Date: Thu, 25 May 1995 18:53:45 -0400 From: ae913@freenet.carleton.ca (Timothy Ross Wilson) To: queerlaw@abacus.oxy.edu, queerplanet@abacus.oxy.edu Subject: Egan v. Canada Reply-To: ae913@freenet.carleton.ca - 1 - Document[5]51 egan v. canada James Egan and John Norris Nesbit Appellants v. Her Majesty The Queen in Right of Canada Respondent and Attorney General of Quebec, Canadian Human Rights Commission, Commission des droits de la personne du Quebec, Equality for Gays and Lesbians Everywhere, Metropolitan Community Church of Toronto, Inter-Faith Coalition on Marriage and the Family and Canadian Labour Congress Interveners Indexed as: Egan v. Canada File No.: 23636. 1994: November 1; 1995: May 25. Present: Lamer C.J. and La Forest, L'Heureux-Dube, Sopinka, Gonthier, Cory, McLachlin, Iacobucci and Major JJ. on appeal from the federal court of appeal Constitutional law -- Charter of Rights -- Equality rights -- Old age security legislation providing for allowance for spouse of pensioner -- Definition of "spouse" restricted to person of opposite sex -- Whether definition of "spouse" infringes s. 15(1) of Canadian Charter of Rights and Freedoms -- If so, whether infringement justifiable under s. 1 of Charter -- Old Age Security Act, R.S.C., 1985, c. O-9, ss. 2, 19(1). The appellants are homosexuals who have lived together since 1948 in a relationship marked by commitment and interdependence similar to that which one expects to find in a marriage. When E became 65 in 1986, he began to receive old age security and guaranteed income supplements under the Old Age Security Act. On reaching age 60, N applied for a spousal allowance under s. 19(1) of the Act, which is available to spouses between the ages of 60 and 65 whose combined income falls below a fixed level. His application was rejected on the basis that the relationship between N and E did not fall within the definition of "spouse" in s. 2, which 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 brought an action in the Federal Court seeking a declaration that the definition contravenes s. 15(1) of the Canadian Charter of Rights and Freedoms on the ground that it discriminates on the basis of sexual orientation and a declaration that the definition should be extended to include "partners in same-sex relationships otherwise akin to a conjugal relationship". The Trial Division dismissed the action. The Federal Court of Appeal, in a majority decision, upheld the judgment. Held (L'Heureux-Dube, Cory, McLachlin and Iacobucci JJ. dissenting): The appeal should be dismissed. The definition of "spouse" in s. 2 of the Old Age Security Act is constitutional. Per Lamer C.J. and La Forest, Gonthier and Major JJ.: The analysis under s. 15 of the Charter involves three steps: the first looks to whether the law has drawn a distinction between the claimant and others; the second questions whether the distinction results in disadvantage, and examines whether the impugned legislation imposes a burden, obligation or disadvantage on a group of persons to which the claimant belongs which is not imposed on others, or does not provide them with a benefit which it grants others; the third step assesses whether the distinction is based on an irrelevant personal characteristic which is either enumerated in s. 15(1) or one analogous thereto. The first step is satisfied in this case, since Parliament has clearly made a distinction between the claimant and others. The second step is also satisfied: while it may be true that the appellants have suffered no prejudice because by being treated as individuals they have received considerably more in combined federal and provincial benefits than they would have received had they been treated as "spouses", there is nothing to show that this is generally the case with homosexual couples. Sexual orientation is a deeply personal characteristic that is either unchangeable or changeable only at unacceptable personal costs, and so falls within the ambit of s. 15 protection as being analogous to the enumerated grounds. All that remains to be considered under the third step is whether the distinction made by Parliament is relevant. In assessing relevancy for this purpose one must look at the nature of the personal characteristic and its relevancy to the functional values underlying the law. A form of comparative analysis must be undertaken to determine whether particular facts give rise to inequality. This comparative analysis must be linked to an examination of the larger context, and in particular with an understanding that the Charter was not enacted in a vacuum, but must be placed in its proper linguistic, philosophic and historical contexts. The singling out of legally married and common law couples as the recipients of benefits necessarily excludes all sorts of other couples living together, whatever reasons these other couples may have for doing so and whatever their sexual orientation. What Parliament clearly had in mind was to accord support to married couples who were aged and elderly, for the advancement of public policy central to society. Moreover, in recognition of changing social realities, s. 2 was amended so that whenever the term "spouse" was used in the Act it was to be construed to extend beyond legal married couples to couples in a common law marriage. Marriage has from time immemorial been firmly grounded in our legal tradition, one that is itself a reflection of long-standing philosophical and religious traditions. But its ultimate raison d'etre transcends all of these and is firmly anchored in the biological and social realities that heterosexual couples have the unique ability to procreate, that most children are the product of these relationships, and that they are generally cared for and nurtured by those who live in that relationship. In this sense, marriage is by nature heterosexual. It would be possible to legally define marriage to include homosexual couples, but this would not change the biological and social realities that underlie the traditional marriage. Many of the underlying concerns that justify Parliament's support and protection of legal marriage extend to heterosexual couples who are not legally married. Many of these couples live together indefinitely, bring forth children and care for them in response to familial instincts rooted in the human psyche. These couples have need for support just as legally married couples do in performing this critical task, which is of benefit to all society. Faced with the social reality that increasing numbers choose not to enter a legal marriage but live together in a common law relationship, Parliament has elected to support these relationships. Parliament is wholly justified in extending support to heterosexual couples like this, which is not to say, however, that it is obligated to do so and may not treat married and unmarried couples differently. Neither in its purpose nor in its effect does the legislation constitute an infringement of the fundamental values sought to be protected by the Charter. None of the couples excluded from benefits under the Act are capable of meeting the fundamental social objectives thereby sought to be promoted by Parliament. While these couples undoubtedly provide mutual support for one another, and may occasionally adopt or bring up children, this is exceptional and in no way affects the general picture. Homosexual couples differ from other excluded couples in that their relationships include a sexual aspect, but this sexual aspect has nothing to do with the social objectives for which Parliament affords a measure of support to married couples and those who live in a common law relationship. The distinction adopted by Parliament is relevant here to describe a fundamental social unit to which some measure of support is given. The impugned legislation, even had it infringed s. 15, would have been upheld for the reasons given in McKinney v. University of Guelph and for those mentioned in the discussion of discrimination in this case. Per Sopinka J.: The impugned legislation infringes s. 15(1) of the Charter, for the reasons given by Cory J. Such infringement, however, is saved under s. 1. Government must be accorded some flexibility in extending social benefits and does not have to be pro-active in recognizing new social relationships. It is not realistic for the Court to assume that there are unlimited funds to address the needs of all. A judicial approach on this basis would tend to make a government reluctant to create any new social benefit schemes because their limits would depend on an accurate prediction of the outcome of court proceedings under s. 15(1). This Court has recognized that it is legitimate for the government to make choices between disadvantaged groups and that it must be provided with some leeway to do so. When the definition of "spouse" in the Old Age Security Act is measured against overall objectives of alleviation of poverty of elderly spouses, it should not be judged on the basis that Parliament has made this choice for all time. The history of the legislation shows an evolving expansion of the definition of the intended recipients of the benefits. The Attorney General of Canada has taken the position that the means chosen does not have to be necessarily the solution for all time. Hence, since the impugned legislation can be regarded as a substantial step in an incremental approach to include all those who are shown to be in serious need of financial assistance due to the retirement or death of a supporting spouse, it is rationally connected to the objective. With respect to minimal impairment, the legislation represents the kind of socio-economic question in respect of which the government is required to mediate between competing groups rather than being the protagonist of an individual. In these circumstances, the Court will be more reluctant to second-guess the choice which Parliament has made. There is also proportionality between the effects of the legislation on the protected right and the legislative objective. The proper balance was struck by Parliament in providing financial assistance to those who were shown to be in the greatest need of assistance. Per Cory and Iacobucci JJ. (dissenting): In determining whether a s. 15(1) right to equality has been violated, the first step is to determine whether, owing to a distinction created by the questioned law, a claimant's right to equality has been denied. During this first step, the inquiry should focus upon whether the challenged law has drawn a distinction between the claimant and others, based on personal characteristics. The second step is to determine whether the distinction created by the law results in discrimination. In order to make this determination, it is necessary to consider first, whether the equality right was denied on the basis of a personal characteristic which is either enumerated in s. 15(1) or which is analogous to those enumerated, and second, whether that distinction has the effect on the claimant of imposing a burden, obligation or disadvantage not imposed upon others or of withholding or limiting access to benefits or advantages which are available to others. Any search for either equality or discrimination requires comparisons to be made between groups of people. Whether or not discrimination exists must be assessed in a larger social, political and legal context. The resolution of the question as to whether there is discrimination under s. 15(1) must be kept distinct from the determination as to whether or not there is justification for that discrimination under s. 1 of the Charter. This analytical separation between s. 15(1) and s. 1 is important since the Charter claimant must satisfy the onus of showing only that there exists in the legislation a distinction which is discriminatory. Only after the court finds a breach of s. 15(1) does the government bear the onus of justifying that discrimination. Since the law challenged draws a clear distinction between opposite-sex couples and same-sex couples, this case presents a situation of direct discrimination. As a result of the definition of a common law spouse as a "person of the opposite sex", homosexual common law couples are denied the benefit of the spousal allowance which is available to heterosexual common law couples. This distinction amounts to a clear denial of equal benefit of the law. In addition to being denied an economic benefit, homosexual couples are denied the opportunity to make a choice as to whether they wish to be publicly recognized as a common law couple because of the definition of "spouse" set out in the Old Age Security Act. The public recognition and acceptance of homosexuals as a couple may be of tremendous importance to them and to the society in which they live. To deny homosexual couples the right to make that choice deprives them of the equal benefit of the law. The distinction in the Act is based on a personal characteristic, namely sexual orientation. Sexual orientation is analogous to the grounds of discrimination enumerated in s. 15(1). The historic disadvantage suffered by homosexual persons has been widely recognized and documented. Sexual orientation is more than simply a "status" that an individual possesses: it is something that is demonstrated in an individual's conduct by the choice of a partner. Just as the Charter protects religious beliefs and religious practice as aspects of religious freedom, so too should it be recognized that sexual orientation encompasses aspects of "status" and "conduct" and that both should receive protection. The distinction drawn by s. 2 of the Old Age Security Act on the basis of sexual orientation constitutes discrimination. The legislation denies homosexual couples equal benefit of the law not on the basis of merit or need, but solely on the basis of sexual orientation. The definition of "spouse" as someone of the opposite sex reinforces the stereotype that homosexuals cannot and do not form lasting, caring, mutually supportive relationships with economic interdependence in the same manner as heterosexual couples. The appellants' relationship vividly demonstrates the error of that approach. The discriminatory impact cannot be deemed to be trivial when the legislation reinforces prejudicial attitudes based on such faulty stereotypes. The impugned legislation is not saved under s. 1 of the Charter. While the objective of the spousal allowance, which is geared toward the mitigation of poverty among "elderly households", is of pressing and substantial importance, the allowance in its present form is not rationally connected to its legislative goals. A program which included the appellants would better achieve the intended goal while respecting the Charter rights of gays and lesbians. Nor is the denial of the appellants' s. 15 rights through the ineligibility for receipt of the spousal allowance minimally impaired simply because the appellants' joint income would have roughly been the same because of N's receipt of provincial support supplementing his income for a completely unrelated reason. The provincial and federal programs are clearly not co-extensive, and even if they were part of the same overlapping legislative scheme, this is not sufficient to ground a s. 1 justification. Finally, the attainment of the legislative goal is outweighed by the abridgment of the right in this case. The importance of providing relief to some elderly couples does not justify an infringement of the equality rights of the elderly couples who do not benefit for constitutionally irrelevant reasons. The definition of "spouse" in s. 2 of the Act should be read down by deleting the words "of the opposite sex" and reading in the words "or as an analogous relationship" after the words "if the two persons publicly represent themselves as husband and wife". Per L'Heureux-Dube J. (dissenting): A return to the fundamental purpose of s. 15 of the Charter is necessary in order to reconcile the divergent approaches taken by this Court in recent jurisprudence, as well as in the present case and in Miron and Thibaudeau. At the heart of s. 15 is the protection of, and respect for, basic human dignity. "Discrimination" must therefore be at the forefront of the court's analysis. In order for discrimination to be addressed and identified in all of its varied contexts and forms, it is preferable to focus on impact (i.e. discriminatory effect) rather than on constituent elements (i.e. the grounds of the distinction). Discriminatory effects must, moreover, be evaluated from the point of view of the victim, rather than from that of the state. Considerations of relevance are more properly viewed as justifications under s. 1 than as factors integral to the identification of discrimination in the first place. The following factors must be established by a rights claimant before the impugned distinction will be found to be discriminatory within the meaning of s. 15 of the Charter: (1) there must be a legislative distinction; (2) this distinction must result in a denial of one of the four equality rights on the basis of the rights claimant's membership in an identifiable group; and (3) this distinction must be "discriminatory" within the meaning of s. 15. A distinction is discriminatory within the meaning of s. 15 where it is capable of either promoting or perpetuating the view that the individual adversely affected by this distinction is less capable, or less worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration. The absence or presence of discriminatory impact should be assessed according to a subjective-objective standard -- the reasonably held view of one who is possessed of similar characteristics, under similar circumstances, and who is dispassionate and fully apprised of the circumstances. This determination is arrived at by considering two categories of factors: (1) the nature of the group adversely affected by the impugned distinction, and (2) the nature of the interest adversely affected by the impugned distinction. With respect to the first category, groups that are more socially vulnerable will experience the adverse effects of a legislative distinction more vividly than if the same distinction were directed at a group which is not similarly vulnerable. In evaluating the nature of the group affected by the impugned distinction, it is relevant to inquire into many of the criteria traditionally employed in the Andrews analysis, such as whether the impugned distinction is based upon fundamental attributes that are generally considered to be essential to our popular conception of `personhood' or `humanness', whether the adversely affected group is already a victim of historical disadvantage, whether this distinction is reasonably capable of aggravating or perpetuating that disadvantage, whether group members are currently vulnerable to stereotyping, social prejudice and/or marginalization, and whether this distinction exposes them to the reasonable possibility of future vulnerability of this kind. Membership in a "discrete and insular minority", lacking in political power and thus vulnerable to having its interests overlooked, is another consideration that may be taken into account. The absence or presence of some of these factors will not, however, be determinative of the analysis. However, awareness of, and sensitivity to, the realities of those experiencing the distinction is an important task that judges must undertake when evaluating the impact of the distinction on members of the affected group. Similarly, the more fundamental the interest affected or the more serious the consequences of the distinction, the more likely that the impugned distinction will have a discriminatory impact even with respect to groups that occupy a position of advantage in our society. While the Charter is not a document of economic rights and freedoms, the nature, quantum and context of an economic prejudice or denial of such a benefit are important factors in determining whether the distinction from which the differing economic consequences flow is one which is discriminatory. The discriminatory calibre of a particular distinction cannot, however, be fully appreciated without also evaluating the constitutional and societal significance of the interests adversely affected. Tangible economic consequences are but one manifestation of the more intangible and invidious harms flowing from discrimination, which the Charter seeks to root out. In other cases, the prejudice will be to an important individual interest rather than to one that is economic in nature. Both categories of factors emphasize that it is no longer the "grounds" of the distinction that are dispositive of the question of whether discrimination exists, but rather the social context of the distinction that matters. An effects-based approach to discrimination is the logical next step in the evolution of s. 15 jurisprudence since Andrews. Homosexual couples are denied the equal benefit of the law on the basis of the legislative distinction in s. 2 of the Old Age Security Act, which defines couples as relationships of "opposite sex". That the appellants are able to claim higher benefits as separate individuals does not alter the fact that they have been denied the benefits, both tangible and intangible, of filing for old age benefits as a couple. The impugned distinction excludes the rights claimants because they are homosexual. Consideration of both the nature of the group and the interest affected leads us to conclude that the distinction is discriminatory. Same-sex couples are a highly socially vulnerable group, in that they have suffered considerable historical disadvantage, stereotyping, marginalization and stigmatization within Canadian society. The distinction relates to a fundamental aspect of personhood and affects individuals who, in addition to being homosexuals, are also elderly and poor. Turning to the interest affected, the impugned legislation is a cornerstone in Canada's social security net, which is, in turn, a cherished and fundamental institution in our society. The violation of s. 15(1) of the Charter cannot be salvaged by s. 1, as it is not relevant to a proportionate extent to a pressing and substantial objective. While the objective of the legislation is pressing and substantial, the means chosen to achieve this objective fails all three branches of the proportionality test. The legislation excludes couples who would fill all of the other criteria in the Act except the requirement that they are of the opposite sex. To find that this distinction is rationally connected to the objective of the legislation requires us to conclude that same-sex couples are so different from married couples that it would be unreasonable to make the same benefits available to both. At best, the government has only demonstrated that this is its assumption. The presumption that same-sex relationships are somehow less interdependent then opposite-sex relationships is, itself, a fruit of stereotype rather than one of demonstrable, empirical reality. Nor is s. 15 minimally impaired. A reasonable alternative remedy is available: the discriminatory effect would be eliminated without prejudice to the rights or interests of any other group by extending coverage to same-sex couples who otherwise fulfil all of the other non-discriminatory criteria required in the Act. Deference under this branch of the s. 1 test is not appropriate when there is a reasonable alternative that is readily available, that is not the subject of conflicting social science views, and that could not result in a concomitant prejudice to another group. Finally, the deleterious effects of the impugned distinction outweigh its salutary effects. Per McLachlin J. (dissenting): The reasons of Cory and Iacobucci JJ. were substantially agreed with. On the basis of the principles outlined in Miron v. Trudel, released concurrently, the impugned legislation infringes s. 15(1) of the Charter and the infringement is not saved under s. 1. Cases Cited By La Forest J. Referred to: Canada (Attorney General) v. Ward, [1993] 2 S.C.R. 689; Miron v. Trudel, S.C.C., No. 22744, May 25, 1995; Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; Symes v. Canada, [1993] 4 S.C.R. 695; R. v. Turpin, [1989] 1 S.C.R. 1296; R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295; United States of America v. Cotroni, [1989] 1 S.C.R. 1469; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Mahe v. Alta. (Gov't) (1987), 54 Alta. L.R. (2d) 212. By Sopinka J. Referred to: McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927. By Cory and Iacobucci JJ. (dissenting) R. v. Oakes, [1986] 1 S.C.R. 103; Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; R. v. Turpin, [1989] 1 S.C.R. 1296; Symes v. Canada, [1993] 4 S.C.R. 695; Ontario Human Rights Commission v. Simpsons-Sears Ltd., [1985] 2 S.C.R. 536; Tetreault-Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22; Brown v. Board of Education of Topeka, 347 U.S. 483 (1954); Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219; Schachter v. Canada, [1992] 2 S.C.R. 679; Janzen v. Platy Enterprises Ltd., [1989] 1 S.C.R. 1252; R. v. Swain, [1991] 1 S.C.R. 933; Douglas v. Canada (1992), 58 F.T.R. 147; Haig v. Canada (1991), 5 O.R. (3d) 245 (Gen. Div.), aff'd (1992), 9 O.R. (3d) 495 (C.A.); Vriend v. Alberta (1994), 152 A.R. 1; Veysey v. Canada (Correctional Service) (1989), 44 C.R.R. 364; Brown v. British Columbia (Minister of Health) (1990), 42 B.C.L.R. (2d) 294; Knodel v. British Columbia (Medical Services Commission) (1991), 58 B.C.L.R. (2d) 356; Leshner v. Ontario (1992), 16 C.H.R.R. D/184; Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872; R. v. Hess, [1990] 2 S.C.R. 906; Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554; R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295. By L'Heureux-Dube J. (dissenting) R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295; Miron v. Trudel, S.C.C., No. 22744, May 25, 1995; Thibaudeau v. Canada, S.C.C., No. 24154, May 25, 1995; Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; R. v. Turpin, [1989] 1 S.C.R. 1296; Ontario Human Rights Commission v. Simpsons-Sears Ltd., [1985] 2 S.C.R. 536; R. v. Genereux, [1992] 1 S.C.R. 259; Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872; R. v. Hess, [1990] 2 S.C.R. 906; Schachtschneider v. Canada, [1994] 1 F.C. 40; R. v. Oakes, [1986] 1 S.C.R. 103; Dagenais v. Canadian Broadcasting Corp., [1994] 3 S.C.R. 835; Reference re ss. 193 and 195.1(1)(c) of the Criminal Code (Man.), [1990] 1 S.C.R. 1123; Symes v. Canada, [1993] 4 S.C.R. 695; Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554; Bliss v. Attorney General of Canada, [1979] 1 S.C.R. 183; Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219; Schachter v. Canada, [1992] 2 S.C.R. 679. By McLachlin J. (dissenting) Miron v. Trudel, S.C.C., No. 22744, May 25, 1995. Statutes and Regulations Cited B.C. Reg. 479/76, ss. 3(2), 10. Canadian Charter of Rights and Freedoms, ss. 1, 7, 15. Constitution Act, 1982, s. 52(1). Guaranteed Available Income for Need Act, R.S.B.C. 1979, c. 158, s. 8. N.B. Reg. 82-227. Old Age Security Act, R.S.C. 1970, c. O-6. Old Age Security Act, R.S.C., 1985, c. O-9, ss. 2, 19. Resolution on Equal Rights for Homosexuals and Lesbians in the European Community, European Parliament. A3-0028/94. Saskatchewan Assistance Act, R.S.S. 1978, c. S-8. Sask. Reg. 78/66. Social Welfare Act, R.S.N.B. 1973, c. S-11. Welfare Assistance Act, R.S.P.E.I. 1988, c. W-3. Welfare Assistance Act Regulations, EC746/84. Authors Cited Abella, Rosalie Silberman. Report of the Commission on Equality in Employment. Ottawa: Supply and Services Canada, 1984. 