of the Old Age Security Act cannot be said to be reasonable. Accordingly, the minimal impairment condition of the proportionality test is not satisfied, even if we apply the modified approach to minimal impairment. The violation of subsection 15(1) brought about by section 2 of the Old Age Security Act is not, therefore, permitted under section 1 of the Charter.

Having said that the impugned provision fails the minimal impairment condition of the proportionality requirement under section 1, there is no need to continue to the final stage of the analysis. Nevertheless, were that analysis required, I would conclude that the effect of the measure on the guaranteed Charter right is not proportional to the objective of the program. The effect of section 2 of the Old Age Security Act is to deny equal benefit of the law to gay and lesbian partners by denying them spouse’s allowance benefits completely. This is not an instance in which a Charter right is marginally affected. The violation in this case is clear and direct. And, as important as providing these benefits to heterosexual partners may be, the denial of those benefits to gay and lesbian partners can be no less significant. Thus, the effects of the measure on the right to receive benefits are not proportional to the objective of the legislation. The impugned provision cannot, therefore, be saved under section 1 of the Charter.

5.         WHAT REMEDY IS APPROPRIATE IN THIS CASE?

Having found a contravention of subsection 15(1) of the Charter which is not saved by section 1, I must now determine what remedial options would be appropriate in this case. The range of available options is limited by the fact that the appellants are relying primarily on subsection 52(1) of the Constitution. Subsection 52(1) reads as follows:

52. (1) The Constitution of Canada is the supreme law of Canada, and any law that is inconsistent with the provisions of the Constitution is, to the extent of the inconsistency, of no force or effect.

There are two issues which must be addressed under this subsection: (1) what is the “extent of the inconsistency”?; and (2) what is of “no force or effect”? Resolving the first issue involves determining the extent to which the impugned provision is inconsistent with subsection 15(1) of the Charter. Answering the second issue requires us to consider whether that inconsistency may be resolved through reading down (which the Supreme Court describes in Schachter as being the same thing as severing or striking out part of a provision), reading in, or a combination of reading down and reading in, or whether the provision must be struck down in its entirety. There is also the possibility, in appropriate cases, of temporarily suspending whichever of these remedial options is selected. However, as Chief Justice Lamer makes clear, the issue of whether to temporarily suspend a remedial declaration “is an entirely separate question from whether reading in or [reading down (i.e., severance or striking out) or striking down] is the appropriate route under s. 52 of the Constitution Act, 1982” (Schachter, supra, at page 716).

In defining the extent to which the impugned provision is inconsistent with the Constitution, we may take some guidance from our section 1 analysis (Schachter, supra). A word of caution may be wise in this regard since it must be recognized that the Court’s analysis under section 1 is determined in part by the evidence that is brought before the Court and by the arguments submitted by the parties. However, in this case the extent of the inconsistency with the Charter is obvious. The issue before us on this appeal does not focus on a problem fundamental to the spouse’s allowance program; rather, we are concerned exclusively with the definition of spouse in section 2 of the Old Age Security Act. That definition fails the minimal impairment stage as well as the final proportionality component of the Oakes test. The definition in section 2 is inconsistent with the Charter to the extent that it excludes gay and lesbian partners from eligibility for spouse’s allowance benefits. That exclusion specifically results from the wording of the definition of spouse in the Act. In particular, the requirements that partners be “of the opposite” sex and that couples “represent themselves as husband and wife” inevitably exclude lesbian and gay partners from eligibility. Those two requirements written into the definition of spouse create the inconsistency with the Charter.

Having defined the extent of the inconsistency, we are now faced with the question of whether the entire program should be struck down or whether we should employ the techniques of reading down (i.e., severance or striking out) or reading in (as described in Schachter) or both of these techniques. In circumstances where a benefit conferring scheme is involved, the preferred course of action is to read down or read in an inconsistent provision rather than striking down the program. This is pertinent to the case before us which plainly involves a benefit conferring scheme. In this case, the right under subsection 15(1) which has been violated is a positive right: the right to equal benefit of the law. Referring to just such a situation, Chief Justice Lamer (Schachter, supra, at page 721) observed:

Positive rights by their very nature tend to carry with them special considerations in the remedial context. It will be a rare occasion when a benefit conferring scheme is found to have an unconstitutional purpose. Cases involving positive rights are more likely to fall into the remedial classifications of reading down/reading in or striking down and suspending the operation of the declaration of invalidity than to mandate an immediate striking down.

The Chief Justice also stated at page 716, however, that:

A delayed declaration is a serious matter from the point of view of the enforcement of the Charter. A delayed declaration allows a state of affairs which has been found to violate standards embodied in the Charter to persist for a time despite the violation. There may be good pragmatic reasons to allow this in particular cases. However, reading in is much preferable where it is appropriate, since it immediately reconciles the legislation in question with the requirements of the Charter.

Thus, it follows from these two passages that, where appropriate, reading down (i.e., severance or striking out) or reading in are the preferred remedial options when a benefit conferring scheme is underinclusive.

The Supreme Court has suggested a number of touchstones for determining whether reading in or reading down (i.e., severance or striking out) are, indeed, appropriate. The factors which the Court identified as relevant to this consideration include: remedial precision; interference with the legislative objective; the effect on the significance of the remaining portion; and the significance of the remaining portion.

The first factor, remedial precision, refers to whether the question of how the statute should be extended in order to comply with the Constitution can be answered with a sufficient degree of precision. In this case, the appropriate method for extending spouse’s allowance benefits is relatively straightforward. The two references to heterosexual relationships in the definition of spouse—“of the opposite sex” and “as husband and wife—”must be read down (i.e., severed or struck out) or in some way modified through reading in so as to bring gay and lesbian partners within that definition.

But while it is clear that the words “of the opposite sex” must be read down (i.e., severed or struck out), it might be argued that the words “as husband and wife” do not necessarily exclude gay and lesbian partners. The definition of spouse does not require a couple to be husband and wife; they are merely required to represent themselves as husband and wife (Knodel, supra, at page 745). Nevertheless, it may be offensive and demeaning for a lesbian or gay couple to be compelled to represent their relationship in that way and to represent themselves in those roles. In our society, the terms husband and wife are gendered concepts which may not accurately describe roles within gay and lesbian relationships. Accordingly, requiring lesbian and gay partners to represent themselves as husband and wife in order to be eligible for spouse’s allowance benefits imposes an unfair burden upon them not faced by heterosexual partners. If this burden were imposed as a condition for eligibility, gay and lesbian partners would not be treated equally and might still be discriminatorily excluded from receiving spouse’s allowance benefits. It goes without saying that a remedy for a subsection 15(1) violation should not have inegalitarian implications. Thus, rather than requiring gay and lesbian couples to represent themselves as husband and wife, they should instead be allowed to establish that they are in an analogous relationship to husbands and wives.

