M. v. H., [1999] 2 SCR 3

M. v. H., [1999] 2 S.C.R. 3

The Attorney General for Ontario                                                    Appellant

v.

M.                                                                                                       Respondent

and

H.                                                                                                        Respondent

and

The Foundation for Equal Families, the Women’s

Legal Education and Action Fund (LEAF),

Equality for Gays and Lesbians Everywhere (EGALE),

the Ontario Human Rights Commission,

the United Church of Canada, the Evangelical

Fellowship of Canada, the Ontario Council of Sikhs,

the Islamic Society of North America,

Focus on the Family and REAL Women of Canada                        Interveners

Indexed as:  M. v. H.

File No.:  25838.

1998:  March 18; 1999:  May 20.

Present:  Lamer C.J. and L’Heureux‑Dubé, Gonthier, Cory, McLachlin, Iacobucci, Major, Bastarache and Binnie JJ.

on appeal from the court of appeal for ontario

Constitutional law -- Charter of Rights  -- Equality rights -- Definition of “spouse” -- Family Law Act extending right to seek support to members of unmarried opposite-sex couples -- Whether failure to provide same rights to members of same-sex couples infringes right to equality -- If so, whether infringement justified -- Canadian Charter of Rights and Freedoms, ss. 1 , 15(1)  -- Family Law Act, R.S.O. 1990, c. F.3, s. 29.

Family law -- Spousal support -- Definition of “spouse” -- Family Law Act extending right to seek support to members of unmarried opposite-sex couples -- Whether failure to provide same rights to members of same-sex couples infringes right to equality guaranteed by Canadian Charter of Rights and Freedoms  -- Family Law Act, R.S.O. 1990, c. F.3, s. 29.

M. and H. are women who lived together in a same-sex relationship beginning in 1982.  During that time they occupied a home which H. had owned since 1974.  In 1982, M. and H. started their own advertising business.  The business enjoyed immediate success and was the main source of income for the couple during the relationship.  H.’s contribution to this company was greater than that of M.  In 1983, M. and H. purchased a business property together.  In 1986, they purchased as joint tenants a vacation property in the country.  They later sold the business property and used the proceeds to finance the construction of a home on the country property.  As a result of a dramatic downturn in the advertising business in the late 1980s, the parties’ debt increased significantly.  H. took a job outside the firm and placed a mortgage on her home to pay for her expenses and those of M.  By the fall of 1992, M. and H.’s relationship had deteriorated.  M. left the common home and sought an order for partition and sale of the house and other relief.  M. later amended her application to include a claim for support pursuant to the provisions of the Family Law Act (“FLA”), and served notice of a constitutional question challenging the validity of the definition of “spouse” in s. 29, which includes a person who is actually married and also “either of a man and woman who are not married to each other and have cohabited . . . continuously for a period of not less than three years”.  Section 1(1) defines “cohabit” as “to live together in a conjugal relationship, whether within or outside marriage”.  The motions judge held that s. 29 of the FLA offends s. 15(1)  of the Canadian Charter of Rights and Freedoms , and that it is not saved by s. 1.  She declared that the words “a man and woman” were to be read out of the definition of “spouse” in s. 29 and replaced with the words “two persons”.   H. appealed the decision and was joined in the appeal by the Attorney General for Ontario.  The Court of Appeal upheld the decision, without costs, but suspended implementation of the declaration of invalidity for one year, to give the Ontario legislature time to amend the FLA.  Neither M. nor H. appealed this decision.  Leave to appeal to this Court was ultimately granted to the Attorney General for Ontario on the condition that M.’s costs were to be paid regardless of the outcome.  M. was also granted the right to cross‑appeal with respect to the Court of Appeal’s one‑year suspension of the declaration, and its refusal to award costs.

Held (Gonthier J. dissenting on the appeal):  The appeal and the cross-appeal should be dismissed, but the remedy should be modified.  Section 29 of the FLA is declared of no force or effect.  The effect of that declaration is temporarily suspended for a period of six months.

1.  Constitutional Issue

Per Lamer C.J. and L’Heureux-Dubé, Cory, McLachlin, Iacobucci and Binnie JJ.:  The proper approach to analyzing a claim of discrimination under s. 15(1)  of the Charter , as set out in Law, requires the court to make the following three broad inquiries.  First, does the impugned law (a) draw a formal distinction between the claimant and others on the basis of one or more personal characteristics, or (b) fail to take into account the claimant’s already disadvantaged position within Canadian society resulting in substantively differential treatment between the claimant and others on the basis of one or more personal characteristics?  If so, there is differential treatment for the purpose of s. 15(1).  Second, was the claimant subject to differential treatment on the basis of one or more of the enumerated and analogous grounds?  And third, does the differential treatment discriminate in a substantive sense, bringing into play the purpose of s. 15(1)  of the Charter  in remedying such ills as prejudice, stereotyping, and historical disadvantage?

The FLA draws a distinction by specifically according rights to individual members of unmarried cohabiting opposite-sex couples, which by omission it fails to accord to individual members of cohabiting same-sex couples.  The legislature drafted s. 29 to allow either a man or a woman to apply for support, thereby recognizing that financial dependence can arise in an intimate relationship in a context entirely unrelated either to child rearing or to any gender‑based discrimination existing in our society.  The obligation to provide spousal support was extended to include relationships which exist between a man and a woman, have a specific degree of permanence and are conjugal.   Since gay and lesbian individuals are capable of being involved in conjugal relationships, and since their relationships are capable of meeting the FLA’s temporal requirements, the distinction of relevance to this appeal is between persons in an opposite-sex, conjugal relationship of some permanence and persons in a same‑sex, conjugal relationship of some permanence.  It is thus apparent that the legislation has drawn a formal distinction between the claimant and others, on the basis of a personal characteristic, namely sexual orientation.  Sexual orientation has already been determined to be an analogous ground to those enumerated in s. 15(1)  of the Charter .

The central question of this appeal is whether the differential treatment imposed by the impugned legislation on an enumerated or analogous ground is discriminatory within the meaning of s. 15(1).  This inquiry is to be undertaken in a purposive and contextual manner, focussing on whether the differential treatment imposes a burden upon or withholds a benefit from the claimant in a manner that reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration. 

Section 29 of the FLA creates a distinction that withholds a benefit from the respondent M.  The type of benefit salient to the s. 15(1) analysis cannot encompass only the conferral of an economic benefit, but must also include access to a process that could confer an economic or other benefit.  Further, the spousal support provisions of the FLA help protect the economic interests of individuals in intimate relationships.

This denial of a benefit violates the purpose of s. 15(1).  One factor which may demonstrate that legislation that treats the claimant differently has the effect of demeaning the claimant’s dignity is the existence of pre-existing disadvantage, stereotyping, prejudice, or vulnerability experienced by the individual or group at issue.  In this case, there is significant pre-existing disadvantage and vulnerability and these circumstances are exacerbated by the impugned legislation.  The legislative provision in question draws a distinction that prevents persons in a same‑sex relationship from gaining access to the court‑enforced and ‑protected support system.  The denial of that potential benefit, which may impose a financial burden on persons in the position of the claimant, contributes to the general vulnerability experienced by individuals in same‑sex relationships.  A second relevant factor is that the legislation at issue fails to take into account the claimant’s actual situation.  Being in a same‑sex relationship does not mean that it is an impermanent or a non‑conjugal relationship.  Third, while the existence of an ameliorative purpose or effect may help to establish that human dignity is not violated where the person or group that is excluded is more advantaged with respect to the circumstances addressed by the legislation, the allegedly ameliorative purpose of this legislation does not do anything to lessen the charge of discrimination in this case.   Fourth, the nature of the interest protected by s. 29 of the FLA is fundamental.  The exclusion of same-sex partners from the benefits of s. 29 promotes the view that M., and individuals in same-sex relationships generally, are less worthy of recognition and protection.  It implies that they are judged to be incapable of forming intimate relationships of economic interdependence as compared to opposite-sex couples, without regard to their actual circumstances.  Such exclusion perpetuates the disadvantages suffered by individuals in same‑sex relationships and contributes to the erasure of their existence.

The first stage of the justification test under s. 1  of the Charter , as outlined in Oakes, asks whether the legislation limiting a Charter  right furthers a pressing and substantial objective.  Where a law violates the Charter  owing to underinclusion, this stage is properly concerned with the object of the legislation as a whole, the impugned provisions of the Act, and the omission itself.  The purpose of the FLA (Parts I to IV) is to provide for the equitable resolution of economic disputes that arise when intimate relationships between individuals who have been financially interdependent break down.  This is also one of the objectives of the impugned spousal support provisions in Part III; the other objective is to alleviate the burden on the public purse by shifting the obligation to provide support for needy persons to those parents and spouses who have the capacity to provide support to these individuals.  Neither the FLA in general, nor Part III in particular, was designed to remedy the disadvantages suffered by women in opposite-sex relationships.  Moreover, since the spousal support provisions in Part III are not primarily concerned with the protection of children, this is not part of the objective of s. 29 of the FLA.  Providing for the equitable resolution of economic disputes when intimate relationships between financially interdependent individuals break down, and alleviating the burden on the public purse to provide for dependent spouses, are pressing and substantial objectives.  These objectives promote both social justice and the dignity of individuals, which are values that underlie a free and democratic society.

Even if it were accepted that Part III of the FLA is meant to address the systemic sexual inequality associated with opposite‑sex relationships, the required nexus between this objective and the chosen measures, required by the second stage of the s. 1 analysis, is absent in this case.  A gender‑neutral support system cannot be rationally connected to the goal of improving the economic circumstances of heterosexual women upon relationship breakdown.  In addition, there is no evidence to demonstrate that the exclusion of same‑sex couples from the spousal support regime of the FLA in any way furthers the objective of assisting heterosexual women.  Although there is evidence to suggest that same‑sex relationships are not typically characterized by the same economic and other inequalities which affect opposite‑sex relationships, this does not explain why the right to apply for support is limited to heterosexuals.  The infrequency with which members of same‑sex relationships find themselves in circumstances resembling those of many heterosexual women is no different from heterosexual men, who, notwithstanding that they tend to benefit from the gender‑based division of labour and inequality of earning power, have as much right to apply for support as their female partners.  The protection of children, even if it were accepted as an objective, also fails the rational connection test.  It would have to be concluded that the spousal support provisions in Part III of the FLA are simultaneously underinclusive and overinclusive.  They are overinclusive because members of opposite-sex couples are entitled to apply for spousal support irrespective of whether or not they are parents and regardless of their reproductive capabilities or desires.  They are also underinclusive since an increasing percentage of children are being conceived and raised by lesbian and gay couples as a result of adoption, surrogacy and donor insemination.  Nor is the exclusion of same-sex couples from s. 29 of the FLA rationally connected to the dual objectives of the spousal support provisions of providing for the equitable resolution of economic disputes that arise upon the breakdown of financially interdependent relationships and reducing the burden on the public purse.  If anything, the goals of the legislation are undermined by the impugned exclusion.  The inclusion of same‑sex couples in s. 29 of the FLA would in fact better achieve the objectives of the legislation while respecting the Charter  rights of individuals in same-sex relationships.

The appellant’s case also fails the minimal impairment branch of the second stage of the Oakes test.  The argument that the exclusion of same‑sex couples from s. 29 of the FLA minimally impairs M.’s s. 15 rights since reasonable alternative remedies are available where economic dependence does occur in such relationships cannot be accepted.  Compared to awards of spousal support, the remedies available under the equitable doctrine of unjust enrichment are less flexible, impose more onerous requirements on claimants, and are available under far narrower circumstances.  They therefore do not provide an adequate alternative to spousal support under the FLA.  The law of contract is an equally unacceptable alternative to the spousal support scheme under the FLA.  The voluntary assumption of mutual support obligations is not equivalent to a statutory entitlement to apply for a support order.  Moreover, these alternative regimes do not address the fact that exclusion from the statutory scheme has moral and societal implications beyond economic ones.  As no group will be disadvantaged by granting members of same‑sex couples access to the spousal support scheme under the FLA, the notion of deference to legislative choices in the sense of balancing claims of competing groups has no application here.  Moreover, government incrementalism cannot constitute a reason to show deference to the legislature in this case.

The impugned legislation also fails to survive the final branch of the s. 1 analysis.  Where, as here, the impugned measures actually undermine the objectives of the legislation it cannot be said that the deleterious effects of the measures are outweighed by the promotion of any laudable legislative goals, or by the salutary effects of those measures.

If the remedy adopted by the court below is allowed to stand, s. 29 of the FLA will entitle members of same‑sex couples who otherwise qualify under the definition of "spouse" to apply for spousal support.  However, any attempt to opt out of this regime by means of a cohabitation agreement provided for in s. 53 or a separation agreement set out in s. 54 would not be recognized under the Act.  As this option is available to opposite-sex couples, and protects the ability of couples to choose to order their own affairs in a manner reflecting their own expectations, “reading in” would in effect remedy one constitutional wrong only to create another, and thereby fail to ensure the validity of the legislation.  Severing s. 29 of the Act such that it alone is declared of no force or effect is thus the most appropriate remedy here.  This remedy should be temporarily suspended for a period of six months.

Per Major J.:  The purpose of s. 29 of the FLA, as stated by the majority, is to allow persons who become financially dependent on one another in the course of a lengthy “conjugal” relationship some relief from financial hardship resulting from the breakdown of that relationship.  The relationship at issue in this case meets those requirements.  The exclusion of same-sex couples from the scheme to determine and redress this financial dependence on the basis of their sexual orientation denies them equal benefit of the law contrary to s. 15(1)  of the Charter .  In order to dispose of the appeal it is unnecessary to consider whether other types of long-term relationships may also give rise to dependency and relief.  The statute’s categorical exclusion of an individual whose situation falls squarely within its mandate, and who might be entitled to its benefits and protection, denies that person the equal concern and respect to which every Canadian is entitled, and constitutes discrimination.  The exclusion at issue undermines the intention of the legislation, which was designed in part to reduce the demands on the public welfare system.  By leaving potentially dependent individuals without a means of obtaining support from their former partners, s. 29 burdens the public purse with their care.  It is therefore not rationally connected to the valid aims of the legislation and cannot be justified under s. 1  of the Charter .

Per Bastarache J.:  Section 29 of the FLA has drawn a distinction between opposite-sex partners and same-sex partners in relationships of permanence.  The comparison is best made, not with married couples, whose status was consensually acquired, but with unmarried cohabiting couples.  Same-sex couples are capable of meeting all of the statutory prerequisites set out in ss. 29 and 1(1) of the FLA, but for the requirement that they be a man and a woman.  It is now well established that sexual orientation is a personal characteristic analogous to those found in s. 15(1)  of the Charter .  Discrimination exists because of the exclusion of persons from the regime on the basis of an arbitrary distinction, sexual orientation.  Since this exclusion suggests that their union is not worthy of recognition or protection, there is a denial of equality within the meaning of s. 15 .

There is no need to be deferential to the legislative choice in this case.  The nature of the interest affected by the exclusion is fundamental; the group affected is vulnerable; it is possible to isolate the challenged provision from the complex legislative scheme; there is no evidence of the government establishing priorities or arbitrating social needs; the legislative history indicates that there was no consideration given to the Charter  right to equal concern and respect; and the government’s interest in setting social policies can be met without imposing a burden on non-traditional families.  The traditional Oakes test should therefore be strictly applied.

The infringement is not justifiable under s. 1.  Since the s. 1 analysis places a burden on the government to justify the legislative incursion on the Charter  right, it is appropriate that governmental intention should, where possible, be considered and evaluated on its own terms to explain why the restriction on a Charter  right is justifiable.  The “limit” on the equality right in this case is the failure to treat individuals in same-sex relationships that otherwise meet the criteria of Part III of the FLA in the same way as individuals in opposite-sex relationships exhibiting similar characteristics.  The primary legislative purpose in extending support obligations outside the marriage bond was to address the subordinated position of women in non‑marital relationships.  The FLA fundamentally altered the nature of support obligations by extending them beyond marriage and therefore beyond the realm of consensual undertaking.  It did so because of a pressing social concern:  that many women found themselves in a legal vacuum at the end of a relationship; that their economic dependence was worsened as a result of those relationships; and that the economic position of women generally in society relative to men placed them in a position where they might be induced to enter or remain in those relationships without the legal safeguards of marriage even when they might otherwise want the mutual rights and obligations which would thereby be imposed.  The legislative purpose of the definition in Part III of the FLA is to impose support obligations upon partners in relationships in which they have consciously signalled a desire to be so bound (i.e. through marriage); and upon those partners in relationships of sufficient duration to indicate permanence and seriousness, and which involve the assumption of household responsibilities, or other career or financial sacrifices, by one partner for the common benefit of the couple and cause or enhance an economic disparity between the partners.

The need for imposition of support payments by one party to a relationship to the other, in the case of traditional family relationship breakdown, is a pressing and substantial objective in Canadian society.  It is clear that with respect to at least one category captured by the government’s legislative purposes, gendered relationships of some permanence, there is a high likelihood of serious economic detriment to women on the breakdown of those relationships.  The justification for legislative intervention affecting the autonomy of heterosexual couples does not however explain the pressing need to exclude all other family relationships from the governmental regime.  Although same-sex couples do not generally share the imbalance in power that is characteristic of opposite-sex couples and which causes economic dependency in the course of a relationship, there is no evidence that their inclusion would cause any particular difficulty.  The context in which the s. 1 analysis is made is also one where there is ascription of family status to same-sex relationships in society in general, as in some legislation and government policies.  In order to be consistent with Charter  values, the purpose of the definition in s. 29 must be respectful of the equality of status and opportunity of all persons.  It would be consistent with Charter  values of equality and inclusion to treat all members in a family relationship equally and all types of family relationships equally.

Even if one were to accept the argument that the exclusion is serving a valid purpose by not imposing on same-sex couples a reduction in freedom and autonomy that is mandated by economic imperatives largely irrelevant to same-sex couples, there would be no rational connection between that purpose and the total exclusion dictated by s. 29.  The exclusion of same-sex couples is not a valid means of achieving the positive purpose of s. 29, economic equality within the family.  When, as here, the exclusion specifically detracts from the general legislative purpose, the objective of the restriction cannot be considered pressing and substantial.  Even if the primary purpose of s. 29 was simply to recognize and promote the traditional family, and not to secure economic equality within the couple, which could be considered simply a means to an end, the exclusion of same‑sex partners could not be demonstrably justified.  Denial of status and benefits to same‑sex partners does not a priori enhance respect for the traditional family, nor does it reinforce the commitment of the legislature to the values in the Charter .  No evidence was adduced showing any beneficial impact of the exclusion on society, or what Charter  values would be served by the exclusion; but the detrimental effects clearly exist, both for the individual without recourse to the family law regime, and for society, faced with the prospect of giving social assistance to that aggrieved individual in need.  As for the protection of the freedom and autonomy of persons engaged in a same-sex relationship, s. 29 will only affect those who are in fact in a situation of economic imbalance analogous to that which more commonly occurs in the case of heterosexual relationships.  The entitlement resulting from a wider definition of “spouse” does not create an absolute right to support.  The justification for interference with personal autonomy is therefore the same for same-sex partners and opposite-sex partners.

Per Gonthier J. (dissenting):  The appropriate analytical approach to be taken to claims arising under s. 15(1)  of the Charter  was set out in this Court’s decision in Law.  When applied in this case, that approach leads to the conclusion that s. 29 of the FLA does not infringe s. 15(1)  of the Charter .

This legislation is an exception to the general rule that the law imposes no obligation of support as between persons.  This exception has been historically imposed on married couples, and this legislation is the current-day expression of this historical exception.  Section 29 of the FLA now imposes the burden of a mandatory support regime on a discrete group of people: opposite-sex married couples and certain opposite-sex cohabiting couples.  An analysis of the contextual factors set out in Law reveals that this differentiation is not discriminatory.  As this Court explained in Law, throughout the s. 15(1) analysis, it is necessary to have regard to the purpose of the legislation.  Here, the primary purpose of s. 29 of the FLA is to recognize the social function specific to opposite-sex couples and their position as a fundamental unit in society, and to address the dynamic of dependence unique to men and women in opposite-sex relationships.  In addition to this specific social function, this dynamic of dependence stems from the roles regularly taken by one member of that relationship, the biological reality of the relationship, and the pre-existing economic disadvantage that is usually, but not exclusively, suffered by women.  This purpose is apparent from the text and legislative history of the provision, and the preamble to the statute.

The statute’s preamble refers to the desirability of encouraging and strengthening the role of the family and recognizing marriage as a form of partnership.  The statements made in the legislature when amendments were introduced to extend support obligations to certain unmarried cohabiting opposite-sex couples indicate that these were premised on the social reality that such relationships exhibit a dynamic of dependence, which often arises because the couple has children and the mother is the primary caregiver.  They also confirm that the legislation was not aimed at “need” abstractly defined, or need generated by “interdependence”, but rather addressed the specific need arising when an individual (typically a woman) cohabits in an opposite-sex relationship and, relying upon the expectation of continued and future support from her partner, foregoes existing employment prospects that rendered her self-sufficient before entering into the relationship, and new ones that arise during the relationship.  

The use of gender-neutral language in the FLA provides no support for the implication that the purpose of the FLA’s spousal support provisions was to address “interdependence” unrelated to the social reality of the relationship.  When the amendments were first introduced, the tenor of the times favoured gender-neutral language.  The legislature recognized that in rare cases, a man might be able to make out a claim against a woman, but it expected that the vast majority of claimants would be women.

This proper understanding of the purpose reveals that the s. 15(1) claim must fail.  The impugned legislative provision has drawn a distinction between the claimant and others.  Sexual orientation is recognized as an analogous ground, and even though s. 29 of the FLA does not draw a facial distinction on the basis of sexual orientation, the effect of s. 29's definition of “spouse” is to do so.  Moreover, that distinction has resulted in disadvantage to the claimant.  Part III undoubtedly bestows a benefit -- access to a spousal support regime -- the denial of which results in a disadvantage to M.  At the same time, it imposes a burden on individuals in opposite-sex relationships, as it restricts their financial freedom and liberties.  The distinction drawn by s. 29 does not discriminate because it does not involve the stereotypical application of presumed group or personal characteristics and it does not otherwise have the effect of perpetuating or promoting the view that individuals in same-sex relationships are less deserving of concern, respect, and consideration.