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 Supreme Court L.R. (2d) 503. Canada. House of Commons. Standing Committee on Health, Welfare and Social Affairs. Minutes of Proceedings and Evidence, June 12, 1975, p. 25:7. Canada. House of Commons. Sub-Committee on Equality Rights. Equality For All: Report of the Parliamentary Committee on Equality Rights. Ottawa: Queen's Printer, 1985. Duclos, Nitya. "An Argument for Legal Recognition of Same Sex Marriage". In Carol Rogerson, Family Law Cases and Materials 1991-92, vol. I. Faculty of Law, University of Toronto. Eichler, Margrit. Families in Canada Today, Recent Changes and their Policy Consequences, 2nd ed. Toronto: Gage, 1988. Ginson, 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 U.N.B.L.J. 2. Gibson, Dale. The Law of the Charter: Equality Rights. Toronto: Carswell, 1990. Hogg, Peter W. Constitutional Law of Canada, 3rd ed. Scarborough, Ont.: Carswell, 1992. House of Commons Debates, vol. I, 1st Sess., 31st Parl., October 22, 1979, p. 476. Iyer, N. "Categorical Denials: Equality Rights and the Shaping of Social Identity" (1993), 19 Queen's L.J. 179. Lepofsky, M. David. "The Canadian Judicial Approach to Equality Rights: Freedom Ride or Rollercoaster?" (1992), 1 N.J.C.L. 315. Petersen, Cynthia. "A Queer Response to Bashing: Legislating Against Hate" (1991), 16 Queen's L.J. 237. Quebec. Commission des droits de la personne. Comite de consultation. De l'illegalite a l'egalite: Rapport de la consultation publique sur la violence et la discrimination envers les gais et lesbiennes. Montreal: La Commission, 1994. Smith, C. Lynn. "Judicial Interpretation of Equality Rights under the Canadian Charter of Rights and Freedoms: Some Clear and Present Dangers" (1988), 23 U.B.C. L. Rev. 65. Smith, Carolyn Gibson. "Proud but Cautious": Homophobic Abuse and Discrimination in Nova Scotia. Nova Scotia Public Interest Research Group, July 1994. Vanier Institute of the Family. Profiling Canada's Families. Ottawa: Vanier Institute of the Family, 1994. APPEAL from a judgment of the Federal Court of Appeal, [1993] 3 F.C. 401, 103 D.L.R. (4th) 336, 15 C.R.R. (2d) 310, 153 N.R. 161, affirming a judgment of the Federal Court, Trial Division, [1992] 1 F.C. 687, 87 D.L.R. (4th) 320, 38 R.F.L. (3d) 294, 47 F.T.R. 305, dismissing the appellants' claim. Appeal dismissed, L'Heureux-Dube, Cory, McLachlin and Iacobucci JJ. dissenting. Joseph J. Arvay, Q.C., and Leah Greathead, for the appellants. H. J. Wruck, Q.C., F. E. Campbell, Q.C., and L. M. Hitch, for the respondent. Madeleine Aube, for the intervener the Attorney General of Quebec. William F. Pentney and J. Helen Beck, for the intervener the Canadian Human Rights Commission. Helene Tessier, for the intervener the Commission des droits de la personne du Quebec. Cynthia Petersen, for the intervener Equality for Gays and Lesbians Everywhere. Charles M. Campbell and Susan Ursel, for the intervener Metropolitan Community Church of Toronto. Peter R. Jervis and Iain T. Benson, for the intervener Inter-Faith Coalition on Marriage and the Family. Steven Barrett and Vanessa Payne, for the intervener the Canadian Labour Congress. Solicitors for the appellants: Arvay Finlay, Victoria. Solicitor for the respondent: The Department of Justice, Ottawa. Solicitor for the intervener the Attorney General of Quebec: Madeleine Aube, Sainte-Foy. Solicitor for the intervener the Canadian Human Rights Commission: William F. Pentney, Ottawa. Solicitor for the intervener the Commission des droits de la personne du Quebec: Helene Tessier, Montreal. Solicitor for the intervener Equality for Gays and Lesbians Everywhere: Cynthia Petersen, Ottawa. Solicitors for the intervener Metropolitan Community Church of Toronto: Iler, Campbell, Toronto. Solicitors for the intervener Inter-faith Coalition on Marriage and the Family: Lerner & Associates, Toronto. Solicitors for the intervener the Canadian Labour Congress: Sack Goldblatt Mitchell, Toronto. PARAGRAPH NUMBERING The reasons are published in the SCR by order of precedence and the paragraph numbering included in those reasons has been prepared accordingly. For this case, the reasons will be published in the following order: La Forest J. (and Lamer C.J. and Gonthier and Major JJ. (paras. 1 to 30) L'Heureux-Dube J. (paras. 31 to 101) Sopinka J. (paras. 102 to 111) Iacobucci and Cory JJ. (paras. 112 to 230) McLachlin J. (para. 231) N.B.: The paragraph numbering will now appear in the SCR. LA FOREST J.: 1 This appeal concerns the constitutionality of ss. 2 and 19(1) of the Old Age Security Act, R.S.C., 1985, c. O-9, which accord to spouses of pensioners under the Act whose income falls below a stipulated amount, an allowance when they reach the age of 60, payable until they themselves become pensioners at age 65. The appellants maintain these provisions violate s. 15 of the Canadian Charter of Rights and Freedoms as discriminating against persons living in a homosexual relationship because the effect of the definition of "spouse" in s. 2 is to restrict the allowances to spouses in a heterosexual union, i.e. those who are legally married or who live in a common law relationship. 2 My colleague, Mr. Justice Cory, has set forth the facts and judicial history as well as the applicable constitutional and statutory provisions. It will be sufficient for me, then, to refer only to such of this material as may be essential to an understanding of what follows. 3 The provision providing for the payment of the allowance is s. 19(1), which reads as follows: 19. (1) Subject to this Act and the regulations, for each month in any fiscal year, a spouse's allowance may be paid to the spouse of a pensioner if the spouse (a) is not separated from the pensioner; (b) has attained sixty years of age but has not attained sixty-five years of age; and (c) has resided in Canada after attaining eighteen years of age and prior to the day on which the spouse's application is approved for an aggregate period of at least ten years and, where that aggregate period is less than twenty years, was resident in Canada on the day preceding the day on which the spouse's application is approved. That provision, without more, would be confined -- as the term "spouse" had been interpreted in the Act before the original enactment of s. 19(1) in 1975 --to spouses in a legal marriage. At the time of the enactment of s. 19(1), however, s. 2 of the Act was amended to define the term "spouse" for the purposes of the Act to include common law spouses described in the definition. The definition reads: 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 effect of this new definition was to extend the allowances to spouses living in a common law relationship as well as those in a legal marriage. 4 The appellants, James Egan and John Norris Nesbit, are homosexuals who have since 1948 lived together in a relationship marked by commitment and interdependence similar to that which one expects to find in a marriage. When Egan became 65 in 1986, he began to receive old age security and guaranteed income supplements under the Act. On reaching age 60, Nesbit applied for a spousal allowance under s. 19(1), which as I mentioned is available to spouses as defined in the Act between the ages of 60 and 65 whose combined income falls below a fixed level. His application was rejected because the relationship between Nesbit and Egan did not fall within the Act. 5 The appellants' claim before this Court is that the Act contravenes s. 15 of the Charter in that it discriminates on the basis of sexual orientation. To establish that claim, it must first be determined that s. 15's protection of equality without discrimination extends to sexual orientation as a ground analogous to those specifically mentioned in the section. This poses no great hurdle for the appellants; the respondent Attorney General of Canada conceded this point. While I ordinarily have reservations about concessions of constitutional issues, I have no difficulty accepting the appellants' contention that whether or not sexual orientation is based on biological or physiological factors, which may be a matter of some controversy, it is a deeply personal characteristic that is either unchangeable or changeable only at unacceptable personal costs, and so falls within the ambit of s. 15 protection as being analogous to the enumerated grounds. As the courts below observed, this is entirely consistent with a number of cases on the point. Indeed, there is a measure of support for this position in this Court. In Canada (Attorney General) v. Ward, [1993] 2 S.C.R. 689, at pp. 737-39, speaking for my colleagues as well, I observed that the analogous grounds approach in s. 15 was appropriate to a consideration of the character of "social groups" subject to protection as Convention refugees. These, I continued, encompass groups defined by an innate or unchangeable characteristic which, I added, would include sexual orientation. 6 The only difficulty I have with the Crown's concession is that it would seem difficult in the absence of more precise argument to consider the point mentioned by Gonthier J. in Miron v. Trudel, S.C.C., No. 22744, issued concurrently, regarding the need to consider the nature of a ground, be it enumerated in s. 15 or analogous, as a basis for discrimination and its necessary limitation where the distinction drawn by legislation merely reflects or is the expression of a fundamental reality or value. I need not pursue this, however, since I do not think that, in the circumstances of this case, the appellants have suffered discrimination under s. 15 as that term has been defined by this Court. 7 The nature of discrimination within the meaning of s. 15(1) of the Charter was first discussed by this Court in the seminal case of Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143. In the principal reasons in that case, McIntyre J., at p. 175, underlined the importance in a constitutional document, which is not easily modified, of achieving a workable balance that permits government to perform effectively its function of making ongoing choices in the interests of society and the work of the courts in ensuring protection for the equality rights described in s. 15. As he stated, what we must do is "to provide a continuing framework for the legitimate exercise of governmental power and, at the same time, for the unremitting protection" of equality rights. And he warned (see p. 168), as I did in my separate reasons, that not all distinctions resulting in disadvantage to a particular group will constitute discrimination. It would bring the legitimate work of our legislative bodies to a standstill if the courts were to question every distinction that had a disadvantageous effect on an enumerated or analogous group. This would open up a s. 1 enquiry in every case involving a protected group. As I put it in Andrews, at p. 194, "it was never intended in enacting s. 15 that it become a tool for the wholesale subjection to judicial scrutiny of variegated legislative choices in no way infringing on values fundamental to a free and democratic society". 8 What then is discrimination? There are several comments in the course of McIntyre J.'s remarks in Andrews that go a long way towards clarifying the concept. Thus, at p. 174, he stated: . . . 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. This statement cannot, however, be looked at in isolation. It must be read in conjunction with McIntyre J.'s earlier comment, at p. 165, as follows: In other words, the admittedly unattainable ideal [of equality] 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. Similarly in my separate reasons, at p. 193, I observed that "the relevant question . . . is . . . whether the impugned provision amounts to discrimination in the sense in which my colleague has defined it, i.e. on the basis of `irrelevant personal differences' such as those listed in s. 15 . . .". 9 As Gonthier J. has noted in Miron v. Trudel, this involves a three-step analysis, which he puts this way (at pp. 8-9): The first step looks to whether the law has drawn a distinction between the claimant and others. The second step then questions whether the distinction results in disadvantage, and examines whether the impugned law imposes a burden, obligation or disadvantage on a group of persons to which the claimant belongs which is not imposed on others, or does not provide them with a benefit which it grants other (Andrews, supra). . . . The third step assesses whether the distinction is based on an irrelevant personal characteristic which is either enumerated in s. 15(1) or one analogous thereto. 10 There is no question that the first step is satisfied in this case. Parliament has clearly made a distinction between the claimant and others. This, of course, does not carry one very far. Parliament is in the business of making such distinctions in developing programs and policies which is the task assigned to it in our democratic system. Further ingredients must be added to warrant a distinction being discriminatory. 11 The second step will also, in general at least, not be of great assistance. Ordinarily decisions do result in advantages or disadvantages to individuals and groups, sometimes intentionally, sometimes unintentionally. Parliament, as I mentioned, is in the business of making choices, and this inevitably involves the distribution of benefits and burdens in our society. 12 In this case, however, the respondent contends that the appellants have suffered no prejudice because by being treated as individuals they have received considerably more in combined federal and provincial benefits than they would have received had they been treated as "spouses". I would simply dispose of this argument on the ground that, while this may be true in this specific instance, there is nothing to show that this is generally the case with homosexual couples, which is the point the respondent must establish. My colleague Cory J. also makes this point in his reasons. 13 I turn then to the third step of the analysis described by Gonthier J. Since it has already been accepted that "sexual orientation" is an analogous ground under s. 15(1), all that remains to be considered under this step is whether the distinction made by Parliament is relevant, what Gonthier J. describes in Miron v. Trudel, at p. 10, as the second aspect of this third step. He there notes that in assessing relevancy for this purpose one must look at "the nature of the personal characteristic and its relevancy to the functional values underlying the law". At this stage, he adds, one must necessarily undertake a form of comparative analysis to determine whether particular facts give rise to inequality; see also Symes v. Canada, [1993] 4 S.C.R. 695, at p. 754. This proposition, too, derives from McIntyre J.'s reasons in Andrews, supra, who, at p. 164, states: It [equality] 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. 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. Gonthier J. adds, citing Wilson J. in R. v. Turpin, [1989] 1 S.C.R. 1296, at pp. 1331-32, that this comparative analysis must be linked to an examination of the larger context, and in particular with an understanding that the Charter was, in Dickson J.'s words, "not enacted in a vacuum", but must "be placed in its proper linguistic, philosophic and historical contexts" if we are to avoid mechanical and sterile categorization; see R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295, at p. 344; see also United States of America v. Cotroni, [1989] 1 S.C.R. 1469, at pp. 1490-91. 14 In embarking upon this comparative analysis, I shall begin with an examination of the statute with a view to determining "the functional values underlying the law". I shall then examine the personal characteristic here in issue to determine its relevancy to these functional values. 15 In undertaking an examination of the statute, it is now settled that one must not focus narrowly on a provision that has the effect of depriving a group of a benefit that others who initially appear to be in a similar position are accorded. This is the "similarly situated test" which has been categorically rejected by this Court; see Andrews; see also McKinney v. University of Guelph, [1990] 3 S.C.R. 229, at p. 279. Among the reasons for its rejection is that put forth by Kerans J.A. in Mahe v. Alta. (Gov't) (1987), 54 Alta. L.R. (2d) 212, at p. 244 (in a passage approved by McIntyre J. in Andrews, at p. 168), namely: . . . the test accepts an idea of equality which is almost mechanical, with no scope for considering the reason for the distinction. In consequence, subtleties are found to justify a finding of dissimilarity which reduces the test to a categorization game. Moreover, the test is not helpful. After all, most laws are enacted for the specific purpose of offering a benefit or imposing a burden on some persons and not on others. The test catches every conceivable difference in legal treatment. [Emphasis added.] 16 Having quoted this passage, McIntyre J. in Andrews immediately stated that such a fixed rule or formula cannot be accepted, and then concluded by setting forth how the relevant statute must be approached in the following remarks, at p. 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. It is in that spirit that I propose to examine the Old Age Security Act. 17 Mahoney J.A. in the Court of Appeal has usefully described the broad structure of the Act in the following passage (at p. 411): 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. As is evident from this description, Parliament, in addition to providing greater benefits to the elderly in need, long ago took special account of married couples in need; as I mentioned earlier, before 1975 the term "spouse" only applied to persons who were legally married. This special interest is clearly expressed by the Minister of Health and Welfare, the Honourable Marc Lalonde when testifying before the Standing Committee on Health, Welfare and Social Affairs in relation to the amendment adding the spousal allowance in 1975. He stated: 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 up on 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 the Bill, no more than that, no less than that. See Minutes of Proceedings and Evidence, June 12, 1975, at p. 25:7. 18 I add that other evidence of Parliament's continuing concern for the needs of married couples is the benefits for widows. To complete the picture, I repeat that in 1975 Parliament, by defining spouse as above described, extended the benefits under the Act beyond those who were legally married to common law couples, and it is that definition that has formed the principal focus of the appellants' attack. As I earlier noted, however, it is dangerous to focus narrowly on a particular provision. A more comprehensive contextual approach must be taken to determine the relevancy of the personal characteristic in question to the functional values underlying the law. That is how I propose to proceed. 19 The singling out of legally married and common law couples as the recipients of benefits necessarily excludes all sorts of other couples living together such as brothers and sisters or other relatives, regardless of sex, and others who are not related, whatever reasons these other couples may have for doing so and whatever their sexual orientation. Mahoney J.A., at p. 412 of his reasons in the Court of Appeal, lucidly describes these various couples in the following passage: 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. 20 What reason or purpose, then, can be assigned to the distinction made by Parliament? It seems to me that it is both obvious and deeply rooted in our fundamental values and traditions, values and traditions that could not have been lost on the framers of the Charter. Simply stated, what Parliament clearly had in mind was to accord support to married couples who were aged and elderly, and this for the advancement of public policy central to society. Moreover, in recognition of changing social realities, s. 2 was amended so that whenever the term "spouse" was used in the Act it was to be construed to extend beyond legal married couples to couples in a common law marriage. 21 My colleague Gonthier J. in Miron v. Trudel has been at pains to discuss the fundamental importance of marriage as a social institution, and I need not repeat his analysis at length or refer to the authorities he cites. Suffice it to say that marriage has from time immemorial been firmly grounded in our legal tradition, one that is itself a reflection of long standing philosophical and religious traditions. But its ultimate raison d'etre transcends all of these and is firmly anchored in the biological and social realities that heterosexual couples have the unique ability to procreate, that most children are the product of these relationships, and that they are generally cared for and nurtured by those who live in that relationship. In this sense, marriage is by nature heterosexual. It would be possible to legally define marriage to include homosexual couples, but this would not change the biological and social realities that underlie the traditional marriage. 22 The marital relationship has special needs with which Parliament and the legislatures and indeed custom and judge-made law have long been concerned. The legal institution of marriage exists both for the protection of the relationship and for defining the obligations that flow from entering into a legal marriage. Because of its importance, legal marriage may properly be viewed as fundamental to the stability and wellbeing of the family and, as such, as Gonthier J. argued in Miron v. Trudel, Parliament may quite properly give special support to the institution of marriage. It is spouses in legal marriage who constitute the bulk of the beneficiaries of spousal allowances. 23 But many of the underlying concerns that justify Parliament's support and protection of legal marriage extend to heterosexual couples who are not legally married. Many of these couples live together indefinitely, bring forth children and care for them in response to familial instincts rooted in the human psyche. These couples have need for support just as legally married couples do in performing this critical task, which is of benefit to all society. Language has long captured the essence of this relationship by the expression "common law marriage". 24 Faced with the social reality that increasing numbers choose not to enter a legal marriage but live together in a common law relationship, Parliament has elected to support these relationships. The legal institution of marriage has long been viewed as the fundamental instrument to promote the underlying values I have referred to. But Parliament cannot force people to get married, and I see no reason why it should not take the necessary means to promote the basic social interests and policies that inform the institution of legal marriage through other instrumentalities. Support of common law relationships with a view to promoting their stability seems well devised to advance many of the underlying values for which the institution of marriage exists. For example, children brought up by single parents more often end up in poverty and impose greater burdens on society. Parliament, it seems to me, is wholly justified in extending support to heterosexual couples like this, which is not to say, however, that it is obligated to do so and may not treat married and unmarried couples differently (see the reasons of Gonthier J. in Miron). 25 Viewed in the larger context, then, there is nothing arbitrary about the distinction supportive of heterosexual family units. And for the reasons set forth by Gonthier J. in Miron, I am not troubled by the fact that not all these heterosexual couples in fact have children. It is the social unit that uniquely has the capacity to procreate children and generally cares for their upbringing, and as such warrants support by Parliament to meet its needs. This is the only unit in society that expends resources to care for children on a routine and sustained basis. As counsel for the intervener the Interfaith Coalition on Marriage and the Family put it, whether the mother or the father leaves the paid workforce or whether both parents are paying after-tax dollars for daycare, this is the unit in society that fundamentally anchors other social relationships and other aspects of society. I add that I do not think the courts should attempt to require meticulous line drawing that would ensure that only couples that had children were included. This could impose on Parliament the burden of devising administrative procedures to ensure conformity that could be both unnecessarily intrusive and difficult to administer, thereby depriving Parliament of that "reasonable room to manoeuvre" which this Court has frequently recognized as necessary; see United States of America v. Cotroni, supra, at p. 1495, and the cases there cited. This I think is wholly consistent with the workable balance between Parliament and the courts sought to be achieved in Andrews, supra, to which I have earlier referred. 26 Neither in its purpose or effect does the legislation constitute an infringement of the fundamental values sought to be protected by the Charter. None of the couples excluded from benefits under the Act are capable of meeting the fundamental social objectives thereby sought to be promoted by Parliament. These couples undoubtedly provide mutual support for one another, and that, no doubt, is of some benefit to society. They may, it is true, occasionally adopt or bring up children, but this is exceptional and in no way affects the general picture. I fail to see how homosexuals differ from other excluded couples in terms of the fundamental social reasons for which Parliament has sought to favour heterosexuals who live as married couples. Homosexual couples, it is true, differ from other excluded couples in that their relationships include a sexual aspect. But this sexual aspect has nothing to do with the social objectives for which Parliament affords a measure of support to married couples and those who live in a common law relationship. 27 In a word, the distinction made by Parliament is grounded in a social relationship, a social unit that is fundamental to society. That unit, as I have attempted to explain, is unique. It differs from all other couples, including homosexual couples. Other excluded couples, it is true, do not have to be described by reference to sex or sexual preferences, but this is of no moment. The distinction adopted by Parliament is relevant, indeed essential, to describe the relationship in the way the statute does so as to differentiate the couples described in the statute from all couples that do not serve the social purposes for which the legislature has made the distinction. Homosexual couples are not, therefore, discriminated against; they are simply included with these other couples. 