Having found that we can say with a relatively high degree of precision what remedial action is required to bring the spouse’s allowance program in line with the Charter, we can move on to consider the implications of ordering such a remedy. In considering the remedies of reading in and reading down (i.e., severance or striking out), a court should be mindful of the degree to which a particular remedy intrudes into the legislative sphere. A subsection 52(1) remedy should avoid trenching unduly upon the legislative objective of a program or the means chosen by the legislature to achieve that objective. The objective of the spouse’s allowance program is to ensure that, when one partner in a couple retires, the couple will continue to receive income equivalent to the amount that would be earned if both members of the couple were retired. Extending spouse’s allowance benefits to lesbian and gay partners does not compromise the objective of the program and may be seen as furthering that objective by making the coverage of the program more thorough. In addition, including gay and lesbian partners within the ambit of the program does not alter the basic means selected for achieving the program’s objective. However, this does not end our inquiry into the degree to which reading in/reading down may intrude into the legislative sphere. In Schachter, Chief Justice Lamer, at pages 709-710 explained:

Even where extension by way of reading in can be used to further the legislative objective through the very means the legislature has chosen, to do so may, in some cases, involve an intrusion into budgetary decisions which cannot be supported.

Any remedy granted by a court will have some budgetary repercussions whether it be a saving of money or an expenditure of money…. In determining whether reading in is appropriate then, the question is not whether courts can make decisions that impact on budgetary policy; it is to what degree they can appropriately do so. A remedy which entails an intrusion into this sphere so substantial as to change the nature of the legislative scheme in question is clearly inappropriate.

As proposed, extending the spouse’s allowance program to include lesbian and gay partners requires a combination of reading in and reading down (i.e., severance or striking out). In this case, since the combination of these remedies results in the extension of monetary benefits, it is proper to point out that the extension of those benefits cannot be said to be so substantial as to change the nature of the legislative scheme. The evidence submitted to the Court regarding the cost of extending benefits to eligible gay and lesbian partners was equivocal at best. Information about the number of interdependent gay and lesbian relationships and the financial circumstances of those relationships is lacking. There is, however, enough information to allow us to conclude that extending benefits to eligible lesbian and gay partners would not involve a significant intrusion into Parliament’s budgetary decision-making. As compared to the current budgetary outlay for the spouse’s allowance program, the cost of bringing gay men and lesbians into the program would not, to use the language of Chief Justice Lamer in Schachter, supra, be “so substantial as to change the nature of the legislative scheme.”

This leads us directly to our next consideration, the effect of the proposed remedy on the remaining portion of the legislation. Is the significance of the remaining portion changed so markedly, without the inconsistent portion, that the assumption that the legislature would have enacted it is unsafe? In explaining the proper approach to answering this question, Chief Justice Lamer in Schachter, supra, at page 711, remarked:

In cases where the issue is whether to extend benefits to a group not included in the statute, the question of the change in significance of the remaining portion sometimes focuses on the relative size of the two relevant groups. For instance, in Knodel, supra, Rowles J. extended the provision of benefits to spouses to include same-sex spouses. She considered this course to be far less intrusive to the intention of the legislature than striking down benefits to heterosexual spouses since the group to be added was much smaller than the group already benefitted.

The same analysis applies in this case as was used by Madam Justice Rowles in Knodel. As in Knodel, this case involves extending benefits to a group—eligible gay and lesbian partners—which is much smaller than the group receiving benefits—eligible heterosexual partners. This supports the assumption that Parliament would likely have enacted the spouse’s allowance program without the language excluding lesbians and gay men (Schachter, supra, at page 712).

The final consideration is the significance or long standing nature of the remaining portion of the Act. As was outlined by Chief Justice Lamer in Schachter, supra, at page 713:

It is sensible to consider the significance of the remaining portion when asking whether the assumption that the legislature would have enacted the remaining portion is a safe one. If the remaining portion is very significant, or of a long standing nature, it strengthens the assumption that it would have been enacted with the impermissible portion.

The legislation involved in this case can properly be described as being of a long standing nature. The Old Age Security Act has its origins in the The Government Annuities Act, 1908, S.C. 1908, c. 5. The Old Age Security Act officially came into existence in 1951 [S.C. 1951, c. 18] when it replaced the Old Age Pensions Act, R.S.C. 1927, c. 156. The particular program in question, the spouse’s allowance program, was introduced in 1975 [S.C. 1974-75-76, c. 58, s. 5]. For eighteen years, this program has played an integral role in providing benefits to dependent heterosexual partners. Both the Old Age Security Act generally, and the spouse’s allowance program in particular, are a significant and durable aspect of Canadian society. The long history and obvious importance of the legislation in question lends credence to the assumption that Parliament would have enacted the spouse’s allowance program in any event, if there had been no language excluding gay and lesbian partners.

Indeed, all of the factors considered support the assumption of legislative approval for both reading in and reading down (i.e., severance or striking out) rather than striking down in this case. In my view, the appropriate remedy to bring section 2 into compliance with subsection 15(1), therefore, would be to order the words “of the opposite sex” read down (i.e., severed or struck out) of the definition of spouse in section 2 of the Old Age Security Act and to read into the phrase “if the two persons publicly represent themselves as husband and wife” the words “or as in an analogous relationship”. Hence, to avoid violating subsection 15(1), the definition of spouse in section 2 of the Act should read:

2. …

“spouse”, in relation to any person, includes a person who is living with that person, having lived with that person for at least one year, if the two persons have publicly represented themselves as husband and wife or as in an analogous relationship;

Under this definition, I am of the view that lesbian and gay partners would be eligible to apply for spouse’s allowance benefits. Whether they receive those benefits should, of course, depend upon whether they meet the age, income and other requirements, but in no way should it depend upon their sexual orientation.

I would, therefore, have allowed this appeal with costs both here and at trial. The words “of the opposite sex” should be read out and the words “or as in an analogous relationship” should be read in. I would also have issued a declaration to the effect that gay men and lesbians cannot be denied old age security spouse’s allowance benefits so long as they meet the usual eligibility requirements.

* * *

The following are the reasons for judgment rendered in English by

Robertson J.A.: This appeal stems from the refusal of the respondent to pay the appellant, John Nesbit, a spouse’s allowance under the Old Age Security Act, R.S.C., 1985, c. O-9 (hereinafter the “Act”). The refusal was based on the fact that the person Mr. Nesbit identified as his spouse is not a person of the opposite sex as required by the Act. That other person is the appellant James Egan. The question we must address is whether the exclusion of homosexual or same-sex couples from the definition of “spouse” contained in the Act is discriminatory within the meaning of subsection 15(1) of the Canadian Charter of Rights and Freedoms (hereinafter the “Charter”). If the response is positive, then it is incumbent to determine whether the impugned legislation can be demonstrably justified as a reasonable limit under section 1 of the Charter. The relevant facts and legislation are relatively straightforward.

BACKGROUND

In September of 1986, the appellant Egan reached 65 years of age and became entitled to receive old age security and guaranteed income supplement benefits. The Act also provides for a spouse’s allowance to be paid to the spouse of a pensioner where the former is between 60 and 65 years of age and their combined income falls below a fixed amount. Following an exchange of correspondence with the Department of National Health and Welfare, the appellant Nesbit (the only litigant that would be entitled to receive payments) applied for a spouse’s allowance. In that application, he described the appellant Egan as his spouse. It was rejected solely on the basis that the relationship of Mr. Nesbit with Mr. Egan did not meet the definition of “spouse” prescribed by section 2 of the Act:

2. …

“spouse”, in relation to any person, includes a person of the opposite sex who is living with that person, having lived with that person for at least one year, if the two persons have publicly represented themselves as husband and wife;

The appellants have resided together, in an intimate relationship, since 1948. The Trial Judge found that they share bank accounts, credit cards and property ownership and, by their wills, had appointed each other as their respective executors and beneficiaries. To their families and friends, they refer to themselves as partners and at one point publicly exchanged rings. A brief comment on this aspect of the evidence is warranted.