The concept of “stereotyping” is often linked to the contextual factor set out in Law of “correspondence”.  Where the legislation takes into account the actual need, capacity or circumstances of the claimant and the comparator, then it is unlikely to rest on a stereotype.  The jurisprudence recognizes that in the absence of reliance on a stereotype, discrimination will be difficult to establish.  Distinctions drawn on enumerated or analogous grounds usually rely upon the stereotypical application of presumed characteristics, rather than an accurate account of the true situation, or actual abilities, circumstances, or capacities.  Where the legislation takes into account the claimant’s actual situation in a manner that respects his or her human dignity, it will be less likely to infringe s. 15(1).

Our system of family law is, to a great degree, based upon the legal rights and duties flowing from marriage.  While the legislature has restricted the meaning of “spouse” in Part III of the FLA, the restriction has not been made on the basis of stereotypical assumptions regarding group or personal characteristics.  On the contrary, s. 29's definition of “spouse” corresponds with an accurate account of the actual needs, capacity and circumstances of opposite-sex couples as compared to others, including same-sex couples.  This appeal is concerned with the support obligation, which is an essential feature of marriage itself, and more recently, of certain other opposite-sex couples.  Cohabiting opposite-sex couples are the natural and most likely site for the procreation and raising of children.  This is their specific, unique role.  This unique social role of opposite-sex couples has two related features.  First, women bear a disproportionate share of the child care burden in Canada, and this burden is borne both by working mothers and mothers who stay at home with their children.  Section 29 of the FLA is specifically structured to meet this reality.  This can be seen from the fact that the necessary cohabitation period to come within the scope of the section is reduced where there is a child of the relationship.  The second feature is that one partner (most often the woman) tends to become economically dependent upon the other.  Even in the absence of children, women in cohabiting opposite-sex relationships often take on increased domestic responsibilities which limit their prospects for outside employment, precisely because their lower average earnings make this an efficient division of labour for the couple.  If the relationship breaks down, the woman is usually left in a worse situation, probably with impaired earning capacity and limited employment opportunities.  Finally, even when the female partner is not suffering from this dynamic of dependence, the male partner often is.  When the woman is the primary wage-earner, the man adopts a primary role in household responsibilities.  This, coupled with other forms of dependency that follow for men in such relationships, are all reflective of a dynamic of dependence unique to opposite-sex couples.

It is this dynamic of dependence that the legislature has sought to address by way of Part III of the FLA.  The restricted definition of “spouse” in Part III of the FLA has not been made on the basis of stereotypical assumptions regarding group or personal characteristics.  The Charter does not compel the extension of the legislature’s efforts to address this problem to long-term same-sex couples.  While long-term same sex relationships may manifest many of the features of long-term opposite-sex relationships, the same dynamic of dependence is not present.  Lesbian relationships are characterized by a more even distribution of labour, a rejection of stereotypical gender roles, and a lower degree of financial interdependence than is prevalent in opposite-sex relationships.  Mere need in an individual case, unrelated to systemic factors, is insufficient to render the FLA’s scheme constitutionally underinclusive.  This is especially true because the scheme involves, in counterpart to an access to support, a restriction on freedom and a burden.

While individuals must be treated with equal respect and must not be discriminated against on the basis of the stereotypical application of irrelevant personal characteristics, the state is not barred from recognizing that some relationships fulfil different social roles and have specific needs, and responding to this reality with positive measures to address those differences.   

None of the remaining contextual factors set out in Law establish that the legislation perpetuates or promotes the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration.  Although the claimant is a member of a group that suffers pre-existing, historical disadvantage, the claimant’s group is relatively advantaged in relation to the subject-matter of the legislation.  Individuals in same-sex relationships do not carry the same burden of fulfilling the social role that those in opposite-sex relationships do.  They do not exhibit the same degree of systemic dependence.  They do not experience a structural wage differential between the individuals in the relationship.  In this sense, individuals in same-sex relationships are an advantaged group as compared to individuals in opposite-sex relationships, and there is thus no need to consider whether the legislation aggravates or exacerbates any pre-existing disadvantage.  Nor can it be said that the ameliorative legislation excludes a group which is disadvantaged in relation to the subject-matter of the legislation.  The main targets of the ameliorative legislation, partners in opposite-sex relationships (particularly women), suffer a structural disadvantage which is unknown to individuals in same-sex relationships.  With respect to the nature and scope of the affected interest,  the non-inclusion of individuals in same-sex relationships in the mandatory support regime does not result in severe and localized consequences.  While the legislature does not force individuals in same-sex relationships to provide support, it also does not prevent them from doing so by way of contract or otherwise.  This may result in some additional expenses, but it is difficult to see how this possible expense results in discriminatory non-recognition of the group. 

A reasonable person in the circumstances of the claimant who has regard to all of these contextual factors would find that the restrictive definition of “spouse” in s. 29 of the FLA does not constitute a violation of his or her human dignity.  To the contrary, acknowledging individual personal traits is a means of fostering human dignity.  By recognizing individuality, and rejecting forced uniformity, the law celebrates differences, fostering the autonomy and integrity of the individual. 

2.  Costs

Since the Court of Appeal did not err in failing to grant costs to M., the successful party, no order should be made with respect to costs in that court.  Costs are a discretionary determination and absent any clear error, this Court should be loath to interfere.

Cases Cited

By Cory and Iacobucci JJ.

Applied:  Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; referred to:  Egan v. Canada, [1995] 2 S.C.R. 513; Miron v. Trudel, [1995] 2 S.C.R. 418; Thibaudeau v. Canada, [1995] 2 S.C.R. 627; Forget v. Quebec (Attorney General), [1988] 2 S.C.R. 90; Borowski v. Canada (Attorney General), [1989] 1 S.C.R. 342; Molodowich v. Penttinen (1980), 17 R.F.L. (2d) 376; Vriend v. Alberta, [1998] 1 S.C.R. 493; R. v. Oakes, [1986] 1 S.C.R. 103; Eldridge v. British Columbia (Attorney General), [1997] 3 S.C.R. 624; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; R. v. Butler, [1992] 1 S.C.R. 452; RJR‑MacDonald Inc. v. Canada (Attorney General), [1995] 3 S.C.R. 199; Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877; Moge v. Moge, [1992] 3 S.C.R. 813; Pettkus v. Becker, [1980] 2 S.C.R. 834; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Tétreault‑Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22; Dagenais v. Canadian Broadcasting Corp., [1994] 3 S.C.R. 835;

Schachter v. Canada, [1992] 2 S.C.R. 679; Attorney‑General for Alberta v. Attorney‑General for Canada, [1947] A.C. 503; Canadian Pacific Ltd. v. Matsqui Indian Band, [1995] 1 S.C.R. 3.

By Bastarache J.

Applied:  Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; referred to:  Vriend v. Alberta, [1998] 1 S.C.R. 493; Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877; Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; Ross v. New Brunswick School District No. 15, [1996] 1 S.C.R. 825; R. v. Keegstra, [1990] 3 S.C.R. 697; R. v. Butler, [1992] 1 S.C.R. 452; Egan v. Canada, [1995] 2 S.C.R. 513; Pettkus v. Becker, [1980] 2 S.C.R. 834; Peter v. Beblow, [1993] 1 S.C.R. 980; Winnipeg Child and Family Services (Northwest Area) v. G. (D.F.), [1997] 3 S.C.R. 925; R. v. Edwards Books and Art Ltd., [1986] 2 S.C.R. 713; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Romer v. Evans, 116 S.Ct. 1620 (1996); Heydon’s Case (1584), 3 Co. Rep. 7a, 76 E.R. 637; Pepper v. Hart, [1993] A.C. 593; R. v. Oakes, [1986] 1 S.C.R. 103; R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295; R. v. Morgentaler, [1988] 1 S.C.R. 30; Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; Eldridge v. British Columbia (Attorney General), [1997] 3 S.C.R. 624; Benner v. Canada (Secretary of State), [1997] 1 S.C.R. 358; Miron v. Trudel, [1995] 2 S.C.R. 418; Moge v. Moge, [1992] 3 S.C.R. 813; Symes v. Canada, [1993] 4 S.C.R. 695.

By Gonthier J. (dissenting)

Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; Reference re Authority to Perform Functions Vested by the Adoption Act, [1938] S.C.R. 398; Taylor v. Rossu (1998), 161 D.L.R. (4th) 266; R. v. Heywood, [1994] 3 S.C.R. 761; Vriend v. Alberta, [1998] 1 S.C.R. 493; Athabasca Tribal Council v. Amoco Canada Petroleum Co., [1981] 1 S.C.R. 699; Canada Labour Relations Board v. City of Yellowknife, [1977] 2 S.C.R. 729; R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295; Moge v. Moge, [1992] 3 S.C.R. 813; Egan v. Canada, [1995] 2 S.C.R. 513; Miron v. Trudel, [1995] 2 S.C.R. 418; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143; Andrews v. Ontario (Minister of Health) (1988), 64 O.R. (2d) 258; Dean v. District of Columbia, 653 A.2d 307 (1995); Layland v. Ontario (Minister of Consumer and Commercial Relations) (1993), 104 D.L.R. (4th) 214; Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219; Edwards v. Attorney-General for Canada, [1930] A.C. 124; R. v. Turpin, [1989] 1 S.C.R. 1296; Symes v. Canada, [1993] 4 S.C.R. 695; Grant v. South-West Trains Ltd., [1998] I.C.R. 449; Thibaudeau v. Canada, [1995] 2 S.C.R. 627; Corbière v. Canada (Minister of Indian and Northern Affairs), [1999] 2 S.C.R. 203.

Statutes and Regulations Cited

Absentees Act, R.S.O. 1990, c. A.3.

Act respecting certain separate rights of property of Married Women, C.S.U.C. 1859, c. 73.

Act to establish a Court of Chancery in this Province, S.U.C. 1837, 7 Wm. IV, c. 2, s. 3.

Adoption Act, R.S.B.C. 1996, c. 5, s. 5(1).

Bankruptcy and Insolvency Act , R.S.C., 1985, c. B-3  [am. 1992, c. 27], ss. 121(4) [ad. 1997, c. 12, s. 87], 136(1)(d.1) [idem, s. 90], 178(1)(c) [rep. & sub. c. 3 (2nd Supp.), s. 28].

Business Corporations Act, R.S.O. 1990, c. B.16.

Canadian Charter of Rights and Freedoms , ss. 1 , 2( a ) , (b), (d), 7 -14 , 15(1) .

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

Change of Name Act, R.S.O. 1990, c. C.7.

Child and Family Services Act, R.S.O. 1990, c. C.11.

Consent to Treatment Act, 1992, S.O. 1992, c. 31.

Constitution Act, 1867 , ss. 91(26) , 92(12) , (13) .

Constitution Act, 1982 , s. 52 .

Coroners Act, R.S.O. 1990, c. C.37.

Criminal Code , R.S.C., 1985, c. C-46 , s. 718.2 (a)(i) [ad. 1995, c. 22, s. 6].

Deserted Wives’ and Children’s Maintenance Act, R.S.O. 1937, c. 211 [am. 1954, c. 22; 1958, c. 23].

Deserted Wives’ and Children’s Maintenance Act, R.S.O. 1960, c. 105.

Deserted Wives’ and Children’s Maintenance Act, R.S.O. 1970, c. 128 [am. 1971, c. 98; 1973, c. 133].

Divorce Act , R.S.C., 1985, c. 3 (2nd Supp .), s. 15(2), (3), (7).

Domestic Relations Act, R.S.A. 1980, c. D-37.

Election Act, R.S.O. 1990, c. E.6.

Estates Act, R.S.O. 1990, c. E.21.

Evidence Act, R.S.O. 1990, c. E.23.

Execution Act, R.S.O. 1990, c. E.24.

Family Law Act, R.S.N. 1990, c. F-2, s. 35(c).

Family Law Act, R.S.O. 1990, c. F.3, preamble, Part I, s. 1(1) “cohabit”, “spouse”, (2),  Parts II, III, ss. 29 “spouse”, 30, 31, 32, 33(1), (2), (7), (8), (9), 34(4), (5), 35, 41, Part IV, ss. 53, 54, Part V.

Family Law Act, S.N.W.T. 1997, c. 18, s. 1.

Family Law Act, S.P.E.I. 1995, c. 12, ss. 1(g), 29.

Family Law Act, 1986, S.O. 1986, c. 4, s. 29.

Family Law Reform Act, 1975, S.O. 1975, c. 41.

Family Law Reform Act, 1978, S.O. 1978, c. 2, s.14.

Family Maintenance Act, R.S.M. 1987, c. F20, s. 4(3).

Family Maintenance Act, R.S.N.S. 1989, c. 160, s. 2(m).

Family Maintenance Act, S.S. 1990-91, c. F-6.1, s. 2(l)(iii).

Family Maintenance Enforcement Amendment Act, 1997, S.B.C. 1997, c. 19, s. 1(d).

Family Property and Support Act, R.S.Y. 1986, c. 63, s. 35.

Family Relations Act, R.S.B.C. 1996, c. 128, s. 1.

Family Relations Amendment Act, 1997, S.B.C. 1997, c. 20, s. 1(c).

Family Responsibility and Support Arrears Enforcement Act, 1996, S.O. 1996, c. 31.

Family Services Act, S.N.B. 1980, c. F-2.2, s. 112(3).

Family Support Plan Act, R.S.O. 1990, c. S.28 [am. 1991, c. 5 , s. 1(1)].

Income Tax Act, R.S.O. 1990, c. I.2.

Interpretation Act, R.S.O. 1990, c. I.11, ss. 8, 17.

Juries Act, R.S.O. 1990, c. J.3.

Land Transfer Tax Act, R.S.O. 1990, c. L.6.

Marriage Act, R.S.O. 1990, c. M.3.

Married Woman’s Real Estate Act, 1873, S.O. 1873, c. 18.

Married Women’s Property Act, R.S.O. 1970, c. 262.

Married Women’s Property Act, S.O. 1872, c. 16.

Married Women’s Property Act, S.O. 1884, c. 19.

Matrimonial Causes Act, 1931, S.O. 1931, c. 25 [subsequently R.S.O. 1970, c. 265].

Members’ Conflict of Interest Act, R.S.O. 1990, c. M.6.

Mental Health Act, R.S.O. 1990, c. M.7.

Municipal Act, R.S.O. 1990, c. M.45.

Municipal Elections Act, R.S.O. 1990, c. M.53.

Nunavut Act , S.C. 1993, c. 28 , s. 29(1)  [am. 1998, c. 15, s. 4].

Nursing Homes Act, R.S.O. 1990, c. N.7.

Ontario Guaranteed Annual Income Act, R.S.O. 1990, c. O.17.

Partnerships Act, R.S.O. 1990, c. P.5.

Pension Benefits Act, R.S.O. 1990, c. P.8.

Rent Control Act, 1992, S.O. 1992, c. 11.

Retail Sales Tax Act, R.S.O. 1990, c. R.31.

Rules of Civil Procedure, R.R.O. 1990, Reg. 194, Rules 20, 21.

Securities Act, R.S.O. 1990, c. S.5.

Small Business Development Corporations Act, R.S.O. 1990, c. S.12.

Succession Law Reform Act, R.S.O. 1990, c. S.26.

Trustee Act, R.S.O. 1990, c. T.23.

Workers’ Compensation Act, R.S.O. 1990, c. W.11.

Authors Cited

Blackstone, William.  Commentaries on the Laws of England, 4th ed., Book I.  Oxford:  Clarendon Press, 1770.

Blumstein, Philip, and Pepper Schwartz.  American Couples.  New York:  Morrow, 1983.

Canada.  Citizenship and Immigration Canada.  Building on a Strong Foundation for the 21st Century:  New Directions for Immigration and Refugee Policy and Legislation.  Ottawa:  Citizenship and Immigration Canada, 1998.

Canada.  Statistics Canada.  Family Expenditure in Canada 1992.  Ottawa:  Statistics Canada, 1994.

Canada.  Statistics Canada.  Labour and Household Surveys Analysis Division.  Family Income After Separation.  Ottawa:  Statistics Canada, 1997.

Canada.  Statistics Canada.  Infomat: A Weekly Review, October 17, 1997. “More lone-parent and common-law families”, p. 4.

Cardell, Mona, Stephen Finn and Jeanne Marecek.  “Sex-Role Identity, Sex-Role Behavior, and Satisfaction in Heterosexual, Lesbian, and Gay Male Couples”, Psychology of Women Quarterly, 5(3) (Spring 1981): 488-94.

Chambers, Lori.  Married Women and Property Law in Victorian Ontario.  Toronto:  University of Toronto Press, 1997.

Cossman, Brenda, and Bruce Ryder.  Gay, Lesbian and Unmarried Heterosexual Couples and the Family Law Act:  Accommodating a Diversity of Family Forms, a Research Paper prepared for the Ontario Law Reform Commission, June 1993.

Cretney, S. M., and J. M. Masson.  Principles of Family Law, 6th ed.  London:  Sweet & Maxwell, 1997.

Dworkin, Ronald.  Freedom’s Law:  The Moral Reading of the American Constitution.  Cambridge, Mass.:  Harvard University Press, 1996.

Eldridge, Natalie S., and Lucia A. Gilbert.  “Correlates of Relationship Satisfaction in Lesbian Couples”, Psychology of Women Quarterly, 14 (1990): 43-62.

Ely, John Hart.  Democracy and Distrust:  A Theory of Judicial Review.  Cambridge, Mass.:  Harvard University Press, 1980.

Evans, R. R.  The Law and Practice Relating to Divorce and Other Matrimonial Causes.  Calgary:  Burroughs, 1923.

Herek, Gregory M.  “Myths About Sexual Orientation:  A Lawyer’s Guide to Social Science Research” (1991), 1 Law & Sexuality 133.

Hogg, Peter W.  Constitutional Law of Canada, loose-leaf ed., vol. 2.  Scarborough, Ont.:  Carswell, 1992 (updated 1997, release 2).

Hogg, Peter W., and Allison A. Bushell.  “The Charter Dialogue Between Courts and Legislatures (Or Perhaps The Charter of Rights Isn’t Such A Bad Thing After All)” (1997), 35 Osgoode Hall L.J. 75.

Jackman, Martha.  “Protecting Rights and Promoting Democracy:  Judicial Review Under Section 1 of the Charter ” (1996), 34 Osgoode Hall L.J. 661.

Lynch, Jean M., and Mary Ellen Reilly.  “Role Relationships:  Lesbian Perspectives”, Journal of Homosexuality, 12(2) (Winter 1985/86): 53-69.

MacDougall, Donald J.  “Alimony and Maintenance”.  In D. Mendes da Costa, ed., Studies in Canadian Family Law, vol. 1.  Toronto:  Butterworths, 1972, 283.

McCamus, John D.  “Family Law Reform in Ontario”.  In Special Lectures of the Law Society of Upper Canada 1993:  Family Law:  Roles, Fairness and Equality.  Scarborough, Ont.:  Carswell, 1994, 451.

Mendes, Errol P.  “The Crucible of the Charter :  Judicial Principles v. Judicial Deference in the Context of Section 1”.  In Gérald-A. Beaudoin and Errol P. Mendes, The Canadian Charter of Rights and Freedoms , 3rd ed.  Scarborough, Ont.:  Carswell, 1996, 3-1.

Ontario.  Law Reform Commission.  Report on Family Law, Part IV, “Family Property Law”.  Toronto:  Ministry of the Attorney General, 1974.

Ontario.  Law Reform Commission.  Report on Family Law, Part VI, “Support Obligations”.  Toronto:  Ministry of the Attorney General, 1975.

Ontario.  Law Reform Commission.  Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act.  Toronto:  The Commission, 1993.

Ontario.  Legislative Assembly.  Official Report of Debates, 3rd Sess., 35th Parl., June 1, 1994, p. 6583.

Ontario.  Legislature of Ontario Debates, 3rd Sess., 30th Parl., October 26, 1976, pp. 4102-3.

Ontario.  Legislature of Ontario Debates, 3rd Sess., 30th Parl., November 18, 1976, pp. 4793, 4801.

Ontario.  Legislature of Ontario Debates, 3rd Sess., 30th Parl., November 22, 1976, pp. 4890-91, 4898.

Ontario.  Legislature of Ontario Debates, 1st Sess., 31st Parl., October 18, 1977, pp. 901, 904.

Ontario.  Ministry of the Attorney General.  Family Law Reform.  Toronto:  Ministry of the Attorney General, 1976.

Peplau, Letitia Anne.  “Lesbian and Gay Relationships”.  In John C. Gonsiorek and James D. Weinrich, eds., Homosexuality:  Research Implications for Public Policy.  Newbury Park:  Sage, 1991, 177.

Schneider, Margaret S.  “The Relationships of Cohabiting Lesbian and Heterosexual Couples:  A Comparison”, Psychology of Women Quarterly, 10 (1986):  234-39.

Slater, Suzanne.  The Lesbian Family Life Cycle.  New York:  Free Press, 1995.

Strauss, Peter L.  “The Courts and the Congress:  Should Judges Disdain Political History?” (1998), 98 Colum. L. Rev. 242.

Sullivan, Ruth.  Driedger on the Construction of Statutes, 3rd ed.  Toronto:  Butterworths, 1994.

APPEAL and CROSS-APPEAL from a judgment of the Ontario Court of Appeal (1996), 31 O.R. (3d) 417, 142 D.L.R. (4th) 1, 96 O.A.C. 173, 25 R.F.L. (4th) 116, 40 C.R.R. (2d) 240, [1996] O.J. No. 4419 (QL), affirming a decision of the Ontario Court (General Division) (1996), 27 O.R. (3d) 593, 132 D.L.R. (4th) 538, 17 R.F.L. (4th) 365, 35 C.R.R. (2d) 123, [1996] O.J. No. 365 (QL), declaring s. 29 of the Family Law Act to be unconstitutional.  Appeal dismissed, Gonthier J. dissenting.  Cross-appeal dismissed.

Robert E. Charney and Peter C. Landmann, for the appellant.

Martha A. McCarthy and Lynn D. Lovell, for the respondent M.

Christopher D. Bredt, for the respondent H.