28 I add that this distinction exists in a plethora of statutes, both federal and provincial, and, indeed, it directly or indirectly forms the substratum of an abundance of legal principles and rules at common law and under the civil law system. I realize, of course, that the distinction could in certain circumstances be used in a discriminatory manner, but that is not this case. It is relevant here to describe a fundamental social unit, indeed the fundamental unit in society, to which some measure of support is given. I add, interstitially, that this support does not exacerbate an historic disadvantage; rather it ameliorates an historic economic disadvantage, both for couples who are legally married and those who live in a common law relationship. If the distinction is thought to be irrelevant here, this would, in my view, mean that the courts would have to embark upon a s. 1 justification every time a distinction was made on the basis of marriage, legal or common law. Moreover, it would interfere with the appropriate balance between legislative and judicial power described in Andrews, which I have discussed earlier in these reasons. 29 Had I concluded that the impugned legislation infringed s. 15 of the Charter, I would still uphold it under s. 1 of the Charter for the considerations set forth in my reasons in McKinney v. University of Guelph, [1990] 3 S.C.R. 229, especially at pp. 316-18, some of which are referred to in the reasons of my colleague Justice Sopinka, as well as for those mentioned in my discussion of discrimination in the present case. 30 I would dismiss the appeal and would answer the constitutional questions as follows: 1.Does the definition of "spouse" in s. 2 of the Old Age Security Act, R.S.C., 1985, c. O-9, infringe or deny s. 15(1) of the Canadian Charter of Rights and Freedoms? No. 2.If the answer to question 1 is yes, is the infringement or denial demonstrably justified in a free and democratic society pursuant to s. 1 of the Canadian Charter of Rights and Freedoms? Assuming it is an infringement, it is justifiable under s. 1 of the Charter. L'HEUREUX-DUBE J.: 31 This appeal raises the question of whether a legislative distinction that limits eligibility for a spousal supplement under the Old Age Security Act, R.S.C., 1985, c. 0-9, to "opposite sex" spouses is discriminatory within the meaning of s. 15 of the Canadian Charter of Rights and Freedoms and, if so, whether it is saved by s. 1 of the Charter. Although I agree with much of what is said by my colleagues Justices Cory and Iacobucci, as well as with the result they reach, I have some concerns as to the proper approach to be taken to s. 15 of the Charter, which I shall outline below. I. Analysis 32 There is no more important task in approaching any Charter right than that of characterizing properly its purpose: R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295. Defining with accuracy and sensitivity the purpose of a particular right is, in short, the starting point for rights analysis. By implication, a right is said to be violated when the purpose of that right is denied, undermined, or frustrated by legislative action. Disagreement, no matter how small, at the foundational level of establishing the right's purpose will only magnify over time in terms of how that right is applied. More difficult cases, quite naturally, will only make these differences more apparent. I believe that this phenomenon is beginning to manifest itself in the divergent approaches to s. 15 taken in recent cases before this Court, of which this case, Miron v. Trudel, No. 22794, and Thibaudeau v. Canada, No. 24154, released concurrently, are no exception. The emergence of these differences suggests to me that we may not necessarily be operating with the same underlying purpose in mind. For s. 15 jurisprudence to continue to develop along principled lines, I believe that two things are necessary: (1) we must revisit the fundamental purpose of s. 15; and (2) we must seek out a means by which to give full effect to this fundamental purpose. I set out below my efforts to begin this dialogue. A. The Purpose of s. 15 33 Section 15 does not guarantee the complete absence of distinctions. Nor, for that matter, does it guarantee equality in the abstract. Indeed, as this Court recognized in Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, at p. 169, "for the accommodation of differences, which is the essence of true equality, it will frequently be necessary to make distinctions". The key, then, is to define when those distinctions are constitutionally permissible, and when they are not. 34 As an important starting point to evaluating the purpose of s. 15(1), we need look no further than its text. It reads: 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. [Emphasis added.] It is plain from the language of s. 15 that its fundamental purpose is to guarantee to all individuals a certain kind of equality: equality without discrimination. By implication, where "discrimination" is not present, then the Charter guarantee of equality is satisfied. The nine "grounds" enumerated after this basic guarantee to freedom from discrimination are particular applications and illustrations of the ambit of s. 15. They are not the guarantee itself. 35 If the fundamental purpose of s. 15 is to guarantee equality "without discrimination", then it follows that the pivotal question is, "How do we define 'discrimination'"? Under the approach set out in Andrews, this Court has sought to define "discrimination" by reference to the nine grounds enumerated in s. 15(1) as well as by reference to "analogous grounds", which embody characteristics seen to be held in common by the enumerated grounds. As several commentators have suggested, and as I shall argue shortly, this is an indirect means by which to define discrimination. A preferable approach would be to give independent content to the term "discrimination", and to develop s. 15 along the lines of that definition. In so doing, however, we must heed the warning sounded by Professor Lynn Smith in "Judicial Interpretation of Equality Rights under the Canadian Charter of Rights and Freedoms: Some Clear and Present Dangers" (1988), 23 U.B.C. L. Rev. 65, at p. 86: ... I think that there is a real risk that if the courts continue to review all claims which the ingenious legal mind can create, and if the courts decide that the standard of review will be the same no matter what the claim, then that standard will tend toward the lowest common denominator, or at least toward the standard appropriate to the bulk of the cases being brought. Such a standard will not be a high one- -that is, it will be an easy one for government to pass. This tendency, if it occurs, will then result in a watered-down level of protection for all claims. [Emphasis added.] In short, the standard must not be so broad or vague as to risk undermining in practice the very purposes that s. 15 is intended to further. 36 This Court has recognized that inherent human dignity is at the heart of individual rights in a free and democratic society: Big M Drug Mart Ltd., supra, at p. 336 (per Dickson J. (as he then was)). More than any other right in the Charter, s. 15 gives effect to this notion. Building upon this foundation, I believe that the essence of "discrimination" was largely captured by McIntyre J., speaking for the majority of the Court on this point, in Andrews, supra, at p. 171: 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 as human beings equally deserving of concern, respect and consideration. [Emphasis added.] Equality, as that concept is enshrined as a fundamental human right within s. 15 of the Charter, means nothing if it does not represent a commitment to recognizing each person's equal worth as a human being, regardless of individual differences. Equality means that our society cannot tolerate legislative distinctions that treat certain people as second-class citizens, that demean them, that treat them as less capable for no good reason, or that otherwise offend fundamental human dignity. In a similar vein, I refer to the words of Wilson J. in McKinney v. University of Guelph, [1990] 3 S.C.R. 229, at p. 387 (dissenting, but not on this point): It is, I think, now clearly established that what lies at the heart of s. 15(1) is the promise of equality in the sense of freedom from the burdens of stereotype and prejudice in all their subtle and ugly manifestations. However, the nature of discrimination is such that attitudes rather than laws or rules may be the source of the discrimination. [Emphasis added.] 37 We can further inform our understanding of the purpose of s. 15 by recognizing what it is not. The Charter is a document of civil, political and legal rights. It is not a charter of economic rights. This is not to say, however, that economic prejudices or benefits are irrelevant to determinations under s. 15 of the Charter. Quite the contrary. Economic benefits or prejudices are relevant to s. 15, but are more accurately regarded as symptomatic of the types of distinctions that are at the heart of s. 15: those that offend inherent human dignity. 38 Finally, we must bear in mind that it has been recognized by this Court that an important, though not necessarily exclusive, purpose of s. 15 is the prevention or reduction of distinctions that may worsen the circumstances of those who have already suffered marginalization or historical disadvantage in our society: Andrews, supra; R. v. Turpin, [1989] 1 S.C.R. 1296. See also Lepofsky, "The Canadian Judicial Approach to Equality Rights: Freedom Ride or Rollercoaster?" (1992), 1 N.J.C.L. 315. In many ways, this recognition flows from the fact that discrimination is as much an effect as a purpose, and that those individuals and groups that are more vulnerable in society are also more likely to experience the effects of a distinction more severely. This important relationship was recognized in Judge Abella's groundbreaking Report of the Commission on Equality in Employment (1984), at p. 9: The impact of behaviour is the essence of "systemic discrimination". It suggests that the inexorable, cumulative effect on individuals or groups of behaviour that has an arbitrarily negative impact on them is more significant than whether the behaviour flows from insensitivity or intentional discrimination.... The effect of the system on the individual or group, rather than its attitudinal sources, governs whether or not a remedy is justified. [Emphasis added.] As such, s. 15 requires that we place both the group and the particular distinction adversely affecting that group within the "social, political and legal fabric of our society": Andrews, supra, at p. 152. 39 To summarize, at the heart of s. 15 is the promotion of a society in which all are secure in the knowledge that they are recognized at law as equal human beings, equally capable, and equally deserving. A person or group of persons has been discriminated against within the meaning of s. 15 of the Charter when members of that group have been made to feel, by virtue of the impugned legislative distinction, that they are less capable, or less worthy of recognition or value as human beings or as members of Canadian society, equally deserving of concern, respect, and consideration. These are the core elements of a definition of "discrimination"-- a definition that focuses on impact (i.e. discriminatory effect) rather than on constituent elements (i.e. the grounds of the distinction). 40 Three clarifications should be made at this juncture. First, I acknowledge that the above definition essentially tries to put into words the notion of fundamental human dignity. Dignity being a notoriously elusive concept, however, it is clear that this definition cannot, by itself, bear the weight of s. 15's task on its shoulders. It needs precision and elaboration. I shall attempt to demonstrate shortly how this approach to discrimination can find more concrete and principled expression using many of the criteria that have in the past proven themselves to be highly apposite under the approach taken by this Court in Andrews. As such, it will become evident that the approach I suggest is far less a departure from that developed in Andrews than may appear at first blush. I believe many of those analytical tools to be valid. The problem, in my mind, lies not with the tools but with the framework within which they have in the past been employed. In short, if the framework is not perfectly suited for the tools, then we do not use the tools to their full potential. 41 Second, I note that although the utopian ideal would be a society in which nobody is made to feel debased, devalued or denigrated as a result of legislative distinctions, such an ideal is clearly unrealistic. The guarantee against discrimination cannot possibly hold the state to a standard of conduct consistent with its most sensitive citizens. Clearly, a measure of objectivity must be incorporated into this determination. This being said, however, it would be ironic and, in large measure, self- defeating to the purposes of s. 15 to assess the absence or presence of discriminatory impact according to the standard of the "reasonable, secular, able-bodied, white male". A more appropriate standard is subjective- objective -- the reasonably held view of one who is possessed of similar characteristics, under similar circumstances, and who is dispassionate and fully apprised of the circumstances. The important principle, however, which this Court has accepted, is that discriminatory effects must be evaluated from the point of view of the victim, rather than from the point of view of the state. 42 Third, I feel compelled to comment on the role of "relevance" in s. 15. Relevance has been advocated by commentators such as Professor Gibson as a means of rendering s. 15 more open-ended, so as to extend its ambit beyond that endorsed by this Court in Andrews. If I understand him correctly, he would treat "irrelevance" as an integral aspect of a finding of discrimination. See Gibson, "Equality for Some" (1991), 40 U.N.B.L.J. 2; "Analogous Grounds of Discrimination Under the Canadian Charter: Too Much Ado About Next to Nothing" (1991), 29 Alta. L. Rev. 772; The Law of the Charter: Equality Rights (1990). Although I commend his efforts to ensure that s. 15 not be overly limited in its scope, I cannot agree with the means by which he seeks to achieve such open-endedness. 43 Briefly put, my concern is that "relevance" is a double- edged sword. Although there will certainly be instances in which irrelevant distinctions will trigger s. 15 scrutiny where s. 15 might not otherwise have been engaged, I find it equally easy to envision situations where relevance will be used to deny protections under s. 15 to groups that are otherwise deserving of it. In particular, where a distinction is relevant to the purpose of the legislation, then it is not discriminatory for the purposes of s. 15. In my view, such an approach takes far too narrow a view of discrimination. Relevance can, by definition, only be evaluated as against the purpose of the impugned legislation. Consequently, it fails to take into account the possibility that a distinction that is relevant to the purpose of the legislation may nonetheless still have a discriminatory effect. If s. 15 is about recognizing the equal worth and dignity of each human being, it seems counter- productive to say that this sense of equal worth has not been impugned merely because the legislative distinction is relevant to some legitimate legislative purpose. Who are we, for instance, to say to persons over the age of 65 that, because mandatory retirement is relevant to several important social goals, those persons cannot reasonably feel demeaned as human beings by mandatory retirement legislation? 44 Using relevance to define the absence or presence of discrimination raises other difficulties. It is no good, for instance, for a distinction to be relevant to a legislative purpose if that purpose is, itself, discriminatory. Moreover, including a "relevance" test within a s. 15 determination would place the onus on the rights claimant to characterize properly the purpose of the legislation which it claims is irrelevant to the impugned distinction. This strikes me as undesirable, since between the rights claimant and the government, the government is clearly in the superior position to characterize properly the purpose of its own legislation. Moreover, it is unclear to me what standard of relevance would be appropriate for such a determination. To find a distinction to be discriminatory, it seems quite reasonable, for instance, to require a lesser degree of irrelevance where the consequences of the distinction on the affected individuals are severe than where they are minimal. Finally, a "relevancy" standard would appear to impose an internal limitation on s. 15 that does not arise naturally from its plain language. Given that the decision to impose a limitation at the level of the right itself has consequences in terms of who must bear the onus of overcoming the limitation, this Court has consistently eschewed imposing internal limitations on rights wherever possible: Andrews, supra, at p. 178. 45 In sum, I believe that it is more accurate and more desirable to treat relevance as, in fact, a justification for distinctions that have a discriminatory impact on persons or groups, to be considered under s. 1 of the Charter. I shall elaborate upon this matter below. B. Giving Effect to the Purpose of Section 15 46 I will first discuss why I believe that the current vehicle of choice for fulfilling the purposes of s. 15, the "grounds" approach, is incapable of giving full effect to this purpose. I will then elaborate upon an approach that I believe to be more capable of enabling s. 15 to realize its full potential. (i) The Imperfect Vehicle of "Grounds" 47 In order to realize fully the purpose of Charter rights, it is necessary to look to the effects of impugned legislative actions. In the context of s. 15, no intention to discriminate need be demonstrated in order to render a particular distinction discriminatory. In Andrews, supra, in the course of his discussion on the nature of discrimination, McIntyre J. referred to the conclusions of this Court in the human rights case of Ontario Human Rights Commission v. Simpsons-Sears Ltd., [1985] 2 S.C.R. 536. I believe the following observation, at p. 173 of Andrews, to be at the core of this Court's philosophy regarding the notion of discrimination: ... [n]o intent was required as an element of discrimination, for it is in essence the impact of the discriminatory act or provision upon the person affected which is decisive in considering any complaint. [Emphasis added.] It is here, however, that the emphasis on grounds and, correspondingly, on "analogous grounds", developed in Andrews runs into the greatest difficulty. Rather than focus on "discrimination" directly, the "grounds" approach focuses courts' attentions on the types of grounds which may be a basis for a finding of discrimination. Because a finding that a ground is either enumerated or analogous is a necessary pre- condition for a finding of discrimination, most analysis is devoted toward characterizing the basis for the distinction and, if the basis is not an enumerated ground, deciding whether the ground is "analogous". This approach inquires into whether the characteristics of the ground are sufficient to constitute a basis for discrimination, rather than into the absence or presence of discriminatory effects themselves. 48 We must remember that the grounds in s. 15, enumerated and analogous, are instruments for finding discrimination. They are a means to an end. By focusing almost entirely on the nature, content and context of the disputed ground, however, we have begun to approach it as an end, in and of itself. Such an approach, in effect, approaches s. 15 not by giving primacy to the word "discrimination", but rather by giving primacy to the nine enumerated grounds. In essence, it defines the preconditions to when discrimination will be present exclusively by reference to qualities seen generally to reside in those grounds. 49 It is obvious that this Court could not have adopted an enumerated and analogous grounds approach if, instead of there being nine enumerated grounds in s. 15(1), there had been none. Would the absence of "particularities" in s. 15(1) have changed the basic guarantee to equality without discrimination? In the alternative, what would have happened under the "analogous grounds" approach if, instead of setting out nine enumerated grounds, s. 15(1) had set out only three or four? What if, furthermore, religion was not one of them? Most would agree that the common characteristics of all of the enumerated grounds other than religion is that they involve so-called "immutable" characteristics. Religion, on the other hand, has been described as being premised on a "fundamental choice". Does this mean that s. 15, despite being consciously left open-ended by the drafters, could never have encompassed discrimination on the basis of religion, or any other characteristic which involves a "fundamental choice"? This result seems absurd, yet it seems to flow inevitably from an approach to "discrimination" that relies exclusively on drawing analogies from the essential characteristics of the enumerated grounds. It also demonstrates, in my mind, why reliance on characteristics "analogous" to those in the enumerated grounds is a potentially unsatisfactory means of giving effect to s. 15's open-ended character. 50 Additional problems arise when certain grounds, particularly grounds based upon legal status (marital status, family status, citizenship, province of residence, etc.) may be said to give rise to discriminatory concerns in certain contexts but not in others. Are these grounds therefore sometimes analogous and sometimes not analogous? In these types of circumstances, the finding of "analogousness" will be driven by the result we want to reach. If we want to conclude that the impugned distinction is discriminatory, then we find the grounds to be analogous. If we want to conclude that a distinction is non-discriminatory, then we simply say that although the ground "may be analogous in some contexts", it is not in this case: see, e.g., Turpin, supra, per Wilson J.; R. v. Genereux, [1992] 1 S.C.R. 259, per Lamer C.J.. 51 In addition to defining discrimination indirectly, a "grounds"-based approach suffers an additional shortcoming. Briefly put, "grounds" are, themselves, an imperfect means for discerning discriminatory conduct. This problem was underlined by Wilson J. in McKinney, supra, at pp. 392-93: The grounds enumerated in s. 15 represent some blatant examples of discrimination which society has at last come to recognize as such .... The listing of sex, age and race, for example, is not meant to suggest that any distinction drawn on these grounds is per se discriminatory. Their enumeration is intended rather to assist in the recognition of prejudice when it exists .... It follows, in my opinion, that the mere fact that the distinction drawn in this case has been drawn on the basis of age does not automatically lead to some kind of irrebuttable presumption of prejudice. Rather it compels one to ask the question: is there prejudice? Is the mandatory retirement policy a reflection of the stereotype of old age? Is there an element of human dignity at issue? [Emphasis added.] If a finding of discrimination does not flow automatically from a finding that a distinction has been drawn on the basis of an enumerated or analogous ground (see Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872; R. v. Hess, [1990] 2 S.C.R. 906), then it follows that reliance on "grounds" may not contemplate the entire picture. An additional dimension of analysis is needed. 52 At this juncture, an important question must be asked. If the purpose of s. 15 is really to provide a broad guarantee of protection against discrimination in all of its forms, then why does it matter if the basis for distinction is abstractly "analogous" to the enumerated categories? The answer, I think, is that it does not matter. As this Court has frequently acknowledged, the essence of discrimination is its impact, not its intention. The enumerated or analogous nature of a given ground should not be a necessary precondition to a finding of discrimination. If anything, a finding of discrimination is a precondition to the recognition of an analogous ground. The effect of the "enumerated or analogous grounds" approach may be to narrow the ambit of s. 15, and to encourage too much analysis at the wrong level. 53 We will never address the problem of discrimination completely, or ferret it out in all its forms, if we continue to focus on abstract categories and generalizations rather than on specific effects. By looking at the grounds for the distinction instead of at the impact of the distinction on particular groups, we risk undertaking an analysis that is distanced and desensitized from real people's real experiences. To make matters worse, in defining the appropriate categories upon which findings of discrimination may be based, we risk relying on conventions and stereotypes about individuals within these categories that, themselves, further entrench a discriminatory status quo. More often than not, disadvantage arises from the way in which society treats particular individuals, rather than from any characteristic inherent in those individuals. 54 For all of these reasons, I am led inevitably to the conclusion that a truly purposive approach to s. 15 must place "discrimination" first and foremost in the Court's analysis. This is not to say that the essential characteristics of the nine enumerated grounds are irrelevant to our inquiry. They are, in fact, highly relevant. I turn now to a discussion of their important role in an approach that looks to groups rather than grounds, and discriminatory impact rather than discriminatory potential. (ii) Putting "Discrimination" First 55 In my view, for an individual to make out a violation of their rights under s. 15(1) of the Charter, he or she must demonstrate the following three things: (1) that there is a legislative distinction; (2) that this distinction results in a denial of one of the four equality rights on the basis of the rights claimant's membership in an identifiable group; (3) that this distinction is "discriminatory" within the meaning of s. 15. The following remarks are devoted to elaborating upon the last criterion. 56 A distinction is discriminatory within the meaning of s. 15 where it is capable of either promoting or perpetuating the view that the individual adversely affected by this distinction is less capable, or less worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration. This examination should be undertaken from a subjective-objective perspective: i.e. from the point of view of the reasonable person, dispassionate and fully apprised of the circumstances, possessed of similar attributes to, and under similar circumstances as, the group of which the rights claimant is a member. 