The purpose of the kind of evidence adduced by the appellants is to establish that same-sex relationships can be similar to opposite-sex ones, both in their diversity and interaction.[12] The appellants’ relationship can be characterized in terms of commitment, duration and interdependency—attributes found in the ideal opposite sex spousal relationship. However, this Court has not been requested to adjudicate in terms of stereotypical relationships, the degree of required intimacy or such other considerations that the terms “spousal status” or “spouse” might entail. Rather, we have been asked to determine whether the refusal to extend to the appellants, and such otherwise qualified same-sex couples, a benefit available to opposite sex couples violated their guarantee of equal benefit of the law without discrimination.[13] That determination must be based on the Trial Judge’s finding that: “had [the appellant] Nesbit been a woman cohabiting with Egan substantially on the same terms as he in fact cohabited with Egan he would have been eligible for the spouse’s allowance.”[14]

At trial, the appellants maintained that, as a matter of statutory construction, the term “spouse” as defined in section 2 of the Act “includes” persons living in a same-sex relationship with a pensioner. The Trial Judge rejected that construction. In this regard, his decision is consistent with cases of similar import.[15] The Trial Judge then rejected the argument that the appellants were discriminated against on the basis of “sex”. This conclusion is also consistent with existing authorities.[16] (Both arguments were abandoned on appeal.) Hence, it remained to be determined whether the appellants had been discriminated against on the basis of an analogous ground—sexual orientation. In pursuing that question the Trial Judge concluded: (1) that there was nothing discriminatory in a law which restricts a benefit to opposite sex couples; and (2) that the benefit which the appellants sought was denied because of their “non-spousal status” and not because of their “sexual orientation”. At page 704, he stated:

Parliament has chosen to address the needs of persons of the opposite sex who live together in a conjugal state, either statutory or common law, as husband and wife. This unit has traditionally been treated as the basic unit of society upon which society depends for its continued existence. I can see nothing discriminatory against the plaintiffs in a law which provides certain benefits to this group and which law does not provide the same benefits to a homosexual couple in the position of the plaintiffs. The plaintiffs as an homosexual couple, just as a bachelor and a spinster who live together or other types of couples who live together, do not fall within the traditional meaning of the conjugal unit or spouses. When compared to the unit or group which benefits by the challenged law the plaintiffs fall into the general group of non-spouses and do not benefit because of their non-spousal status rather than because of their sexual orientation. [Emphasis is mine.]

To the Trial Judge, the appellants are being treated “just like” all other non-spousal couples whose relationship is marked by commitment, duration and interdependency (hereinafter “other non-spousal relationships”). The appellants maintain that that comparison is misdirected. They argue that they are “just like” and “no less deserving” than opposite-sex partners who qualify for the spouse’s allowance.

At the outset, I think it is important to recognize that these conflicting views of “likeness” have been expressed, in one form or another, in all of the cases involving spousal benefits. That is true despite the fact that the similarly situated test had been rejected earlier by the Supreme Court. This is but one of the issues that must be acknowledged and addressed herein.

THE CHARTER—DISCRIMINATION DEFINED

Subsection 15(1) of the Canadian Charter of Rights and Freedoms provides:

15. (1) Every individual is equal before and under the law and has the right to the equal protection and equal benefit of the law without discrimination and, in particular, without discrimination based on race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.

In order for the appellants to succeed, they must demonstrate that the restrictive definition of “spouse” as found in the Act is discriminatory within the meaning of subsection 15(1) of the Charter on the basis of sexual orientation.[17]

Although sexual orientation is not expressly referred to in subsection 15(1), it is trite law that the list of enumerated grounds is not finite and that the protections available under that section can be extended to those who establish that they are members of a disadvantaged group and can show discrimination on analogous grounds. The classic test is found in the reasons of Madam Justice Wilson, first articulated in Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, and restated in R. v. Turpin, [1989] 1 S.C.R. 1296, at page 1332:

… the determination of whether a group falls into an analogous category to those specifically enumerated in s. 15 is “not to be made only in the context of the law which is subject to challenge but rather in the context of the place of the group in the entire social, political and legal fabric of our society” (p. 152). If the larger context is not examined, the s. 15 analysis may become a mechanical and sterile categorization process conducted entirely within the four corners of the impugned legislation. A determination as to whether or not discrimination is taking place, if based exclusively on an analysis of the law under challenge is likely, in my view, to result in the same kind of circularity which characterized the similarly situated similarly treated test clearly rejected by this Court in Andrews.

Though the Supreme Court has yet to adjudicate on whether homosexuals constitute a disadvantaged group entitled to Charter protection, I take it to be settled law that sexual orientation can be invoked as an analogous ground of discrimination under subsection 15(1).[18] The respondent conceded this point and, in my opinion, rightly so.

It can be readily accepted that the definition of “spouse” in section 2 has a differential impact on same-sex couples. The critical issue is whether that distinction is based on sexual orientation and is discriminatory within the meaning ascribed to that term by the existing jurisprudence. Since not every distinction drawn between individuals or groups violates section 15, the difficulty in discerning these situations where the impugned legislation is discriminatory remains.

The leading decision on the law of discrimination under section 15 of the Charter lies in the reasons of the Supreme Court in Andrews v. Law Society of British Columbia, supra. In fact, it is the only decision of that Court in which a law has been declared discriminatory based on an unenumerated ground (citizenship).[19] I propose to reproduce what are now regarded as the critical passages found in the reasons for judgment of McIntyre J. (speaking for the Court on this issue). Together, they provide the analytical framework on which we are required to resolve the issue before us:

Concept of Equality

Section 15(1) of the Charter provides for every individual a guarantee of equality before and under the law, as well as the equal protection and equal benefit of the law without discrimination. This is not a general guarantee of equality; it does not provide for equality between individuals or groups within society in a general or abstract sense, nor does it impose on individuals or groups an obligation to accord equal treatment to others. It is concerned with the application of the law. (Per McIntyre J., at pages 163-164.)

Purpose

It is clear that the purpose of s. 15 is to ensure equality in the formulation and application of the law. The promotion of equality entails the promotion of a society in which all are secure in the knowledge that they are recognized at law as human beings equally deserving of concern, respect and consideration. It has a large remedial component. (Per McIntyre J., at page 171.)

Discrimination Defined

I would say then that discrimination may be described as a distinction, whether intentional or not but based on grounds relating to personal characteristics of the individual or group, which has the effect of imposing burdens, obligations, or disadvantages on such individual or group not imposed upon others, or which withholds or limits access to opportunities, benefits, and advantages available to other members of society. Distinctions based on personal characteristics attributed to an individual solely on the basis of association with a group will rarely escape the charge of discrimination, while those based on an individual’s merits and capacities will rarely be so classed. (Per McIntyre J., at pages 174-175.)