R. Douglas Elliott, for the intervener the Foundation for Equal Families.

Carol A. Allen, for the intervener the Women’s Legal Education and Action Fund (LEAF).

Cynthia Petersen and Pam McEachern, for the intervener Equality for Gays and Lesbians Everywhere (EGALE).

Joanne D. Rosen, for the intervener the Ontario Human Rights Commission.

Jeff G. Cowan, for the intervener the United Church of Canada.

Peter R. Jervis, Michael Meredith and Danielle Shaw, for the interveners the Evangelical Fellowship of Canada, the Ontario Council of Sikhs, the Islamic Society of North America and Focus on the Family.

David M. Brown, for the intervener REAL Women of Canada.

The judgment of Lamer C.J. and L’Heureux-Dubé, Cory, McLachlin, Iacobucci and Binnie JJ. was delivered by

Cory and Iacobucci JJ. ‑‑

I. Introduction and Overview

1                                   The principal issue raised in this appeal is whether the definition of “spouse” in s. 29 of the Family Law Act, R.S.O. 1990, c. F.3 (“FLA”) infringes s. 15(1)  of the Canadian Charter of Rights and Freedoms , and, if so, whether the legislation is nevertheless saved by s. 1  of the Charter .  In addition, M. was granted leave to cross‑appeal on the issue of the appropriate remedy to be granted and also as to costs.

2                                   Our view on this principal issue may be summarized as follows. Section 15(1)  of the Charter  is infringed by the definition of “spouse” in s. 29 of the FLA.  This definition, which only applies to Part III of the FLA, draws a distinction between individuals in conjugal, opposite-sex relationships of a specific degree of duration and individuals in conjugal, same-sex relationships of a specific degree of duration. We emphasize that the definition of “spouse” found in s. 1(1) of the FLA, and which applies to other parts of the FLA, includes only married persons and is not at issue in this appeal.  Essentially, the definition of “spouse” in s. 29 of the FLA extends the obligation to provide spousal support, found in Part III of the FLA,  beyond married persons to include individuals in conjugal opposite-sex relationships of some permanence. Same-sex relationships are capable of being both conjugal and lengthy, but individuals in such relationships are nonetheless denied access to the court-enforced system of support provided by the FLA. This differential treatment is on the basis of a personal characteristic, namely sexual orientation, that, in previous jurisprudence, has been found to be analogous to those characteristics specifically enumerated in s.15(1).

3                                   The crux of the issue is that this differential treatment discriminates in a substantive sense by violating the human dignity of individuals in same-sex relationships.  As Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497, established, the inquiry into substantive discrimination is to be undertaken in a purposive and contextual manner. In the present appeal, several factors are important to consider. First, individuals in same-sex relationships face significant pre-existing disadvantage and vulnerability, which is exacerbated by the impugned legislation. Second, the legislation at issue fails to take into account the claimant’s actual situation. Third, there is no compelling argument that the ameliorative purpose of the legislation does anything to lessen the charge of discrimination in this case. Fourth, the nature of the interest affected is fundamental, namely the ability to meet basic financial needs following the breakdown of a relationship characterized by intimacy and economic dependence. The exclusion of same-sex partners from the benefits of the spousal support scheme implies that they are judged to be incapable of forming intimate relationships of economic interdependence, without regard to their actual circumstances. Taking these factors into account, it is clear that the human dignity of individuals in same-sex relationships is violated by the definition of “spouse” in s. 29 of the FLA.

4                                   This infringement is not justified under s.1  of the Charter  because there is no rational connection between the objectives of the spousal support provisions and the means chosen to further this objective. The objectives were accurately identified by Charron J.A., in the court below, as providing for the equitable resolution of economic disputes when intimate relationships between financially interdependent individuals break down, and alleviating the burden on the public purse to provide for dependent spouses. Neither of these objectives is furthered by the exclusion of individuals in same-sex couples from the spousal support regime. If anything, these goals are undermined by this exclusion.

5                                   In this case, the remedy of reading in is inappropriate, as it would unduly recast the legislation, and striking down the FLA as a whole is excessive.  Therefore the appropriate remedy is to declare s. 29 of no force and effect and to suspend the application of the declaration for a period of six months.

6                                   In our elaboration of this position in these joint reasons, Cory J. has addressed the issues of mootness and the breach of s. 15(1)  of the Charter .  Iacobucci J. has addressed s. 1  of the Charter , the appropriate remedy, costs and the disposition.

Cory J. ‑‑

7                                   At the outset, it must be stressed that the questions to be answered are narrow and precise in their scope.  The FLA provides a means whereby designated persons may apply to the court for support from a spouse or, if unmarried, from a man or woman with whom they lived in an opposite‑sex conjugal relationship.  The Act specifically extends  the obligation for support beyond married persons who, as a result of their married status, have additional rights under the Act.

8                                   The question to be resolved is whether the extension of the right to seek support to members of unmarried opposite‑sex couples infringes s. 15(1)  of the Charter  by failing to provide the same rights to members of same‑sex couples.

II.  Factual Background

9                                   M. and H. are women who met while on vacation in 1980.  It is agreed that in 1982 they started living together in a same‑sex relationship that continued for at least five years.  That relationship may have lasted ten years, but that figure is disputed by H.  During that time they occupied a home which H. had owned since 1974.  H. paid for the upkeep of the home, but the parties agreed to share living expenses and household responsibilities equally.  At the time, H. was employed in an advertising firm and M. ran her own company.

10                               In 1982, M. and H. started their own advertising business.  The business enjoyed immediate success and was the main source of income for the couple during the relationship.  H.’s contribution to this company was greater than that of M.  Epstein J., the trial judge, observed that this disparity was probably due to the fact that M. had no previous experience in advertising, and, as time went on, she was content to devote more of her time to domestic tasks rather than the business.  Nevertheless, the parties continued to be equal shareholders in the company.

11                               In 1983, M. and H. purchased a business property together.  In 1986, they purchased as joint tenants a vacation property in the country.  They later sold the business property and used the proceeds to finance the construction of a home on the country property.

12                               As a result of a dramatic downturn in the advertising business in the late 1980s, the parties’ debt increased significantly.  H. took a job outside the firm and placed a mortgage on her home to pay for her expenses and those of M.  M. also tried to find employment but was unsuccessful.  Her company, which she had continued to run on a casual basis throughout the relationship, produced very little income.

13                               By September of 1992, M. and H.’s relationship had deteriorated.  H. was concerned about what she perceived to be an unfair disparity in their relative financial contributions.  H. presented M. with a draft agreement to settle their affairs.  The same day that the agreement was presented, M. took some of her personal belongings and left the common home.  Upon M.’s departure, H. changed the locks on the house.

14                               The parties did not divide the personal property or household contents.  M. alleged that she encountered serious financial problems after the separation.  In October 1992, M. sought an order for partition and sale of the house; a declaration that she was the beneficial owner of certain lands and premises owned by H. and by the companies M. named as defendants; and an accounting of the transactions carried out by the companies.  By Notice of Cross‑Application, H. and the corporate defendants sought damages for slander of title, partition and sale of property, the repayment of certain loans, and other relief.  M. then amended her application to include a claim for support pursuant to the provisions of the FLA, and served Notice of a Constitutional Question challenging the validity of the definition of “spouse” in s. 29 of the Act.

15                               H. brought a motion under Rule 20 of the Ontario Rules of Civil Procedure, R.R.O. 1990, Reg. 194, for summary judgment or, alternatively, for the determination of a question of law under Rule 21.  The motions were heard by Epstein J., who dismissed the motion for summary judgment in February 1994, but adjourned the Rule 21 motion until this Court’s judgment in Egan v. Canada, [1995] 2 S.C.R. 513, was released.

16                               In February 1996, Epstein J. released her judgment on the constitutional issues.  She held that s. 29 of the FLA offends s. 15(1)  of the Charter , and that it is not saved by s. 1.  H. appealed the judgment and was joined in the appeal by the intervener, the Attorney General for Ontario.

17                               The Ontario Court of Appeal ultimately upheld this decision, but suspended implementation of the declaration of invalidity for one year, to give the Ontario legislature time to amend the FLA.  Neither of the respondents appealed this decision.  Leave to appeal to this Court was ultimately granted to the Attorney General for Ontario on the condition that M.’s costs were to be paid regardless of the outcome.  M. was also granted the right to cross‑appeal with respect to the Court of Appeal’s one‑year suspension of the declaration, and the issue of costs.

18                               Shortly before the appeal was heard in this Court, M. and H. concluded a settlement of the financial issues raised in the proceedings.

III. Relevant Statutory Provisions

19                               Family Law Act, R.S.O. 1990, c. F.3

1.--(1) In this Act,

                                                                    ...

“cohabit” means to live together in a conjugal relationship, whether within or outside marriage;

                                                                   . . .

“spouse” means either of a man and woman who

(a) are married to each other, or

(b) have together entered into a marriage that is voidable or void, in good faith on the part of the person asserting a right under this Act.

(2)  In the definition of “spouse”, a reference to marriage includes a marriage that is actually or potentially polygamous, if it was celebrated in a jurisdiction whose system of law recognizes it as valid.

29.  In this Part,

                                                                    ...

“spouse” means a spouse as defined in subsection 1 (1), and in addition includes either of a man and woman who are not married to each other and have cohabited,

(a)       continuously for a period of not less than three years, or

(b)       in a relationship of some permanence, if they are the natural or adoptive parents of a child.

30.  Every spouse has an obligation to provide support for himself or herself and for the other spouse, in accordance with need, to the extent that he or she is capable of doing so.

31.--(1) Every parent has an obligation to provide support, in accordance with need, for his or her unmarried child who is a minor or is enrolled in a full time program of education, to the extent that the parent is capable of doing so.

(2) The obligation under subsection (1) does not extend to a child who is sixteen years of age or older and has withdrawn from parental control.

33.  (1) A court may, on application, order a person to provide support for his or her dependants and determine the amount of support.

(2) An application for an order for the support of a dependant may be made by the dependant or the dependant's parent.

                                                                   . . .

(7) An order for the support of a child should,

(a)  recognize that each parent has an obligation to provide support for the child;

(b) recognize that the obligation of a natural or adoptive parent outweighs the obligation of a parent who is not a natural or adoptive parent; and

(c) apportion the obligation according to the capacities of the parents to provide support.

(8) An order for the support of a spouse should,

(a)  recognize the spouse's contribution to the relationship and the economic consequences of the relationship for the spouse;

(b)  share the economic burden of child support equitably;

(c)  make fair provision to assist the spouse to become able to contribute to his or her own support; and

(d)  relieve financial hardship, if this has not been done by orders under Parts I (Family Property) and II (Matrimonial Home).

(9) In determining the amount and duration, if any, of support in relation to need, the court shall consider all the circumstances of the parties, including,

(a)       the dependant's and respondent's current assets and means;

(b)       the assets and means that the dependant and respondent are likely to have in the future;

(c)       the dependant's capacity to contribute to his or her own support;

(d)       the respondent's capacity to provide support;

(e)       the dependant's and respondent's age and physical and mental health;

(f)        the dependant's needs, in determining which the court shall have regard to the accustomed standard of living while the parties resided together;

(g)       the measures available for the dependant to become able to provide for his or her own support and the length of time and cost involved to enable the dependant to take those measures;

(h)       any legal obligation of the respondent or dependant to provide support for another person;

(i)        the desirability of the dependant or respondent remaining at home to care for a child;

(j)        a contribution by the dependant to the realization of the respondent's career potential;

                          (k)       if the dependant is a child,

(i) the child’s aptitude for and reasonable prospects of obtaining an       education, and

(ii) the child’s need for a stable environment;

(l)        if the dependant is a spouse,

(i)    the length of time the dependant and respondent cohabited,

(ii)   the effect on the spouse's earning capacity of the responsibilities assumed during cohabitation,

(iii)whether the spouse has undertaken the care of a child who is of             the age of eighteen years or over and unable by reason of illness,  disability or other cause to withdraw from the charge of his or her parents,

(iv)  whether the spouse has undertaken to assist in the continuation of a program of education for a child eighteen years of age or over who is unable for that reason to withdraw from the charge of his or her parents,

(v)   any housekeeping, child care or other domestic service performed by the spouse for the family, as if the spouse were devoting the time spent in performing that service in remunerative employment and were contributing the earnings to the family's support,

(vi)  the effect on the spouse's earnings and career development of the responsibility of caring for a child; and

(m)      any other legal right of the dependant to support, other than out of public money.

                                                                   . . .

53.-- (1) A man and a woman who are cohabiting or intend to cohabit and who are not married to each other may enter into an agreement in which they agree on their respective rights and obligations during cohabitation, or on ceasing to cohabit or on death, including,

(a) ownership in or division of property;

(b) support obligations;

(c)       the right to direct the education and moral training of their children, but not the right to custody of or access to their children; and

(d) any other matter in the settlement of their affairs.

(2) If the parties to a cohabitation agreement marry each other, the agreement shall be deemed to be a marriage contract.

54.  A man and a woman who cohabited and are living separate and apart may enter into an agreement in which they agree on their respective rights and obligations, including,

(a) ownership in or division of property;

(b) support obligations;

(c) the right to direct the education and moral training of their children;

(d) the right to custody of and access to their children; and

(e) any other matter in the settlement of their affairs.

Divorce Act ,   R.S.C., 1985, c. 3 (2nd Supp .)

15. ...

 (2) A court of competent jurisdiction may, on application by either or both spouses, make an order requiring one spouse to secure or pay, or to secure and pay, such lump sum or periodic sums, or such lump sum and periodic sums, as the court thinks reasonable for the support of

(a) the other spouse;

(b) any or all children of the marriage; or

(c) the other spouse and any or all children of the marriage.

(3) Where an application is made under subsection (2), the court may, on application by either or both spouses, make an interim order requiring one spouse to secure or pay, or to secure and pay, such lump sum or periodic sums, or such lump sum and periodic sums, as the court thinks reasonable for the support of

(a) the other spouse,

(b) any or all children of the marriage, or

(c) the other spouse and any or all children of the marriage,

pending determination of the application under subsection (2).

                                                                   . . .

(7) An order made under this section that provides for the support of a spouse should

(a) recognize any economic advantages or disadvantages to the spouses arising from the marriage or its breakdown;

(b) apportion between the spouses any financial consequences arising  from the care of any child of the marriage over and above the obligation  apportioned between the spouses pursuant to subsection (8);

(c) relieve any economic hardship of the spouses arising from the breakdown of the marriage; and

(d) in so far as practicable, promote the economic self‑sufficiency of each spouse within a reasonable period of time.

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.

IV.  Judgments Below

A.  Ontario Court (General Division) (1996), 27 O.R. (3d) 593

20                               Epstein J. found it was neither necessary nor appropriate to determine whether financial dependency actually existed between M. and H. Instead, she held that there was sufficient evidence to support the conclusion reached by Iacobucci J. in Egan, supra, at para. 197, that same‑sex relationships are capable of giving rise to economic interdependence.

21                               Turning to consider the Charter  issues, Epstein J. noted that all parties conceded that s. 29 of the FLA violated s. 15(1)  of the Charter .  Nevertheless, the parties had submitted ample evidence to allow her to conduct a full analysis of the s. 15(1) issue on its merits. Citing this Court’s decisions in Egan, Miron v. Trudel, [1995] 2 S.C.R. 418, and Thibaudeau v. Canada, [1995] 2 S.C.R. 627, she followed a three‑step approach to s. 15(1):  (1) does the law draw a distinction between the claimant and others; (2) if so, does this distinction result in the imposition of a burden, obligation or disadvantage not placed on others or the failure to provide a benefit given to others; (3) is the distinction based on a personal characteristic enumerated in s. 15(1) or an analogous ground?

22                               Following this approach, Epstein J. concluded first that s. 29 of the FLA draws a distinction between opposite‑sex partners and same‑sex partners.  She further held that the exclusion of same‑sex couples from the definition of “spouse” in s. 29 denies a dependent same‑sex spouse the benefit and protection of the law that is afforded a dependent opposite‑sex spouse, as well as the choice of being publicly recognized as a common‑law couple.  Epstein J., relying on this Court’s holding in Egan that sexual orientation is an analogous ground, ultimately concluded that the definition of “spouse” in s. 29 of the FLA violates s. 15(1)  of the Charter .

23                               Epstein J. also noted that some members of this Court would, as part of the s. 15 analysis, determine whether the distinction made by the legislation was relevant to the functional values underlying the law.  She found that the support provisions of the FLA were intended to address the economic consequences of the breakdown of a relationship of some permanence and held that the exclusion of same‑sex couples from s. 29 is not relevant to those values.  She concluded therefore that under either analytical approach, s. 29 of the FLA contravened s. 15(1)  of the Charter .

24                               Epstein J. then turned to consider whether the legislation is saved by s. 1  of the Charter .  She found that the spousal support provisions in the FLA are intended to protect those who become economically dependent in a relationship marked by marriage or intimate cohabitation, and who require assistance in becoming self‑sufficient upon the breakdown of that relationship.  She held that the objective of the provisions was pressing and substantial.

25                               Turning next to the proportionality test, Epstein J. reasoned that the objective of the provisions is not attained by denying otherwise eligible persons access to a means of redress on the basis of their sexual orientation.  She therefore held that the exclusion of same‑sex couples is not rationally connected to the legislative objective.  She reasoned that to justify this exclusion, those who seek to uphold the legislation would have to show that economically interdependent family units typically involve opposite‑sex couples, and that economically interdependent relationships between same‑sex partners are an anomaly.  Epstein J. found that the evidence before her did not discharge this onus.  Further, she rejected the Attorney General’s argument that M.’s s. 15 rights were minimally impaired because she was still fully entitled to litigate her property claims.  She found that other forms of redress are not relevant to a Charter  claim:  “discrimination is discrimination, regardless of what the aggrieved person may be able to do to ease the pain” (p. 613).

26                               In evaluating next the proportionality between the effects of the legislation and its objective, Epstein J. concluded that the importance of providing an opportunity for some couples to claim support does not justify the infringement of the equality rights of other couples who are prevented from doing so for constitutionally irrelevant reasons.  Epstein J. further noted that while some people may hold strong feelings about traditional family forms, granting same‑sex couples access to the courts to pursue benefit claims will not affect the formation of opposite‑sex unions.

27                               Epstein J. considered the issue of legislative deference.  She noted that unlike Egan, the possibility of increased demand for public funds is not an issue in this case.  Rather, extending the spousal support provisions to cover same‑sex couples results in decreased government expenditures because fewer people will need to look to social assistance upon the breakdown of a relationship.  She also found that the Ontario legislature made it clear that it cannot or will not move to redress this inequality.  She noted that a bill that would have made the changes requested by M. was voted down in 1994, and that the Attorney General for Ontario opposed such an extension in this case.  She concluded it was not realistic to regard the current state of Ontario law on this issue as part of a process of legislative reform.

28                               Epstein J. held that s. 29 of the FLA is of no force and effect to the extent that it excludes same‑sex couples from its definition of “spouse”.  She declared that the words “a man and woman” are to be severed from s. 29, and in their place the words “two persons” are to be read into the definition.

B.  Ontario Court of Appeal (1996), 31 O.R. (3d) 417

1.    Charron J.A., Doherty J.A. concurring

29                               At the outset, Charron J.A. determined that a consideration of the relevance of the discriminatory distinction to the functional values of the legislation is better left to the s. 1 analysis.  She adopted the approach to the s. 15(1) analysis set out by Cory J. in Egan, supra, and by McLachlin J. in Miron, supra.

30                               Charron J.A. found that the distinction created by the legislation was between members of same‑sex couples and members of unmarried opposite‑sex couples.  As such, the ability to marry was not at issue; instead, the inclusion of unmarried opposite‑sex couples in the spousal support scheme opened the door to M.’s inequality claim.

31                               Charron J.A. went on to conclude that same‑sex couples are capable of meeting all the requirements in ss. 1(1) and 29 of the FLA, except the requirement that a spouse must be “either of a man and woman”.  As such, she held that s. 29 draws a distinction based on sexual orientation.  In light of this Court’s finding in Egan that sexual orientation is an analogous ground, she held that this distinction must constitute discrimination.

32                               Turning to s. 1  of the Charter , Charron J.A. found that the general objective of the FLA was to provide for the equitable resolution of economic disputes that arise when intimate relationships between individuals who have become financially interdependent break down.  Charron J.A. found that the support provisions in Part III of the FLA were aimed at furthering this objective, and were also intended to alleviate the burden on the public purse by shifting the obligation to support needy persons from the state to “spouses”, as defined by the Act, who had the capacity to provide support.

33                               Charron J.A. also concluded that access to the spousal support scheme was extended beyond married couples because it was neither fair nor effective to choose marriage as the exclusive marker to identify those intimate relationships which give rise to economic interdependence or those individuals who should bear the burden of providing support to the other member of a failed relationship. Charron J.A. concluded that the underlying objective of both the legislation and the support provisions was clearly pressing and substantial.

34                               Charron J.A. next considered the proportionality test.  She noted that while the legislative provision could be found to be rationally connected to the objectives of the legislation, she also found it was incumbent on those supporting the legislation to show that the exclusion is also rationally connected to these objectives.  Charron J.A. could not identify any evidence to support this proposition.  Instead, she found that the inclusion of same‑sex cohabitants within the s. 29 definition would only serve to further the goals of the legislation in that a greater number of persons would have access to the FLA’s dispute‑resolution mechanism and fewer people would look to the public purse to meet their needs.

35                               Finally, Charron J.A. found that the provision does not minimally impair M.’s Charter  rights, and that alternatives substantially less invasive of Charter  rights might have been found.  She concluded there was no proportionality between the effect of the measure and its objective, and therefore held that the infringement of s. 15(1) was not justified under s. 1  of the Charter .

36                               Charron J.A. would have granted the same remedy as Epstein J. ‑‑ i.e. severing the words “a man and woman” from the s. 29 definition and adding in their place the words “two persons”.  However, Charron J.A. ordered that the declaratory remedy be temporarily suspended for one year to allow the legislature an opportunity to address the unconstitutionality of s. 29 of the FLA, and of other legislative provisions not before the court that might be affected by the ruling.