57 The means by which courts may give principled expression to this notion is perhaps best illustrated by a simple analogy. If a projectile were thrown against a soft surface, then it would leave a larger scar than if it were thrown against a resilient surface. In fact, the depth of the scar inflicted will generally be a function of both the nature of the affected surface and the nature of the projectile used. In my view, assessing discriminatory impact is, in principle, no different. In order for a court to determine from a subjective- objective perspective whether the impugned distinction will leave a non-trivial discriminatory "scar" on the group affected, it is instructive to consider two categories of factors: (1) the nature of the group adversely affected by the distinction and (2) the nature of the interest adversely affected by the distinction. In my view, neither is completely meaningful without the other. The nature of the group affected 58 No one would dispute that two identical projectiles, thrown at the same speed, may nonetheless leave a different scar on two different types of surfaces. Similarly, groups that are more socially vulnerable will experience the adverse effects of a legislative distinction more vividly than if the same distinction were directed at a group which is not similarly socially vulnerable. As such, a distinction may be discriminatory in its impact upon one group yet not discriminatory in its impact upon another group. While it may be discriminatory against women to prohibit female guards from searching male prisoners, it may not be discriminatory against men to prohibit male guards from searching female prisoners: Weatherall v. Canada (Attorney General), supra. While it may be discriminatory to define a particular criminal offence as only applying to women, it may not be discriminatory to restrict the applicability of the offence of sexual assault of a minor to men: R. v. Hess, supra. In the same way that it does not really matter why the affected surface is soft, it is not necessary that there be a formal nexus between the social vulnerability of the affected group and the prejudice flowing from the impugned distinction in order for that vulnerability to be relevant to determining whether the distinction is discriminatory. Put another way, it is merely admitting reality to acknowledge that members of advantaged groups are generally less sensitive to, and less likely to experience, discrimination than members of disadvantaged, socially vulnerable or marginalized groups. See, by analogy, Schachtschneider v. Canada, [1994] 1 F.C. 40 (C.A.), per Linden J.A.. 59 Most of the factors identified in Andrews under the "analogous grounds" approach as characteristic of the enumerated grounds in s. 15 are, not surprisingly, integral to evaluating the nature of the group affected by the impugned distinction. It is highly relevant, for instance, to inquire into whether the impugned distinction is based upon fundamental attributes, such as those enumerated in s. 15, that are generally considered to be essential to our popular conception of 'personhood' or 'humanness'. Furthermore, it is important to ask ourselves questions such as "Is the adversely affected group already a victim of historical disadvantage?"; "Is this distinction reasonably capable of aggravating or perpetuating that disadvantage?"; "Are group members currently socially vulnerable to stereotyping, social prejudice and/or marginalization?"; and "Does this distinction expose them to the reasonable possibility of future social vulnerability to stereotyping, social prejudice and/or marginalization?". Membership in a "discrete and insular minority", lacking in political power and thus susceptible to having its interests overlooked, is yet another consideration that may be taken into account. 60 Consideration of these factors involves the recognition that differently situated groups are starting on different levels of the s. 15 playing field. In my view, our approach to s. 15 must reflect that reality. Indeed, I reiterate McIntyre J.'s words in Andrews, supra, at p. 169 that "for the accommodation of differences, which is the essence of true equality, it will frequently be necessary to make distinctions". Treating historically vulnerable, disadvantaged or marginalized groups in the same manner as groups which do not generally suffer from such vulnerability may not accommodate, or even contemplate, those differences. In fact, ignoring such differences may compound them, by making access to s. 15 relief most difficult for those groups that are the most disempowered of all in Canadian society. 61 To summarize, the more socially vulnerable the affected group and the more fundamental to our popular conception of "personhood" the characteristic which forms the basis for the distinction, the more likely that this distinction will be discriminatory. Of course, these factors only provide half of the picture. We do not enjoy the full spectrum of colours and contexts until we have also turned our attention to a second set of considerations. The nature of the affected interest 62 In the same way that a very dense projectile will impact upon a surface more sharply than a less dense projectile, an examination of the nature of the interest affected by the impugned distinction is helpful in determining whether that distinction is discriminatory. This examination requires an evaluation of both economic and non-economic elements. 63 As I noted earlier, the Charter is not a document of economic rights and freedoms. Rather, it only protects "economic rights" when such protection is necessarily incidental to protection of the worth and dignity of the human person (i.e. necessary to the protection of a "human right"). Nonetheless, the nature, quantum and context of an economic prejudice or denial of such a benefit are important factors in determining whether the distinction from which the differing economic consequences flow is one which is discriminatory. If all other things are equal, the more severe and localized the economic consequences on the affected group, the more likely that the distinction responsible for these consequences is discriminatory within the meaning of s. 15 of the Charter. 64 Although a search for economic prejudice may be a convenient means to begin a s. 15 inquiry, a conscientious inquiry must not stop here. The discriminatory calibre of a particular distinction cannot be fully appreciated without also evaluating the constitutional and societal significance of the interest(s) adversely affected. Other important considerations involve determining whether the distinction somehow restricts access to a fundamental social institution, or affects a basic aspect of full membership in Canadian society (e.g.. voting, mobility). Finally, does the distinction constitute a complete non- recognition of a particular group? It stands to reason that a group's interests will be more adversely affected in cases involving complete exclusion or non-recognition than in cases where the legislative distinction does recognize or accommodate the group, but does so in a manner that is simply more restrictive than some would like. 65 Referring back to our analogy once again, if the projectile is dense enough and thrown hard enough, then it will leave a mark on even the most resilient of surfaces. Similarly, the more fundamental the interest affected or the more serious the consequences of the distinction, the more likely that the impugned distinction will have a discriminatory impact even with respect to groups that occupy a position of advantage in society. 66 To summarize, tangible economic consequences are but one manifestation of the more intangible and invidious harms flowing from discrimination, which the Charter seeks to root out. In other cases, the prejudice will be to an important individual interest rather than to one that is economic in nature. The nature of the interest affected is therefore highly relevant to whether the distinction that adversely affects that interest is discriminatory in nature. In all but the most extreme cases, this factor cannot be considered in isolation. It only assumes meaningful proportions when assessed in light of the nature of the group affected. Frameworks vs. rigid legal tests 67 It must be emphasized that there are no absolute preconditions to, or preclusions from, a finding of discrimination. Although the presence of one or more of the aforementioned factors in either of these two categories may tend toward the conclusion that the impugned distinction is discriminatory, it does not inevitably lead that way. Conversely, the absence of one or more of these factors does not necessarily preclude there still being a finding of discrimination. Courts must treat these considerations as a matrix rather than as a single equation, and as the microscope rather than as the object being studied. 68 Equality and discrimination are notions that are as varied in form as they are complex in substance. Attempts to evaluate them according to legal formulas which incorporate rigid inclusionary and exclusionary criteria are doomed to become increasingly complex and convoluted over time as "hard" cases become the rule rather than the exception. I prefer to steer clear of those rocky shoals, if at all possible, and to adopt a pragmatic and functional approach to s. 15. I believe that an analysis that examines both sets of factors in the basic framework set out above will enable courts to arrive in a principled manner at an answer that reflects as closely as possible the experience of those in the affected group. If, after examining the nature of both the group and the interest affected, a court concludes that the impact of the impugned distinction is capable of inflicting a non-trivial discriminatory "scar" on the affected group, then it must conclude that this distinction is discriminatory. 69 Once it is found that s. 15(1) of the Charter has been violated by a distinction which is discriminatory, examination must pass on to s. 1. C. Section 1 of the Charter 70 Section 15 of the Charter only guarantees freedom from discrimination subject to reasonable limits, demonstrably justifiable in a free and democratic society. In my view, there is no possible justification for a discriminatory distinction other than that it is relevant to an important objective. As such, a distinction found to violate s. 15(1) may only be saved under s. 1 if it is found to be relevant to a proportionate extent to a pressing and substantial objective. This is accomplished by reference to the framework to s. 1 analysis set down in R. v. Oakes, [1986] 1 S.C.R. 103 and modified by the majority of this Court in Dagenais v. Canadian Broadcasting Corp.,[1994] 3 S.C.R. 835. (i) Pressing and substantial objective 71 In Oakes, it was held that the objective of the legislation, which the measures responsible for a limit on a Charter right or freedom are designed to serve, must be of sufficient importance to warrant overriding a constitutionally protected right or freedom. It therefore must relate to concerns which are pressing and substantial in a free and democratic society. Moreover, in Big M, supra, this Court held that a purpose may never, itself, be unconstitutional. By that same token, where the purpose of impugned legislation is, itself, discriminatory, it cannot be saved under s. 1. I would add, however, that where the court has available to it several possible, and equally likely, interpretations of the purpose of the legislation, then it should prefer one that is consistent with Charter values over one that is not. (ii) Relevant to a proportionate extent 72 Where the purpose of the legislation is found to be pressing and substantial and non-discriminatory, it then becomes necessary to decide whether there is a sufficient degree of proportionality between the impugned distinction (i.e. the means to achieve the purpose) and the rights violation. 73 As I noted earlier, an important concern with including relevance within the ambit of s. 15 analysis is that a court would have little guidance on how relevant/irrelevant a particular distinction need be to the purpose of the legislation in order for that distinction to be discriminatory. Professor Bayefsky, in "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 Supreme Court L.R. (2d) 503, at p. 528, favours a sliding scale of scrutiny which recognizes that all trade- offs between furtherance of a pressing and substantial objective and a rights violation occur along a continuum rather than at a single point. To date, this Court has developed no better approach for evaluating whether this equilibrium has been appropriately respected than through the proportionality test in s. 1. Therefore, for a distinction that is discriminatory to be justified under s. 1, it must be shown to be relevant to a proportionate extent to the purpose of the legislation. This determination flows from the traditional three-part proportionality test set out in Oakes and modified in Dagenais. 74 First, the distinction must be rationally connected to the pressing and substantial objective of the legislation. This standard sets down a basic relevancy requirement. Where a distinction is, essentially, irrelevant to the purpose which the legislation seeks to advance, then the distinction cannot be saved under s. 1. Discrimination on the basis of the legal status of the group affected (e.g. citizenship, province of residence, marital status) may raise special problems in this respect, since legal status always comes attached with specific rights and obligations. Because of the various rights and obligations which differing status-based groups may enjoy, part of the rational connection determination in such instances may require some inquiry into whether the distinction drawn in the impugned legislation is relevant to one or more of those rights and/or obligations. If the distinction does not relate rationally to either a right or an obligation which attaches to the affected status-based group, then I do not see how a distinction drawn on the basis of membership in that status-based group would not be irrelevant. 75 The next stage of the proportionality analysis requires that the government demonstrate that the legislation impairs as little as reasonably possible the claimant's s. 15 rights. Depending on the circumstances, the difficulty or impossibility of finding a workable alternative basis of distinction may be a valid consideration under this branch of the proportionality test. So, too, may be the fact that the government has had to make a reasonable trade-off in a field where conflicting social science views, or conflicting rights and interests between groups, are at stake. The mere fact, however, that legislation is "social" in nature is not, by itself, a reason for increased deference. In fact, to defer to the legislative prerogative in circumstances where social science views do not substantially conflict, and where there is a reasonable, alternative means of fulfilling the legislative objective in a way that would materially lessen the magnitude of the rights violation, would frustrate the purpose of the Charter. 76 Finally, there must be a proportionality between the discriminatory effects of the impugned distinction and the salutary effects of the distinction: Dagenais v. Canadian Broadcasting Corp., supra. Factors such as the importance of the state interest, the extent to which it is furthered by the impugned distinction, the constitutional and societal significance of the interests adversely affected, the severity of the rights deprivation suffered by the individual, and the potential for entrenching marginalization or stigmatization of particular groups will all be relevant considerations to this branch of the s. 1 examination. The government must shoulder a heavier justificatory burden when the Charter infringement is severe: Reference re ss. 193 and 195.1(1)(c) of the Criminal Code (Man.), [1990] 1 S.C.R. 1123, at p. 1190. 77 It should be noted, finally, that neither s. 1 nor s. 15 calls for a balance sheet approach to discrimination (i.e. summing up all direct and incidental economic benefits to a particular distinction and comparing them against the sum of the economic prejudices, in order to see if there is a net economic prejudice). Such an approach to discrimination loses the forest for the trees. In the context of s. 15, it must be recalled that actual economic prejudice is only one (albeit sometimes significant) ingredient in the more fundamental determination of whether a particular legislative distinction is discriminatory. Similarly, when conducting a s. 1 analysis, it must be recalled that the rights deprivation that triggered s. 15 is not economic. It is something which is fundamental to the person. As such, the mere fact that other economic benefits may flow from the same distinction does not necessarily render that rights-violation any less severe, any less impairing, or any less worthy of condemnation. D. Implications and "Adverse Effects" 78 Adopting an effects-based approach to s. 15 that looks to groups rather than grounds recognizes the importance of adverse effects discrimination in s. 15 without requiring us to resolve some of the intractable issues that have sprung up around that doctrine. A good example is that raised in Bayefsky, supra, at pp. 518-19: Restricting the possible grounds of discrimination puts litigants in the position of first having to prove that their distinguishing feature is caught. This marks the introduction of the same sort of counter-productive, formalistic and artificial debates that have been conducted under anti-discrimination legislation on such issues as the following: does "sex" include differentiation on the basis of "pregnancy" or does "sex" include distinctions made on the basis of sexual orientation? The Court, for example, will have to ask whether legislation involving differential treatment of domestic workers differentiates on the basis of sex because most domestic workers are women. In order to apply section 15 they will be required to find that a differential impact on women should be classed as sex discrimination, rather than facing squarely the issue of differential treatment of domestic workers. And if legislation discriminating against domestic workers can be caught by proving such differential impact upon an enumerated ground, what about similar legislation drawing distinctions that disadvantage farm workers? If one cannot prove that such legislation has a differential impact upon an enumerated or analogous ground, are farm workers excluded from making a section 15 argument, while domestic workers are covered? What logic can there be to this approach to interpreting section 15? [Emphasis added.] See also N. Iyer, "Categorical Denials: Equality Rights and the Shaping of Social Identity" (1993), 19 Queen's L.J. 179. 79 To expand briefly upon the example of domestic workers, under traditional adverse effects doctrine, what percentage of the group would have to have been women in order to succeed in a sex-based discrimination claim? Fifty percent? Ninety percent? As this Court found in Symes v. Canada, [1993] 4 S.C.R. 695, it is difficult to draw a principled distinction along such lines. I believe that it is both easier and more intellectually honest to examine the effect of the distinction on the group affected. In this case, that group would be domestic workers, and the only decision is: does the distinction discriminate against domestic workers? 80 As I noted in Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554 at p. 645, categories of discrimination cannot be reduced to watertight compartments, but rather will often overlap in significant measure. When assessing the social context of the impugned distinction, it is therefore of relevance that a significant majority of domestic workers are immigrant women, a subgroup that has historically been both exploited and marginalized in our society. Awareness of, and sensitivity to, the realities of those experiencing the distinction is an important task that judges must undertake when evaluating the impact of the distinction on members of the affected group. Discrimination cannot be fully appreciated or addressed unless courts' analysis focuses directly on the issue of whether these workers are victims of discrimination, rather than becoming distracted by ancillary issues such as "grounds", be they enumerated or analogous. 81 In conclusion, the approach to s. 15 set out in Andrews was an extremely good start. The ensuing jurisprudence has contributed in diverse and important respects to giving content to s. 15 of the Charter. As cases become more and more difficult, however, I believe that it is becoming increasingly evident that we may have been putting the cart before the horse. Although s. 15 is a general guarantee of "equality without discrimination", we have failed to put "discrimination", itself, at the forefront of our analysis. Instead, we have begun to define ourselves into boxes by making "grounds" a precondition to discrimination. As such, we may be denying s. 15 relief to persons who are victims of legislatively sanctioned discrimination, but who are unable to fit themselves into an established or analogous "ground". 82 In this day and age, discrimination is rarely an explicit purpose, and all too frequently an effect. The shortcomings of relying upon "grounds of distinction" are becoming increasingly evident -- we are coming to realize more and more that some "grounds" may give rise to discrimination in some contexts and not in others. In reality, it is no longer the "grounds" that are dispositive of the question of whether discrimination exists, but the social context of the distinction that matters. Given the growing reliance on "context", and the declining role of the "grounds of distinction", I believe that it is a natural next step in s. 15 jurisprudence to admit that, in fact, context is of primary importance and that abstract "grounds of distinction" are simply an indirect method to achieve a goal which could be achieved more simply and truthfully by asking the direct question: "Does this distinction discriminate against this group of people?" See Wilson J. in McKinney, supra, at pp. 392-93. There is no need to use the vehicle of "analogous grounds" to make s. 15 open- ended. An effects-based approach to the notion of "discrimination" will, itself, be open-ended. At the same time, as long as we view "discrimination" through a principled framework of analysis, the open-endedness will be constrained along reasonable and principled lines. 83 This approach bears far greater similarity than difference to jurisprudence inspired by Andrews. It will not, for instance, apply very differently to the vast number of distinctions drawn according to place of residence, or between professionals, employees, or different types of occupations. It is not enough to demonstrate that one has been treated differently by the legislation, and that this differential treatment has been less advantageous than that accorded to other parties. The distinction must, itself, be shown somehow to impugn a basic right to equal human dignity and worth. This determination, as I discussed above, will largely be a function of the nature of both the group and the interest adversely affected by the impugned distinction. As a practical matter, all other things being equal, the greater the degree of unfettered choice involved in membership in the group that is adversely affected, the more difficult it will be to demonstrate that differential treatment undermines basic human rights. As well, in many such situations it will be fair to say that the nature of the interest being adversely affected by the distinction, as well as the constitutional and societal significance of that interest, will generally be comparatively low. 84 It is hoped that this approach offers to individuals and groups that may otherwise be denied the protections of s. 15 the opportunity to claim the fullest protections of our constitution. In my view, s. 15 will never realize its full potential until we adopt an approach that is truly effects-sensitive. The courts of this land must not rest until all individuals are able to enjoy the fundamental right to "equality without discrimination". II. Application to the Facts 85 Our first task is to determine whether there is a legislative distinction between the rights claimants, and Nesbit, and others, and whether this distinction results in the denial of one of the claimants' four basic equality rights (equality before the law, equality under the law, equal protection of the law, and equal benefit of the law) on the basis of their membership in an identifiable group. 86 For the reasons he sets out, I agree with Cory J. that it is clear that homosexual couples are denied the equal benefit of the law on the basis of the legislative distinction in s. 2 of the Old Age Security Act, which defines couples as relationships of "opposite sex". That Egan and Nesbit are able to claim higher benefits as separate individuals does not alter the fact that they have been denied the benefits, both tangible and intangible, of filing for old age benefits as a couple. It would take too narrow a view of the phrase "benefit of the law" to define it strictly in terms of economic interests. Official state recognition of the legitimacy and acceptance in society of a particular type of status or relationship may be of greater value and importance to those affected than any pecuniary gain flowing from that recognition. 87 Furthermore, I share my colleague's reasoning and analysis to the effect that the impugned distinction in s. 2 of the Old Age Security Act is made on the basis of sexual orientation. It would be inconsistent with an effects-oriented approach to the Charter to require as a precondition for protection on this basis that sexual orientation be directly identified in the impugned legislation. Rather, it is sufficient that the couple's sexual orientation cannot in any meaningful way be separated from the "opposite sex" requirement in the Act. This reasoning translates equally well into the framework that I have developed above. In short, the impugned distinction excludes the rights claimants because they are homosexual. To the extent that I frame the problem somewhat more narrowly than Cory J., it is because I prefer to analyze the problem from the point of view of the group actually affected by the distinction, rather than in light of the somewhat illusory neutrality afforded by speaking of the "ground" of "sexual orientation". 88 The last step within s. 15 is to ask ourselves if the distinction is one which is capable of either promoting or perpetuating a view that the appellants and Nesbit are, by virtue of their homosexuality, less capable or less worthy of recognition or value as human beings or as members of Canadian society, equally deserving of concern, respect, and consideration. This evaluation is accomplished by reference to both the nature of the group affected and the nature of the interest affected. These two factors constitute, in short, the all-important socio- economic context of the impugned distinction. 89 In my view, consideration of these two factors leads inevitably to the conclusion that the impugned distinction is discriminatory. My colleague Cory J. cites extensive evidence that same-sex couples are a highly socially vulnerable group, in that they have suffered considerable historical disadvantage, stereotyping, marginalization and stigmatization within Canadian society. The distinction, moreover, is on the basis of an aspect of 'personhood' that is quite possibly biologically based and that is at the very least a fundamental choice. Finally, I cannot help but note that, in addition to being homosexuals, those individuals directly affected by the distinction are all, by definition, also elderly and poor. They are therefore at the margins of an already marginalized group within society. 