Distinctions Based on Irrelevant Personal Differences

Recognizing that there will always be an infinite variety of personal characteristics, capacities, entitlements and merits among those subject to a law, there must be accorded, as nearly as may be possible, an equality of benefit and protection and no more of the restrictions, penalties or burdens imposed upon one than another. In other words, the admittedly unattainable ideal should be that a law expressed to bind all should not because of irrelevant personal differences have a more burdensome or less beneficial impact on one than another. (Per McIntyre J., at page 165; and see also La Forest J., at page 193.)

Contextual Approach

Consideration must be given to the content of the law, to its purpose, and its impact upon those to whom it applies, and also upon those whom it excludes from its application. The issues which will arise from case to case are such that it would be wrong to attempt to confine these considerations within such a fixed and limited formula. (Per McIntyre J., at page 168.)

Burden of Proof

A complainant under s. 15(1) must show not only that he or she is not receiving equal treatment before and under the law or that the law has a differential impact on him or her in the protection or benefit accorded by law but, in addition, must show that the legislative impact of the law is discriminatory. (Per McIntyre J., at page 182.)

Unfortunately, the ease with which one may extrapolate the various tenets of the Supreme Court’s test for determining discrimination is matched only by the difficulty encountered in applying it to the facts of a particular case.

In view of the fact that the Supreme Court has not had the opportunity to express any opinion on the Charter in the context of gay and lesbian rights, it is instructive to review the limited jurisprudence emanating from Canada’s other superior courts. At worst, the decisions reveal diversity in judicial opinion. At best, they provide a better understanding of the issues that must be confronted.

THE JURISPRUDENCE

To date the cases in which gay and lesbian rights have been advanced fall into one of two categories. First, there are those in which an individual has been denied access or become disentitled to a “benefit”. The second category is comprised of cases in which one of the criteria of entitlement is spousal status; that is to say the benefit is restricted to those legally married and, when recognized by statute, partners to a common law marriage. In both categories, the discrimination complained of can be either direct (intentional) or indirect (unintentional).

Cases falling within the first category can be decided against the backdrop of provincial and federal human rights legislation. So-called benefits of general entitlement, whether they be classified as private or public law benefits, inevitably involve matters of employment,[20] accommodation and services and have been made the subject of anti-discrimination laws. With respect to human rights legislation, five provinces and one territory ban discrimination based on sexual orientation; they are Ontario, Quebec, Nova Scotia, New Brunswick, Manitoba and Yukon.[21] Until recently, the Canadian Human Rights Act did not include sexual orientation as a prohibited ground of discrimination (presently in Bill form; see Bill C-108, 1992). Prior to the legislative amendment, it had been determined in Haig v. Canada (1992), 9 O.R. (3d) 495 (C.A.), that that omission constituted discrimination offending subsection 15(1) of the Charter.[22]

The decision of the Ontario Court of Appeal in Haig is instructive and represents one of two cases in which a court has had the opportunity to deal with the issue of discrimination, in the context of sexual orientation, and the aggrieved party has been successful.[23]

In Haig, the respondent Birch, a captain in the Canadian Armed Forces, decided to inform his commanding officer that he was gay. In response, Captain Birch was informed that he was subject to a policy directive relating to “homosexuals” which rendered him ineligible for promotions, postings or further military training. Ultimately, he was released from the Forces on medical grounds which included the fact that he was unable to continue working under the career restrictions imposed on him by the policy directive and therefore he was unfit for further service. Eventually, Captain Birch sought a declaration that the absence of sexual orientation from the list of prescribed grounds of discrimination in section 3 of the Canadian Human Rights Act was itself discriminatory. Both the Trial Judge [(1991), 5 O.R. (3d) 245] and the Court of Appeal agreed. The latter Court had no difficulty in concluding that (per Krever J.A., at page 498): “It is difficult to imagine a clearer case of invidious discrimination in employment by reason only of sexual preference than the experience of the respondent Birch.”

The discrimination found in Haig illustrates the irrelevancy of a personal characteristic—sexual orientation—in determining one’s entitlement or disentitlement to employment. However, in some instances, discriminatory practices are not as transparent as that which arose in Haig and, hence, careful scrutiny is required when determining whether the intent or effect of an impugned law is discriminatory within the meaning of the Charter. In this respect, the decision of the British Columbia Supreme Court in Brown v. British Columbia (Minister of Health) (1990), 66 D.L.R. (4th) 444, is illustrative. In Brown, the issue did not centre on access to employment, but rather access to a costly drug believed helpful in the treatment of a potentially fatal disease.

The allegation of discrimination arose from the decision of the provincial government to place the drug AZT, the only drug which at that time proved effective in treating AIDS, under the provincial drug plan. The effect of that decision was to require all AIDS patients, other than those on welfare or in long-term care facilities, to pay a portion of the cost of the drug. That funding policy affected an identifiable group, of which 90% were “homosexual” or “bisexual” males. The majority of these men did not qualify for the exemption and were faced with drug payments in excess of $2,000 per year, which the Court held was likely to cause a “serious financial burden” (at page 457). However, this funding decision was not challenged until the government decided subsequently to fully fund another equally expensive drug, cyclosporin, required by other seriously ill persons: transplant patients and those suffering from cancer. Nonetheless, the Court found no discrimination within the meaning of subsection 15(1) of the Charter.

While it is not my intent to review critically the decision in Brown, it can be a matter of personal judgment whether the reasons proffered by the government, and accepted by the Court, to explain the differential approach to funding drugs, for cancer and transplant patients and those suffering from AIDS, are persuasive.[24] For our purposes, it is important to recognize that this is precisely the type of case where a “distinction” can exacerbate conditions of prejudice for members of a disadvantaged minority by restricting access to a basic and needed benefit. To the extent that a policy decision (having the force of law) has both a disproportionate and burdensome impact on such a group, a finding of discrimination should not, in my opinion, be easily displaced.

The following discussion focuses on two conflicting decisions involving Charter challenges falling into the second category—benefits based on spousal status. In both cases, the claimants sought to show that they had been denied a benefit because of their sexual orientation. As will be explained, the true complaint was not denial of a benefit, but rather denial of a benefit on the same terms available to opposite-sex couples.

In Andrews v. Ontario (Minister of Health), supra, the provincial health insurance legislation denied dependant coverage to the same-sex partner of the plaintiff. “Dependant” was defined to include the spouse of an insured person which the provincial authorities interpreted as meaning a person of the opposite sex. The Trial Judge concluded that the legislation was not discriminatory and hence did not violate subsection 15(1) of the Charter. First, he concluded that “homosexual” couples are not like “heterosexual” couples because the former do not procreate nor raise children. Second, the Trial Judge ruled out discrimination because the plaintiff was being treated in exactly the same manner as all unmarried cohabiting persons. Finally, he was swayed by the fact that the premium being paid by an opposite-sex couple equalled the combined premiums paid by partners to a same-sex relationship.