2.    Finlayson J.A., dissenting

37                               Finlayson J.A. began by holding that the distinction created by s. 29 is between opposite‑sex and same‑sex couples.  As such, he held that the provision must be scrutinized for any discriminatory impact on same‑sex couples and not on the individual members of such a couple.  Accordingly, Finlayson J.A. found that the motions judge made a serious error in finding that the impact of the legislative distinction could be considered with respect to M. only. He instead held that M. and H. had to be looked at collectively to determine whether, as a couple, they were denied equal protection and benefit of the law.

38                               Finlayson J.A. distinguished this case from Egan, supra, and Miron, supra, by noting that the FLA does not confer an economic gain on one member of a couple; rather, it simply provides for the possible redistribution of wealth within a couple. Moreover, in Egan, the two partners made a joint decision to apply for the pensioner’s spousal allowance; here, H. vigorously objects to being deemed M.’s spouse for the purpose of the FLA’s support provisions.  Finally, Finlayson J.A. held that, unlike Egan, it cannot be said that the FLA somehow sends a message that same‑sex relationships are less worthy of respect or support. He reasoned that the FLA did not offer monetary support to opposite‑sex relationships but, rather, attempted to make one member of the relationship responsible, to some degree, for the social costs of marriage.

39                               Accordingly, Finlayson J.A. held that, while the support provisions do treat same‑sex couples differently from opposite‑sex couples, same‑sex couples are not denied either the material benefits or the dignity accorded to other groups.  Cohabiting same‑sex couples simply cannot obtain the marital status accorded to opposite‑sex couples. Finlayson J.A. was not convinced that the formal attributes of marital status, unaccompanied by any material advantage, constituted a “benefit” of the law.

40                               Finlayson J.A. also found that the s. 15(1) issue in this case impacts directly on the societal concept of marriage, since he concluded that the definition of “spouse” in s. 29 of the FLA is designed to extend marital status to couples who have married formally or who cohabit as husband and wife.  Because he found that neither Egan nor Miron addressed this aspect of the s. 15(1) issue, Finlayson J.A. adopted the four‑judge dissent in Egan wherein La Forest J. addressed the meaning of marriage as a basic social institution.  For the same reasons, he concluded there was no s. 15(1) breach in this case, since the legislature chose to regulate opposite‑sex unions recognizing that they are the traditional and basic social structure for the procreation of children.  He held that the legislature recognized the historical fact of the interrelationship between child rearing and the dependency of the female spouse, and accordingly provided for mutual support obligations between spouses in this context.

41                               Finlayson J.A. went on to note that had he found a s. 15(1) violation, he would have upheld the validity of the provision under s. 1.  He accepted and applied this Court’s holding in Egan to the effect that governments must be permitted to act incrementally in social policy areas.  He also held that the court below characterized the purpose of the FLA’s support provisions too narrowly.  He noted that under s. 30 of the FLA, support obligations rest on both spouses and are not contingent on the breakdown of the relationship.  Further, entitlement to support is not conditional on proof that the applicant’s need is causally related to the relationship.  He concluded that the motions judge, in confining her Charter  analysis to a finding that the state denied M. a benefit, did not appreciate the nature or extent of the burden she was proposing to place on H.

42                               Finally, Finlayson J.A. noted that the term “cohabit” in s. 1(1) of the FLA  means “to live together in a conjugal relationship, whether within or outside marriage”.  He noted that the court below had not determined how same‑sex cohabitants could come within the definition of a “conjugal” couple, that is to say, two persons holding themselves out as husband and wife.  Similarly, he found the court below had failed to address the fact that Part IV of the FLA would not permit same‑sex cohabitants to opt out of the statutory support regime, leaving same‑sex couples in a different position than their opposite‑sex counterparts.  Finlayson J.A. would have allowed the appeal and declared s. 29 of the FLA constitutionally valid.

V.  Analysis

A.  Mootness

43                               Shortly before this appeal was heard in this Court, the respondents M. and H. reached a settlement of the financial issues that gave rise to this litigation.  The original parties to this dispute thus no longer have a vested interest in its outcome, and the appeal is moot so far as they are concerned.  However, leave to appeal was not granted to either of the original parties; rather, the Attorney General for Ontario alone sought and was granted leave to appeal the judgment of the court below.  This case is similar to Forget v. Quebec (Attorney General), [1988] 2 S.C.R. 90, and the conclusions reached by Lamer J. (as he then was) in that case at p. 97 are apposite and determinative of this issue:

... even if the appeal were pointless so far as respondent is concerned, it is not respondent who is appealing the Court of Appeal judgment.  It is the Attorney General of Quebec who is appealing from that decision:  the question of whether the issue is moot is determined in light of the appellant’s interests.  If this Court refuses to address the issue, the Court of Appeal judgment declaring void the relevant sections . . . will stand. . . . Moreover, these are generally applicable provisions: the problem of discrimination does not affect respondent alone, but may arise in respect of every professional candidate.  In my view, therefore, the issue is not moot so far as Quebec is concerned and it is our duty to consider it on its merits.

44                               Moreover, even if the appeal were moot, it would be appropriate for the Court to exercise its discretion in order to decide these important issues.  The social cost of leaving this matter undecided would be significant.  The appeal has come to this Court in an adversarial context.  The record is ample and complete and all points of view were very well presented.  A consideration of all these factors confirms that it would be appropriate for this Court to exercise its discretion to hear this appeal.  See Borowski v. Canada (Attorney General), [1989] 1 S.C.R. 342, at pp. 361‑62.

B.  Does Section 29 of the FLA Infringe Section 15(1) of the Charter ?

45                               The Attorney General for Ontario, displaying great candour, very fairly conceded that s. 29 of the FLA contravenes the provisions of s. 15  of the Charter .  His entire argument was directed at demonstrating that the section was nonetheless justifiable and saved by s. 1  of the Charter .  The Court is certainly not bound by this concession. Although, in my view, he was correct in taking this position, it would not be appropriate in this appeal to undertake only a s. 1 analysis without considering whether s. 15  has in fact been violated.  The s. 15(1) issue in this case is important not only to the parties but also to many Canadians.  It was the subject of extensive submissions by the respondent H. and many of the interveners.

1.    Approach to Section 15(1)

46                               In the recent decision of this Court in Law, supra, Iacobucci J. summarized some of the main guidelines for analysis under s. 15(1) to be derived from the jurisprudence of this Court.  He emphasized that these guidelines do not represent a strict test, but rather should be understood as points of reference for a court that is called upon to decide whether a claimant’s right to equality without discrimination under the Charter  has been infringed:  see para. 88.

47                               Iacobucci J. explained that the s. 15(1) equality guarantee is to be interpreted and applied in a purposive and contextual manner, in order to permit the realization of the provision’s strong remedial purpose, and to avoid the pitfalls of a formalistic or mechanical approach.  Following a review of this Court’s jurisprudence regarding the fundamental purpose of s. 15(1), he stated this purpose in the following terms, at para. 88:

In general terms, the purpose of s. 15(1) is to prevent the violation of essential human dignity and freedom through the imposition of disadvantage, stereotyping, or political or social prejudice, and to promote a society in which all persons enjoy equal recognition at law as human beings or as members of Canadian society, equally capable and equally deserving of concern, respect and consideration.

Iacobucci J. stated that the existence of a conflict between the purpose or effect of an impugned law, on the one hand, and this fundamental purpose of the equality guarantee, on the other, is essential in order to found a discrimination claim.

48                               In Law, Iacobucci J. reviewed various articulations of the proper approach to be taken in analyzing a s. 15(1) claim, as expressed in the jurisprudence of this Court.  At para. 39, he summarized the basic elements of this Court’s approach as involving three broad inquiries, in the following terms:

In my view, the proper approach to analyzing a claim of discrimination under s. 15(1)  of the Charter  involves a synthesis of these various articulations.  Following upon the analysis in Andrews, supra, and the two‑step framework set out in Egan, supra, and Miron, supra, among other cases, a court that is called upon to determine a discrimination claim under s. 15(1) should make the following three broad inquiries.  First, does the impugned law (a) draw a formal distinction between the claimant and others on the basis of one or more personal characteristics, or (b) fail to take into account the claimant’s already disadvantaged position within Canadian society resulting in substantively differential treatment between the claimant and others on the basis of one or more personal characteristics?  If so, there is differential treatment for the purpose of s. 15(1).  Second, was the claimant subject to differential treatment on the basis of one or more of the enumerated and analogous grounds?  And third, does the differential treatment discriminate in a substantive sense, bringing into play the purpose of s. 15(1)  of the Charter  in remedying such ills as prejudice, stereotyping, and historical disadvantage? [Emphasis in original.]

2.    The Structure of the Family Law Act

49                               To begin, it may be useful to review briefly the structure of the FLA and the rights and obligations it establishes.  First and foremost, it is of critical importance to recognize that the FLA contains more than one definition of “spouse”.  The first definition is set out in s. 1(1) and includes only persons who are actually married, or who have entered into a void or voidable marriage in good faith.  This definition applies to all parts of the Act.

50                               The second definition is found in s. 29, and extends the meaning of “spouse”, but only for certain purposes.  Specifically, unmarried opposite‑sex couples who have cohabited for at least three years, or who are the natural or adoptive parents of a child and have also cohabited in a relationship of “some permanence”, bear a mutual obligation of support under Part III of the FLA.  They also have the right to enter into cohabitation agreements to regulate their relationship under Part IV, and may bring a claim for dependants’ relief in tort under Part V.

51                               All these rights and obligations are obviously available to married persons as well.  However, married persons have additional rights under the FLA that are denied common law cohabitants, even those who meet the requirements of s. 29.  Under Part I, a husband or wife may apply for an equal share of the wealth generated during the marriage, and of the matrimonial home.  Under Part II, both married spouses have a right to possession of the matrimonial home, regardless of who owns the property.  Moreover, the ability of the owner of the matrimonial home to sell or encumber the property without the consent of the other spouse is severely restricted.  These mutual rights and obligations are denied all unmarried opposite‑sex cohabitants.

52                               These observations on the structure of the FLA serve to emphasize that this appeal has nothing to do with marriage per se.  Much of the FLA is devoted solely to regulating the relationship that exists between married persons, or persons who intend to be married.  They alone are guaranteed certain property rights that are not extended to any unmarried persons.  In some specific instances ‑‑ such as Part III dealing with support obligations ‑‑ the legislature has seen fit to extend the rights and obligations that arise under the FLA beyond married persons to include certain unmarried persons as well.

53                               In other words, the FLA draws a distinction by specifically according rights to individual members of unmarried cohabiting opposite‑sex couples, which by omission it fails to accord to individual members of same‑sex couples who are living together.  It is this distinction that lies at the heart of the s. 15 analysis.  The rights and obligations that exist between married persons play no part in this analysis.  The legislature did not extend full marital status, for the purposes of all the rights and obligations under the FLA, to those unmarried cohabitants included in s. 29 of the Act.  Rather, the definition of “spouse” in s. 29 only applies for certain purposes.  Specifically, it allows persons who became financially dependent in the course of a lengthy intimate relationship some relief from financial hardship resulting from the breakdown of that relationship.  It follows that this provision was designed to reduce the demands on the public welfare system.  This will be discussed more fully in the s. 1 analysis below.

54                               It is true that women in common law relationships often tended to become financially dependent on their male partners because they raised their children and because of their unequal earning power.  But the legislature drafted s. 29 to allow either a man or a woman to apply for support, thereby recognizing that financial dependence can arise in an intimate relationship in a context entirely unrelated  either to child rearing or to any gender‑based discrimination existing in our society.  See discussion of s. 1  of the Charter , below.  Indeed, the special situation of financial dependence potentially created by procreation is specifically addressed in s. 29(b).  This appeal is concerned only with s. 29(a).  That section is aimed at remedying situations of dependence in intimate relationships without imposing any limitation relating to the circumstances that may give rise to that dependence.

55                               It is thus apparent that in this appeal there is no need to consider whether same‑sex couples can marry, or whether same‑sex couples must, for all purposes, be treated in the same manner as unmarried opposite‑sex couples.  The only determination that must be made is whether, in extending the spousal support obligations set out in Part III of the FLA to include unmarried men or women in certain opposite‑sex relationships, the legislature infringed the equality rights of men or women in similar same‑sex relationships, and if so, whether that infringement may be saved by s. 1  of the Charter .

3.    The Existence of Differential Treatment

56                               A consideration of the essence of M.’s claim requires a more detailed review of Part III of the FLA.  In Part III, ss. 30 to 32 impose an obligation on persons to support themselves and their dependants.  A “dependant” can be the spouse, child or parent of the person who must fulfil the support obligation.  The definition of “spouse” in s. 29 applies to all of Part III and includes a person who is actually married, and also:

. . . either of a man and woman who are not married to each other and have cohabited,

(a) continuously for a period of not less than three years, or

(b) in a relationship of some permanence, if they are the natural or adoptive parents of a child.

Section 1(1) defines “cohabit” as “to live together in a conjugal relationship, whether within or outside marriage”.

57                               The definition clearly indicates that the legislature decided to extend the obligation to provide spousal support beyond married persons.  Obligations to provide support were no longer dependent upon marriage.  The obligation was extended to include those relationships which:

(i)    exist between a man and a woman;

(ii)   have a specific degree of permanence;

(iii) are conjugal.

Only individuals in relationships which meet these minimum criteria may apply for a support order under Part III of the FLA.

58                               Same‑sex relationships are capable of meeting the last two requirements.  Certainly same‑sex couples will often form long, lasting, loving and intimate relationships.  The choices they make in the context of those relationships may give rise to the financial dependence of one partner on the other.  Though it might be argued that same‑sex couples do not live together in “conjugal” relationships, in the sense that they cannot “hold themselves out” as husband and wife, on this issue I am in agreement with the reasoning and conclusions of the majority of the Court of Appeal.

59                               Molodowich v. Penttinen (1980), 17 R.F.L. (2d) 376 (Ont. Dist. Ct.), sets out the generally accepted characteristics of a conjugal relationship.  They include shared shelter, sexual and personal behaviour, services, social activities, economic support and children, as well as the societal perception of the couple. However, it was recognized that these elements may be present in varying degrees and not all are necessary for the relationship to be found to be conjugal.  While it is true that there may not be any consensus as to the societal perception of same‑sex couples, there is agreement that same‑sex couples share many other “conjugal” characteristics.  In order to come within the definition, neither opposite‑sex couples nor same‑sex couples are required to fit precisely the traditional marital model to demonstrate that the relationship is “conjugal”.

60                               Certainly an opposite‑sex couple may, after many years together, be considered to be in a conjugal relationship although they have neither children nor sexual relations.  Obviously the weight to be accorded the various elements or factors to be considered in determining whether an opposite‑sex couple is in a conjugal relationship will vary widely and almost infinitely.  The same must hold true of same‑sex couples.  Courts have wisely determined that the approach to determining whether a relationship is conjugal must be flexible.  This must be so, for the relationships of all couples will vary widely.  In these circumstances, the Court of Appeal correctly concluded that there is nothing to suggest that same‑sex couples do not meet the legal definition of “conjugal”.

61                               Since gay and lesbian individuals are capable of being involved in conjugal relationships, and since their relationships are capable of meeting the FLA’s temporal requirements, the distinction of relevance to this appeal is between persons in an opposite‑sex, conjugal relationship of some permanence and persons in a same‑sex, conjugal relationship of some permanence.  In this regard, I must disagree with the dissenting opinion in the court below, which characterized the distinction arising in s. 29 as being between opposite‑sex and same‑sex couples.  This conclusion would require that the section be scrutinized for any discriminatory impact it may have on same‑sex couples, and not on the individual members of that couple.  Section 29 defines “spouse” as “either of a man and woman” who meet the other requirements of the section.  It follows that the definition could not have been meant to define a couple.  Rather it explicitly refers to the individual members of the couple.  Thus the distinction of relevance must be between individual persons in a same‑sex, conjugal relationship of some permanence and individual persons in an opposite‑sex, conjugal relationship of some permanence.

62                               Thus it is apparent that the legislation has drawn a formal distinction between the claimant and others, based on personal characteristics.  As stated in Law, supra, the first broad inquiry in the s. 15(1) analysis determines whether there is differential treatment imposed by the impugned legislation between the claimant and others.  It is clear that there is differential treatment here.  Under s. 29 of the FLA, members of opposite‑sex couples who can meet the requirements of the statute are able to gain access to the court‑enforced system of support provided by the FLA.  It is this system that ensures the provision of support to a dependent spouse.  Members of same‑sex couples are denied access to this system entirely on the basis of their sexual orientation.

4.    Sexual Orientation is an Analogous Ground

63                               Not every legislative distinction is discriminatory.  Before it can be found that it gives rise to discrimination, it must be shown that an equality right was denied on the basis of an enumerated or analogous ground, and that this differential treatment discriminates “in a substantive sense, bringing into play the purpose of s. 15(1)  of the Charter ”:  Law, supra, at para. 39 (emphasis in original).

64                               In Egan, supra, this Court unanimously affirmed that sexual orientation is an analogous ground to those enumerated in s. 15(1).  Sexual orientation is “a deeply personal characteristic that is either unchangeable or changeable only at unacceptable personal costs” (para. 5).  In addition, a majority of this Court explicitly recognized that gays, lesbians and bisexuals, “whether as individuals or couples, form an identifiable minority who have suffered and continue to suffer serious social, political and economic disadvantage”  (para. 175, per Cory J.; see also para. 89, per L’Heureux‑Dubé J.).

5.    The Existence of Discrimination in a Purposive Sense

65                               The determination of whether differential treatment imposed by legislation on an enumerated or analogous ground is discriminatory within the meaning of s. 15(1)  of the Charter  is to be undertaken in a purposive and contextual manner.  The relevant inquiry is whether the differential treatment imposes a burden upon or withholds a benefit from the claimant in a manner that reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration:  Law, supra, at para. 88.

66                               The respondent H. has argued that the differential treatment imposed by s. 29 of the FLA does not deny the respondent M. the equal benefit of the law since same‑sex spouses are not being denied an economic benefit, but simply the opportunity to gain access to a court‑enforced process.  Such an analysis takes too narrow a view of “benefit” under the law.  It is a view this Court should not adopt.  The type of benefit salient to the s. 15(1) analysis cannot encompass only the conferral of an economic benefit.  It must also include access to a process that could confer an economic or other benefit:  Egan, supra, at paras. 158-59; Vriend v. Alberta, [1998] 1 S.C.R. 493, at para. 87.  Further, the spousal support provisions of the FLA help protect the economic interests of individuals in intimate relationships.  When a relationship breaks down, the support provisions help to ensure that a member of a couple who has contributed to the couple’s welfare in intangible ways will not find himself or herself utterly abandoned.  This protective aspect of the spousal support provisions is properly considered in relation to s. 15(1).  Thus it is appropriate to conclude that s. 29 of the FLA creates a distinction that withholds a benefit from the respondent M.  The question is whether this denial of a benefit violates the purpose of s. 15(1).

67                               In Law, Iacobucci J. explained that there are  a variety of contextual factors that may be referred to by a s. 15(1) claimant in order to demonstrate that legislation demeans his or her dignity.  The list of factors is not closed, and there is no specific formula that must be considered in every case.  In Law itself, Iacobucci J. listed four important contextual factors in particular which may influence the determination of whether s. 15(1) has been infringed.  He emphasized, at paras. 59-61, that in examining these contextual factors, a court must adopt the point of view of a reasonable person, in circumstances similar to those of the claimant, who takes into account the contextual factors relevant to the claim.

68                               One factor which may demonstrate that legislation that treats the claimant differently has the effect of demeaning the claimant’s dignity is the existence of pre‑existing disadvantage, stereotyping, prejudice, or vulnerability experienced by the individual or group at issue.  As stated by Iacobucci J. in Law, supra, at para. 63:

As has been consistently recognized throughout this Court’s jurisprudence, probably the most compelling factor favouring a conclusion that differential treatment imposed by legislation is truly discriminatory will be, where it exists, pre‑existing disadvantage, vulnerability, stereotyping, or prejudice experienced by the individual or group [citations omitted]. These factors are relevant because, to the extent that the claimant is already subject to unfair circumstances or treatment in society by virtue of personal characteristics or circumstances, persons like him or her have often not been given equal concern, respect, and consideration.  It is logical to conclude that, in most cases, further differential treatment will contribute to the perpetuation or promotion of their unfair social characterization, and will have a more severe impact upon them, since they are already vulnerable.

69                               In this case, there is significant pre‑existing disadvantage and vulnerability, and these circumstances are exacerbated by the impugned legislation. The legislative provision in question draws a distinction that prevents persons in a same‑sex relationship from gaining access to the court‑enforced and ‑protected  support system.  This system clearly provides a benefit to unmarried heterosexual persons who come within the definition set out in s. 29, and thereby provides a measure of protection for their economic interests.  This protection is denied to persons in a same‑sex relationship who would otherwise meet the statute’s requirements, and as a result, a person in the position of the claimant is denied a benefit regarding an important aspect of life in today’s society.  Neither common law nor equity provides the remedy of maintenance that is made available by the FLA.  The denial of that potential benefit, which may impose a financial burden on persons in the position of the claimant, contributes to the general vulnerability experienced by individuals in same‑sex relationships.

70                               A second contextual factor that was discussed in Law as being potentially relevant to the s. 15(1) inquiry is the correspondence, or the lack of it, between the ground on which a claim is based and the actual need, capacity, or circumstances of the claimant or others:  para. 70. Iacobucci J. nonetheless cautioned that the mere fact that the impugned legislation takes into account the claimant’s actual situation will not necessarily defeat a s. 15(1) claim, as the focus of the inquiry must always remain upon the central question of whether, viewed from the perspective of the claimant, the differential treatment imposed by the legislation has the effect of violating human dignity.  However, the legislation at issue in the current appeal fails to take into account the claimant’s actual situation.  As I have already discussed, access to the court‑enforced spousal support regime provided in the FLA is given to individuals in conjugal relationships of a specific degree of permanence.  Being in a same‑sex relationship does not mean that it is an impermanent or a non‑conjugal relationship.