90 I turn next to the nature of the interest adversely affected. The impugned legislation, the Old Age Security Act, is a cornerstone in Canada's social security net, which is, in turn, a cherished and fundamental institution in our society. It involves the guarantee of a minimum level of income for elderly couples in conjugal relationships. The interest affected is therefore quite a fundamental one, from both a societal and, quite possibly, a constitutional perspective. Although the claimants cannot be said to suffer any economic prejudice from the distinction since they are each entitled as individuals to a certain minimum income level, it cannot be overlooked that the rights claimants have been directly and completely excluded, as a couple, from any entitlement to a basic shared standard of living for elderly persons cohabiting in a relationship analogous to marriage. This interest is an important facet of full and equal membership in Canadian society. Given the marginalized position of homosexuals in society, the metamessage that flows almost inevitably from excluding same-sex couples from such an important social institution is essentially that society considers such relationships to be less worthy of respect, concern and consideration than relationships involving members of the opposite sex. This fundamental interest is therefore severely and palpably affected by the impugned distinction. 91 Blending my analysis of the nature of the interest together with my conclusions regarding the nature of the group affected by the impugned distinction, I am convinced that this distinction is reasonably capable of exacting a discriminatory toll upon the group affected. I would therefore find the impugned distinction to be discriminatory within the meaning of s. 15(1) of the Charter. 92 A violation of s. 15(1) of the Charter may be saved under s. 1 if the impugned distinction is demonstrably justifiable in our free and democratic society. I am of the opinion, however, that the instant violation cannot be salvaged by s. 1, as it is not relevant to a proportionate extent to a pressing and substantial objective. 93 Like Iacobucci J., I would characterize the purpose of the legislation, which s. 2 of the Act is intended to further, in the way it was framed by Linden J.A. (at p. 446): In general terms ... 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. I believe that this objective is pressing and substantial and that it is not, itself, animated by a discriminatory purpose. I find, however, that the impugned distinction (i.e. the impugned means to achieve this objective) fails all three branches of the proportionality test. 94 With respect to the first branch of the proportionality test set out in Oakes, I have a great deal of difficulty even concluding that the exclusion of same-sex couples is rationally connected to the objective cited above. The legislation excludes couples who would fill all of the other criteria in the Act except the requirement that they are of the opposite sex. To find that this distinction is rationally connected to the objective of the legislation requires us to conclude that same-sex couples are so different from married couples that it would be unreasonable to make the same benefits available to both. At best, the government has only demonstrated that this is its assumption. I am in entire agreement with Iacobucci J.'s observation that the presumption that same-sex relationships are somehow less interdependent then opposite-sex relationships is, itself, a fruit of stereotype rather than one of demonstrable, empirical reality. See Mossop, supra, at pp. 630-31, per L'Heureux- Dube J. It would be strange, indeed, to permit the government to justify a discriminatory distinction on the basis of presumptions which are, themselves, discriminatory. This would defeat the very purpose which s. 15(1) is intended to further. 95 In this context, I would also reject any argument that homosexual relationships have a distinct biological reality -- namely that homosexuality is non-procreative - - as dangerously reminiscent of the type of biologically based arguments that this Court has now firmly rejected. Compare Bliss v. Attorney General of Canada, [1979] 1 S.C.R. 183 at p. 190, with Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219, at pp. 1243-44. Moreover, as Iacobucci J. points out, it is evident that the absence or presence of children has nothing whatsoever to do with eligibility for the old age spousal supplement under the Act. I therefore find that the impugned distinction fails the rational connection branch of the proportionality test. Though it is not necessary to address the remaining branches of the Oakes test, I shall do so briefly to demonstrate that I would reach the same result in any event. 96 I agree with Iacobucci J.'s application of the minimal impairment branch of the test. To this, I would add that it cannot be argued that the fact that greater economic benefits may be available to Nesbit, as a single person, under the provincial Guaranteed Available Income for Need Act, R.S.B.C. 1979, c. 158, is evidence that the infringement of s. 15 is minimally impairing. As I have already noted elsewhere, a balance sheet approach to economic benefits and prejudices trivializes and, in many ways, misconstrues the nature of the human rights violation at issue. 97 The Attorney General of Canada argues that the impugned legislation is social, not punitive, and should thereby enjoy a somewhat more deferential standard of review by this Court. As I have already noted, I do not believe that such deference is appropriate when there is a reasonable alternative, readily available, that is not the subject of conflicting social science views, that would materially lessen the effect of the rights violation to the affected group, and that would not result in a concomitant prejudice to another group. To accord deference merely because the issue is a "social" one would be to issue a licence to discriminate in favour of the status quo. In the present case, I believe that a reasonable alternative remedy is, indeed, available. The discriminatory effect would, in fact, be eliminated without prejudice to the rights or interests of any other group, by extending coverage to same-sex couples who otherwise fulfil all of the other non-discriminatory criteria required in the Act. 98 Finally, the third branch of the test requires that there be a proportionality between the discriminatory and salutary effects of the distinction. At this stage, it is appropriate to recall that the Charter breach (i.e. the discriminatory effect) is quite severe. The discrimination against homosexuals arises on the face of the legislation and flows inevitably from the "opposite sex" requirement in s. 2 of the Act. In addition, the impugned distinction is in an Act that plays an important role in a very important Canadian social institution. The interest at issue is a fundamental one -- the right to a basic level of income for the elderly -- and the non- recognition is complete, rather than partial. Although the claimants are not necessarily economically worse off as a result of their exclusion from the Old Age Security Act, the complete exclusion from the program of same-sex couples has a significant discriminatory impact in terms of perpetuating prejudice, stereotyping, and marginalization of same-sex couples, and homosexuals and lesbians individually. 99 It can be argued that the primary salutary effect of the distinction, on the other hand, is the savings it ostensibly entails to the public purse. The government's expert estimates this saving as ranging between $12 million and $37 million. The appellants' cross- examination at trial of that expert suggests that this figure may be considerably less. I would nonetheless make three observations in relation to this argument. First, by the government's own account, these sums account for only between two and four percent of the total cost of the old age supplement program. Second, I have referred to these savings as "ostensible" because if the affected persons had been in heterosexual relationships instead of homosexual relationships, the government would have to have paid out this money anyway. Finally, I note that the majority of this Court recognized in Schachter v. Canada, [1992] 2 S.C.R. 679, at p. 709, that budgetary considerations should not be determinative of a s. 1 analysis, and should more properly be considered when attempting to formulate an appropriate remedy. On this basis, I conclude that the deleterious effects of the impugned distinction outweigh its salutary effects. 100 It goes without saying that I cannot agree with the novel approach to s. 1 taken by Sopinka J. in this case, particularly in light of the following remarks by Dickson C.J. in R. v. Oakes, [1986] 1 S.C.R. 103 at p. 136: A second contextual element of interpretation of s. 1 is provided by the words "free and democratic society". Inclusion of these words as the final standard of justification for limits on rights and freedoms refers the Court to the very purpose for which the Charter was originally entrenched in the Constitution: Canadian society is to be free and democratic. The Court must be guided by the values and principles essential to a free and democratic society which I believe embody, to name but a few, respect for the inherent dignity of the human person, commitment to social justice and equality, accommodation of a wide variety of beliefs, respect for cultural and group identity, and faith in social and political institutions which enhance the participation of individuals and groups in society. The underlying values and principles of a free and democratic society are the genesis of the rights and freedoms guaranteed by the Charter and the ultimate standard against which a limit on a right or freedom must be shown, despite its effect, to be reasonable and demonstrably justified. [Emphasis added.] There is a first time to every discrimination claim. To permit the novelty of the appellants' claim to be a basis for justifying discrimination in a free and democratic society undermines the very values which our Charter, including s. 1, seeks to preserve. 101 I therefore conclude that the impugned distinction violates s. 15 of the Charter and cannot be saved by s. 1. III. Disposition 102 I would allow the appeal with costs throughout, and adopt the remedy proposed by Iacobucci J. SOPINKA J.: 103 I have read the reasons of my colleagues, Justices La Forest, L'Heureux-Dube and Cory and Iacobucci, and while I agree with Cory J. that the impugned legislation infringes s. 15(1) of the Canadian Charter of Rights and Freedoms, in my opinion such infringement is saved under s. 1. 104 I agree with the respondent the Attorney General of Canada that government must be accorded some flexibility in extending social benefits and does not have to be pro- active in recognizing new social relationships. It is not realistic for the Court to assume that there are unlimited funds to address the needs of all. A judicial approach on this basis would tend to make a government reluctant to create any new social benefit schemes because their limits would depend on an accurate prediction of the outcome of court proceedings under s. 15(1) of the Charter. The problem is identified by Professor Hogg in Constitutional Law of Canada (3rd ed. 1992), at pp. 911-12, where he states: It seems likely that virtually any benefit programme could be held to be under-inclusive in some respect. The effect of Schachter [[1990] 2 F.C. 129 (C.A.)] and Tetreault-Gadoury [[1991] 2 S.C.R. 22] is to subject benefit programmes to unpredictable potential liabilities. These decisions by-pass the normal processes by which a government sets its priorities and obtains parliamentary approval of its estimates. 105 This Court has recognized that it is legitimate for the government to make choices between disadvantaged groups and that it must be provided with some leeway to do so. In McKinney v. University of Guelph, [1990] 3 S.C.R. 229, La Forest J., in the course of his reasons which found that s. 9(a) of the Ontario Human Rights Code, 1981 violated s. 15(1) but was justified under s. 1, stated at pp. 317-18: In looking at this type of issue, it is important to remember that a Legislature should not be obliged to deal with all aspects of a problem at once. It must surely be permitted to take incremental measures. It must be given reasonable leeway to deal with problems one step at a time, to balance possible inequalities under the law against other inequalities resulting from the adoption of a course of action, and to take account of the difficulties, whether social, economic or budgetary, that would arise if it attempted to deal with social and economic problems in their entirety, assuming such problems can ever be perceived in their entirety. This Court has had occasion to advert to possibilities of this kind. In R. v. Edwards Books and Art Ltd., [[1986] 2 S.C.R. 713], Dickson C.J., there dealing with the regulation of business and industry, had this to say, at p. 772: I might add that in regulating industry or business it is open to the legislature to restrict its legislative reforms to sectors in which there appear to be particularly urgent concerns or to constituencies that seem especially needy. In this context, I agree with the opinion expressed by the United States Supreme Court in Williamson v. Lee Optical of Oklahoma, 348 U.S. 483 (1955), at p. 489: Evils in the same field may be of different dimensions and proportions, requiring different remedies. Or so the legislature may think.... Or the reform may take one step at a time, addressing itself to the phase of the problem which seems most acute to the legislative mind.... The legislature may select one phase if one field and apply a remedy there, neglecting the others. The question becomes whether the cut-off point can be reasonably supported. At pages 318-19 he added: But generally, the courts should not lightly use the Charter to second-guess legislative judgment as to just how quickly it should proceed in moving forward towards the ideal of equality. The courts should adopt a stance that encourages legislative advances in the protection of human rights. Some of the steps adopted may well fall short of perfection, but as earlier mentioned, the recognition of human rights emerges slowly out of the human condition, and short or incremental steps may at times be a harbinger of a developing right, a further step in the long journey towards full and ungrudging recognition of the dignity of the human person. 106 It is in this context that the s. 1 criteria must be considered. When s. 2 of the Old Age Security Act, R.S.C., 1985, c. O-9, is measured against overall objectives of alleviation of poverty of elderly spouses, it should not be judged on the basis that Parliament has made this choice for all time. The history of the legislation shows an evolving expansion of the definition of the intended recipients of the benefits. Successive amendments expanded the group entitled to these benefits on the basis of current information as to those in the greatest need of financial assistance resulting from the retirement or death of the "breadwinner". As La Forest J. points out, prior to 1975 the interpretation of "spouse" would not have extended to common law spouses. By virtue of S.C. 1974-75-76, c. 58, s. 1, a definition of "spouse" was added to include common law spouses. In 1979 provisions were added for certain widowed spouses, and in 1985 these benefits were extended to include all widows and widowers aged 60 to 65 who had not remarried. During the second reading of the 1979 amendment to the Old Age Security Act, the Honourable Flora MacDonald stated: 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 Debates, Vol. I, 1st Sess., 31st Parl., October 22, 1979, at p. 476.) 107 With respect to the 1985 amendment, the trial judge found as follows: The government at the time, in 1985, recognized that the measures introduced did not solve all of the problems of all citizens but, to the Minister of National Health and Welfare, the Honourable Jake Epp, the legislation was addressing itself to those in greatest need. [Emphasis added.] ([1992] 1 F.C. 687, at p. 693.) 108 The Attorney General of Canada has taken the position in his factum that "the means chosen does not have to be necessarily the solution for all time. Rather, there may always be a possibility that more acceptable arrangements can be worked out over time". Viewed in this light, the impugned legislation can be regarded as a substantial step in an incremental approach to include all those who are shown to be in serious need of financial assistance due to the retirement or death of a supporting spouse. It is therefore rationally connected to the objective. 109 With respect to minimal impairment, the legislation in question represents the kind of socio-economic question in respect of which the government is required to mediate between competing groups rather than being the protagonist of an individual. In these circumstances, the Court will be more reluctant to second-guess the choice which Parliament has made. See Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927. I would conclude, as La Forest J. did in McKinney, that I am "not prepared to say that the course adopted by the Legislature, in the social and historical context through which we are now passing, is not one that reasonably balances competing social demands which our society must address" (p. 314). 110 It follows that there is proportionality between the effects of the legislation on the protected right and the legislative objective. The proper balance was struck by Parliament in providing financial assistance to those who were shown to be in the greatest need of financial assistance. 111 It may be suggested that the time has expired for the government to proceed to extend the benefits to same-sex couples and that it cannot justify a delay since 1975 to include same-sex couples. While there is some force in this suggestion, it is necessary to keep in mind that only in recent years have lower courts recognized sexual orientation as an analogous ground, and this Court will have done so for the first time in this case. While it is true, as Cory J. observes, that many provincial legislatures have amended human rights legislation to prohibit discrimination on the basis of sexual orientation, these amendments are of recent origin. Moreover, human rights legislation operates in the field of employment, housing, use of public facilities and the like. This can hardly be equated with the problems faced by the federal government which must assess the impact of extending the benefits contained in some fifty federal statutes. Given the fact that equating same-sex couples with heterosexual spouses, either married or common law, is still generally regarded as a novel concept, I am not prepared to say that by its inaction to date the government has disentitled itself to rely on s. 1 of the Charter. 112 In the result, I would dismiss the appeal. I would answer the constitutional questions as follows: Question 1. Does the definition of "spouse" in s. 2 of the Old Age Security Act, R.S.C., 1985, c. O- 9, infringe or deny s. 15(1) of the Canadian Charter of Rights and Freedoms? Answer: Yes. Question 2. If the answer to question 1 is "yes", is the infringement or denial demonstrably justified in a free and democratic society pursuant to s. 1 of the Canadian Charter of Rights and Freedoms? Answer: Yes. CORY and IACOBUCCI JJ.: 113 The issue raised by this appeal is whether the definition of "spouse" in s. 2 of the Old Age Security Act, R.S.C., 1985, c. 0-9, violates s. 15(1) of the Canadian Charter of Rights and Freedoms. In these joint reasons Cory J. has dealt with the issues pertaining to the breach of s. 15(1) of the Charter while Iacobucci J. has considered the applicability of s. 1 of the Charter and the appropriate remedy. CORY J.: 114 The appellants James Egan and John Norris Nesbit are a homosexual couple. They have lived together since 1948 in what is obviously an intimate, caring, mutually supportive relationship. They have shared and continue to share bank accounts, credit cards and property ownership. By their wills they have appointed each other their respective executors and beneficiaries. To their families and friends they refer to themselves as partners. 115 On October 1, 1986, some 38 years after the relationship began, Egan, having reached age 65, became eligible to receive old age security and a guaranteed income supplement, pursuant to the provisions of the Old Age Security Act, R.S.C. 1970, c. O-6 (later R.S.C., 1985, c. 0-9). The same Act provides for a spousal allowance to be paid to the spouse of a pensioner when that spouse is between 60 and 65 years of age and the couple's combined income falls below a fixed level. Upon reaching age 60, Nesbit applied for a spousal allowance describing Egan as his spouse. The application was rejected by the Department of National Health and Welfare solely on the basis that the relationship between Nesbit and Egan was of a homosexual nature and thus did not meet the definition of spouse set out in s. 2 of the Act. 116 The appellants brought an action in the Federal Court seeking a declaration that the definition of "spouse" in s. 2 of the Old Age Security Act contravenes s. 15(1) of the Charter on the grounds that it discriminates on the basis of sexual orientation. They also sought a declaration that the definition of "spouse" should be extended to include "partners in same sex relationships otherwise akin to conjugal relationships". On December 2, 1991, the Trial Division dismissed the appellants' claim: [1992] 1 F.C. 687, 87 D.L.R. (4th) 320, 38 R.F.L. (3d) 294, 47 F.T.R. 305. On April 29, 1993, a majority of the Federal Court of Appeal dismissed the appeal: [1993] 3 F.C. 401, 103 D.L.R. (4th) 336, 15 C.R.R. (2d) 310, 153 N.R. 161. Linden J.A., dissenting, found that s. 2 of the Act did infringe s. 15(1) of the Charter and was not justified under s. 1. Relevant Statutory and Constitutional Provisions Old Age Security Act, R.S.C. 1985, c. O-9 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; "spouse's allowance" means the spouse's allowance authorized to be paid under Part III; 19. (1) Subject to this Act and the regulations, for each month in any fiscal year, a spouse's allowance may be paid to the spouse of a pensioner if the spouse (a) is not separated from the pensioner; (b) has attained sixty years of age but has not attained sixty-five years of age; and (c) has resided in Canada after attaining eighteen years of age and prior to the day on which the spouse's application is approved for an aggregate period of at least ten years and, where that aggregate period is less than twenty years, was resident in Canada on the day preceding the day on which the spouse's application is approved. Canadian Charter of Rights and Freedoms 1. The Canadian Charter of Rights and Freedoms guarantees the rights and freedoms set out in it subject only to such reasonable limits prescribed by law as can be demonstrably justified in a free and democratic society. 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. Constitution Act, 1982 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. Judgments Below (a) Federal Court, Trial Division, [1992] 1 F.C. 687 117 The trial court judge determined that the spousal allowance had been directed "to alleviating the financial plight of elderly married couples, primarily women who were younger than their spouses and who generally did not enter the work force" (p. 693). He found as a fact (at p. 695) 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. 118 The trial court judge decided that the definition of "spouse" created a distinction between heterosexual couples and homosexual couples. However, he was of the view that "that distinction is not made upon the basis of the sexual orientation of the [appellants] and thus does not discriminate against them on that basis". Rather, he held that the distinction was between "spouses" and "non-spouses". He concluded (at pp. 703-4) that: The [appellants] do not fall within the meaning of the word "spouse" any more than heterosexual couples who live together and do not publicly represent themselves as man and wife such as a brother and sister, brother and brother, sister and sister, two relatives, two friends, or parent and child. The single sex couple fall into the same category as those, i.e. the non spousal couple category. Therefore, the trial judge concluded that the Old Age Security Act did not infringe the appellants' s. 15(1) rights on the basis of either their sex or their sexual orientation. (b) Federal Court of Appeal, [1993] 3 F.C. 401 (i) Robertson and Mahoney JJ.A. for the majority 119 Robertson J.A. stated that the issue on appeal was whether the exclusion of homosexual couples from the definition of "spouse" contained in the Old Age Security Act was discriminatory within the meaning of s. 15(1) of the Charter. Mahoney J.A. stated (at p. 410) that it was an "unassailable conclusion of fact based on the evidence" that had Nesbit been a woman cohabiting with Egan, he would have been entitled to the Spousal Allowance. Robertson J.A. further stated (at p. 461) that: . . . I take it to be settled law that sexual orientation can be invoked as an analogous ground of discrimination under subsection 15(1). The respondent conceded this point and, in my opinion, rightly so. 120 The majority reasons held, however, that the impugned legislation did not draw a distinction on the basis of sexual orientation. Rather, they found that "the criterion of entitlement is expressed in terms of spousal status" (p. 478, per Robertson J.A.). They observed that the definition of "spouse" excluded a broad class of non-spouses. In order to show that there was discrimination on the basis of sexual orientation, the appellants would have to show that homosexual couples were adversely affected by the law as compared to other "non-spousal" couples. They concluded that because homosexual couples were not affected more adversely than other "non-spousal" couples, there was no discrimination under s. 15(1) of the Charter. (ii) The dissenting reasons 121 Linden J.A. was of the view that the court had to subject the definition of "spouse" in the Old Age Security Act to a thorough s. 15(1) Charter analysis. He also found that, since the parties were seeking a remedy under s. 52(1) of the Constitution Act, 1982, the analysis should focus on the group, namely homosexuals, rather than on the parties personally. 122 Linden J.A. concluded that the spousal allowance was a benefit of the law under s. 15(1) of the Charter. After reviewing the history of discrimination suffered by homosexuals, Linden J.A. concluded that sexual orientation should be recognized as an analogous ground. 