By itself the latter determination would seem sufficient to dispose of the matter. It is difficult, in the circumstances, to argue convincingly that the plaintiff was denied equal benefit of the law. Medical coverage was available to all and on identical financial terms. For our purposes, what is relevant is that the litigation focused on whether a same-sex couple is “like” an opposite-sex one or whether a same-sex couple is “like” other non-spousal relationships.

The reasoning employed in Andrews v. Ontario did not find its way into the decision rendered in Knodel v. British Columbia (Medical Services Commission), supra. However, the facts of these two cases are virtually identical. In Knodel, the definition of “spouse” contained in the Medical Services Act Regulations [B.C. Reg. 144/68, s. 2.01 (as am. by B.C. Reg. 5/77)] under the Medical Services Act [R.S.B.C. 1979, c. 255] of British Columbia did not extend benefits to same-sex couples and thus the petitioner’s application to include his partner as a dependant spouse under the health care plan was rejected. The Trial Judge held that the Regulations had the effect of treating same-sex couples differently than opposite-sex couples. In fact, her entire reasoning is premised implicitly on the understanding that one is “just like” the other. She went on to hold that the Regulations discriminated against the petitioner on the ground of sexual orientation by imposing an economic penalty arising from the refusal to extend a benefit available to opposite-sex couples. (I shall deal with that particular finding shortly.) She also held that the Regulations had a further impact. They affected the plaintiff’s dignity and self esteem and reinforced the “homophobia” expressed by certain members of society.

Unlike the approach pursued in Andrews v. Ontario, the Trial Judge in Knodel was of the view that the underinclusiveness of the legislation was, in the words of Dickson C.J. in Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219, at page 1240, “simply a backhanded way of permitting discrimination.” Her analysis portrays the issue in terms of a distinction affecting two classes of individuals (at page 756):

Where the state makes a distinction between two classes of individuals, A and B, that has the effect of imposing a greater burden on individuals within class B, and if the individuals within class B fall within the class of individuals protected by s. 15(1) of the Charter, the manner in which the legislative provision or law is drafted is irrelevant for constitutional purposes; i.e., it is immaterial whether the subject law states: (1) A benefits; or (2) Everyone benefits except B. In both cases, the impact upon the individual within group B is the same.

Based on the reasoning in Andrews v. Ontario, and of the Trial Judge in the present case, the above analysis is flawed to the extent that it fails to acknowledge and deal with the fact that class “B” is composed not only of same-sex couples, but also others whose dependant relationship lacks spousal status. In effect, the exclusion from class “A” extends de facto to classes “B” (same-sex couples) and “C” (other non-spousal relationships).[25]

As noted earlier, the Trial Judge in Knodel concluded that same-sex couples were denied a benefit available to opposite-sex couples. It should not be presumed, however, that these same-sex partners had been denied medical coverage. Under the relevant regulations, the combined annual premiums paid by a same-sex couple was $84 greater than that paid by an opposite-sex couple. Technically speaking, the petitioner (and his partner) was denied equal benefit of the law. But it would be misleading to state that the regulations had the effect of denying a benefit to a same-sex couple otherwise available to an opposite-sex one. On the contrary, the same-sex partner was granted medical coverage but had to pay $42 more per year because he did not fall within the definition of “spouse”.[26] In this respect, the facts of Knodel resemble more closely those found in Andrews v. Ontario than those found in Brown.

In Brown, the government’s differential funding policy raised the inference of an intent to exclude gay men from the obvious financial benefit derived from the government’s decision to limit full funding to other seriously ill patients. Alternatively, it could be argued that that funding policy had a direct and more burdensome impact on this disadvantaged group. As well, the fact that the majority of these men faced a “serious financial burden” could have had the effect of denying access to a needed drug. By comparison, both Andrews v. Ontario and Knodel involved a legislative scheme in which a benefit—health coverage—was available to all. In Andrews v. Ontario the cost was the same regardless of spousal status. In Knodel the cost differential could not reasonably be regarded as impeding access to the health care system. I think it fair to conclude that in these two cases the real objection to the legislation stemmed from the distinction that had been drawn between opposite-sex and same-sex couples. What the claimants truly sought was not a benefit per se, but rather legal recognition that their conjugal relationship is no different than an opposite-sex one which qualifies for a spousal benefit.

In summary, an examination of the extant jurisprudence unmasks at least three unresolved issues.

First, is the relationship of a same-sex couple “just like” or the “same as” an opposite-sex one? Correlatively, is the relationship of a same-sex couple any different from other non-spousal relationships? As noted earlier, and discussed below, such an analytical approach remains unavailable. Thus, it is only natural to ask why it is that litigants and the courts have continued to approach the issue from such a perspective.

Second, of what significance, if any, is the fact that the non-spousal category consists not only of same-sex couples but, as well, other non-spousal relationships? The answer to that question is not dependent on a comparison between perceived groups in the excluded category. What is relevant is that members of the disadvantaged group are not the only ones denied the benefit.

Third, is not the true complaint grounded in the failure of Parliament to recognize that same-sex relationships are no different than opposite-sex ones, including those which fall into the spousal category? If that is so, then what we are confronted with is an indirect challenge to the common law and statutory concept of marriage, as reflected in legislation which offers a restricted definition of “spouse”.

ANALYSIS

The appellants’ objection to the judgment below lies in the way in which the Trial Judge characterized the distinction drawn by Parliament. In their view, the comparison is not as between “spouses” and “non-spouses”, but rather as between “heterosexual” and “homosexual” couples. They maintain that: “Homosexual men and women whose partners have retired at 65 are no different from, and no less deserving of the spousal allowance than, heterosexual men and women.” (Appellants’ factum, at page 10.) Discrimination is said to arise because the impugned legislation has the effect of reinforcing the disadvantage endured by a group which, throughout history, has been both “marginalized” and “stigmatized”. Finally, they argue that the proper way to rectify the inequality is to extend the definition of “spouse” to include same-sex couples.

I propose to examine the appellants’ arguments by looking first at the issue of “sameness” which they have deemed to be of fundamental significance. Equally important is the application of the test for determining whether in law a distinction based on a prohibited ground is discriminatory. As will become evident, the positions that the appellants have adopted herein represent a fundamental shift in the direction of gay and lesbian rights in Canada. The question which remains is whether they are in accord with the law as found in Andrews v. Law Society of British Columbia, supra.

1.         SAMENESS

(a)       The Similarly Situated Test

The argument that same-sex couples are the “same” as opposite-sex ones and, therefore, both should be treated alike is grounded in the Aristotelian principle of formal equality restated in the similarly situated test—people that are alike should be treated alike, while people that are unalike should be treated unalike. However, the proper legal response to that argument is a matter of settled law. The test was expressly rejected by the Supreme Court in Andrews. Of course, the response is equally applicable to the argument that the appellants are being treated in exactly the same manner as persons in other non-spousal relationships. Hence, in strict legal theory, neither party to this action is permitted to dwell on whether one type of couple is “like” or “unlike” any other.

At the same time, it is understandable why one is attracted to the notion of formal equality—it accords with common sense.[27] Yet the similarly situated test was rejected in Andrews because if applied literally, and without exception, it would inevitably work a clear injustice.[28] At page 166, McIntyre J. stated:

The [similarly situated] test as stated, however, is seriously deficient in that it excludes any consideration of the nature of the law. If it were to be applied literally, it could be used to justify the Nuremberg laws of Adolf Hitler. Similar treatment was contemplated for all Jews. [Emphasis is mine.]