71                               A third contextual factor referred to by Iacobucci J. in Law, supra, at para. 72,  is the question of whether the impugned legislation has an ameliorative purpose or effect for a group historically disadvantaged in the context of the legislation:

An ameliorative purpose or effect which accords with the purpose of s. 15(1)  of the Charter  will likely not violate the human dignity of more advantaged individuals where the exclusion of these more advantaged individuals largely corresponds to the greater need or the different circumstances experienced by the disadvantaged group being targeted by the legislation. I emphasize that this factor will likely only be relevant where the person or group that is excluded from the scope of ameliorative legislation or other state action is more advantaged in a relative sense. Underinclusive  ameliorative legislation that excludes from its scope the members of an historically disadvantaged group will rarely escape the charge of discrimination: see Vriend, supra, at paras. 94-104, per Cory J.

In other words, the existence of an ameliorative purpose or effect may help to establish that human dignity is not violated where the person or group that is excluded is more advantaged with respect to the circumstances addressed by the legislation. Gonthier J. argues that the legislation under scrutiny in the present appeal is just such ameliorative legislation -- that it is meant to target women in married or opposite-sex relationships. He proceeds to argue that in this legal context, women in same-sex relationships are not similarly disadvantaged. For the reasons expressed elsewhere, we disagree with this characterization of the legislation. Accordingly, we reject the idea that the allegedly ameliorative purpose of this legislation does anything to lessen the charge of discrimination in this case.

72                               A fourth contextual factor specifically adverted to by Iacobucci J. in Law, at para. 74, was the nature of the interest affected by the impugned legislation.  Drawing upon the reasons of L’Heureux‑Dubé J. in Egan, supra, Iacobucci J. stated that the discriminatory calibre of differential treatment cannot be fully appreciated without considering whether the distinction in question restricts access to a fundamental social institution, or affects a basic aspect of full membership in Canadian society, or constitutes a complete non‑recognition of a particular group.  In the present case, the interest protected by s. 29 of the FLA is fundamental, namely the ability to meet basic financial needs following the breakdown of a relationship characterized by intimacy and economic dependence.  Members of same‑sex couples are entirely ignored by the statute, notwithstanding the undeniable importance to  them of the benefits accorded by the statute.

73                               The societal significance of the benefit conferred by the statute cannot be overemphasized.  The exclusion of same‑sex partners from the benefits of s. 29 of the FLA promotes the view that M., and individuals in same‑sex relationships generally, are less worthy of recognition and protection.  It implies that they are judged to be incapable of forming intimate relationships of economic interdependence as compared to opposite-sex couples, without regard to their actual circumstances.  As the intervener EGALE submitted, such exclusion perpetuates the disadvantages suffered by individuals in same‑sex relationships and contributes to the erasure of their existence.

74                               Therefore I conclude that an examination of the  four factors outlined above, in the context of the present appeal, indicate that the human dignity of individuals in same-sex relationships is violated by the impugned legislation.  In light of this, I conclude that the definition of “spouse” in s. 29 of the FLA violates s. 15(1).

Iacobucci J. ‑‑

C.  Is Section 29 of the FLA Justified Under Section 1 of the Charter ?

1.  Stare Decisis and Egan

75                               At the outset, I wish to address the appellant's submission that an independent examination of the s. 1 issues is unnecessary in the present case.  The appellant asserts that the principle of stare decisis binds this Court to the decision in Egan, supra,  and that the s. 1 analysis in that case ought to apply with equal force to the case at bar.  Although I recognize the fundamental role of precedent in legal analysis, I cannot accept this submission.  Granted, Egan, like the case now before this Court, was also concerned with the opposite‑sex definition of "spouse" in provincial legislation.  However, the similar focus of the two cases is not sufficient to bind the Court to the Egan decision.  The instant case is based on entirely different legislation with its own unique objectives and legislative context.  As a result, it must be evaluated on its own merits.

2.  Approach to Section 1

76                               The analytical framework for determining whether a law constitutes a “reasonable” limit that can be “demonstrably justified in a free and democratic society” under s. 1  of the Charter  was first set out by Dickson C.J. in R. v. Oakes, [1986] 1 S.C.R. 103. Although it has since been refined, the general approach is now well established (see, e.g., Egan, supra, per Cory and Iacobucci JJ., at para. 182; Eldridge v. British Columbia (Attorney General), [1997] 3 S.C.R. 624, at para. 84, per La Forest J.; Miron, supra, at para. 163, per McLachlin J., and Vriend, supra, at para. 108, per Cory and Iacobucci JJ.).

77                               However, it is important not to lose sight of the underlying principles animating this general approach.  As Dickson C.J. so eloquently put it in Oakes, supra, at p. 136, the inclusion of the words “free and democratic” as the standard of justification in s. 1  of the Charter 

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.

78                               As noted by this Court in Vriend, supra, at para. 134, the introduction of the Charter  brought about “a redefinition of our democracy”.  Central to this democratic vision is a dialogue of mutual respect between the courts and the legislatures, which includes the idea that:

In carrying out their duties, courts are not to second‑guess legislatures and the executives; they are not to make value judgments on what they regard as the proper policy choice; this is for the other branches.  Rather, the courts are to uphold the Constitution and have been expressly invited to perform that role by the Constitution itself.  But respect by the courts for the legislature and executive role is as important as ensuring that the other branches respect each others’ role and the role of the courts. [Vriend, at para. 136.]

This Court has often stressed the importance of deference to the policy choices of the legislature in the context of determining whether the legislature has discharged its burden of proof under s. 1  of the Charter : see, for example, Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927, at pp. 993‑94, per Dickson C.J. and Lamer J. (now Chief Justice) and Wilson J.; R. v. Butler, [1992] 1 S.C.R. 452, at pp. 502‑4, per Sopinka J.; RJR‑MacDonald Inc. v. Canada (Attorney General), [1995] 3 S.C.R. 199, at paras. 135‑37, per McLachlin J.  However, it is important to note that deference is not a kind of threshold inquiry under s. 1.  As a general matter, the role of the legislature demands deference from the courts to those types of policy decisions that the legislature  is best placed to make.  The simple or general claim that the infringement of a right is justified under s. 1 is not such a decision.  As Cory J. stated in Vriend, supra, at para. 54:  “The notion of judicial deference to legislative choices should not . . . be used to completely immunize certain kinds of legislative decisions from Charter  scrutiny.”

79                               Under s. 1, the burden is on the legislature to prove that the infringement of a right is justified.  In attempting to discharge this burden, the legislature will have to provide the court with evidence and arguments to support its general claim of justification.  Sometimes this will involve demonstrating why the legislature had to make certain policy choices and why it considered these choices to be reasonable in the circumstances.  These policy choices may be of the type that the legislature is in a better position than the court to make, as in the case of difficult policy judgments regarding the claims of competing groups or the evaluation of complex and conflicting social science research:  Irwin Toy, supra, at p. 993, per Dickson C.J. and Lamer and Wilson JJ.  Courts must be cautious not to overstep the bounds of their institutional competence in reviewing such decisions.  The question of deference, therefore, is intimately tied up with the nature of the particular claim or evidence at issue and not in the general application of the s. 1 test; it can only be discussed in relation to such specific claims or evidence and not at the outset of the analysis.

80                               I therefore agree with my colleague, Bastarache J., that an examination of context is essential in determining whether deference is appropriate.  It may also be the case that a discussion of context is appropriate at the outset of a s. 1 analysis, depending on the nature of the evidence at issue, for ease of reference when later applying the various steps of s. 1: see, for example, Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877, at para. 88, per Bastarache J. However, with respect to his reasons in the present appeal, I am concerned that  Bastarache J. implies that the question of deference in a general sense should also be determined at the outset of the inquiry.  For example, Bastarache J. states that the question to ask in this case is whether the Court can “rewrite the boundary in order to include that smaller number of individuals [in same‑sex relationships] who are in such a position [of dependency], or must it defer to legislative determination of the issue?”  (para. 304).  The question of rewriting boundaries is, to my mind, at most a question of the appropriate remedy should the rights infringement be unjustified.  The question of deference to the role of the legislature certainly enters into any discussion of remedy, as discussed in Vriend, supra, and can enter into the discussion of whether the legislature has discharged its burden under any of  the steps of the s. 1 test.  However, the question of deference is not an issue that can be determined prior to engaging in any of these specific inquiries.  Nor should it be determined at the outset of the inquiry, given the court’s important role in applying s. 1  of the Charter  to determine whether the infringement of a guaranteed right can be justified in a free and democratic society.

81                               I will therefore not deal with the question of deference at the outset and will instead discuss it, where appropriate, under the various steps of the s. 1 test.

3.  Pressing and Substantial Objective

82                               Section 29 of the FLA defines “spouse” as being either of a man and woman who are married to each other or cohabiting within the meaning of the Act.  Same‑sex couples are necessarily excluded from this definition, thereby giving rise to the charge that the legislation is underinclusive.  In Vriend, supra, at paras. 109‑11, this Court found that where a law violates the Charter  owing to under‑inclusion, the first stage of the s. 1 analysis is properly concerned with the object of the legislation as a whole, the impugned provisions of the Act, and the omission itself.

83                               As to the objective of the FLA as a whole, I note that the submissions of the appellant do not directly address this issue.  Nevertheless, I am of the view, as was Charron J.A. in the court below, that a suitable starting point for determining the purpose of the FLA is the preamble of the Act:

Whereas it is desirable to encourage and strengthen the role of the family; and whereas for that purpose it is necessary to recognize the equal position of spouses as individuals within marriage and to recognize marriage as a form of partnership; and whereas in support of such recognition it is necessary to provide in law for the orderly and equitable settlement of the affairs of the spouses upon the breakdown of the partnership, and to provide for other mutual obligations in family relationships, including the equitable sharing by parents of responsibility for their children;

84                               Although the preamble of the FLA provides some insight into the objective of the Act, its utility is limited.  For example, the reference to “marriage” is somewhat misleading.  As recognized by members of the legislature during debate concerning the Family Law Reform Act, 1978, S.O. 1978, c. 2 (“FLRA”), this reference does not reflect the full purpose of the amended Act, which accords rights to both married and unmarried couples.  Further, it was also noted in debate that the emphasis on encouraging and strengthening the role of the family is inaccurate as the legislation is actually intended to deal with the breakup of the family: Legislature of Ontario Debates, October 18, 1977, at p. 904.

85                               It seems to me that a more complete and accurate statement of the objective of the current version of the FLA is provided by the Ontario Law Reform Commission (“OLRC”) in the following excerpt from the Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act (1993), at pp. 43‑44:

The purpose of the Family Law Act is to provide for the equitable resolution of economic disputes that arise when intimate relationships between individuals who have been financially interdependent break down (Parts I‑IV).  As well, it ensures that family members have a means to seek redress when an immediate relative is injured or killed through the negligence of a third party (Part V).

This statement is largely consistent with the preamble but better reflects the design of the current legislation.

86                               Turning to the objective of the impugned provisions, s. 29 of the FLA defines the term “spouse” as it appears in the support obligation provisions of Part III of the Act.  There is considerable disagreement between the parties as to the underlying purpose of these provisions.  The appellant submits that their objective is twofold.  First, Part III of the FLA is said to have been designed to remedy the systemic inequality associated with opposite‑sex relationships, including the economic dependence of women on men resulting from women assuming primary responsibility for child care and from gender‑based inequality in earning power.  In his reasons in this appeal, Bastarache J. has identified this same inequality as the “mischief and defect” that the Part III provisions were meant to address.  Second, Part III is said to reflect a concern for children and the conditions under which they are raised.

87                               Although I do not dispute the claim that economically dependent heterosexual women and children are well served by the spousal support provisions in Part III of the FLA, in my view, there is insufficient evidence to demonstrate that the protection of these groups informs the fundamental legislative objectives behind this part of the Act.  In fact, with respect to Bastarache J., it seems to me that the legislative history and the terms of the provisions themselves contradict the appellant’s assertions.

88                               With respect to the first of the proposed objectives described above, the appellant submits that the government accepted the conclusion of the OLRC in 1974‑75 that married women tend to become economically dependent upon their partners owing to the traditional division of labour between husbands and wives.  By enacting Part II of the FLRA (now Part III of the FLA), the government was said to have recognized and addressed the need of such women for spousal support.  However, it seems to me that these submissions overlook the import of the changes introduced by the new legislation.

89                               In contrast to predecessor legislation (i.e., the Deserted Wives’ and Children's Maintenance Act, R.S.O. 1937, c. 211 (as amended by S.O. 1954, c. 22; S.O. 1958, c. 23; R.S.O. 1960, c. 105; R.S.O. 1970, c. 128; S.O. 1971, c. 98; S.O. 1973, c. 133)), the 1978 FLRA abandoned a statutory spousal support regime under which only a wife could oblige her husband to pay support in favour of one which imposed mutual support obligations on both men and women.  Indeed, the thrust of the OLRC’s 1975 remarks which preceded the new legislation emphasize the importance of a gender‑neutral scheme.  Although the Commission recognized the financial dependence of many married women upon their husbands, as is evident from the following passages, the OLRC’s recommendations encouraged the government to premise support obligations on need and actual dependence rather than on the assumption that wives are inherently dependent upon their husbands for support because of the traditional roles assumed by men and women:

   Any reform of the provincial law governing inter‑spousal support obligations should include amongst its primary objectives the elimination of the underlying assumption that a wife is inherently dependent upon her husband for support.

                                                                    ...

[W]e have concluded that the courts, in awarding support, should place much greater emphasis upon the need of dependent spouses and the reasons for their dependency . . . . Circumstances can certainly arise in which a husband becomes dependent upon his wife for support, whether by reason of his infirmity or incapacity, or for reasons related simply to the domestic arrangements which they have adopted. ...

[T]he law might be regarded as lacking in flexibility if it was incapable of recognizing a domestic arrangement between the spouses which was based upon a degree of dependency by the husband upon his wife for support.  An example of this situation is shown by spouses who agree to advance the wife’s career at the expense of the husband’s.  Although this is currently not a common phenomenon in Canada, the law should take account of new developments in society even though they may appear to run counter to the conventions of the day.

Ontario Law Reform Commission, Report on Family Law, Part  VI, “Support Obligations” (1975), at pp. 7 and 10‑11.  Similar sentiments were expressed by a government position paper produced prior to the introduction of the 1978 FLRA:  Ministry of the Attorney General for Ontario, Family Law Reform (1976), at p. 9.

90                               The inaccuracy of the appellant's submissions with respect to the first of the purported objectives is also reflected in the terms of Part III of the Act.  For example, the provisions of the FLA that  establish the right to receive support as well as the obligation to provide it use gender‑neutral language.  Under s. 33(2), an application for an order for support may be made by a “dependant”, who is a “person” to whom another has an obligation to provide support (s. 29).  The support obligation is borne by “spouses” who are defined as “either of a man and woman” in ss. 29 and 1(1) of the Act.

91                               In addition, s. 33(9), which sets out various factors for the courts to consider in determining the amount and duration of support, is also cast in gender‑neutral terms and makes no mention of the position of heterosexual women or their needs.  Moreover, s. 33(8) establishes the purposes of an order for the support of a spouse but is silent with respect to the economic vulnerability of heterosexual women, their tendency to take on primary responsibility for parenting, the greater earning capacity of men, and systemic sexual inequality.  In the face of this clearly gender‑neutral scheme, the fact that a significant majority of the spousal support claimants are women does not, in my view, establish that the goal of Part III of the FLA is to address the special needs of women in opposite‑sex relationships.

92                               The terms of the spousal support provisions of Part III are also inconsistent with the second of the appellant’s proposed objectives, namely, protecting children and ensuring that the conditions under which they are raised are adequate.  Although the provisions of Part III that deal exclusively with child support clearly reflect these legitimate legislative concerns (see ss. 31 and 33(7)), it seems to me that the spousal support provisions do not share the same focus.  Part III of the FLA imposes spousal support obligations on opposite‑sex couples irrespective of whether or not they have children.  Indeed, as noted by the intervener EGALE, cohabiting opposite‑sex partners who are not the parents of a child are expressly included in the s. 29 definition of “spouse” after three years of cohabitation.

93                               As I see the matter, the objectives of the impugned spousal support provisions were accurately identified by Charron J.A. in the court below.  Relying in part on the OLRC description of the goal of the FLA set out above, she identified the objectives of the Part III provisions as both a means to provide “for the equitable resolution of economic disputes that arise when intimate relationships between individuals who have been financially interdependent break down” and to “alleviate the burden on the public purse by shifting the obligation to provide support for needy persons to those parents and spouses who have the capacity to provide support to these individuals” (p. 450).  I find support for this position in the legislative debates, the terms of the provisions, as well as the jurisprudence of this Court.

94                               In keeping with the 1975 OLRC recommendations discussed above, the government moved to introduce legislation that would confer “no privileges” and impose “no disability on either men or women as a group but rather tak[e] account of the individual situation in each matrimonial dispute” (Legislature of Ontario Debates, October 26, 1976, at p. 4102).  With the emphasis of spousal support now fixed on need and actual dependence as opposed to presumptions regarding relations between the sexes, the 1978 FLRA was regarded as “a code of economic relations between the spouses upon the severance of that union” (Legislature of Ontario Debates, October 18, 1977, at p. 901).  In my view, these statements are entirely consistent with the first of the objectives suggested by the majority of the Court of Appeal.

95                               The terms of Part III of the FLA are also consonant with the first objective advanced by the Court of Appeal.  Expressly included among the purposes of a spousal support order in s. 33(8) of the Act are the relief of financial hardship, the recognition of the spouse’s contribution to the relationship, and the economic consequences of the relationship for the spouse.

96                               In addition, this Court has previously used language similar to that chosen by the Court of Appeal to describe the purpose of the spousal support provisions in the Divorce Act .  In Moge v. Moge, [1992] 3 S.C.R. 813, at p. 848, L’Heureux‑Dubé J., writing on behalf of the majority of the Court, stated that “the support provisions of the Act are intended to deal with the economic consequences, for both parties, of the marriage or its breakdown” (emphasis in original).  She continued, noting that “[w]hat the Act requires is a fair and equitable distribution of resources to alleviate the economic consequences of marriage or marriage breakdown for both spouses, regardless of gender” (p. 849).  As these remarks were made with respect to legislation different from that under review in the case at bar, I do not suggest that they are determinative of the objective of Part III of the FLA.  Nevertheless, in light of the virtually identical list of purposes for a support order found in each statute, it seems to me that the objective described in Moge has considerable bearing on the present case (see s. 33(8) of the FLA and s. 15(7)  of the Divorce Act ).

97                               I note that Bastarache J. has also relied on Moge, supra, to support his position that redressing the disadvantages suffered by women in opposite‑sex relationships is the pressing and substantial objective of the s. 29 definition of “spouse” in the FLA.  Indeed, at pp. 853‑54 of that decision L’Heureux‑Dubé J. undertook an extensive examination of the power imbalances that are typical of such relationships.  However, in my view, this general social reality does not detract from the principle that dependencies can and do develop irrespective of gender in intimate conjugal relationships.  It seems to me that this is the true mischief which the gender‑neutral support provisions of the FLA are designed to address.

98                               As to the goal of reducing the strain on the public purse, members of the legislature have complained publicly about the number of dependent people who turn to the welfare rolls upon the breakdown of their relationships.  The notion that the spousal support provisions of the FLA and its predecessors were in large part aimed at shifting the financial burden away from the government and on to those partners with the capacity to provide support for dependent spouses has been voiced several times in legislative debates.  For example, prior to the passing of the 1978 FLRA, the Honourable Mr. McMurtry made the following comments regarding dependent spouses:

They have been induced to enter into the relationship and to stay home and raise the children arising from the union, or children of another union, and have thus been put in a position of total dependency on the person as a result of being out of the labour market for a lengthy period of time.  Many of these people are later abandoned and, under the present law, they have nowhere to turn but to the welfare authorities for support.

This is not a small problem.

Legislature of Ontario Debates, November 18, 1976, at p. 4793.  See also Legislature of Ontario Debates, October 26, 1976, at p. 4103; November 22, 1976, at pp. 4898 and 4890‑91.

99                               My colleague, Bastarache J., argues that the Court of Appeal had no basis for determining that “intimacy” is part of the purpose of s. 29 of the FLA (at para. 347).  With respect, I disagree.  Section 29 refers to individuals who have “cohabited”.  Section 1(1), as noted by Cory J., at para. 56, defines “cohabit” as “to live together in a conjugal relationship, whether within or outside marriage”.  The accepted characteristics of a conjugal relationship, as outlined by Cory J. at para. 59, go to the core of what we would generally refer to as “intimacy”.

100                           Having discussed the objective of the legislation as a whole and of the impugned provision, I turn to the objective of the omission.  As I have already stated, when dealing with underinclusive legislation it is important also to consider the impugned omission when construing the objective.  Often legislation does not simply further one goal but rather strikes a balance among several goals, some of which may be in tension.  This balancing exercise may only become apparent after asking whether, in the case of underinclusive legislation, there is any objective being  furthered by the impugned omission.  A consideration of what is omitted from legislation may also lead a court to refine its interpretation of the objectives of the impugned legislation, perhaps reducing its scope.  I agree with my colleague, Bastarache J., at para. 329, that if the omission is not taken into account in construing the objective then it is more likely that the omission will cause the impugned legislation to fail the rational connection step of the proportionality analysis.

101                           However, the concerns just outlined do not imply that the court must find that there is a separate objective being furthered by the omission.  Even if there is no such objective the omission must still be evaluated as part of the means chosen to further the objective of the specific provision in question, under the proportionality analysis.  Otherwise the court risks collapsing the two stages of the Oakes test (pressing and substantial objective and proportionality) into a general question regarding the reasonableness of the omission.  There may be exceptions to this general approach, such as when there is evidence of a deliberate omission by the legislature that is “on its face the very antithesis of the principles embodied in the legislation as a whole”: Vriend, supra, at para. 116.

102                           With these concerns in mind, I turn to the present appeal.  The appellant does not argue that a separate objective is furthered by the impugned omission.  Rather, the argument is that a proper consideration of the exclusion of same‑sex couples from the definition of “spouse” in s. 29 of the FLA reduces the apparent scope of the objective furthered by that provision.  The appellant made two arguments in this regard.  First, the appellant argued that the FLA is a remedial statute designed to address the power imbalance that continues to exist in many opposite‑sex relationships.  Thus, it was submitted that the inclusion of same‑sex couples in a scheme established to deal with problems that are not typical of their relationships is inappropriate.  Further, the appellant asserted that where persons fall outside the rationale for which a benefit was established, the legislature is justified in withholding it from those persons.