123 Linden J.A. observed that the legislative distinction was explicitly drawn on the basis of whether partners in a relationship were of the same sex or of the opposite sex. Although "being in a same sex relationship is not necessarily the defining characteristic of being gay or lesbian", he found that the distinction was based on a characteristic or matter related to sexual orientation, "since it is lesbians and gay men who may enter into same sex relationships" (p. 432). He found that the exclusion of homosexuals from the spousal allowance constituted discrimination on the basis of sexual orientation. 124 With respect to s. 1 of the Charter, Linden J.A. applied the test in R. v. Oakes, [1986] 1 S.C.R. 103, and found that the objective of the spousal allowance program was pressing and substantial and that the means employed under the program were rationally connected to achieving the objective. However, as a result of the definition of "spouse", the program did not impair the right guaranteed in s. 15(1) as little as possible. Therefore, the violation could not be saved under s. 1. 125 By way of remedy, Linden J.A. would have read down the definition of spouse to exclude the words "of the opposite sex". In addition, after the words "if the two persons publicly represent themselves as husband and wife" in s. 2, he would have read in the words "or as in an analogous relationship". Further, he would have declared that homosexual couples could not be denied the spousal allowance so long as they met the usual eligibility requirements. Issues 126 The constitutional questions which have been stated by this Court are: 1. Does the definition of "spouse" in s. 2 of the Old Age Security Act, R.S.C., 1985, C. 0-9, infringe or deny s. 15(1) of the Canadian Charter of Rights and Freedoms? 2. If the answer to question 1 is "yes", is the infringement or denial demonstrably justified in a free and democratic society pursuant to s. 1 of the Canadian Charter of Rights and Freedoms? Analysis Preliminary Issue 127 One submission must be dealt with at the outset. The respondent contends that by this action the appellants are requesting the Court "to change fundamentally the essential meaning of the societal concept of marriage". I cannot accept that submission. It appears to me to be inaccurate and misleading. This case cannot be taken as constituting a challenge to either the traditional common law or statutory concepts of marriage. Rather, the sole issue presented is whether the state can define a "common law spouse" in a manner which explicitly excludes homosexual couples. Eligibility for payment of the spousal allowance under the Old Age Security Act is not in any way contingent upon being married. Rather, the spousal allowance is specifically made available to common law couples. The only aspect of the Act which is being challenged is the definition of a common law spouse. Thus, any contention that this appeal will affect the societal concept of marriage can be set aside. How Should s. 15(1) be Applied? 128 Section 15(1) of the Charter is of fundamental importance to Canadian society. The praiseworthy object of the section is the prevention of discrimination and 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": Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, at p. 171. It has been recognized that the purpose of s. 15(1) is "to advance the value that all persons be subject to the equal demands and burdens of the law and not suffer any greater disability in the substance and application of the law than others": R. v. Turpin, [1989] 1 S.C.R. 1296, at p. 1329. It is this section of the Charter, more than any other, which recognizes and cherishes the innate human dignity of every individual. It is this section which recognizes that no legislation should treat individuals unfairly simply on the basis of personal characteristics which bear no relationship to their merit, capacity or need. 129 With this background in mind, it is appropriate to consider the principles which should guide a court in an interpretation of s. 15(1) and then to apply those principles to the situation presented in this case. 130 In Andrews, supra, and Turpin, supra, a two-step analysis was formulated to determine whether a s. 15(1) right to equality had been violated. The first step is to determine whether, due to a distinction created by the questioned law, a claimant's right to equality before the law, equality under the law, equal protection of the law or equal benefit of the law has been denied. During this first step, the inquiry should focus upon whether the challenged law has drawn a distinction between the claimant and others, based on personal characteristics. 131 Not every distinction created by legislation gives rise to discrimination. Therefore, the second step must be to determine whether the distinction created by the law results in discrimination. In order to make this determination, it is necessary to consider first, whether the equality right was denied on the basis of a personal characteristic which is either enumerated in s. 15(1) or which is analogous to those enumerated, and second, whether that distinction has the effect on the claimant of imposing a burden, obligation or disadvantage not imposed upon others or of withholding or limiting access to benefits or advantages which are available to others. 132 In Andrews, supra, it was recognized that any search for either equality or discrimination requires comparisons to be made between groups of people. At p. 164, McIntyre J. stated: It [equality] 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. 133 It is true that, in that same case, the so-called `similarly situated test' was rejected on the grounds that its reasoning was unduly formalistic and circular: it uncritically accepted the distinction drawn by the questioned statute and then proceeded to rely upon that same categorization in order to justify the distinction drawn. Nonetheless, any discussion of equality or discrimination requires an element of comparison. The fact that a comparison must be made does not mean that courts will be returning to the similarly situated test, as suggested by the respondent. Rather, making the comparison recognizes that discrimination cannot be identified in a vacuum. For example, in Symes v. Canada, [1993] 4 S.C.R. 695, at p. 754, the following appears in the reasons of Iacobucci J.: . . . it is important to realize that, in order to determine whether particular facts demonstrate equality or inequality, one must necessarily undertake a form of comparative analysis. 134 In Andrews, supra, the basic definition of discrimination was set out in these words (at pp. 174-75): . . . 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 will rarely escape the charge of discrimination, while those based on an individual's merits and capacities will rarely be so classed. 135 In Turpin, supra, it was noted that whether or not discrimination exists must be assessed in a larger social, political and legal context. At pages 1331-32, Wilson J. wrote: In determining whether there is discrimination on grounds relating to the personal characteristics of the individual or group, it is important to look not only at the impugned legislation which has created a distinction that violates the right to equality but also to the larger social, political and legal context. . . . A finding that there is discrimination will, I think, in most but perhaps not all cases, necessarily entail a search for disadvantage that exists apart from and independent of the particular legal distinction being challenged. 136 Finally, in Turpin it was emphasized that the resolution of the question as to whether there is discrimination under s. 15(1) must be kept distinct from the determination as to whether or not there is justification for that discrimination under s. 1 of the Charter. At page 1325, Wilson J. stated: In defining the scope of the four basic equality rights it is important to ensure that each right be given its full independent content divorced from any justificatory factors applicable under s. 1 of the Charter. This analytical separation between s. 15(1) and s. 1 is important since the Charter claimant must satisfy the onus of showing only that there exists in the legislation a distinction which is discriminatory. Only after the court finds a breach of s. 15(1) does the government bear the onus of justifying that discrimination. Adverse Effect Discrimination or Direct Discrimination 137 The respondent contends that the majority of the Court of Appeal was correct when it found that this was a case of adverse effect discrimination. I cannot agree with that argument. 138 Direct discrimination involves a law, rule or practice which on its face discriminates on a prohibited ground. Adverse effect discrimination occurs when a law, rule or practice is facially neutral but has a disproportionate impact on a group because of a particular characteristic of that group. The distinction between direct discrimination and adverse effect discrimination was set out in Ontario Human Rights Commission v. Simpsons-Sears Ltd., [1985] 2 S.C.R. 536, at p. 551 in these words: A distinction must be made between what I would describe as direct discrimination and the concept already referred to as adverse effect discrimination in connection with employment. Direct discrimination occurs in this connection where an employer adopts a practice or rule which on its face discriminates on a prohibited ground. For example, "No Catholics or no women or no blacks employed here." ... It [adverse effect discrimination] arises where an employer for genuine business reasons adopts a rule or standard which is on its face neutral, and which will apply equally to all employees, but which has a discriminatory effect upon a prohibited ground on one employee or group of employees in that it imposes, because of some special characteristic of the employee or group, obligations, penalties, or restrictive conditions not imposed on other members of the work force. Although that case dealt with the Ontario Human Rights Code, the same definition has been adopted in s. 15(1) cases: see Andrews, supra, at p. 165. 139 The law challenged in this case is, quite simply, not facially neutral. Section 2 of the Act defines "spouse" as being "a person of the opposite sex". It thereby draws a clear distinction between opposite-sex couples and same-sex couples. Thus, this case presents a situation of direct discrimination. The Application of s. 15(1) to the Situation Presented in This Case 1. Denial of Equal Benefit of the Law 140 The appellants submit that they have been denied the right to equal benefit of the law. The issue at this threshold stage of the s. 15(1) inquiry, then, is whether the definition of "spouse" draws a distinction which confers a benefit on some while denying that same benefit to others. It is important to remember that this stage of the inquiry is not concerned with whether or not that basis of distinction was either an enumerated or analogous ground or whether there has been discrimination. Nor is the inquiry at this stage concerned with any possible justification for the distinction. (a) Whom Does the Act Seek To Benefit? 141 Looking at the plain wording of the Act, as opposed to any proposed objective of the legislation, it is clear that, in circumstances where the combined income of the pensioner and the opposite-sex spouse falls below a certain level, the Act confers a spousal benefit upon the opposite-sex spouse who is between the ages of 60 and 64. It is not necessary that the spouses be married. The only two requirements for eligibility are that the spouses have lived together for one year and that their combined income falls below the fixed level. 142 The respondent seems to contend that the Act was not one of general application. It is argued that the appellants were not denied equal benefit of the law because the legislation was only intended to confer a benefit upon either heterosexual couples who have raised children or upon dependent female spouses. These submissions cannot be accepted. 143 The Act makes no reference to children. Further, the minimal requirements pertaining to common law couples make it apparent that it would apply to heterosexual couples who have never had children or those who have had children in relationships other than their present one. It is clear, then, that the Act is not concerned with benefiting those who have raised and nurtured children. 144 Similarly, it cannot be said that the Act was designed to benefit only women. A concern about dependent female spouses may have motivated the creation of the spousal allowance. However, from its inception, the spousal allowance has been available equally to male and female spouses. Moreover, the contention that the allowance is for women only does not accord with current demographics. In today's society, the situation in which both spouses are working can no longer be looked upon as rare. Rather, a study by the Vanier Institute of the Family reveals that this tends to be the rule and not the exception: Profiling Canada's Families (1994). 145 The Institute records that in 1975, 42 percent of married women and 59 percent of single women worked in the paid labour force. In 1992, 61 percent of married women and 65 percent of single women worked in the paid labour force. Today, most women work in the paid labour force before, during and after marriage: Profiling Canada's Families (1994), at p. 67. 146 In 1970, in 30 percent of married couples with children under the age of 19, both spouses worked in the paid labour force. In 1990, in 70 percent of such married couples both spouses worked in the paid labour force. In 1990, both parents were full-time employees in 51 percent of such families. Overall, only 15 percent of families followed the "conventional" heterosexual pattern of financial interdependence with a full-time male breadwinner and an unpaid female spouse working in the home: Profiling Canada's Families, at pp. 71 and 74. 147 In bar admission courses and chartered accounting courses, women have made up almost half the graduates in recent years. The figures for medical graduates are not far different. All this indicates that the earnings of women in those professions, as in all fields, may soon equal those of men. Professional women may not be members of the segment of society that would ordinarily benefit from the Act. Yet, the advance of women in these areas serves to indicate that it can no longer be automatically assumed that the male spouse will always earn more than the female or that the female spouse will always be in greater need of a spousal allowance than the male. This very situation seems to have been foreseen by the legislation which does not distinguish between the entitlement of male and female spouses. Rather, either spouse may receive the spousal allowance. 148 Thus, it can be seen that the Act is designed to benefit either the male or female member of a heterosexual common law couple who have lived together for a period of one year and have a net income which is below the fixed level. Payment of the spousal allowance has nothing to do with the recognition of the contribution made by the couple in raising children nor has it anything to do with the gender of the spouse. 149 Does the Act, by its provisions, make a distinction between different groups of people? The answer to that question must be derived from and based upon a review of the challenged legislation. In this case, the challenged Act specifically defines a common law spouse as a "person of the opposite sex" and requires the couple to publicly represent themselves as husband and wife. Linden J.A. observed in his reasons that the very expression "husband and wife" is based upon notions of gender and that the words may not be separable from their heterosexual origins. It is clear, then, that as a result of this definition of "spouse", homosexual common law couples are denied the benefit of the spousal allowance which is available to heterosexual common law couples. It must now be considered whether this distinction amounts to a denial of equal benefit of the law. (b) Does the Distinction Constitute a Denial of Equal Benefit of the Law? 150 In order to reach a decision in this case, it is not necessary to define precisely what constitutes equal benefit of the law. In Turpin, supra, the caution was expressed (at p. 1326) that: . . . it would be unwise, if not foolhardy, to attempt to provide exhaustive definitions of phrases which by their nature are not susceptible of easy definition and which are intended to provide a framework for the "unremitting protection" of equality rights in the years to come. 151 Yet, it seems clear that the denial of the spousal allowance to homosexual couples constitutes a clear denial of equal benefit of the law. The spousal allowance confers an economic benefit which, as a result of the statute's definition of spouse, is denied to homosexual common law couples. Thus, they have been denied equal benefit of the law. 152 The respondent argues that the appellants were not denied a benefit and put forward two grounds in support of that position. First, it is alleged that Nesbit had access to a government subsidy provided by a provincial statute, the Guaranteed Available Income for Need Act, R.S.B.C. 1979, c. 158, as amended ("GAIN"). Second, it was contended that Egan and Nesbit could receive more government income support payments by claiming separately as individuals under the Old Age Security Act and GAIN than they could receive as a couple receiving a pension and a spousal allowance under the Old Age Security Act. There are three reasons for rejecting these submissions of the respondent. 153 First, the relief sought in this action is a finding pursuant to s. 52(1) of the Constitution Act, 1982 that a portion of the Act is unconstitutional. Section 52(1) operates to invalidate all or a part of any Act when it is found to be inconsistent with the Constitution. The appellants are not alleging that the discrimination is unique or particular to their personal situation but, rather, that the Act discriminates against all homosexual common law couples who are living in a state which is comparable to heterosexual common law couples. It follows that the appellants must demonstrate that homosexual couples in general are denied equal benefit of the law, not that they themselves are suffering a particular or unique denial of a benefit. The precise mathematical calculation of benefits which could be paid to couples either as individuals or as a couple is of little assistance as it will inevitably vary from case to case depending upon the particular economic circumstances of each couple and each member of that couple. Rather, a reading of the legislation reveals that it denies the spousal allowance to all homosexual common law couples and thus, it is established that the Act has denied equal benefit of the law. 154 Second, in seeking the answer as to whether or not there has been a denial of equal benefit of the law, it is of course appropriate to have regard to the entire statute which has been called into question. Obviously a benefit which is denied in one portion of an Act may be replaced by compensation provided for in another portion of the same Act. It may, as well, be appropriate and indeed necessary to look at other legislation from the same jurisdiction to determine the issue. Clearly a benefit denied in one federal statute may be replaced by compensation or a benefit provided in another federal statute. 155 However, it is inappropriate to look to provincial legislation to correct or rectify the denial of a benefit set out in a federal Act. Provincial legislatures have exclusive control over matters within their jurisdiction. It follows that the benefits which are enacted by those legislatures may well vary from province to province. Thus, it would only be appropriate to have regard to provincial legislation if the federal Act in question explicitly stated that the provincial law was incorporated into its provisions or that the benefits conferred under the federal and provincial statutes were to be coordinated. 156 Most importantly, the question as to how federal and provincial statutes interact should not be considered in a s. 15(1) analysis. It is a question which goes to the possible justification for an act which can only be addressed under s. 1 of the Charter: see Tetreault-Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22, at p. 42; and Symes v. Canada, supra, at pp. 773-74. Postponing this inquiry to s. 1 is appropriate because, if a claimant has established that the challenged legislation has denied an equal benefit of the law, then the government would, under s. 1 of the Charter, bear the onus of demonstrating that the denial was offset and justified by benefits provided under other provincial legislation. 157 In any event, it would appear that the figures upon which the respondent bases its contention that the appellants are better off when treated individually are highly speculative and may well be incorrect. Certainly, if the calculations were made now on the basis of the present provincial policy, which considers the combined income of homosexual spouses when assessing the quantum of benefits due, the appellants would receive less than they would if they had received a spousal allowance. 158 Third, the concept of equal benefit of the law should not be restricted to a simple calculation of economic profit or loss. The equality right set out in s. 15(1) is not phrased as guaranteeing an equal right to a benefit but, rather, it is expressed as guaranteeing equal benefit of the law. The manner of expression is significant and furthers the aim of this section to foster respect for the innate dignity of every individual. 159 A law may well confer a benefit by providing individuals with the opportunity to make a significant choice. In Turpin, supra, this very issue was considered. The Criminal Code provided that in certain situations an accused could choose to be tried by a judge and jury or before a judge alone. The sections of the Code providing this choice had not been proclaimed in Alberta and this failure was alleged to infringe s. 15 by denying an equal benefit of the law. In her reasons, Wilson J. observed that the word "benefit" should be interpreted so as to recognize that the benefit provided by the section was the opportunity to the accused to choose the form of trial. It was pointed out that to force the accused to be tried by a judge and jury because the court had determined that it was better for an accused to do so would be to "imprison a man in his privileges and call it the Constitution" (p. 1313). 160 That same analysis should be undertaken and applied in this case. To force homosexual common law couples to claim federal and provincial support as individuals because they would get more money would be to imprison them in their privileges. Heterosexual couples might also be better off financially if they claimed government subsidies as individuals rather than as a couple. Yet, cohabiting heterosexual persons have the right to make a choice as to whether they wish to be publicly recognized as a common law couple. Homosexual couples, on the other hand, are denied the opportunity because of the definition of "spouse" set out in the challenged Act. The public recognition and acceptance of homosexuals as a couple may be of tremendous importance to them and to the society in which they live. To deny homosexual couples the right to make that choice deprives them of the equal benefit of the law. 161 The law confers a significant benefit by providing state recognition of the legitimacy of a particular status. The denial of that recognition may have a serious detrimental effect upon the sense of self-worth and dignity of members of a group because it stigmatizes them even though no economic loss is occasioned. This principle has been recognized in the cases of the U.S. Supreme Court dealing with the segregation of races. See, for example, Brown v. Board of Education of Topeka, 347 U.S. 483 (1954). In equality cases, "the main consideration must be the impact of the law on the individual or the group concerned": see Andrews, supra, at p. 165. The choice of a spouse is a matter of great importance to the individuals involved. A very real benefit which is derived from the payment of the spousal allowance is the recognition by the state of the societal benefits which flow from supporting a couple who, for at least a year, have established a stable relationship which involves cohabitation, commitment, intimacy and economic interdependence. This benefit of the law is very significant. Its importance can be seen by considering what the result might be if, for example, the benefit were to be denied to couples because the individuals were of different races or different religions. The public outcry would, I think, be immediate and well merited. Such legislation would clearly infringe s. 15(1) because its provisions would indicate that the excluded groups were inferior and less deserving of benefits. Similarly, an Act which denies equal benefits to homosexual couples who live in a loving and stable common law relationship as a result of their sexual orientation would appear to equally infringe s. 15(1) of the Charter. 162 In our democratic society, every individual is recognized as important and deserving of respect. Each individual is unique and distinct. Because of the uniqueness of individuals, their tastes will vary infinitely from matters as prosaic as food and clothing to matters as fundamental as religious belief. Religious belief and the form of worship are personal characteristics. These characteristics may seem extremely peculiar and vastly perplexing to the majority. Yet, so long as the form of worship is not unlawful, it must be not only tolerated but also protected by the Charter. Similarly, individuals, because of their uniqueness, are bound to vary in those personal characteristics which may be manifested by their sexual preferences whether heterosexual or homosexual. So long as those preferences do not infringe any laws, they should be tolerated. In its attempt to prohibit discrimination, the Charter seeks to reinforce the concept that all human beings, however different they may appear to be to the majority, are all equally deserving of concern, respect and consideration. 163 It follows that, in my view, s. 2 of the Old Age Security Act denies equal benefit of the law to homosexual couples. It does so by denying them an economic benefit and by denying them the right to make a choice in a matter which affects them deeply and personally in a manner that denies their inherent dignity. With the denial of the benefit established, it is now necessary to determine whether that denial is discriminatory. 2. Does the Distinction Result in Discrimination? (a) Distinction On The Basis of a Personal Characteristic 164 In Andrews, supra, at p. 174, McIntyre J. observed that one of the prime characteristics of discrimination is that it involves a "distinction . . . based on grounds relating to personal characteristics of the individual or group". The first question to be resolved, then, is whether the distinction set out in s. 2 of the Old Age Security Act is one "based on" personal characteristics. It is my view that the distinction in the Act is indeed based on a personal characteristic, specifically, sexual orientation. 