Any hope that the test would survive Andrews, so long as it was not being applied mechanically, was summarily quashed in McKinney v. University of Guelph, [1990] 3 S.C.R. 229. Writing for the majority, La Forest J. stated (at page 279): “Simply put, I do not believe that the similarly situated test can be applied other than mechanically, and I do not believe that it survived Andrews v. Law Society of British Columbia.”

The decision of the Supreme Court to reject the similarly situated test has had only a limited effect on eliminating formal comparisons between perceived groups. Interestingly enough, when the test has been applied expressly it was for the purpose of refuting the argument that same-sex couples are no different than other non-spousal relationships (see Knodel, supra). Yet, counsel seem prepared to argue cases, and courts decide them, on the presupposition that same-sex couples are either “like” or “unlike” opposite-sex ones, as the case may be. In so doing, they effectively sidestep the similarly situated test.

It is naive to pretend that the “sameness” issue disappeared with the Supreme Court’s rejection of the similarly situated test. Nor should it have been expected that the issue would disappear completely. So long as discrimination is defined in terms of a distinction based on “personal characteristics”, there will always be a temptation to make a comparison between those who do and do not possess that difference or characteristic.

In virtually every discrimination case involving spousal benefits, the claimants have introduced evidence focussing on just how similar the lifestyles of same-sex couples are when compared with opposite-sex ones.[29] This case is no exception. It differs from others only to the extent that the evidence before us is limited and pales in comparison with the testimony of academics, sociologists and health care professionals documented in the case reports.[30] Such evidence has but one obvious purpose—to establish the “similarity” between the two kinds of relationships. The evidence given by a clinical psychiatrist in Knodel is on point (at page 735):

Dr. Myers reports that “there is a high degree of similarity between homosexual and heterosexual life partners and that they are much more the same in their attitudes, expectations, and values than are different.”

Nonetheless, it must be formally recognized that the appellants’ case is built on the “similarity” between same-sex and opposite-sex couples. The Supreme Court’s decision in Andrews destroys the very premise on which their claim of discrimination rests. Moreover, the inapplicability of the similarly situated test is not the only legal impediment to the success of this appeal. Correlatively, by arguing similarity of relationships the appellants run afoul of the “irrelevancy” component of the discrimination test set out in Andrews. In rejecting the similarly situated test, the Supreme Court has defined inequality in terms of differential treatment based on “irrelevant personal differences”.[31]

(b)       Irrelevant Personal Differences

It will be recalled that in Andrews, McIntyre J. held (at page 165) that a person is not to be deprived of a benefit because of a distinction based on irrelevant personal differences.[32] On reflection, it is evident that sexual orientation is both relevant and essential to the appellants’ argument. It is the characteristic which makes them similar to an opposite-sex (common law) couple and distinguishes them from other non-spousal relationships.

The significance of the legal requirement that a distinction be based on an irrelevant personal difference cannot be ignored or dispensed with summarily.[33] In my view, it is basic to our present understanding of the law of discrimination under section 15 of the Charter.

The appellants have not argued discrimination because of a distinction based on an irrelevancy. Rather, they have identified the distinction in terms of differential treatment being accorded same-sex and opposite-sex couples and then deemed that irrelevant. Consequently, sexual orientation remains highly relevant. Homosexuality does not distinguish a same-sex couple from an opposite-sex one. It is the basis on which the two can be deemed “alike” and, at the same time, the basis on which other non-spousal relationships can be distinguished. If, however, sexual orientation were to remain an irrelevant consideration or criterion, the situation of same-sex couples could be viewed no differently than other non-spousal relationships. Hence, it is apparent that the appellants’ claim for spousal benefits represents a fundamental shift in the nature of the rationale underscoring gay and lesbian rights in Canada.

Cases such as Haig, discussed supra, illustrate that the pursuit of gay and lesbian rights has, until recently, been based on the understanding that sexual orientation must remain an irrelevant consideration when determining access to benefits of general entitlement. The guarantee of equality and freedom from discrimination is clearly tied to the belief that certain personal differences cannot be regarded as a valid criterion of disentitlement.

Indeed, the decriminalization in Canada in 1967 of “homosexual” sexual activity between consenting adults and the decision of a number of provincial legislatures to ban discrimination based on sexual orientation (there are exceptions) was premised on an unarticulated right to privacy and on the belief that it is an irrelevant factor. The legislatures and courts were to remain outside “the bedrooms of the nation”. Today, however, we are being invited to adjudicate on evidence solely for the purpose of establishing the similarity between same-sex households and opposite-sex ones.

When placed in this perspective, it is apparent that the Charter, as interpreted in Andrews, reflects the rationale originally proffered in support of recognizing gay and lesbian rights in this country.[34] That earlier rationale is also compatible with the decision to reject the similarly situated test. One commentator explains the competing views of homosexuality in terms which, in my view, reveal an incisive appreciation of the reasons underlying wide ranging and deeply held differences of opinion and, as well, the varying rationales on which gay and lesbian rights may be premised:[35]

Views toward those labelled homosexual range from condemnation to pity to indifference to respect. This range also appears in the legal community. Although some courts and commentators are very sympathetic to claims by lesbians and gay men for equal treatment, a substantial portion of the legal community retains negative views of those with minority sexual orientations.

Differing views along these two dimensions—whether sexual orientation is part of identity and whether gay men and lesbians are viewed positively, negatively or neutrally—combine to create four competing conceptions of homosexuality. The “sin” conception views homosexual acts as immoral and wrong; it generally does not ascribe to the view of homosexuality as an intrinsic part of identity. The “illness” viewpoint similarly sees homosexuality negatively; this framework, however, sees it as part of the affected individual’s personality, albeit a potentially curable component. The “neutral difference” approach, like the illness approach, embraces the concept of sexual orientation as identity but views it merely as a difference that should not be a basis for discriminatory treatment. Finally, the “social construct” conception rejects categorizing individuals by sexual orientation and views same-sex acts and relationships as not materially different from opposite-sex ones.

There is little doubt that the “neutral difference” model, outlined above, is reflected in the Supreme Court’s interpretation of the Charter adopted in Andrews. In effect, the appellants argue that the Charter should mirror the “social construct” model, which is at some distance from the reasoning of the Supreme Court in Andrews. There it was held that non-citizens constitute a “discrete and insular” minority not to be denied access to employment solely because of an irrelevant personal difference—citizenship. The present law, as I understand it, dictates that sexual orientation is neither a valid criterion of disentitlement, nor a valid criterion of entitlement. It is not a criterion of any sort. Rather, it must be regarded as an irrelevant personal difference.

To the extent that the Andrews v. Law Society of British Columbia decision defines discrimination in terms of a distinction based on an irrelevant personal difference, and has ruled out the application of the similarly situated test, the appellants’ claim that the definition of “spouse” in section 2 of the Act is discriminatory must fail.