103                           With respect, I disagree with these submissions.  As I stated above, I do not believe that the purpose of the FLA in general, nor Part III in particular, is to remedy the disadvantages suffered by women in opposite‑sex relationships.

104                           The second objective for the omission advanced by the appellant is the promotion of opposite‑sex relationships to ensure the protection of children.  Having found that neither the FLA as a whole nor the spousal support provisions in Part III of the Act are primarily concerned with the protection of children, I must also reject the submission that this is part of the  objective of s. 29 of the FLA.

105                           Finally, I note that Bastarache J. accepts that the rejection of the Equality Rights Statute Law Amendment Act, 1994 by the Ontario legislature can provide evidence regarding the objective of s. 29 of the FLA.  In particular, he argues, at para. 349:  “It can therefore be inferred that the legislature’s purpose was also to exclude all types of relationships not typically characterized by the state of economic dependency apparent in traditional family relationships.”  With respect, I cannot agree that a failed amendment can provide evidence as to the objective of the legislation that was to have been amended.  Section 17 of the Interpretation Act, R.S.O. 1990, c. I.11, provides:  “The repeal or amendment of an Act shall be deemed not to be or to involve any declaration as to the previous state of the law.”  If the amendment of an Act may not be used to interpret the meaning of the Act prior to the amendment, then I do not see how a failed amendment may be used in this manner.

106                           Therefore I endorse the description of the objectives of the impugned provisions provided by Charron J.A. in the court below.  These objectives are consonant with the overall scheme of the FLA and are not plausibly reinterpreted through examining the omission of same‑sex spouses.  Providing for the equitable resolution of economic disputes when intimate relationships between financially interdependent individuals break down, and alleviating the burden on the public purse to provide for dependent spouses, are to my mind pressing and substantial objectives.  These objectives promote both social justice and the dignity of individuals -- values Dickson C.J. identified in Oakes, supra, at p. 136, as values underlying a free and democratic society.

107                           In saying this, I wish to note my disagreement with my colleague, Bastarache J., who argues, at para. 354, that s. 29 of the FLA “must be respectful of the equality of status and opportunity of all persons” in order to be consistent with Charter  values and therefore pass this stage of the s. 1 analysis.  While I agree that an objective must be consistent with the principles underlying the Charter  in order to pass the first stage of the s. 1 analysis, I find Bastarache J.’s approach unnecessarily narrow.  It may be that a violation of s. 15(1) can be justified because, although not designed to promote equality, it is designed to promote other values and principles of a free and democratic society.  This possibility must be left open, as the inquiry into Charter  values under s. 1 is a broad inquiry into the values and principles that, as Dickson C.J. stated in Oakes, supra, at p. 136, “are the genesis of the rights and freedoms guaranteed by the Charter ” (emphasis added).

4.  Proportionality Analysis

(a)  Rational Connection

108                           At the second stage of the s. 1 analysis, the focus shifts from the objective alone to the nexus between the objective of the provisions under attack and the means chosen by the government to implement this objective.  As I have already stated, the means chosen include both the impugned provision and the omission in question. It falls to the party invoking s. 1 to demonstrate that there is a rational connection between the objective and the means (see, e.g., Oakes, supra, at p. 141; Vriend, supra, at para. 118).  I concluded above that the dual objectives put forth by the appellant do not reflect the true purposes of the spousal support provisions in Part III of the FLA and relied instead on those set out by the court below.  Nevertheless, it seems to me that no rational connection exists irrespective of which of the objectives is relied upon for this analysis.

109                           Even if I were to accept that Part III of the Act is meant to address the systemic sexual inequality associated with opposite‑sex relationships, the required nexus between this objective and the chosen measures is absent in this case.  In my view, it defies logic to suggest that a gender‑neutral support system is rationally connected to the goal of improving the economic circumstances of heterosexual women upon relationship breakdown.  In addition, I can find no evidence to demonstrate that the exclusion of same‑sex couples from the spousal support regime of the FLA in any way furthers the objective of assisting heterosexual women.

110                           Although there is evidence to suggest that same‑sex relationships are not typically characterized by the same economic and other inequalities which affect opposite‑sex relationships (see, e.g., M. S. Schneider, "The Relationships of Cohabiting Lesbian and Heterosexual Couples:  A Comparison", Psychology of Women Quarterly, 10 (1986), at p. 237, and J. M. Lynch and M. E. Reilly, "Role Relationships:  Lesbian Perspectives", Journal of Homosexuality, 12(2) (Winter 1985/86), at pp. 53‑54, 66), this does not, in my mind, explain why the right to apply for support is limited to heterosexuals.  As submitted by LEAF, the infrequency with which members of same‑sex relationships find themselves in circumstances resembling those of many heterosexual women is no different from heterosexual men who, notwithstanding that they tend to benefit from the gender‑based division of labour and inequality of earning power, have as much right to apply for support as their female partners.

111                           Put another way, it is important to recall that the ability to make a claim for spousal support does not automatically translate into a support order.  To the extent that any relationship is characterized by more or less economic dependence, this will affect the amount and duration, if any, of an award under s. 33(9) of the FLA.  Thus, it is no answer to say that same‑sex couples should not have access to the spousal support scheme because their relationships are typically more egalitarian.  In the case at bar, the respondent does not seek a support order, but rather only access to the support structure provided by the Act.  In much the same way, the appellant in Vriend, supra, did not ask this Court to comment upon the primacy of gay and lesbian concerns over other competing interests.  The legislation had an existing internal balancing mechanism to deal with such issues and the appellant only sought access to Alberta’s human rights machinery.  In both cases, it is the denial of access to the legislative schemes that cannot be justified.

112                           The second of the objectives put forth by the appellant, namely, the protection of children, also fails the rational connection test.  The appellant submits that the exclusion of same‑sex partners from Part III of the FLA is rationally connected to this objective as such couples are far less likely to engage in parenting than opposite‑sex couples.  I have several comments to make by way of response.

113                           Even if I were to accept that the object of the legislation is the protection of children, I would have to conclude that the spousal support provisions in Part III of the FLA are simultaneously underinclusive and overinclusive.  They are overinclusive because members of opposite‑sex couples are entitled to apply for spousal support irrespective of whether or not they are parents and regardless of their reproductive capabilities or desires.  Thus, if the legislation was meant to protect children, it would be incongruous that childless opposite‑sex couples were included among those eligible to apply for and to receive the support in question.

114                           The impugned provisions are also underinclusive.  An increasing percentage of children are being conceived and raised by lesbian and gay couples as a result of adoption, surrogacy and donor insemination.  Although their numbers are still fairly small, it seems to me that the goal of protecting children cannot be but incompletely achieved by denying some children the benefits that flow from a spousal support award merely because their parents were in a same‑sex relationship.  As Cory J. and I noted in Egan, supra, at para. 191, "[i]f 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.”

115                           The same result follows from the objectives identified by Charron J.A. in the court below.  No evidence has been supplied to support the notion that the exclusion of same‑sex couples from the spousal support regime furthers the objective of providing for the equitable resolution of economic disputes that arise upon the breakdown of financially interdependent relationships.  Similarly, it is nonsensical to suggest that the goal of reducing the burden on the public purse is advanced by limiting the right to make private claims for support to heterosexuals.  The impugned legislation has the deleterious effect of driving a member of a same‑sex couple who is in need of maintenance to the welfare system and it thereby imposes additional costs on the general taxpaying public.

116                           If anything, the goals of the legislation are undermined by the impugned exclusion.  Indeed, the inclusion of same‑sex couples in s. 29 of the FLA would better achieve the objectives of the legislation while respecting the Charter  rights of individuals in same‑sex relationships.  In these circumstances, I conclude that the exclusion of same‑sex couples from s. 29 of the Act is simply not rationally connected to the dual objectives of the spousal support provisions of the legislation.

117                           Given this lack of a rational connection, s. 29 of the FLA is not saved by s. 1  of the Charter . Although it is therefore not strictly necessary to consider the other two branches of the second stage of the Oakes test, I will discuss them briefly in order to clarify some fundamental misunderstandings advanced in this appeal.

(b)  Minimal Impairment

118                           When legislative action impairs constitutional rights, the government must demonstrate that the impairment is no more than is reasonably necessary to achieve its goals (see, e.g., Eldridge, supra, at para. 86, Miron, supra, at para. 163).  The appellant suggests that the exclusion of same‑sex couples from s. 29 of the FLA minimally impairs the respondent's s. 15 rights since reasonable alternative remedies are available where economic dependence does occur in such relationships.  I cannot accept these submissions.

119                           The appellant's arguments on this point are based on the remedies available under the equitable doctrine of unjust enrichment (e.g. constructive trust) and the law of contract.  Turning first to the equitable remedies, the doctrine of unjust enrichment allows claimants to found an action on indirect or non‑financial contributions to the acquisition, maintenance, or preservation of an asset held by the other spouse.  However, to be successful, the applicant must demonstrate his or her spouse's enrichment, a corresponding personal deprivation and the absence of any juristic reason for the enrichment.  See Pettkus v. Becker, [1980] 2 S.C.R. 834; OLRC Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act, supra, at pp. 10‑11.

120                           Moreover, I note, as did the motions judge, that equitable common law remedies such as a constructive trust are proprietary in nature and that not all relationships will give rise to property claims.  Indeed, as submitted by LEAF, the FLA expressly recognizes that entitlement to the division of property is in addition to, and not in lieu of entitlement to support.  Thus, it seems to me that compared to awards of spousal support, the equitable remedies are less flexible, impose more onerous requirements on claimants, and are available under far narrower circumstances.  I do not accept that they provide an adequate alternative to spousal support under the FLA.

121                           In my view, the law of contract is an equally unacceptable alternative to the spousal support scheme under the FLA.  The appellant emphasizes that the impugned provisions of the Act do not preclude same‑sex partners from contracting for mutual support obligations.  However, the voluntary assumption of such obligations is not equivalent to a statutory entitlement to apply for a support order.

122                           Firstly, the FLA establishes a default system of support rights.  Opposite‑sex partners who have not turned their minds to the economic consequences of relationship breakdown are automatically protected under the  statute, while those who have considered the issue and prefer alternative arrangements are free to contract out of the regime.  In contrast, same‑sex partners are denied the protection that default legislation inherently provides.  Those who want to resolve support issues before the relationship breaks down are forced either to expend resources to devise a suitable contractual arrangement or risk being left without a remedy in law.

123                           Secondly, as noted by EGALE, the protection that a domestic contract provides to economically vulnerable individuals is markedly inferior to that offered by the FLA.  For example, support orders issued under the Act are subject to bankruptcy provisions and special enforcement mechanisms that protect the recipient should the payor default on support payments.  These same protections are not available to individuals who have only a contractual right to spousal support (Bankruptcy and Insolvency Act , R.S.C., 1985, c. B-3, ss.  121(4) , 136(1) (d.1), 178(1) (c); the Family Responsibility and Support Arrears Enforcement Act, 1996, S.O. 1996, c. 31).

124                           In sum, neither the common law equitable remedies nor the law of contract are adequate substitutes for the FLA's spousal support regime.  Indeed, if these remedies were considered satisfactory there would have been no need for the spousal support regime, or its extension to unmarried, opposite‑sex couples.  It must also be remembered that the exclusion of same‑sex partners from this support regime does not simply deny them a certain benefit, but does so in a manner that violates their right to be given equal concern and respect by the government.  The alternative regimes just outlined do not address the fact that exclusion from the statutory scheme has moral and societal implications beyond economic ones, as discussed by my colleague, Cory J., at paras. 71-72. Therefore the existence of these remedies fails to minimize sufficiently the denial of same‑sex partners' constitutionally guaranteed equality rights.

125                           However, the appellant asserts that the circumstances of this case call for a measure of deference to the decision of the Ontario legislature.  In this context, it is argued that it was reasonable for the government to conclude that it had impaired the rights of same‑sex partners as little as possible.

126                           As I see the matter, the deferential approach advocated by the appellant is inappropriate in the case at bar.  This Court has resorted to such an approach where the impugned legislation involves the balancing of claims of competing groups (see, e.g., Irwin Toy, supra, at pp. 999‑1000; McKinney v. University of Guelph, [1990] 3 S.C.R. 229, at pp. 317‑19, per La Forest J.; and Egan, supra, at para. 29, per La Forest J. and at paras. 105-8, per Sopinka J.).  As Dickson C.J. and Lamer and Wilson JJ. stated in Irwin Toy, supra, at p. 993:

When striking a balance between the claims of competing groups, the choice of means, like the choice of ends, frequently will require an assessment of conflicting scientific evidence and differing justified demands on scarce resources.  Democratic institutions are meant to let us all share in the responsibility for these difficult choices.  Thus, as courts review the results of the legislature’s deliberations, particularly with respect to the protection of vulnerable groups, they must be mindful of the legislature’s representative function.

This is not such a case.  As no group will be disadvantaged by granting members of same‑sex couples access to the spousal support scheme under the FLA, the notion of deference to legislative choices in the sense of balancing claims of competing groups has no application to this case.

127                           I acknowledge that some individuals in same-sex relationships, including H. herself, have expressed reservations about being treated as "spouses" within the family law system (see, e.g., OLRC Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act, supra; B. Cossman and B. Ryder, Gay, Lesbian and Unmarried Heterosexual Couples and the Family Law Act:  Accommodating a Diversity of Family Forms (1993), a Research Paper prepared for the OLRC, at pp. 135‑39).  However, these differences of opinion within the same constitutionally relevant group do not constitute a reason to defer to the choices of the legislature.  Indeed, as noted by EGALE, given that the members of equality‑seeking groups are bound to differ to some extent in their politics, beliefs and opinions, it is unlikely that any s. 15 claims would survive s. 1 scrutiny if unanimity with respect to the desired remedy were required before discrimination could be redressed.

128                           In addition, the deferential approach is not warranted, as submitted by the appellant, on the basis that Part III of the FLA and s. 29 thereof are steps in an incremental process of reform of spousal support.  As this Court noted in Vriend, supra, government incrementalism, or the notion that government ought to be accorded time to amend discriminatory legislation, is generally an inappropriate justification for Charter  violations.  However, even if I were to accept that such a justification might be suitable in the present case, it seems to me that its application to the facts of the case at bar cannot legitimize the continued exclusion of same‑sex couples from the FLA’s spousal support regime.

129                           The appellant contends that the decision to provide equal status to both sexes under the FLRA, followed by the extension of the right to claim support to opposite‑sex common‑law couples and the further broadening of the definition of "spouse" under the FLA by reducing the requisite period of cohabitation from five to three years, is significant evidence of incremental progress toward the ideal of equality.  Therefore, it is submitted that this Court ought to be wary of interfering with the existing legislation.  I disagree.  None of the reforms cited by the appellant has addressed the equal rights and obligations of individuals in same‑sex relationships.  In fact, there is no evidence of any progress with respect to this group since the inception of the spousal support regime. If the legislature refuses to act so as to evolve towards Charter  compliance then deference as to the timing of reforms loses its raison d’être.

130                           Moreover, in contrast to Egan, supra, where Sopinka J. relied in part on incrementalism in upholding the impugned legislation under s. 1  of the Charter , there is no concern regarding the financial implications of extending benefits to gay men and lesbians in the case at bar.  As already pointed out, rather than increasing the strain on the public coffers, the extension will likely go some way toward alleviating those concerns because same‑sex couples as a group will be less reliant on government welfare if the support scheme is available to them.  Thus, I conclude that government incrementalism cannot constitute a reason to show deference to the legislature in the present case.

131                           Finally, as this Court has emphasized on other occasions, "[d]eference must not be carried to the point of relieving the government of the burden which the Charter  places upon it of demonstrating that the limits it has imposed on guaranteed rights are reasonable and justifiable":  RJR‑MacDonald, supra, at para. 136, per McLachlin J. See also Eldridge, supra; Tétreault‑Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22; and Vriend, supra.

132                           In the present case, the government has failed to show that it had a reasonable basis for concluding that the rights of same‑sex couples were impaired no more than was reasonably necessary to achieve its goals.  The exclusion from the s. 29 definition of "spouse", and consequently from the FLA spousal support regime, is absolute.  No effort has been made to tailor the limiting measure.  I conclude that the appellant’s case also fails at the minimal impairment stage of the s. 1 analysis.

(c)  Proportionality Between the Effect of the Measure and the Objective

133                           In order for the impugned legislation to survive the final stage of the s. 1 analysis, there "must be a proportionality between the deleterious effects of the measures which are responsible for limiting the rights or freedoms in question and the objective, and there must be a proportionality between the deleterious and the salutary effects of the measures":  Dagenais v. Canadian Broadcasting Corp., [1994] 3 S.C.R. 835, at p. 889 (emphasis in original).  The damaging effects engendered by the exclusion of same‑sex couples from s. 29 of the FLA, as noted by Cory J., are numerous and severe.  Such harms cannot be justified where the statute has not achieved what it set out to do.  Where, as here, the impugned measures actually undermine the objectives of the legislation it cannot be said that the deleterious effects of the measures are outweighed by the promotion of any laudable legislative goals, nor by the salutary effects of those measures.

134                           I therefore conclude that the exclusion of same‑sex couples from s. 29 of the FLA cannot be justified as a reasonable limit on constitutional rights under s. 1  of the Charter .  Before turning to a discussion of the appropriate remedy, I wish to emphasize, like Cory J., that the sole issue presented by this case is whether the Charter  mandates that same‑sex couples be accorded the right to apply for spousal support under the FLA.  This appeal does not challenge traditional conceptions of marriage, as s. 29 of the Act expressly applies to unmarried opposite‑sex couples.  That being said, I do not wish to be understood as making any comment on marriage or indeed on related issues.

135                           In addition, despite the contentions of the appellant, the facts of this case do not require me to consider whether financially interdependent individuals who live together in non‑conjugal relationships, such as friends or siblings, ought to be constitutionally entitled to apply for support upon the breakdown of their relationships.  Any such claims would require an independent constitutional analysis, the outcome of which cannot be predicted in advance.  Thus, arguments based on the possible extension of the definition of "spouse" beyond the circumstances of this case are entirely speculative and cannot justify the violation of the constitutional rights of same‑sex couples in the case at bar.

VI. Remedy

136                           Having found that the exclusion of same‑sex couples from s. 29 of the FLA is unconstitutional and cannot be saved under s. 1  of the Charter , I must now consider the issue of remedy under s. 52  of the Constitution Act, 1982 .  In the court below, the words "a man and woman" were read out of the definition of "spouse" in s. 29 of the FLA and replaced with the words "two persons".  The application of the order was suspended for a period of one year.  With respect, I am not convinced that that is a suitable remedy in the circumstances of the present case.

137                           In the leading case on constitutional remedies, Schachter v. Canada, [1992] 2 S.C.R. 679, and more recently in Vriend, supra, this Court stated that the first step in selecting the appropriate remedial course is to determine the extent of the inconsistency between the impugned legislation and the Charter .  In the case at bar, the inconsistency emanates from the underinclusive definition of "spouse" in s. 29 of the FLA.  As I have concluded above, the exclusion of same‑sex partners from this definition violates the equality rights guaranteed in s. 15  of the Charter  and cannot survive any of the stages of review that comprise the s. 1 analysis.

138                           Having identified the extent of the inconsistency, the Court must determine the appropriate remedy.  Schachter provides several options in the present case:  (1) "striking down":  the Court may hold that the FLA in its entirety is of no force or effect; or (2) "severance":  the Court may hold that only the offending portion of the statute, namely, s. 29 is of no force or effect and that the rest of the Act remains in force; or (3) "reading in/reading down":  the Court may engage in some combination of reading in and reading down so as to replace the offending words with language that will include the wrongly excluded group (as the inconsistency in the instant case stems from an omission, reading down alone is inappropriate); or (4) striking down, severance, or reading in/reading down with a temporary suspension of the Court's order so that the government has an opportunity to enact a constitutionally valid spousal support scheme.

139                           In determining whether the reading in/reading down option is more appropriate than either striking down or severance, the Court must consider how precisely the remedy can be stated, budgetary implications, the effect the remedy would have on the remaining portion of the legislation, the significance or long‑standing nature of the remaining portion and the extent to which a remedy would interfere with legislative objectives (see Schachter, supra; Vriend, supra).  As to the first of these criteria, the remedy of reading in is only available where the court can direct with a sufficient degree of precision what is to be read in to comply with the Constitution.  Remedial precision requires that the insertion of a handful of words will, without more, ensure the validity of the legislation and remedy the constitutional wrong (see Egan, supra, at para. 223, per Cory and Iacobucci JJ.; Vriend, supra, at para. 155).

140                           In the present case, the defect in the definition of "spouse” can be precisely traced to the use of the phrase “a man and woman”, which has the effect of excluding same‑sex partners from the spousal support scheme under the FLA.  I recognize that there is remedial precision in so far as reading down this phrase and reading in the words “two persons” will, without more “remedy the constitutional wrong”.  However, I am not persuaded that reading in will also “ensure the validity of the legislation”.

141                           If the remedy adopted by the court below is allowed to stand, s. 29 of the FLA will entitle members of same‑sex couples who otherwise qualify under the definition of "spouse" to apply for spousal support.  However, any attempt to opt out of this regime by means of a cohabitation agreement provided for in s. 53 or a separation agreement set out in s. 54 would not be recognized under the Act.  Both ss. 53 and 54 extend to common‑law cohabitants but apply only to agreements entered into between "a man and woman".  Any extension of s. 29 of the Act would have no effect upon these Part IV domestic contract provisions of the FLA, which do not rely upon the Part III definition of "spouse".  Thus, same‑sex partners would find themselves in the anomalous position of having no means of opting out of the default system of support rights.  As this option is available to opposite‑sex couples, and protects the ability of couples to choose to order their own affairs in a manner reflecting their own expectations, reading in would in effect remedy one constitutional wrong only to create another, and thereby fail to ensure the validity of the legislation.