165 The respondent argues that the distinction was not drawn in reliance upon a personal characteristic but rather on the basis of "spousal" as opposed to "non-spousal" status. The respondent submits that homosexual common law couples are non-spousal couples just as are siblings, parent-child relationships, roommates or any other non-spousal household excluded from the Act. This position was adopted by the Trial Division and the majority of the Court of Appeal. With respect, I cannot accept that position. To say that the distinction is between "spouses" and "non-spouses" is to avoid the very issue which is presented by the legislation in this case, namely the definition of a "spouse". 166 Section 2 of the Act provides that an allowance is available to "spouses". It is worth repeating that the appellants are not challenging the Parliamentary decision to confer benefits on spousal as opposed to non-spousal households. What is in dispute is whether, having decided to confer a benefit on common law spouses, the legislation may then employ a definition of spouse which discriminates on the basis of sexual orientation. It is clear that Parliament does not have any constitutional obligation to provide benefits. However, once the decision has been made to confer a benefit, it cannot be applied in a discriminatory manner. See Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219, at p. 1240; Schachter v. Canada, [1992] 2 S.C.R. 679, at pp. 721-22. The fact that, in the past, the term "spouse" had only a heterosexual connotation must not serve to prevent an inquiry into whether the statutory definition limiting "spouse" only to heterosexual couples violates s. 15(1) of the Charter. This principle was emphasized in Turpin where at p. 1328 Wilson J. wrote: The argument that s. 15 is not violated because departures from its principles have been widely condoned in the past and that the consequences of finding a violation would be novel and disturbing is not, in my respectful view, an acceptable approach to the interpretation of Charter provisions. 167 The words of the statutory definition are clear and can have but one meaning. The legislation defines spouse as "a person of the opposite sex". To treat persons of the same sex who represent themselves as a common law couple differently from persons of the opposite sex representing themselves as a common law couple is a differentiation which must be based upon sexual orientation. I would add that, although the statute appears to do so, it is not necessary for the challenged legislation to directly identify sexual orientation as a criterion for eligibility. For example, it has been held that a distinction made on the basis of pregnancy constitutes discrimination on the basis of sex. Similarly, differential treatment in the form of sexual harassment constitutes discrimination on the basis of sex: see Brooks, supra; and Janzen v. Platy Enterprises Ltd., [1989] 1 S.C.R. 1252. What is relevant in resolving the issue is whether the difference in treatment affects the individual or group in a manner which is related to their personal characteristics. To put it another way, the question is whether the difference in treatment is closely related to a personal characteristic of a group to which the claimant belongs. On this issue, I think the reasoning of Linden J.A. was appropriate and bears repetition. He stated (at p. 431): 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 . . . . 168 In this case, there can be no doubt that the distinction is related to the personal characteristic of sexual orientation. It may be correct to say that being in a same-sex relationship is not necessarily the defining characteristic of being homosexual. Yet, only homosexual individuals will form a part of a same-sex common law couple. It is the sexual orientation of the individuals involved which leads to the formation of the homosexual couple. The sexual orientation of the individual members cannot be divorced from the homosexual couple. To find otherwise would be as wrong as saying that being pregnant had nothing to do with being female. The words "of the opposite sex" in the definition of "spouse" specifically exclude homosexual couples from claiming a spousal allowance. It is not without significance that, when it rejected the appellants' application for a spousal allowance, Health and Welfare Canada's Old Age Security department specifically stated that the reason for ineligibility was the definition of spouse. Indeed on this point the trial judge found (at p. 695) that: . . . 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. Similarly, Mahoney J.A. held that this finding was "an unassailable conclusion of fact based on the evidence" (p. 410). 169 In this case, a great deal of time was spent demonstrating the nature of the warm, compassionate, caring relationship that very evidently existed between the appellants. In passing, it is, I think, worth mentioning that this need not be done in every case. It is not necessary that the evidence demonstrate that a homosexual relationship bears all the features of an ideal heterosexual relationship for the relationship of many heterosexual couples is sometimes far from ideal. The relationships between heterosexuals must vary as infinitely as do the personalities of the individuals involved. In this case, it would have sufficed to prove that the homosexual relationship had existed for more than one year during which time the partners had publicly indicated their relationship and that their combined income was below the statutory limit. This is the same evidence that would be sufficient to qualify a heterosexual common law couple for the spousal allowance. 170 It is therefore evident that the ground of distinction between the appellants and heterosexual couples who qualify for the spousal allowance is their sexual orientation. Since sexual orientation is not an enumerated ground in s. 15(1), it must be determined whether it is analogous to those which are enumerated in the section. (b) Is Sexual Orientation an Analogous Ground? 171 The reasons in Andrews, supra, and Turpin, supra, indicate that in order to determine whether the basis of distinction is analogous to the enumerated grounds, it is first necessary to identify the group which is affected. It is true that in some cases it may be useful to determine whether or not the affected group forms a "discrete and insular minority" which is lacking in political power and, thus, vulnerable to having its interests overlooked or its rights to equal concern and respect violated. Yet, that search is not really an end in itself. While historical disadvantage or a group's position as a discrete and insular minority may serve as indicators of an analogous ground, they are not prerequisites for finding an analogous ground. They may simply be of assistance in determining whether the interest advanced by a claimant is the sort of interest that s. 15(1) was designed to protect. The fundamental consideration underlying the analogous grounds analysis is whether the basis of distinction may serve to deny the essential human dignity of the Charter claimant. Since one of the aims of s. 15(1) is to prevent discrimination against groups which suffer from a social or political disadvantage it follows that it may be helpful to see if there is any indication that the group in question has suffered discrimination arising from stereotyping, historical disadvantage or vulnerability to political and social prejudice. 172 The respondent argued that sexual orientation should only be considered an analogous ground if the appellants could show that homosexuals suffered a specific form of economic disadvantage which was exacerbated by the legislation in question. This argument cannot succeed. It would fragment our concept of discrimination and would seem to be illogical since discrimination, whether it is based on historical, political or societal disadvantage, will almost always have adverse economic consequences. Conversely, economic discrimination is inherently connected to discriminatory social and political attitudes which have prevailed in the past. Yet, the basic issue to be resolved is whether the challenged Act has made a distinction on the basis of an analogous ground. The resolution of that issue must be made "in the context of the place of the group in the entire social, political and legal fabric of our society": see Andrews, supra, at p. 152; Turpin, supra, at p. 1332; and R. v. Swain, [1991] 1 S.C.R. 933, at p. 991. 173 The historic disadvantage suffered by homosexual persons has been widely recognized and documented. Public harassment and verbal abuse of homosexual individuals is not uncommon. Homosexual women and men have been the victims of crimes of violence directed at them specifically because of their sexual orientation: Equality For All: Report of the Parliamentary Committee on Equality Rights (1985), at p. 26; Cynthia Petersen, "A Queer Response to Bashing: Legislating Against Hate" (1991), 15 Queen's L. J. 237; Nova Scotia Public Interest Research Group, "Proud but Cautious": Homophobic Abuse and Discrimination in Nova Scotia (1994); Bill C-41 (1994). They have been discriminated against in their employment and their access to services. They have been excluded from some aspects of public life solely because of their sexual orientation: Equality For All, supra, at pp. 30-32; Douglas v. Canada (1992), 58 F.T.R. 147 (T.D.). The stigmatization of homosexual persons and the hatred which some members of the public have expressed towards them has forced many homosexuals to conceal their orientation. This imposes its own associated costs in the work place, the community and in private life. 174 For example, a study by the Quebec Human Rights Commission has indicated that the isolation, harassment and violence imposed by the public and the rejection by their families has caused young homosexuals to have a higher rate of attempted and successful suicide than heterosexual youths: De l'illegalite a l'egalite: Rapport de la consultation publique sur la violence et la discrimination envers les gais et lesbiennes (Commission des droits de la personne du Quebec, May 1994), at p. 125. Until 1969, certain forms of homosexual sexual intercourse were criminal offences. Until 1973, the American Psychiatric Association labelled homosexuality a psychiatric disorder and the World Health Organization considered it a psychiatric disorder until as recently as 1993. 175 Homosexual couples as well as homosexual individuals have suffered greatly as a result of discrimination. Sexual orientation is more than simply a "status" that an individual possesses. It is something that is demonstrated in an individual's conduct by the choice of a partner. The Charter protects religious beliefs and religious practice as aspects of religious freedom. So, too, should it be recognized that sexual orientation encompasses aspects of "status" and "conduct" and that both should receive protection. Sexual orientation is demonstrated in a person's choice of a life partner, whether heterosexual or homosexual. It follows that a lawful relationship which flows from sexual orientation should also be protected. The European Parliament, in its legislation prohibiting discrimination on the basis of sexual orientation, specifically sought to address the discrimination faced by homosexuals not only as individuals but as couples: Resolution on Equal Rights for Homosexuals and Lesbians in the European Community (A3-0028/94). These studies serve to confirm overwhelmingly that homosexuals, whether as individuals or couples, form an identifiable minority who have suffered and continue to suffer serious social, political and economic disadvantage. 176 Quite apart from the evidence of historical social, political and economic disadvantage suffered by homosexuals, it is apparent that a legislative consensus is emerging which recognizes that sexual orientation is an analogous and prohibited ground of discrimination. The human rights legislation in New Brunswick, Nova Scotia, Quebec, Ontario, Manitoba and the Yukon all prohibit discrimination on the basis of sexual orientation. As a result of Charter challenges, protection against discrimination on the basis of sexual orientation is also available in Alberta and in the federal jurisdiction. The Ontario Court of Appeal in Haig v. Canada, supra, found that sexual orientation was an analogous ground to those enumerated in s. 15 of the Charter and held that the Canadian Human Rights Act violated s. 15 because it failed to prohibit discrimination on the basis of sexual orientation. The Alberta human rights legislation was similarly found to infringe the Charter in Vriend v. Alberta (1994), 152 A.R. 1 (Q.B.), at p. 14. In both cases, the courts read sexual orientation into the respective Acts. 177 It is significant that a number of courts have, in my view correctly, concluded that sexual orientation is an analogous ground of discrimination in the context both of discrimination against homosexual individuals and of discrimination against homosexual couples. See for example Veysey v. Canada (Correctional Service) (1989), 44 C.R.R. 364; Brown v. British Columbia (Minister of Health) (1990), 42 B.C.L.R. (2d) 294 (S.C.); Knodel v. Britich Columbia (Medical Services Commission) (1991), 58 B.C.L.R. (2d) 356 (S.C.); Leshner v. Ontario (1992), 16 C.H.R.R. D/184 (Bd. of Inq.). Nor can it be forgotten that all three members of the Federal Court of Appeal in this case recognized sexual orientation as an analogous ground and held that the respondent had acted correctly in conceding this point. 178 From the foregoing review, it can be seen that many legislators have recognized sexual orientation as a prohibited ground of discrimination. Similarly, judicial opinion has overwhelmingly recognized that sexual orientation is an analogous ground to those set out in s. 15(1). In my view, there can be no doubt that sexual orientation is indeed a ground of discrimination analogous to those enumerated in s. 15(1). It now remains to be seen whether the distinction on the basis of this analogous ground constitutes discrimination. (c) Is There Discrimination? 179 In my opinion, the distinction drawn by s. 2 of the Old Age Security Act on the basis of sexual orientation does constitute discrimination. The principle was stated with simple clarity by McIntyre J. in Andrews at pp. 174-75: 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. However, cases such as Weatherall v. Canada (Attorney General), [1993] 2 S.C.R. 872, and R. v. Hess, [1990] 2 S.C.R. 906, make it clear that it is not simply the fact that a distinction is drawn on the basis of either an enumerated or analogous ground which gives rise to discrimination. Rather, the existence of discrimination is determined by assessing the prejudicial effect of the distinction against s. 15(1)'s fundamental purpose of preventing the infringement of essential human dignity. The legislature's reliance upon stereotypical reasoning may very well be an extremely significant factor in determining whether discrimination exists. However, in light of the facts presented in this appeal, it is not necessary to elaborate upon other considerations which may also give rise to discrimination. Ultimately, it must be remembered that the question as to whether or not there is discrimination should be addressed from the perspective of the person claiming a Charter violation. 180 In the present appeal, looking at the Act from the perspective of the appellants, it can be seen that the legislation denies homosexual couples equal benefit of the law. The Act does this not on the basis of merit or need, but solely on the basis of sexual orientation. The definition of "spouse" as someone of the opposite sex reinforces the stereotype that homosexuals cannot and do not form lasting, caring, mutually supportive relationships with economic interdependence in the same manner as heterosexual couples. The appellants' relationship vividly demonstrates the error of that approach. The discriminatory impact can hardly be deemed to be trivial when the legislation reinforces prejudicial attitudes based on such faulty stereotypes. The effect of the impugned provision is clearly contrary to s. 15's aim of protecting human dignity, and therefore the distinction amounts to discrimination on the basis of sexual orientation. Conclusion With Regard to s. 15(1) 181 The Act denies common law homosexual couples equal benefit of the law based upon the analogous ground of sexual orientation. The effect of this is discriminatory and so infringes the appellants' rights provided by s. 15(1) of the Charter. It follows that the first constitutional question should be answered as follows: Question: 1. Does the definition of "spouse" in s. 2 of the Old Age Security Act, R.S.C., 1985, c. O-9, infringe or deny s. 15(1) of the Canadian Charter of Rights and Freedoms? Answer: Yes. It remains now to be determined whether the discrimination can be justified under s. 1 of the Charter. IACOBUCCI J.: Section 1 of the Charter 182 Section 1 allows Charter violations to be upheld if these violations are reasonably justifiable in a free and democratic society. The test to establish whether a statutory provision constitutes a "reasonable limit" was first advanced by former Chief Justice Dickson in R. v. Oakes, supra, at pp. 138-39. A limitation to a constitutional guarantee will be sustained once two conditions are met. First, the objective of the legislation must be pressing and substantial. Second, the means chosen to attain this legislative end must be reasonable and demonstrably justifiable in a free and democratic society. In order to satisfy the second requirement, three criteria must be satisfied: (1) the rights violation must be rationally connected to the aim of the legislation; (2) the impugned provision must minimally impair the Charter guarantee; and (3) there must be a proportionality between the effect of the measure and its objective so that the attainment of the legislative goal is not outweighed by the abridgement of the right. In all s. 1 cases the burden of proof is with the government to show on a balance of probabilities that the violation is justifiable. (a) The First Element of the Oakes Test: Is the Legislative Goal Pressing and Substantial? (i) What is the Goal of the Old Age Security Act, R.S.C. 1985, c. O-9? 183 As noted by Cory J., there appears to be some confusion as to the objective of the impugned legislation. In fact, both of the parties to the dispute ascribe different goals to the Act; so, too, do some of the interveners. 184 I am of the view, as was Linden J.A. in the court below, that the objective of the spousal allowance 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 were both members of the couple to be retired, provided, of course, that the non-retired spouse be between the ages of 60 and 64. To this end, the Act is geared towards the mitigation of poverty among "elderly households". I reach this conclusion after reviewing the design of the legislation, as well as the legislative debates and policy statements accompanying the introduction of the spousal allowance. 185 For example, when the spousal allowance was first enacted in 1975, the then Minister of National Health and Welfare, the Honourable Marc Lalonde, described its raison d'etre as follows: 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 bread winner 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. (Minutes of Proceedings and Evidence of the Standing Committee on Health, Welfare and Social Affairs, June 12, 1975, at p. 25:7) 186 Similarly, in a news release dated June 3, 1975, the federal government identified the purpose of the allowance in the following manner: The purpose of this amendment is to provide relief in situations where two persons would otherwise have to live on the pension of one. 187 I note that, owing to the traditional dynamic of heterosexual spousal relationships in which the woman would generally withdraw from the labour market for purposes of raising children, 87 percent of those receiving the spousal allowance are women. This stems from the fact that the retiring "breadwinner" would usually be the husband; the wife would generally be younger than her spouse and would tend to have little, if any, income of her own since she would not be eligible for the old age pension for a few years. The end result would be that the income of the family unit would drop drastically until the wife reached 65 and was awarded her pension. However, the fact that a disproportionately large percentage of the recipients are women does not, in my mind, establish, as submitted by the respondent, that the goal of the Act is the reduction of poverty among elderly women in spousal relationships. I arrive at this conclusion for a number of reasons, most of which have already been alluded to by Cory J. 188 First, both men and women can apply for the allowance. Second, there is no requirement for the couple to have raised children in order to trigger eligibility for the spousal benefit and it is not a prerequisite that the female spouse had to have been a homemaker or out of the paid labour force. Third, the distribution of the benefit is contingent upon the joint family income of the household falling below a certain level. Fourth, the growing presence of women in the labour market is such that the traditional "mother spouse in the house" is becoming less common. Fifth, the legislation makes no reference to the position of women, or to women at all: it is a spouse's pension, not a woman's pension. Finally, given that the allowance is available after simply one year of cohabitation, it does not appear tailored to remedying the disempowered position of women flowing from the dynamic of long-term traditional heterosexual relationships since a woman need not be in such a relationship in order to qualify for the benefit: a woman may be single until the age of 60, then enter a relationship with a man receiving the old age security pension, cohabit for one year, and thereby become eligible to receive the spousal allowance. (ii)Is This Goal Pressing and Substantial? 189 The appellants concede that the alleviation of poverty in elderly households is a goal of pressing and substantial importance. I agree. Moreover, as noted by Lamer C.J. in Schachter v. Canada, supra, at p. 721, "[i]t will be a rare occasion when a benefit conferring scheme is found to have an unconstitutional purpose". The legislation thus satisfies the first component of the Oakes test. (b) The Second Element of Oakes: Proportionality Analysis 190 I conclude that the underinclusiveness of the Act is not a reasonable limit. Although the purpose of the legislation is laudable, it has been implemented in a discriminatory manner in that an equally deserving group meeting the criteria established by the law is denied benefits based on an irrelevant personal characteristic. (i) The Legislation Is Not Rationally Connected to Its Objective 191 If the goal of the legislation is the alleviation of poverty among cohabiting elderly "spouses", then how can this be but incompletely attained by denying otherwise eligible households the spousal allowance merely because of discrimination based on sexual orientation? The exclusion of same-sex partners is simply not rationally connected to the goal of alleviating poverty among elderly couples. If there is an intention to ameliorate the position of a group, it cannot be considered entirely rational to assist only a portion of that group. A more rationally connected means to the end would be to assist the entire group, as that is the very objective which is sought. 192 It is unfortunate that decades of endemic discrimination have resulted in little information being available regarding the numbers of same-sex households in Canada, let alone the numbers of same-sex households that would be eligible for the allowance. Estimates vary widely. The expert for the Crown, Mr. Melvin Hagglund, determined that the number of same-sex couples across Canada that would be entitled to the allowance is somewhere in the range of 15,000 to 30,000. However, upon cross-examination, he conceded that the 1986 Canadian census indicated that there were only 2,700 two-person households across Canada consisting of one person aged 60 to 64 and one person aged 65 or older where the individuals in the household were neither married nor common-law spouses. 193 The respondent Crown submits the cost of such an extension of benefits constitutes grounds for upholding the s. 15 limitation. Mr. Hagglund has estimated the cost of including same-sex spousal cohabitants as ranging from $12 million to $37 million per annum (see Case on Appeal, at p. 123). This evidence is highly speculative and statistically weak and thus accordingly incorporates guesswork. For example, it is based on his generous estimates of the number of eligible same-sex couples and fails to take into account the fact that many of these households will be ineligible because they surpass the maximum income criteria. However, assuming arguendo that Mr. Hagglund's figures are valid, I find, as a question of law, that they do not justify the denial of the appellants' right to equality. 194 The jurisprudence of this Court reveals, as a general matter, a reluctance to accord much weight to financial considerations under a s. 1 analysis. In Schachter, supra, at p. 709, the Chief Justice noted that "[t]his Court has held, and rightly so, that budgetary considerations cannot be used to justify a violation under s. 1". This is certainly the case when the financial motivations are not, as in the case at bar, supported by more persuasive arguments as to why the infringement amounts to a reasonable limit. 195 Despite bearing the onus of proof, the respondent has not supplied evidence demonstrating why the patterns of economic interdependence among same-sex couples are sufficiently different from those in heterosexual relationships to indicate why excluding same-sex couples from the scheme would still enable the legislation to be rationally connected to its goal of mitigating poverty among elderly households. In fact, much of the evidence that exists attests to the fact that same-sex relationships involve similar levels of economic dependence, mutual responsibilities and emotional commitment than heterosexual relationships. For example, in Knodel v. British Columbia (Medical Services Commission), supra, at p. 363, expert testimony was led 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 [they] are different". 196 In Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554, at pp. 630-31, L'Heureux-Dube J., in dissent, canvasses much of this expert and academic authority and agrees with the conclusion that it is in the nature of humankind to form family bonds, this desire not being dependent upon heterosexual orientation. She also approvingly refers to some comments made by the Human Rights Tribunal regarding Mr. Mossop's discrimination claim, holding at p. 630 that "it is to be bound by myth to assume that only heterosexual couples are capable of forming loving caring stable relationships". 