Despite the foregoing conclusion, I propose to deal with the general issue of discrimination. It may well be that my understanding of Andrews is mistaken. Moreover, the fact that spousal status is a condition precedent to receipt of a benefit does not preclude a finding of discrimination on the basis of sexual orientation simpliciter. As is well known, discrimination can occur either directly (intentional) or indirectly (unintentional) and is to be measured in terms of the adverse impact which the legislation has on those affected by the impugned legislation.

2.         DISCRIMINATION

Subject to considerations outlined below, the discrimination test in Andrews is as follows (at page 165):

… the admittedly unattainable ideal should be that a law expressed to bind all should not because of irrelevant personal differences have a more burdensome or less beneficial impact on one than another. [Emphasis is mine.]

The immediate task is to assess the unintended impact of the failure of Parliament to extend the spouse’s allowance program to same-sex couples and to determine whether the consequences constitute discrimination. Before pursuing that part of the analysis, there are two outstanding issues to be addressed. One is to determine the relevance, if any, of the fact that members of the disadvantaged group are not the only persons denied a benefit. The other is concerned with the basis on which we resolve whether a distinction is based on spousal status as opposed to sexual orientation. As noted above, it is extremely important to recognize that a finding of discrimination on the basis of sexual orientation is not automatically precluded because a criterion of entitlement is spousal status. There can be instances where the claimant will not be asserting the relevance of his or her sexual orientation as a basis of entitlement. I shall deal with this issue first.

(a)       Spousal Status v. Sexual Orientation

I take it for granted that discriminatory laws cannot be rendered immune from challenge merely because legislators have chosen to limit access to a benefit by prescribing conditions precedent (e.g., spousal status) which on their face are benign.[36] The following hypothetical, albeit an extreme one, illustrates my point.

Assume that a drug effective in the treatment of AIDS is discovered, but that it can only be produced in small quantities such that there is an immediate shortage. Assume also that the drug is extremely expensive and, thus, the state is required to intervene. It does so by establishing a drug program in which all married persons, as well as common law spouses, and their children are entitled to receive the drug at no cost. All others must pay the full cost.

In such circumstances, I cannot imagine a court preempting a challenge under subsection 15(1) of the Charter simply because the criterion of entitlement is expressed in terms of spousal status. Hence, attention would have to focus on whether the intent and/or effect of the legislation is to discriminate against homosexuals on the basis of sexual orientation. As a claimant’s personal relationship with another, homosexual or otherwise, is irrelevant to that determination neither the similarly situated test, nor the irrelevancy component of the definition of discrimination, are applicable. Discrimination on the basis of spousal status will be ruled out provided the consequences of the impugned legislation are such that claimants satisfy the Supreme Court’s adverse impact test.

(b)       The Relevance of “Other Non-Spousal Relationships”

The above hypothetical brings into question the legal position of those not members of a recognized disadvantaged minority (e.g., those who acquired AIDS through blood transfusions), but who nevertheless would also be denied the benefit. In the instant case, one cannot ignore the fact that the definition of “spouse” found in the Act excludes a broad class of non-spouses. Also excluded are persons cohabiting with another in a non-conjugal relationship. Within this group can be found siblings, friends and relatives. Their relationships are similar to others within the excluded category, both in terms of diversity and interdependency.

Though same-sex couples are subjected to various forms of “invidious” discrimination, it is only proper to ask whether their rights should be determined in isolation of those also affected by the impugned legislation. Certainly, no one would suggest that same-sex couples are more deserving of the spouse’s allowance than those whose financial needs may be just as substantial as the appellants, but who fall into another non-spousal category. Hence, we are forced to ask whether the issue of entitlement can be restricted to a particular group (same-sex couples) within the excluded category. The appellants’ response to this question is brief (appellants’ factum, at page 20):

Whether Parliament would be required to extend the benefit beyond homosexual couples to others such as siblings or parent or child is not before the Court and may depend in part on whether family status is a prohibited ground of discrimination and if so what it means.

I do not find this response persuasive for two reasons. The first is a matter of precedent. In Andrews, McIntyre J. stated: “Consideration must be given to the content of the law … and its impact … upon those whom it excludes from its application” (at page 168). My second reason raises what I consider to be a fundamental issue.

To accede to the appellants’ argument could, in my opinion, generate an inequality on the basis of sexual orientation. Those disentitled to the benefit, and unable to allege discrimination on that ground, would have no recourse unless able to successfully raise another unenumerated ground of challenge, be it expressed in terms of “family”, “spousal” or “marital” status. In the instant case, it is extremely difficult to characterize other non-spousal (relationships) persons as members of a group warranting Charter protection.

If legislation is to be challenged in circumstances where there is more than one group which is denied the benefit, then should not the ground of discrimination be one common to all who come within the excluded category? While I need not answer that question, the response offered by the appellants raises fundamental and unanswered concerns for me. In my view, if differential treatment between the members of two excluded “subgroups” is to be sustained, it is because compelling reasons are offered for extending a benefit to one and not the other. Such compelling reasons must lie in the extent to which the distinction has a disparate and more burdensome impact on one excluded group than another. To me it is axiomatic that courts cannot ignore the fact that others have been excluded and, as well, the effect which the impugned legislation may have on their equality rights. For example, in the drug funding hypothetical I cannot imagine the grounds on which one could ignore the plight of all AIDS sufferers who would be denied the benefit under the legislative scheme.[37]

In my opinion, where the impugned legislation excludes persons other than those who are members of a particular disadvantaged group, proper consideration must be given to the significance of that reality when deciding whether the impact of a distinction is discriminatory. Legislation denying benefits to members of another group may detract from a finding of discrimination contrary to subsection 15(1) of the Charter. It may reveal that one is not dealing with a benefit freely accorded all others in society and, hence, the impact of the legislation might not necessarily have the effect of reinforcing disadvantage or reinforcing disadvantage to the degree necessary to invite judicial intervention.

(c)        The Adverse Impact Test

What is the threshold of disadvantage that must be demonstrated before the impugned legislation can be declared discriminatory? It is apparent that there are two conflicting views as to what is sufficient to meet the adverse impact test stated in Andrews. In my opinion, only one is consistent with the reasoning in that case.

One approach effectively presumes discrimination once it has been found that a distinction is based on a prohibited ground. The consequences of the denial are not measured, for example, in economic or material terms, but in light of the effect which the denial has in reinforcing existing prejudices suffered by members of an historically disadvantaged group. Most recently, this approach was alluded to in Knodel, discussed supra, and now by the appellants. In Knodel the impact of the legislation was assessed in terms of affecting the claimants’ dignity and self-esteem, as well as reinforcing “homophobic” views held by some members of society. What distinguishes this approach from the other is that a finding of discrimination is made without regard to the purposes and objects of the impugned legislation, nor the fact that others may also have been denied the benefit. This approach can be labelled the “abstract” test.

The other approach is a contextual one and requires a full analysis of the immediate impact of the legislation on members of the disadvantaged group, as well as others who are excluded, as measured against the purposes and objects of the legislation. In effect, what the Court is looking for is disadvantage independent of that arising from membership in the group. This interpretation can be labelled the “concrete” test. In my opinion, this is the one mandated by Andrews. There it was held that the Charter does not provide for a general guarantee of equality between individuals and groups in a “general or abstract sense” (at pages 163-164) [underlining added]. Moreover, consideration must be “given to the content of the law, to its purpose and impact upon those to whom it applies” (at page 168) [underlining added]. The fairest way of evaluating this conclusion is to examine critically the position advocated by the appellants. It is consistent with the “abstract” test.