142                           In addition, reading into the definition of "spouse" in s. 29 of the Act will have the effect of including same‑sex couples in Part V of the FLA (Dependants’ Claim for Damages), as that part of the Act relies upon the definition of "spouse" as it is defined in Part III.  In my opinion, where reading in to one part of a statute will have significant repercussions for a separate and distinct scheme under that Act, it is not safe to assume that the legislature would have enacted the statute in its altered form.  In such cases, reading in amounts to the making of ad hoc choices, which Lamer C.J. in Schachter, supra, at p. 707, warned is properly the task of the legislatures, not the courts.

143                           In cases where reading in is inappropriate, the court must choose between striking down the legislation in its entirety and severing only the offending portions of the statute.  As noted by Lamer C.J. in Schachter, at p. 697, “[w]here the offending portion of a statute can be defined in a limited manner it is consistent with legal principles to declare inoperative only that limited portion.  In that way, as much of the legislative purpose as possible may be realized.”

144                           In the case at bar, striking down the whole of the FLA would be excessive as only the definition of “spouse” in Part III of the Act has been found to violate the Charter .  This is not a case where the parts of the legislative scheme which do offend the Charter  are so inextricably bound up with the non‑offending portions of the statute that what remains cannot independently survive.  As a result, it would be safe to assume that the legislature would have passed the constitutionally sound parts of the statute without the unsound parts.  See Attorney‑General for Alberta v. Attorney‑General for Canada, [1947] A.C. 503, at p. 518; Schachter, supra, at p. 697.

145                           On the basis of the foregoing, I conclude that severing s. 29 of the Act such that it alone is declared of no force or effect is the most appropriate remedy in the present case.  This remedy should be temporarily suspended for a period of six months.  Although we have been advised against the imposition of a suspension by both the appellant and the respondent, for the reasons which follow, I find that a suspension is necessary.

146                           In Egan, supra, at para. 226, writing in dissent on behalf of myself and Cory J., I would have granted a suspension of the remedy on the basis that “the extension of the spousal allowance, while certainly a legal issue, is also a concern of public policy”.  In this respect, I noted that “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 ”.  These same concerns arise in the case at bar with respect to the spousal support scheme under the FLA.

147                           In addition, I note that declaring s. 29 of the FLA to be of no force or effect may well affect numerous other statutes that rely upon a similar definition of the term "spouse".  The legislature may wish to address the validity of these statutes in light of the unconstitutionality of s. 29 of the FLA. On this point, I agree with the majority of the Court of Appeal which noted that if left up to the courts, these issues could only be resolved on a case‑by‑case basis at great cost to private litigants and the public purse.  Thus, I believe the legislature ought to be given some latitude in order to address these issues in a more comprehensive fashion.

VII. Costs

148                           At trial, the motions judge ordered the Attorney General for Ontario (the appellant before this Court) to pay M.'s costs.  This decision was unanimously upheld on appeal.  However, Charron J.A. in the Court of Appeal made no order as to costs in the proceedings before her, finding that because the appeal raised a constitutional issue of significant public importance, each party and the interveners should bear their own costs.  This Court gave the Attorney General leave to appeal on the condition that it pay M.'s costs at this level regardless of the result.

149                           M. now submits on cross‑appeal that the Court of Appeal erred in failing to grant costs to the successful party.  I disagree.  Costs are a discretionary determination and absent any clear error, this Court should be loath to interfere: Canadian Pacific Ltd. v. Matsqui Indian Band, [1995] 1 S.C.R. 3, at para. 39.  As I can find no such error in the case at bar, I would therefore make no order with respect to costs in the Court of Appeal.

150                           As to costs in this Court, I would order that the appellant pay the respondent M.'s costs on a solicitor‑and‑client scale.  The appellant chose to pursue this matter despite the fact that the original parties to the action, M. and H., had settled their claim prior to the commencement of proceedings before this Court.  As this case no longer relates to the lis between the parties, the proceedings were in large measure an attempt by the appellant to obtain judicial clarification of the state of the law.  In these circumstances, the respondent should not be put to any expense for costs.

151                           For the same reasons outlined above, namely that there is no longer a lis between the parties and the Attorney General brought this appeal to determine an issue of public importance, the Attorney General has agreed that the respondent H. should be afforded the same treatment as M. and consents to an order to pay H.’s costs of this appeal. I would therefore make such an order.  As I have ordered costs on a solicitor‑client basis to M., I would also do so for H.

VIII.  Conclusions and Disposition

152                           For the reasons set out by Cory J., I conclude that the exclusion of same‑sex couples from the definition of "spouse" in s. 29 of the FLA violates the equality rights guaranteed in s. 15  of the Charter .  Further, for the reasons outlined above, I find that the impugned limitation is not saved under s. 1  of the Charter .  I would declare s. 29 of no force or effect but temporarily suspend the effect of that declaration for a period of six months.  Accordingly, I would dismiss the appeal and the cross‑appeal with solicitor‑and‑client costs to both M. and H. in the proceedings before this Court.

153                           I would thus answer the constitutional questions as follows:

1.  Does the definition of "spouse" in s. 29 of the Family Law Act, R.S.O. 1990, c. F.3, infringe or deny s. 15(1)  of the Canadian Charter of Rights and Freedoms ?

Answer:  Yes.

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.

The following are the reasons delivered by

Gonthier J. (dissenting) ‑‑

I.  Introduction

154                           Does the inability of a woman to bring a claim for spousal support under provincial family legislation against her same-sex partner upon the breakdown of their relationship violate the equality guarantee contained in the Canadian Charter of Rights and Freedoms ?  That is the question raised by this appeal.  The answer depends upon whether s. 29 of Ontario’s Family Law Act, R.S.O. 1990, c. F.3 (“FLA”), and in particular, the definition of “spouse” contained in that section, infringes s. 15(1)  of the Charter  because it is underinclusive.  If so, the Court must consider whether that infringement can be upheld as a demonstrably justified limit under s. 1  of the Charter .  For the reasons which follow, I believe that the impugned section is constitutionally sound.  In my view, s. 29 of the FLA does not infringe s. 15(1)  of the Charter , and consequently, no such claim may be brought.

155                           Plainly, this appeal raises elemental social and legal issues.  Indeed, it is no exaggeration to observe that it represents something of a watershed.  I have had the benefit of reading the reasons of my colleagues Cory and Iacobucci JJ.  I gratefully adopt their account of the facts of this appeal.  However, I am unable to agree with my colleagues’ disposition of this appeal or their underlying reasons for so doing.  I believe that the stance adopted by the majority today will have far-reaching effects beyond the present appeal.  The majority contends, at para. 135,  that it need not consider whether a constitutionally mandated expansion of the definition of “spouse” would open the door to a raft of other claims, because such a concern is “entirely speculative”.  I cannot agree.  The majority’s decision makes further claims not only foreseeable, but very likely.  Because the nature of my disagreement with the majority in this case is basic, I have felt it necessary to set out my views at some length.

156                           The disagreement in this appeal arises from differing views on the purpose of the legislation.  Cory and Iacobucci JJ. ascribe a purpose to s. 29 of the FLA that centres on the interdependency of “intimate” relationships which they refer to as “conjugal” relationships of a specific degree of duration.  In contrast, Bastarache J. believes that this legislation deals with individuals in “permanent and serious” relationships which cause or enhance economic disparity between the partners.  In my opinion, this legislation seeks to recognize the specific social function of opposite-sex couples in society, and to address a dynamic of dependence unique to both men and women in opposite-sex couples that flows from three basic realities.  First, this dynamic of dependence relates to the biological reality of the opposite-sex relationship and its unique potential for giving birth to children and its being the primary forum for raising them.  Second, this dynamic relates to a unique form of dependence that is unrelated to children but is specific to heterosexual relationships.  And third, this dynamic of dependence is particularly acute for women in opposite-sex relationships, who suffer from pre-existing economic disadvantage as compared with men.  Providing a benefit (and concomitantly imposing a burden) on a group that uniquely possesses this social function, biological reality and economic disadvantage, in my opinion, is not discriminatory.  Although the legislature is free to extend this benefit to others who do not possess these characteristics, the Constitution does not impose such a duty on that sovereign body. 

157                           These differing views on the purpose of the legislation are determinative in this appeal.   Because Cory and Iacobucci JJ. (and Major J. by reference) suggest that the FLA targets intimate relationships, they understandably conclude that the legislation is unnecessarily underinclusive.  Bastarache J. concludes that the FLA targets relationships typically characterized by permanence and economic dependency, and notes that same-sex relationships are not typically characterized by economic dependency.  Nonetheless, Bastarache J. sees no reason to exclude individuals in same-sex relationships. In my judgment, the respondent M.’s claim fails because the legislation targets individuals who are in relationships which are fundamentally different from same-sex relationships.  The legislation “corresponds” to the actual need, capacity, and circumstances of the claimant and those of the group the legislation targets.  As such, I find that there is no “discrimination” within the meaning of that term in s. 15(1)  of the Charter  on the facts of this appeal.  Accordingly, I would allow the Attorney General’s appeal.

158                           I begin these reasons with a brief review of the approach this Court has taken in analysing s. 15(1) claims.  With this understanding, I move to explain the purpose and practical operation of the FLA’s spousal support regime, reviewing the history of the law of spousal support generally and the support regime legislated in Ontario specifically.  I explain why the purpose of the legislation is to target a specific group of people who possess unique characteristics, and I contrast this interpretation with those of my colleagues.   Establishing the proper purpose of the legislation is critical to the s. 15(1) analysis.  I begin this analysis by considering the proper comparison group, and then ask whether there has been a distinction created by the legislation between the claimant and this comparison group.  Having established a distinction, I then analyse whether that distinction rests on enumerated or analogous grounds.  Finally, I examine whether this distinction rests on the stereotypical application of presumed group or individual characteristics, or otherwise has the effect of demeaning the claimant’s human dignity.  It is my conclusion that there is no stereotype at work:  the legislation corresponds with the claimant’s need, capacity and circumstances.  A full analysis of all of the contextual factors discussed in Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497, reveals that this differential treatment does not discriminate, and the claim must fail. 

II. Analysis

A.  Approach to Section 15(1)  of the Charter 

159                           The appropriate analytical approach to be taken to claims arising under s. 15(1)  of the Charter  has been addressed by this Court in Law, supra.  This Court unanimously held that in analysing a s. 15(1) claim, one must take into account certain contextual factors, having regard to the purpose of s. 15(1) in protecting human dignity.  There are three broad inquiries which a court must make in its analysis.  First, does the impugned legislative provision draw a distinction between the claimant and others who are relevant comparators based on one or more personal characteristics, or fail to take into account the claimant’s already disadvantaged position within Canadian society, resulting in substantively differential treatment between the claimant and others?  That distinction may be apparent on the face of the legislation, or it may be that the effect of the legislation is to draw such a distinction.  Either type of distinction will suffice.  The court must then consider whether that distinction is based on one or more enumerated or analogous grounds.  The third step in this exercise is to establish discriminatory treatment.  Here, the court considers a variety of contextual factors in establishing whether a burden has been imposed or a benefit withheld in a manner that “reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration” (Law, supra, at para. 88).

160                           At this stage of the analysis, several contextual factors are considered.  Among them are:  pre-existing disadvantage, stereotyping, prejudice, or vulnerability; the correspondence between the ground of distinction and the actual need, capacity, or circumstances of the claimant and others; the ameliorative purpose of the legislation (if any); and the nature and scope of the interest affected.  These factors are to be taken into account in light of the purpose of s. 15(1), the protection of human dignity. Discrimination will generally exist only where an enumerated or analogous ground has been relied upon in a stereotypical fashion as the basis for a legislative distinction.  In the absence of reliance upon a stereotype, discrimination is unlikely to have been made out.  If reliance on a stereotype is found to exist, and the claimant’s human dignity has been violated, then an infringement of s. 15(1) will have been established, and the court must then consider whether the infringement may be justified under s. 1  of the Charter .

B.  The Family Law Act’s Spousal Support Scheme

161                           A proper understanding of both the theory and practical operation of the FLA’s spousal support regime is essential both for establishing appropriate comparison groups and for determining whether there has been discrimination in this case.  At para. 57 of Law, supra, Iacobucci J. explained that in order to establish the appropriate comparator, “the purpose and the effect of the legislation must be considered” (emphasis added).  Locating the appropriate comparator will affect the analysis of many of the contextual factors contained in the discrimination analysis:  Law,  at para. 56.  It is also clear that the purpose of the legislation is also relevant in other areas of the s. 15(1) analysis.  In particular, we are to ask whether the legislation has a discriminatory purpose (Law, at para. 80).  We are also to determine if the legislation has an ameliorative purpose in order to decide whether it is underinclusive of a comparatively disadvantaged group (Law, at para. 72).  Therefore, determining the purpose of the legislation will be relevant for both the s. 15(1) and the s. 1 analysis, as each of the reasons in this case aptly demonstrates.   

162                           In this appeal, the four differing sets of reasons of my colleagues ascribe different purposes to the legislation, and each utilizes a different comparator, which results in a different understanding of the contextual factors.  In my reasons, I provide a brief review of the history of support obligations at common law, and the development of the statutory scheme that eventually replaced most of the common law in this area.  In my view, it is against this background that the purpose and effect of the FLA’s spousal support regime should be scrutinized under the Charter .  The Court has recognized on many occasions -- most recently in Law -- that such a contextual examination is essential.  I continue to adhere to that approach here.

Historical Overview of the Law Governing Spousal Support

163                           I begin with the observation that our law imposes no general obligation of support between persons.  As a general rule, individuals are expected to provide for themselves.  Although governments intervene in myriad ways to provide assistance to those in need, that does not affect the principle advanced here.  Government intervention is, as a general rule, limited to assisting those in need, not imposing legal obligations upon third parties to assist them.  To the general rule that individuals have no obligation to support each other, our law has long made specific exceptions.  I address one of those exceptions in some detail below, but I wish to emphasize here that it has always been considered to be an exception to a general rule.  Where exceptions have been made, they have been narrowly tailored to achieve specific purposes.

164                           The history of family law is, in many ways, the history of the gradual emancipation of women from legal impediments to full equality.  At common law, marriage was a status with reciprocal (though not identical) rights and duties.   Husbands were under a duty to support their wives and this duty continued for so long as the parties remained married.  The obligation upon husbands to support their wives was a counterpart to the control of all real property held by their wives, and outright ownership of their wives’ personal property, which husbands acquired upon marriage.  The capacity of a married woman to own or dispose of property was severely constrained.  The doctrine of coverture dictated that a woman’s separate legal identity disappeared upon marriage.  Blackstone thus observed that, “[b]y marriage, the husband and wife are one person in law . . . [and] the very being or legal existence of the woman is suspended during the marriage, or at least is incorporated and consolidated into that of the husband”:  W. Blackstone, Commentaries on the Laws of England (4th ed. 1770), Book I, at p. 442.  Upon marriage, women suffered “civil death”:  L. Chambers, Married Women and Property Law in Victorian Ontario (1997), at p. 3.

165                           Wives were thus in a perilous situation in the event of marital breakdown.  Marriage was, as a practical matter, indissoluble, because the ecclesiastical courts -- which were the sole curial fora for divorce petitions in England -- were not established in Canada.  A wife, being without property, had no means to support herself independently.  Evidently, the maintenance obligation owed to a wife by her husband was of considerable importance to her.  While theoretically enforceable, in practice a wife’s ability to enforce the maintenance obligation owed to her was severely constrained.  The law was not designed to address the breakdown of marriages, and so, unsurprisingly, it proved inadequate to the task.  This left married women in an extremely vulnerable position.

166                           First, as long as the husband and wife lived together, their standard of living was considered to be a matter entirely within the discretion of the husband.  Second, a wife was deemed to have waived her right to maintenance if, for whatever reason, she and her husband were not living together, even by agreement.  Third, the mechanism for enforcing the maintenance obligation was entirely inadequate.  Indeed, the common law, relying upon the unity of their legal personality, did not permit an action between husband and wife.  Finally, while in theory, the court had the power to order maintenance where a wife had committed adultery, in practice this was rarely done.  This was so even if her husband had himself committed adultery or treated her with great cruelty.  Thus, a wife had to show desertion, adultery, or cruelty on the husband’s part, and yet remain blameless herself.  See the discussion of these points in S. M. Cretney and J. M. Masson, Principles of Family Law (6th ed. 1997), at pp. 81-82; D. J. MacDougall, “Alimony and Maintenance”, c. 6, in D. Mendes da Costa, ed., Studies in Canadian Family Law (1972), vol. 1, 283, at pp. 288-89; R. R. Evans, The Law and Practice Relating to Divorce and Other Matrimonial Causes (1923), at pp. 303-4.

167                           From its establishment in 1837, the Court of Chancery in Upper Canada  intervened on occasion to mitigate the harshness of the common law regime.  The court was authorized by its constituting statute to award alimony:  see An Act to establish a Court of Chancery in this Province, S.U.C. 1837, 7 Wm. IV, c. 2, s. 3.  In addition, a statutory framework slowly developed to address the perceived inadequacies of the common law regime.  A patchwork of legislation, including the Married Women’s Property Acts of 1859, 1872, and 1884 (subsequently R.S.O. 1970, c. 262), and the Married Woman’s Real Estate Act, 1873, S.O. 1873, c. 18, sought to ameliorate the position of married women.  Later, the Deserted Wives’ and Children’s Maintenance Act, R.S.O. 1937, c. 211, as amended (subsequently R.S.O. 1970, c. 128), imposed both a support obligation and a mechanism for enforcing it upon husbands who left their wives, by deeming a wife to have been deserted in cases of cruelty or neglect where the wife had not herself committed uncondoned adultery.  As this Court observed in Reference re Authority to Perform Functions Vested by the Adoption Act, [1938] S.C.R. 398, at p. 419, the purpose of such legislation was to “aim at declaring and enforcing the obligations of husbands and parents to maintain their wives and children”.  See also The Matrimonial Causes Act, 1931, S.O. 1931, c. 25 (subsequently R.S.O. 1970, c. 265).  However, these legislative interventions were plainly insufficient.  As women began to benefit from the social, technological and demographic changes that emerged during and after the Second World War, the spousal support regime, explicitly predicated as it was upon the legal inequality of husbands and wives, was increasingly viewed as an anachronism, and an impediment to gender equality.

168                           In the face of this anachronistic and jumbled system of family law, pressure mounted for a comprehensive overhaul of family law legislation more generally, and marital property and support obligations in particular.  In 1964, the Ontario Law Reform Commission was charged with the task of producing a comprehensive report and recommendations on the reform of family law in Ontario.  That Report, in six volumes, was released in 1975.  In the foreword to its report on family property law, the Commission made note of the context in which it had been asked to make recommendations (Report on Family Law, Part IV, “Family Property Law” (1974), at pp. 4-5):

The common law doctrine of unity of husband and wife, with the consequent extinguishment of the wife’s legal personality and vesting of her property and right to her income in her husband, was necessarily accompanied by the concept of the dependency of the married woman.  The Married Women’s Property Acts have terminated most of the incidents of the unity of legal personality in the marital relation, but the dependency of the wife is still very much a fact in twentieth-century Ontario.  On a secular basis, marriage can be characterized as an economic arrangement that assumes this dependency of the wife upon the husband with a legal framework that is designed to provide a remedy for the wife if the husband’s obligation to maintain her is not properly discharged.  [Footnotes omitted, emphasis added.]

169                           In its recommendations, the Commission consciously rejected the “rigidities and inequalities” that characterized the existing law, which determined property rights as between spouses on the basis of their respective status as either a husband or a wife.  The Commission took the view (at pp. 3-4) that:

Much of the present law presupposes and leads to a view of the parties to a marriage as the recipients of predetermined socio-economic roles rather than as autonomous and independent human beings.  The Commission believes that the law should be shaped consciously to allow all to choose their own roles within the modern society; avoiding as far as is possible past legal complacency in viewing married persons of either sex as stereotypes rather than individuals who are expected to conform to some pre-ordained social or economic condition.  [Footnotes omitted.]

170                           Consequently, the Commission’s recommendations were based on the premise that formal equality between spouses should be the legal norm in Ontario.  However, the Commission did not stop there.  Its recommendations were made with full awareness that according formal legal equality to women would not, in itself, be sufficient to address the obvious substantive inequalities faced by women.  In the particular context of spousal support, the Commission was very much alive to the underlying social and economic realities of women:  see generally Ontario Law Reform Commission, Report on Family Law, Part VI, “Support Obligations” (1975).  Against this background, I turn to a discussion of the statutory scheme enacted by the legislature.

Ontario’s Spousal Support Scheme

171                            The spousal support obligation at issue in this appeal is contained in s. 30 of the FLA:

30.  Every spouse has an obligation to provide support for himself or herself and for the other spouse, in accordance with need, to the extent that he or she is capable of doing so.

172                           It was argued before us that the nature of the spousal support obligation itself is not at issue in this appeal, but rather, simply the question of standing to invoke the spousal support provisions of the FLA.  For reasons discussed in some detail below, I doubt that the nature of the spousal support obligation and the issue of standing to enforce it can be separated so neatly.  That being said, the question of standing is governed by s. 29 of the FLA, which provides, in relevant part, that:

29.  In this Part,

                                                                    ...

“spouse” means a spouse as defined in subsection 1 (1), and in addition includes either of a man and woman who are not married to each other and have cohabited,

(a) continuously for a period of not less than three years, or

(b) in a relationship of some permanence, if they are the natural or adoptive parents of a child.

173                           The effect of ss. 29 and 30 of the FLA is to impose support obligations on members of married couples (because the reference in s. 29 to subs. 1(1) of the Act incorporates married couples, and those individuals who have entered into void or voidable marriages in good faith, within the scope of the definition of “spouse” in s. 29) and on members of unmarried opposite-sex couples who have cohabited continuously for at least three years, or, if they are the natural or adoptive parents of a child, “in a relationship of some permanence”.  The challenge brought by the respondent M. to s. 29 of the FLA is that the section is underinclusive, because the definition of “spouse” therein is restricted to opposite-sex couples, thus, it is claimed, denying her equal benefit of the law under s. 15(1)  of the Charter .