197 I note that the simple fact that, in the case at bar, there is such a focus on the level of commitment in homosexual relationships is in and of itself indicative of the extent to which such couples suffer discrimination. The spousal allowance is provided to heterosexual couples regardless of the existence of a dependency pattern in their relationships while all same-sex couples, including all those sharing economic interdependence, are excluded. Whereas there is a presumption of interdependence in heterosexual relationships, there is a presumption against interdependence in same-sex relationships. The latter presumption is not only incorrect, but it is also the fruit of stigmatizing stereotype. 198 Just as the law has come to acknowledge that differential treatment between married and common-law spouses is constitutionally suspect, so too is differential treatment of relationships based on sexual orientation. In sum, the spousal allowance in its present form is not rationally connected to its legislative goals. A program which included the appellants would better achieve the intended goal while respecting the Charter rights of gays and lesbians. Moreover, the financial consequences thereof are certainly not by themselves sufficient to justify the discriminatory legislation. (ii)There is No Minimal Impairment 199 The respondent suggests that the appellants' s. 15 rights are minimally impaired since Nesbit has received financial assistance through the provincial GAIN program enacted under the auspices of the Guaranteed Available Income for Need Act. This argument was raised earlier in terms of ascertaining whether there was in fact a discriminatory denial of a benefit, and Cory J. dismissed the argument. Similarly, I conclude that the denial of the appellants' s. 15 rights through the ineligibility for receipt of the spousal allowance is not minimally impaired by the fact that Nesbit had been receiving disability income assistance owing to a degenerative back condition which had precluded him from working. 200 In Egan and Nesbit's case, the receipt of the disability insurance has nothing to do with their status as a couple nor with their sexual orientation. The grant is simply designed to allay the hardship inflicted upon persons when they are no longer employable because of a disability. Further, whereas the spousal allowance flows from the federal treasury, the disability benefit is supplied by the British Columbia government. In my mind, the discriminatory effect of the legislation should not be considered to be a reasonable limit simply because the appellants' joint income would have roughly been the same because of Nesbit's receipt of provincial support supplementing his income for a completely unrelated reason. There is no evidence that the provincial GAIN legislation was even designed with the federal spousal allowance in mind. Nor is there any evidence that the aggregate intent of the legislatures was to have denial of the spousal allowance for same-sex couples offset by the disability assistance. As pointed out by Linden J.A., the objectives of the GAIN disability income supplement are simply not co-extensive with those of the spousal allowance; in fact the benefits paid out under GAIN are not "substantially similar" to benefits paid under the spouse's allowance program. It cannot be said that the Guaranteed Available Income for Need Act either attempts to or succeeds in "filling the gap" created by the denial of the spouse's allowance benefits to lesbian and gay partners. 201 If anything, as submitted by the Canadian Labour Congress, a recent series of modifications to provincial social assistance eligibility criteria is such that a large gap is being created in this supposedly co-extensive system. In many provinces, persons involved in same-sex cohabitations are no longer treated as "individuals" under social assistance legislation. However, the federal spousal allowance remains limited to opposite sex heterosexual couples. For the purposes of social assistance legislation such as GAIN, ss. 3(2) and 10 of the GAIN Regulations, B.C. Reg. 479/76, as amended, provide that the income of all "dependents" will be taken into account in determining eligibility for and the amount of assistance. "Dependents" include, as per s. 2(d), an individual who resides with another individual, sharing with that person income and household responsibilities associated with family living. This includes same-sex couples. Both the appellants and the respondent agreed, at the hearing of the appeal, as to the nature and timing (as of July 1992) of these changes in British Columbia. See also Sask. Reg. 78/66, as amended, under The Saskatchewan Assistance Act, R.S.S. 1978, c. S-8, as amended; N. B. Reg. 82-227, as amended, under the Social Welfare Act, R.S.N.B. 1973, c. S-11, as amended; Prince Edward Island Welfare Assistance Act Regulations, EC 746/84, under the Welfare Assistance Act, R.S.P.E.I. 1988, c. W-3, as amended. 202 The following situation is created by all of this legislation: if Egan or Nesbit had been treated as dependents under GAIN (they were actually treated as single persons, since the dependent provision was not in force at the time Nesbit was between 60 and 64), they would have been economically worse off by being deprived of the spousal allowance. This is because Egan's income (i.e. old age pension) would be taken into account in determining Nesbit's eligibility for the GAIN payments. As a result, like a heterosexual spouse in the same position, Nesbit would not have received any GAIN payment but, unlike a heterosexual spouse, he would be deprived of the economic benefit of the spousal allowance. The Canadian Labour Congress also points out that, under s. 8 of the GAIN Regulations, an individual's assets could reduce entitlement to GAIN benefits in circumstances where the person, as long as she was involved in a heterosexual relationship, would still qualify for the spousal allowance. The co-extensive interrelationship of these pieces of legislation thus appears to be somewhat illusory. In my view, they are more disjointed than interconnected. 203 Furthermore, even if the provincial GAIN and the federal Old Age Security Act were part of the same overlapping legislative "scheme", this is not sufficient to ground a s. 1 justification. As held by La Forest J. in Tetreault-Gadoury v. Canada (Employment and Immigration Commission), supra, at p. 42: It is fair to take into account the possibility that a group deprived of benefits under one Act may be receiving equal, or even greater, benefits under another. . . . Still, I doubt whether the objective of fitting the Act within the government's particular legislative scheme of social programs could, in itself, be sufficiently important to justify the infringement of a Charter right. 204 In a manner similar to that noted by Cory J. in his reasons, I find myself uncomfortable with basing the constitutionality of federal legislation upon the actions of a provincial legislature over which Parliament has no control. The possibility arises that a federal initiative might be constitutional in certain provinces yet not in others. Such an approach undermines the effectiveness and uniformity of the Charter. Although there might be cases in which provincial legislation or law could be relied upon in preserving the constitutionality of federal legislation, this would only be in a situation where all of the provinces have specifically ensured that the discriminatory effect of federal legislation be eliminated through provincial enactments or law. Such is clearly not the situation in the case at bar. 205 Furthermore, given the fact that the GAIN and spousal allowance programmes are clearly not, in intent or by effect, co-extensive, the present case can be distinguished from the majority's comments in Symes, supra. In Symes (at p. 773), it was noted that, under a s. 1 analysis, it is important to consider the operation of the impugned legislation (i.e. the Income Tax Act's refusal to permit child-care expenses to be deducted as business expenses) in light of the operation of other governmental systems relating to child care. In the case at bar, the GAIN disability or other provincial social assistance legislation has no relation to the spousal allowance. There is no dovetailing, nor any incorporation by the impugned federal law of the provincial scheme or, for that matter, vice versa. 206 At this point, it is important to emphasize that, although the appellants may have financially benefited from governmental generosity because of Nesbit's unfortunate health situation, not all same-sex couples discriminatorily denied the spousal allowance are in a similar position. As between a heterosexual and homosexual couple without any other governmental assistance sufficient to increase their incomes past the eligibility criteria of the spousal allowance, the homosexual couple, based on the irrelevancy of their sexual orientation, would be denied the benefit the heterosexual couple would receive. It is important to keep this broader perspective in mind. 207 In conclusion, I repeat the following passage from the judgment of Martin J. at trial (at p. 698): 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. 208 The fact that Nesbit may by default claim support because of a physical disability is not a reasonable alternative minimizing the infringement of the appellants' s. 15 rights. (iii) The Attainment of the Legislative Goal Is Not Outweighed by the Abridgement of the Right 209 The importance of providing relief to some elderly couples does not justify an infringement of the equality rights of the elderly couples who do not benefit for constitutionally irrelevant reasons. I echo the finding of Linden J.A. in dissent in the court below (at p. 449): 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. 210 The only way that, at a conceptual level, this aspect of the Oakes test might be satisfied in the appeal at bar is if the purpose of the legislation would be construed as ameliorating the situation and fostering the existence of elderly heterosexual couples only. Neither the appellants nor the respondent suggest that this amounts to the purpose of the legislation. The interveners Equality for Gays and Lesbians Everywhere (EGALE), however, do offer such a suggestion. It is clear that, were this to be the goal of the legislation, such a goal would itself be discriminatory. The law in this area is unequivocal: a constitutionally impermissible purpose will not save a law under s. 1 of the Charter: R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295. Such a goal simply cannot be construed as pressing or substantial. 211 On a broader note, it eludes me how according same-sex couples the benefits flowing to opposite-sex couples in any way inhibits, dissuades or impedes the formation of heterosexual unions. Where is the threat? In the absence of such a threat, the denial of the s. 15 rights of same-sex couples is anything but proportional to the policy objective of fostering heterosexual relationships. In dissenting reasons in Mossop, supra, at p. 634, L'Heureux-Dube J. made the following observation, which I believe to be on point: It is possible to be pro-family without rejecting less traditional family forms. 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. See also Nitya Duclos, "An Argument for Legal Recognition of Same-Sex Marriage", in Carol Rogerson, Family Law Cases and Materials 1991-92, Faculty of Law, University of Toronto, vol. I, at p. 194; Margrit Eichler, Families in Canada Today: Recent Changes and their Policy Consequences (2nd ed. 1988). 212 I also note, as does Cory J. in his reasons in the present case, that the issue before us is whether the Charter mandates that same-sex couples be accorded the benefit of the spousal allowance. Despite suggestions by the respondent and the interveners the Attorney General of Quebec and the Inter-Faith Coalition on Marriage and the Family, the facts of this case do not require us to explore whether same-sex couples ought to be constitutionally entitled to adopt children or get married, or whether benefits given to heterosexual families in recognition of having children violate the Charter. Nor do I have to consider whether other cohabitation arrangements (brother-sister, two friends, uncle-nephew) ought to be entitled to state benefits. These issues are not raised by this appeal. Should such claims arise in the future, they will be evaluated on their own merits, both in terms of s. 15 as well as s. 1 analysis. 213 Since preparing these reasons, I have read the reasons of my colleague, Justice Sopinka. I note that, although he finds the impugned statute to violate the appellants' equality rights, he finds this violation to be justifiable in a free and democratic society under s. 1. In reaching this conclusion, he relies heavily on select passages from this Court's judgment in McKinney v. University of Guelph, [1990] 3 S.C.R. 229, at pp. 317-19, per La Forest J. McKinney involved a s. 15 challenge launched by several professors to a university's mandatory retirement policy and a provision in the Ontario Human Rights Code which limited the protection of the Code in the area of employment to those under 65. These passages from McKinney may seem to support the extremely deferential approach to s. 1 adopted by Sopinka J. However, a close examination of the McKinney decision reveals that La Forest J.'s comments therein can be said to be limited to Charter review of provincial human rights legislation governing private relations only. At page 318 of McKinney, immediately before one of the passages cited by Sopinka J., the following appears: The Charter, we saw earlier, was expressly framed so as not to apply to private conduct. It left the task of regulating and advancing the cause of human rights in the private sector to the legislative branch. This invites a measure of deference for legislative choice. As counsel for the Attorney General for Saskatchewan colourfully put it, this "should lead us to ensure that the Charter doesn't do through the back door what it clearly can't do through the front door". 214 Furthermore, I find that the context of McKinney is wholly distinguishable from the present appeal. This appeal involves a closely held personal characteristic (potentially only shared by a minority) upon which a distinction is drawn without the array of competing interests that animated the s. 1 analysis in McKinney. The only competing interest in the case at bar is budgetary in nature. The abolition of a mandatory retirement age, on the other hand, affects many factors,including: the entire composition of the workforce; the ability of younger people to secure jobs; access to university resources; promotion of academic freedom, excellence and renewal; collective bargaining rights; and the structure of pension plans. 215 However, what causes me greater concern is my colleague's position that, because the prohibition of discrimination against gays and lesbians is "of recent origin" and "generally regarded as a novel concept" (p. 6), the government can be justified in discriminatorily denying same-sex couples a benefit enuring to opposite-sex couples. Another argument he raises is that the government can justify discriminatory legislation because of the possibility that it can take an incremental approach in providing state benefits. 216 With respect, I find both of these approaches to be undesirable. Permitting discrimination to be justified on account of the "novelty" of its prohibition or on account of the need for governmental "incrementalism" introduces two unprecedented and potentially undefinable criteria into s. 1 analysis. It also permits s. 1 to be used in an unduly deferential manner well beyond anything found in the prior jurisprudence of this Court. The very real possibility emerges that the government will always be able to uphold legislation that selectively and discriminatorily allocates resources. This would undercut the values of the Charter and belittle its purpose. I also find that many of the concerns raised by Sopinka J. -- such as according the legislature some time to amend discriminatory legislation -- ought to inform the remedy, and should not serve to uphold or legitimize discriminatory conduct: Schachter, supra. Remedy 217 Having found s. 2 of the Old Age Security Act to be unconstitutional, I now turn to the issue of remedy under s. 52 of the Constitution Act, 1982. The decision of this Court in Schachter, supra, leaves the Court with several options: (1) the Court may strike down the legislation, thereby scuttling the spousal allowance; or (2) the Court may strike down yet suspend that declaration for a specific period of time so that the government may enact a constitutionally legitimate spousal allowance program; or (3) given the discretion to intervene only "to the extent of the inconsistency", the Court may "read in" or "read out/read down" the impugned legislation, either immediately or under a suspensive declaration. 218 The appellants request a remedy akin to the third option above. They pray this Court to: declar[e] that for the purpose of section 2 of the Act the definition of "spouse" should be read down by deleting the words "of the opposite sex" and reading in the words "or as an analogous relationship" after the words "if the two persons publicly represent themselves as husband and wife"; 219 The reading out of the "opposite sex" requirement would ensure that same-sex couples are entitled to benefit from the legislation and reading in "or as an analogous relationship" means they can demonstrate their eligibility without having to misrepresent their relationship as being "of husband and wife". 220 Although the appellants' request is ambitious, I would grant it. The only proviso is that the remedy be temporarily suspended for one year so that Parliament can itself attend to ensuring that the spousal allowance be distributed to same-sex couples. Should Parliament not do so in a constitutionally satisfactory manner within this time frame, the appellants' construction of s. 2 shall be read into the Act. The granting of the appellants' request for remedy is consonant with the principles of "reading in" developed by this Court in Schachter. 221 In Schachter, Lamer C.J. noted that the first step in choosing a remedial course under s. 52 is to define the extent of the inconsistency which must be struck down. In the case at bar, I am faced with a situation where the purpose of the legislation is valid. Its inability to pass constitutional muster stems from the fact that it is underinclusive, and this underinclusiveness emanates from an exclusionary distinction based on the irrelevant ground of sexual orientation. 222 Having isolated the nub of the constitutional inconsistency, I must determine whether it is appropriate to apply the "reading in" approach. Totally invalidating the spousal allowance simply because it is underinclusive legislation would make little sense. In fact, to borrow the Chief Justice's language from Schachter, it would be "absurd" (p. 721), nothing short of "equality with a vengeance" (p. 702). Instead, this appears to be a case in which the offending portion of the statute (the "opposite-sex requirement") can be defined in a limited manner and, consequently, it is consistent with legal principles to declare inoperative only that limited portion. After all, as noted by the Chief Justice at p, 721 of the Schachter decision: Cases involving positive rights [i.e. the conferral of benefits] are more likely to fall into the remedial classification of reading down/reading in. . . . 223 Upon closer perusal of the criteria established by Schachter, it becomes apparent that there is no reason to deny the appellants' request to both read in and read out: (i) There is "remedial precision" insofar as the insertion of a handful of words can ensure the validity of the legislation and remedy the constitutional wrong. (ii) Reading same-sex couples into the statutory definition would not trench upon the goal of the spousal allowance nor interfere with the legislative objective; in fact, it would better promote the attainment of this goal. (iii) There would be no deleterious effect whatsoever on the thrust of the legislation; in other words, heterosexual couples would still acquire the same benefits in the same way that they had before the reading in. Thus, it can be concluded that Parliament, if faced with the choice of not having a spousal allowance program or one that extends to same-sex couples, would have chosen the latter. The legislation has not been changed so markedly that it cannot be assumed that the legislature would not have enacted it. On this point, Linden J.A. noted in the court below that old age security and the spousal allowance were significant and durable aspects of Canadian society. According to the Schachter test, such a finding strengthens the assumption that the legislation would have been enacted without the impermissible omission (at p. 712). (iv) I would dispel the concerns regarding the budgetary considerations that may arise by "reading in" in the instant case by affirming the reasons of Linden J.A. (at p. 454): 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. 224 On a final note, I am buttressed in my conclusion by the fact that the "reading in" approach has already been used in Canada to remedy underinclusive opposite-sex definitions of "spouse" in benefit-conferring legislation. In Knodel, supra, Rowles J. (as she then was) was faced with a remedial question identical to that before us (involving the underinclusiveness of s. 2.01 of the Medical Services Act Regulations, B.C. Reg 144/68, as amended). She elected not to strike down under s. 52(1). Her remedy was to use s. 52(1) to insert same-sex couples into the statutory definition. It is also interesting to note that in Haig v. Canada (1991), 5 O.R. (3d) 245 (Gen. Div.), aff'd (1992), 9 O.R. (3d) 495 (C.A.), and Vriend v. Alberta, supra, courts read "sexual orientation" into human rights legislation. In fact, in Haig the Ontario Court of Appeal remarked (at p. 508) that it was inconceivable . . . that Parliament would have preferred no human rights Act over one that included sexual orientation as a prohibited ground of discrimination. To believe otherwise would be a gratuitous insult to Parliament. 225 Given this judicial authority, I do not hesitate to follow Rowles J.'s approach to read same-sex couples into a statutory definition according benefits to spouses. Nor did Lamer C.J. in Schachter hesitate to look approvingly upon the Knodel decision, as is demonstrated from the following passage, which is directly applicable to the case at bar, at pp. 711-12: 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 couples. She considered this course to be far less intrusive to the intention of the legislature than striking down the benefits to heterosexual spouses since the group to be added was much smaller than the group already benefitted (at p. 391): In the present case, it would clearly be far more intrusive to strike the legislation and deny the benefits to the individuals receiving them than it would be to extend the benefits to the small minority who demonstrated their entitlement to them. . . . Where the group to be added is smaller than the group originally benefitted, this is an indication that the assumption that the legislature would have enacted the benefit in any case is a sound one. 226 The grant of the one-year suspension flows from the fact that the extension of the spousal allowance, while certainly a legal issue, is also a concern of public policy. In this respect, some latitude ought to be given to Parliament to address the issue and devise its own approach to ensuring that the spousal allowance be distributed in a manner that conforms with the equality guarantees of the Charter. I am particularly concerned by the fact that, as mentioned in the earlier discussion regarding the GAIN legislation, the clash between federal and provincial approaches to same-sex eligibility for programs can result (and presently results) in many gay and lesbian households "falling through the cracks". Making incremental changes to the allocation of benefits to same-sex spouses without any effort at co-ordinating the responses of all governments to these changes can result in the very real possibility of economic loss befalling such couples. 227 To this end, according the federal government a one-year period in which to amend the legislation will give both levels of government time to co-ordinate and harmonize their approaches to same-sex benefits, bearing in mind the reality that distinctions based on sexual orientation run the very real risk of offending the Charter. This "grace period" can ensure that a consistent approach to same-sex benefits be developed. However, in specifying this outer limit to the suspensive declaration, I draw once again from the words of the Chief Justice in Schachter, at p. 716: 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. 228 The appellants also seek an individual remedy, namely the retroactive receipt of the spousal allowance. I would not grant this relief. Under the circumstances, the appropriate remedy is to grant the suspensive declaration as outlined above. Conclusions and Disposition 229 For the reasons set out by Cory J., I find that the exclusion of same-sex couples from eligibility for the spousal allowance amounts to a violation of s. 15 of the Charter because the denial of the benefit is rooted in an irrelevant distinction based upon sexual orientation, which is an analogous ground of discrimination. Further, for the reasons outlined above, the impugned legislation is not saved under s. 1 of the Charter. I would allow the appeal and set aside the judgment of the Federal Court of Appeal with costs throughout. 230 I would thus answer the constitutional questions as follows: Question: 1. Does the definition of "spouse" in s. 2 of the Old Age Security Act, R.S.C., 1985, c. O-9, infringe or deny s. 15(1) of the Canadian Charter of Rights and Freedoms? Answer: Yes. Question: 2. If the answer to question 1 is "yes", is the infringement or denial demonstrably justified in a free and democratic society pursuant to s. 1 of the Canadian Charter of Rights and Freedoms? Answer: No. 231 One year from now, barring Parliamentary activity to ensure the constitutionality of s. 2 of the Old Age Security Act, the following definition of the term "spouse" will be inserted into the legislation by this Court: "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. McLACHLIN J.: 232 I am in substantial agreement with the reasons of Cory and Iacobucci JJ. Applying the principles which I discuss in Miron v. Trudel (No. 22744), I would allow the appeal. Appeal dismissed, L'Heureux-Dube, Cory, McLachlin and Iacobucci JJ. dissenting. Solicitors for the appellants: Arvay Finlay, Victoria. Solicitor for the respondent: The Department of Justice, Ottawa. Solicitor for the intervener the Attorney General of Quebec: Madeleine Aube, Sainte-Foy. Solicitor for the intervener the Canadian Human Rights Commission: William F. Pentney, Ottawa. Solicitor for the intervener the Commission des droits de la personne du Quebec: Helene Tessier, Montreal. Solicitor for the intervener Equality for Gays and Lesbians Everywhere: Cynthia Petersen, Ottawa. Solicitors for the intervener Metropolitan Community Church of Toronto: Iler, Campbell, Toronto. Solicitors for the intervener Inter-faith Coalition on Marriage and the Family: Lerner & Associates, Toronto. Solicitors for the intervener the Canadian Labour Congress: Sack Goldblatt Mitchell, Toronto. -- Julio Ducat P. 1962-1995