(d)       The Abstract Test

As I understand the appellants’ argument, the denial of the spousal allowance is discriminatory because it has the effect of reinforcing prejudice endured by a group which historically has been subjected to discriminatory practices. While I can accept readily the latter proposition, there are several legal impediments to this Court arriving at the desired conclusion.

First, this is the very test stated by Madam Justice Wilson in Andrews, and restated in R. v. Turpin (outlined earlier in these reasons), for establishing whether the protections available under subsection 15(1) of the Charter can be extended to those who show discrimination on analogous grounds. Thus, the appellants’ test has the effect of making the term “discrimination” synonymous with “distinction”. In effect, the appellants ask that we declare those provisions of approximately fifty federal statutes which define rights and obligations on the basis of spousal status as being a violation of subsection 15(1) of the Charter (see respondent’s factum, Appendix C, at pages 34-35). Consequently, whenever a benefit is based on spousal status the legal analysis would have to focus on the reasonableness of the impugned legislation pursuant to section 1 of the Charter. Without question, that result is antithetical to the reasoning advanced in Andrews.

Second, and a corollary of the first, the test has the effect of ignoring altogether the content of the law subject to challenge and once again the reasoning adopted by the Supreme Court in Andrews. By comparison, Madam Justice Wilson in Turpin, supra, was concerned with the possibility that a determination regarding discrimination based on an unenumerated ground would be based exclusively on an analysis of the impugned legislation. The appellants are asking that the pendulum be moved to the other extreme. Finally, the interpretation of the adverse impact test being offered brings into play the distinct possibility that an inequality between the disadvantaged group and those also excluded by the legislation will be sanctioned.

I turn now to the alternative approach.

(e)       The Concrete Test

Arguably, the disadvantage experienced by Captain Birch (see Haig, supra, involving constructive dismissal from employment on the ground of sexual orientation) may be said to represent the extreme and, therefore, should not be viewed as establishing the threshold test to be met in all cases. The issue we must face is whether the impact of the impugned legislation can in any reasonable sense meet the test contemplated in Andrews. That assessment cannot be made without first examining the reasons underlying Parliament’s decision to establish the spouse’s allowance program.

I should emphasize that it is not my purpose to engage in a section 1 Charter analysis as to the reasonableness of the legislation. In my view, what we are trying to fathom is whether the Act has a more burdensome impact on same-sex couples having proper regard to the objects of the legislation under consideration.

The Trial judgment offers a concise introduction to the legislative scheme (at page 692):

Provision for the allowance was first enacted by Parliament in 1975 [S.C. 1974-75-76, c. 58]. At that time Parliament addressed the problem frequently faced by aging married couples who found themselves in the position in which one spouse, generally the husband, who was usually the breadwinner in the family unit and who was usually older than his spouse, retired at age 65. The problem was caused by the fact that his wife, who frequently had been the unpaid homemaker, had no income and would not be eligible for the old age pension for a few years, being younger than her retired husband. The unfortunate result was that the income of the two-spouse family unit dropped drastically until the wife reached 65 years of age and became eligible for the old age pension.

The objective of the spouse’s allowance program was elaborated on by the then Minister of National Health and Welfare, the Honourable Marc Lalonde, who made it clear that this target class of married elderly and near-elderly women, who had jeopardized their own earning potential in order to raise a family, is but one part of the larger class of needy elderly and near-elderly. The spouse’s allowance program was never intended to address, by itself, the needs of all elderly and near-elderly persons.[38] During the second reading of the 1979 amendment to the Old Age Security Act [Bill C-6 (An Act to amend the Old Age Security Act, S.C. 1979, c. 4)], the Honourable Flora MacDonald remarked (see Trial judgment at page 692):

Statistics have shown that in 90 per cent of marriages the younger spouse is female and that females live longer than males. These women, who in their younger years remained in the home looking after children, with no access to continuing income or pension plans, are the same women who in their later years too often became the victims of a society which has not yet come to terms with equality in the work place. [House of Commons Debate, October 22, 1979, at p. 476].

Against this backdrop can it be said that the legislative scheme has a disproportionate or more burdensome effect on same-sex couples? I think not. Before us is a case in which a benefit has been conferred on a narrow class of persons who can be readily identified and who are in financial need because of a pattern of financial interdependency, characteristic of heterosexual couples, and which cannot in any reasonable way be deemed relevant to same-sex couples or, for that matter, other non-spousal relationships. Moreover, while it might have been argued that the impugned legislation has a more burdensome impact on same-sex couples because it denies a benefit to an economically disadvantaged group per se, the appellants led no evidence on this point. In any event, I suspect that such a characterization would be found inapt. It is widely recognized that gay men and lesbians “are found in all geographical areas, social classes, education levels, races, and religions.”[39]

Can it be said that the restricted scope of the spouse’s allowance program has the effect of perpetuating or reinforcing homophobic views, or of inflicting emotional or psychological harm on same-sex couples, as was contemplated in Knodel, supra? Or does the real objection lie in the restricted definition of “spouse” found throughout provincial and federal statutes of which section 2 of the Act is but an example? If the latter question reflects the true source of the controversy then in reality we are confronted with an indirect challenge to the common law and statutory concept of marriage, as reflected in the definition of spouse prescribed by the Act. In my opinion, this is the case before us.

I hasten to add that this appeal does not turn on whether same-sex couples should be legally recognized or whether otherwise qualified same-sex couples should be entitled to benefits available to opposite-sex ones. Undoubtedly, there are those whose expectations exceed the jurisdiction of this Court because they continue to perceive the issue in terms of whether we are prepared to recognize that same-sex couples are no different than opposite-sex ones; that is to say the former deserve equal dignity and therefore equal treatment with the latter. But once again that is not an issue which this Court can address.

In the instant case, it is clear that this Court must limit its inquiry to determining whether the appellants’ equality guarantee under section 15 of the Charter was breached. That inquiry is not to be conducted in a factual vacuum. It is necessarily confined to the question of entitlement under the spouse’s allowance program and in accordance with the law stated in Andrews v. Law Society of British Columbia.

In my opinion, the denial of the spouse’s allowance does not cross the line from “distinction” to “discrimination”. The adverse impact test articulated in Andrews cannot be unilaterally reformulated by this Court, nor can the applicability of the similarly situated test and the “irrelevancy” criterion be conveniently forgotten. In the words of the Trial Judge: “the [appellants] do not benefit because of their non-spousal status rather than because of their sexual orientation” (at page 704). I agree. In this regard, this case shares much in common with those which have preceded it.[40] What we are confronted with is an indirect challenge to the common law and statutory concept of marriage. A direct challenge to that legal concept (spousal status) and the discrimination model outlined in Andrews would bring the issue to a head.[41]

For the foregoing reasons, I would dismiss the appeal. Given the unique history of this litigation, which I need not detail, I have every confidence that the respondent will not pursue the costs to which she is entitled.

---