174                           The Family Law Reform Act, 1975, S.O. 1975, c. 41 (“FLRA”), was the first modern legislation in Ontario to reform the law applicable to marriage generally and spousal support in particular.  The FLRA abolished the unity of legal personality between husband and wife and other restraints on the capacity of married persons.  Formal equality between the sexes was thus established.  Amendments were introduced in 1976 with a view to an expansion of the scope of family law legislation.  The Legislative Assembly determined that cohabiting opposite-sex couples were also coming to play the social role previously filled almost exclusively by married couples.  It was in response to the recognition that cohabiting opposite-sex couples were coming to serve as the functional equivalent of actual marriage that support obligations were first extended to cohabiting opposite-sex spouses. 

175                           The amendments became law in 1978 (Family Law Reform Act, 1978, S.O. 1978, c. 2), extending support obligations to members of opposite-sex couples who had cohabited continuously for at least five years, or, where there was a child of the relationship, when the parties had a “relationship of some permanence” (s. 14).  The 1978 amendments also provided for the recognition of domestic contracts, thus enabling married couples or unmarried cohabiting opposite-sex couples to enter into such contracts relating to marital property and support obligations rather than rely upon the default provisions contained in the FLRA itself.  Subsequent amendments in 1986 (Family Law Act, 1986, S.O. 1986, c. 4, s. 29) reduced the period of cohabitation required to trigger the spousal support provisions contained in the FLRA from five to three years.

176                           I note parenthetically that the legislative framework governing spousal support varies considerably from province to province.  In most provinces, and in the three territories, members of cohabiting opposite-sex couples are included within the scope of support obligations, although the precise definition varies, particularly with regard to the duration of cohabitation required to trigger the obligation:  see Family Relations Act, R.S.B.C. 1996, c. 128, s. 1; Family Maintenance Act, R.S.M. 1987, c. F20, s. 4(3); Family Services Act, S.N.B. 1980, c. F-2.2, s. 112(3); Family Law Act, R.S.N. 1990, c. F‑2, s. 35(c); Family Maintenance Act, R.S.N.S. 1989, c. 160, s. 2(m); Family Law Act, S.P.E.I. 1995, c. 12, ss. 1(g), 29; Family Maintenance Act, S.S. 1990-91, c. F-6.1, s. 2(l)(iii); Family Law Act, S.N.W.T. 1997, c. 18, s. 1; Nunavut Act , S.C. 1993, c. 28, s. 29(1) , as am. by S.C. 1998, c. 15, s. 4; Family Property and Support Act, R.S.Y. 1986, c. 63, s. 35.

177                           In Alberta and Quebec, by contrast, cohabiting opposite-sex couples are not within the scope of spousal support obligations.  I note that the exclusion of cohabiting opposite-sex couples from the scope of the spousal support provisions of the Alberta Domestic Relations Act, R.S.A. 1980, c. D-37, was recently challenged successfully in Taylor v. Rossu (1998), 161 D.L.R. (4th) 266 (Alta. C.A.), but the question of whether provincial legislatures are constitutionally barred from restricting access to the spousal support mechanism to married couples has not yet been addressed by this Court, and I will say nothing further about it here.  I simply raise the point to emphasize that there is diversity amongst the approaches taken by the provincial and territorial legislatures as to their spousal support schemes, but a growing political recognition that cohabiting opposite-sex couples should be subject to the spousal support regime that applies to married couples because they have come to fill a similar social role, and, as I discuss below, they exhibit similar needs. 

178                           The FLA and the FLRA were high-profile pieces of legislation, and the extension of spousal support obligations to unmarried opposite-sex couples was the subject of extensive discussion and debate.  Yet there is no evidence to indicate that the Legislative Assembly intended the FLRA or the FLA to extend spousal support obligations to members of same-sex couples.  By contrast, where a province has intended to extend spousal support obligations to members of same-sex couples, it has done so by way of explicit language:  see, e.g., Family Relations Amendment Act, 1997, S.B.C. 1997, c. 20, s. 1(c), and Family Maintenance Enforcement Amendment Act, 1997, S.B.C. 1997, c. 19, s. 1(d).

179                           In the early 1990s, in anticipation of possible amendments to the FLA, the Ontario Law Reform Commission was requested to prepare a report addressing, among other concerns, this very issue.  In its Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act (1993), the Commission indicated, at p. 3, that it was “more hesitant about making recommendations concerning same-sex couples” in respect of support obligations, and cautioned that “better information than the Commission now has before it concerning the attitudes and expectations of cohabiting same-sex couples would be a necessary foundation for any decision to ascribe rights and responsibilities under the Family Law Act to such couples”.  In May and June 1994, the Legislative Assembly of Ontario contemplated extending the scope of s. 29 to include cohabiting same-sex couples.  In the end, however, it declined to do so.  Then Bill 167, the would-be Equality Rights Statute Law Amendment Act, 1994, was defeated by a free vote on second reading on June 9, 1994.

180                           The statutory regime developed since the 1970s to govern spousal support in Ontario is based on an appreciation of the need both to accord formal legal equality to women, and to take into account that formal legal equality would not immediately be reflected in women’s social and economic status.  The legislation, along with the legislative history, the work of the Ontario Law Reform Commission, and the social background, confirm that the Legislative Assembly was well aware of this situation.  In enacting the spousal support provisions of the FLRA, and later the FLA, the Legislative Assembly was aware of the unhappy historical experience of women under the previous family law regime, and had the primary purpose of ameliorating the position of women who had become dependent upon their partners in both married and conjugal opposite-sex relationships. 

Purpose of Section 29 of the Family Law Act

181                           The primary purpose of the FLA is to recognize the social function specific to opposite-sex couples and their position as a fundamental unit in society, and to address the dynamic of dependence unique to men and women in opposite-sex relationships.  This dynamic of dependence stems from this specific social function, the roles regularly taken by one member of that relationship, the biological reality of the relationship, and the pre-existing economic disadvantage that is usually, but not exclusively, suffered by women.  This purpose is apparent from the text of the provision, the preamble to the legislation, and the legislative history of the provision.

182                           In determining the purpose of legislation, the first step must be to look to the language of the legislative provision itself.  The language of the impugned provision must be viewed in the context of the statute as a whole, not taken in isolation.  Where the statutory language, in the context of the statute as a whole, is unclear or ambiguous, resort may then be had to other indicia of legislative intent, such as statements made in the legislature, to inform the court’s understanding of the purpose of the statute.  However, as this Court has cautioned on previous occasions, such indicia of legislative intent, though of assistance in some cases, must be treated with circumspection, given practical concerns as to their reliability, and more theoretical concerns regarding legislative sovereignty:  see R. v. Heywood, [1994] 3 S.C.R. 761, at pp. 787-88.  Ultimately, the language of the statute itself is authoritative:  it may be interpreted in the case of ambiguity by reference to extrinsic sources, but it is the statute, not the sources, which governs. 

183                           Applying these principles of statutory interpretation to the impugned legislation at issue in the present appeal, it is plain that the purpose of s. 29 of the FLA cannot be understood without reference to Part III more generally, and indeed, to the FLA as a whole.  One must begin with the language of the statute itself.  The language of s. 29 itself reveals little.  Section 33(8) indicates the purpose of a support order, as a guideline for the courts:

33.  ...

(8)  An order for the support of a spouse should,

(a)       recognize the spouse’s contribution to the relationship and the economic consequences of the relationship for the spouse;

(b)       share the economic burden of child support equitably;

(c)       make fair provision to assist the spouse to become able to contribute to his or her own support; and

(d)       relieve financial hardship, if this has not been done by orders under Parts I (Family Property) and II (Matrimonial Home).

184                           However, s. 33(8) really only outlines the factors to be taken into account in making an order:  it does not explain why an order should be made in the first place.  A proper analysis of the purpose of the FLA must also take into account the statute’s preamble:

Whereas it is desirable to encourage and strengthen the role of the family; and whereas for that purpose it is necessary to recognize the equal position of spouses as individuals within marriage and to recognize marriage as a form of partnership; and whereas in support of such recognition it is necessary to provide in law for the orderly and equitable settlement of the affairs of the spouses upon the breakdown of the partnership, and to provide for other mutual obligations in family relationships, including the equitable sharing by parents of responsibility for their children;  [Emphasis added.]

185                           Of course, in determining the purpose of a statute, the courts are not necessarily bound by the language of a preamble.  Nevertheless, it must be recognized that the preamble is a valid indicator of the legislative intent animating the statute, and this Court has long held that the preamble to a statute may provide clues as to legislative intent:  Vriend v. Alberta, [1998] 1 S.C.R. 493, at para. 94; Athabasca Tribal Council v. Amoco Canada Petroleum Co., [1981] 1 S.C.R. 699;  Canada Labour Relations Board v. City of Yellowknife, [1977] 2 S.C.R. 729, at p. 734.  Indeed, the Interpretation Act, R.S.O. 1990, c. I.11, s. 8, specifically provides that “[t]he preamble of an Act shall be deemed a part thereof and is intended to assist in explaining the purport and object of the Act”.  Moreover, in the province of Ontario, relatively few statutes contain a preamble.  This may be taken as evidence that where the legislature does choose to make use of a preamble, the language contained therein should not be dismissed as mere hortatory boilerplate, but rather should receive the Court’s careful consideration.

186                           Certainly, resort to the plain language of the statute may be insufficient to determine its purpose.  The courts may then have resort to other extrinsic evidence of legislative intent.  However, this does not mean that the courts may simply accept a purported purpose which sounds acceptable.  The type of extrinsic evidence that may properly be considered is relatively narrow in scope.  Legislative history is an acceptable source, although as indicated above, it must be viewed with caution.

187                           I turn, then, to the legislative history of Part III of the FLA.  In introducing the FLRA in 1976, the then Attorney General, the Hon. Roy McMurtry, explained the “limited obligation” of one common law spouse to support the other.  He stated:

Where two persons have lived together as if married, their relationship often takes on the same financial characteristics as a marriage.  One person frequently becomes dependent on the other, especially if there is a child of the union.  If one of these two people is no longer self-sufficient, it seems reasonable to look to the other to assist in restoring him or her to financial independence.  Certainly, it is more desirable to place a support obligation on common law spouses than to have a large number of persons who are living common law looking to public welfare for support instead. [Emphasis added.]

(Legislature of Ontario Debates, October 26, 1976, at p. 4103).

188                           One cannot read this passage without being struck by two observations.   The first is the obvious intention of the Attorney General that the motive underlying the extension of support obligations to include unmarried cohabiting opposite-sex couples was premised on the social reality that such relationships often exhibit a dynamic of dependence, and that this dynamic of dependence usually arises because the couple has children and the mother is the primary caregiver.  The second observation is that the legislation was not aimed at “need” abstractly defined, or need generated by “interdependence”.  Instead, the legislation addressed the specific need arising when an individual (typically a woman) cohabits in an opposite-sex relationship, and relying upon the expectation of continued and future support from her partner, foregoes existing employment prospects that rendered her self-sufficient before entering into the relationship, and new ones that arise during the relationship.  The above-quoted passage makes specific reference to a party who is “no longer self-sufficient” (emphasis added), implying that she once was, and indicates that the obligation being imposed on the other party is one of “restoring” her to her previous state of self-sufficiency which was eroded by the relationship. 

189                           On second reading, the Attorney General responded to concerns that had been expressed to him that the proposed legislation was overinclusive, in the sense that it would apply to unmarried cohabiting opposite-sex couples who had entered into relationships “with no strings attached” -- that is, in the expectation that no support obligations would be imposed on them -- or that they would be otherwise adversely affected.  The Attorney General stated that:

These are the very people who will be unaffected by this legislation because they do not form a financial dependency on each other.  That is the very essence of their “no strings attached” relationship.  If there is no financial dependency, there will be no need for support by the other party and there will not be a successful claim for support.

By contrast, however, there are many people living together in such relationships who are being exploited by their partner[s].  They have been induced to enter into the relationship and to stay home and raise the children arising from the union, or children of another union, and have thus been put in a position of total dependency on the person as a result of being out of the labour market for a lengthy period of time.  Many of these people are later abandoned and, under the present law, they have nowhere to turn but to the welfare authorities for support.

This is not a small problem.  For example, in September of this year, the government of Ontario has paid out family benefits to over 13,000 unmarried mothers and their 26,000 dependent children, totalling over $3.5 million for that month alone. 

(Legislature of Ontario Debates, November 18, 1976, at p. 4793).

190                           In essence, then, the Attorney General acknowledged that the legislation was, to a certain degree, overinclusive, in that even those individuals in unmarried opposite-sex couples who had not become dependent upon their partners were included within its scope.  He explained this overinclusion on the basis that an order would only actually be made where dependency was demonstrated.  Those who truly did live in a “no strings attached” relationship would have no new obligations imposed upon them.  However, the Attorney General’s position was that the very nature of cohabiting opposite-sex relationships meant that they had in fact generated an appreciable degree of dependency in a significant number of individuals.  Strings may, in fact, have become attached even where the parties did not originally intend that this should be so. 

191                           This passage provides further evidence that the focus of the legislation was to ameliorate the condition of persons, the vast majority of whom were (and are) women, who had become dependent upon their spouses, commonly, though not exclusively, as a result of the mutual decision to raise children.  This legislative history, when combined with the historical survey set out above, and the text of the legislation itself, leaves one in no doubt as to the purpose of the FLA in general, and of the spousal support provisions in particular. 

192                           Having established the legislative purpose, the Court must scrutinize the purpose of the impugned provision to ensure that it is not itself discriminatory:  Law, supra, at para. 80.  We must thus take seriously the respondent M.’s suggestion that the real purpose of the exclusion is discrimination itself.  A legislative purpose may be discriminatory because that was the intention of the legislature.  Here, I see no evidence that the exclusion of same-sex couples (or indeed, any other couples or groups) from the scope of s. 29 of the FLA is motivated by animus toward gays or lesbians.  All the respondent M. can point to is the Legislative Assembly’s adherence to the existing definition.  Yet that contention does not assist the respondent M.  Merely asserting that something is discriminatory or stereotypical is insufficient:  Law, supra, at para. 59.  Even if it was not intended to be discriminatory, the legislative purpose can still objectively be shown to be discriminatory.  To demonstrate that a legislative purpose is objectively discriminatory, we must assess the claim in light of all of the contextual factors set out in Law.  I return to this second aspect later in these reasons.

Competing Views on the Family Law Act

193                           Before considering whether s. 29 of the FLA violates s. 15(1)  of the Charter , I wish to address the competing views on the purpose of the FLA.  My colleagues have chosen to consider the purpose of the FLA only as part of their s. 1 analysis.  Yet, as I explained in para. 162 above, identifying the purpose of the legislation under consideration is essential both to establishing the appropriate comparison group and to assessing the contextual factors at the third stage of the Law analysis.  It is certainly not a task confined to s. 1.  My primary objection to my colleagues’ analysis is that they do not provide a coherent account of the purpose of the impugned legislation.

194                           My colleague Iacobucci J. states that he does not “dispute the claim that economically dependent heterosexual women and children are well served by the spousal support provisions in Part III of the FLA”, but he takes the view (at para. 87) that

there is insufficient evidence to demonstrate that the protection of these groups informs the fundamental legislative objectives behind this part of the Act.  In fact, with respect to Bastarache J., it seems to me that the legislative history and the terms of the provisions themselves contradict the appellant’s assertions.

For the reasons given above, I do not agree.

195                           With respect, my colleagues attribute a purpose to the impugned provision, and the FLA as a whole, that it does not bear.  Rather than analyze the FLA itself, Iacobucci J., like the majority of the Court of Appeal below, ascribes to the impugned legislation a purpose that bears little relation to the actual statute, its structure, or its history.  It comes as no surprise that, having ascribed to the FLA a purpose that its language does not bear, my colleague then strikes down the legislation for failing to fulfil a purpose never intended by the Legislative Assembly.  Given that my colleague’s analysis depends upon the purpose so chosen, my view is that his failure to determine the true purpose of the FLA fatally undermines that analysis. 

196                           Iacobucci J., following the majority in the Court of Appeal below, relies upon a definition of the purpose of the impugned legislation that bears but a tangential relationship to the text of the FLA or its history.  My colleague derives the purpose from the Ontario Law Reform Commission’s Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act, supra, at pp. 43-44:

The purpose of the Family Law Act is to provide for the equitable resolution of economic disputes that arise when intimate relationships between individuals who have been financially interdependent break down (Parts I‑IV).  As well, it ensures that family members have a means to seek redress when an immediate relative is injured or killed through the negligence of a third party (Part V).

197                           The purpose attributed to the FLA in this passage is not necessarily inconsistent with the actual purpose of the FLA.  However, I must emphasize that if mere consistency were the sole test, nothing would prevent the courts from inventing a supposed legislative purpose.  This would be radical in theory, highly undesirable in practice, and conducive to indeterminacy in its consequences.  Many supposed purposes are conceivably consistent with the purpose of the FLA.  That does not, however, transform them into the purpose(s) of the FLA.  To take but one example, the phrase “intimate relationship” nowhere appears in the FLA.  This Court rejected the doctrine of shifting purpose in R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295, at pp. 334-35, confirming that a law’s purpose is set on its enactment, and cannot be subsequently altered other than by amendment.

198                           Iacobucci J., himself relying upon certain legislative history, contends that the references in the FLA’s preamble to the goal of encouraging and strengthening the role of the family have no application to the present appeal because “the legislation is actually intended to deal with the breakup of the family” (emphasis added) (para. 84).  With respect, I do not agree.  To begin, the support obligation addressed in Part III of the FLA does not arise solely upon the breakdown of a relationship.  To the contrary, that obligation is fundamental throughout the course of a marriage or common law relationship.  In practice, few cases may involve applications for support during the subsistence of a relationship, but that does not affect the general proposition that the obligation of spousal support exists throughout the relationship, and is not merely an incident of its breakdown.

199                           Second, I am not persuaded that the legislative provisions creating a spousal support obligation that may be invoked during or upon the breakdown of a conjugal relationship cannot be said to encourage and strengthen the role of the family, in accordance with the language of the preamble to the FLA.  Surely an individual contemplating marriage or a spousal relationship would be encouraged, rather than discouraged, from entering such a relationship by the knowledge that should the relationship break down in the future, he or she would be able to rely upon an efficient mechanism to provide him or her with support from his or her partner in the event that the individual had become dependent on his or her partner during the course of the relationship.  One is not discouraged from a sea voyage by the presence of lifeboats on board.  Consequently, I do not accept that this argument has any merit.

200                           Iacobucci J. also suggests that the gender-neutral language of the spousal support provisions of the FLA weighs in favour of the view that the purpose of those provisions was to address economic dependency more generally.  I do not agree.  First, as a general proposition, I doubt that the use of gender-neutral language in legislation necessarily manifests such an attitude.  The statute books are full of examples of legislation which, though framed in gender-neutral language, are intended as a practical matter to apply in the vast majority of cases to one gender.  Second, in the particular circumstances of this appeal, the use of gender-neutral language in the impugned legislation cannot be advanced as evidence that the purpose of the legislation was to address dependency generally, rather than, as I believe, to address (primarily, at least) the particular dynamic of dependence unique to certain heterosexual relationships, which has particular effects for women cohabiting in opposite-sex relationships.  I concur with Bastarache J. (para. 342) that the use of gender-neutral language in the FLA provides no support for the implication that the purpose of the FLA’s spousal support provisions was to address “interdependence” unrelated to the social reality of the relationship.  This Court specifically rejected such a theory in Moge v. Moge, [1992] 3 S.C.R. 813, at p. 874, and I can see no reason why it should be exhumed here.

201                           It is beyond dispute that at the time the 1978 amendments to the FLRA were enacted, the overwhelming majority of spouses claiming support were women.  The same would likely be true today, although due to the relative improvement of the economic status of women as compared to men, I expect that the number of claims made by men against women will since have increased both in relative and in absolute terms.  Nonetheless, even today, the vast majority of claims made under Part III of the FLA are made by women against men.  When the 1978 amendments to the FLRA were first introduced, the tenor of the times favoured gender-neutral language.  The Legislative Assembly recognized that in rare cases, a man might be able to make out a claim against a woman.  However, it expected that the vast majority of claimants would be women.  To argue that the use of gender-neutral language in the FLA manifests an intention on the part of the legislature to ignore historical and social fact is entirely unsubstantiated.  It would be a strange triumph of form over substance.  Iacobucci J. argues that “[i]n the face of this clearly gender-neutral scheme, the fact that a significant majority of the spousal support claimants are women does not ... establish that the goal of Part III of the FLA is to address the special needs of women in opposite-sex relationships” (para. 91).  See also the reasons of Cory J., at para. 54.  In response, it must be observed that even in the small number of cases in which men make claims against their female spouses, those claims are still claims arising out of opposite-sex relationships, which generate their own dynamic of dependence regardless of who is making a claim.

202                           In this appeal, the Attorney General has submitted that the purpose of the spousal support provisions of the FLA is to remedy the systemic sexual inequality associated with opposite-sex relationships, including the economic dependence of women on men arising from the fact that women usually assume primary responsibility for parenting, and from gender-based inequalities in earning power and labour force participation. A review of the FLA provisions, properly interpreted, reveals that this is their purpose.  Indeed, as our Court recognized in Moge, supra, at pp. 849-50, family law legislation such as the FLA, though facially gender-neutral in its provisions, is, as a practical matter, generally (though not exclusively) invoked by women upon the breakdown of a marriage.  This Court also recognized that “[w]omen have tended to suffer economic disadvantages and hardships from marriage or its breakdown because of the traditional division of labour within that institution” (Moge, at p. 861).  The Court went on to observe, at p. 867, that “[t]he most significant economic consequence of marriage or marriage breakdown, however, usually arises from the birth of children”.  Moreover, the Court considered, at p. 869, that “[i]t is important to note that families need not fall strictly within a particular marriage model in order for one spouse to suffer disadvantages.  For example, even in childless marriages, couples may also decide that one spouse will remain at home.”  It will be seen that this is a gendered assumption, and so may occur even when there are no children.

203                           A further argument advanced by my colleague is that one of the purposes of Part III of the FLA is to reduce provincial expenditures on social assistance, by relieving the province of the burden if a partner were to seek recourse to social assistance.  That such a purpose is reflected in the legislation is undoubted, and there are certainly instances of support for such a view in the legislative history (e.g. Legislature of Ontario Debates, October 26, 1976, at p. 4103).  It is of interest that dower, the old life interest held by a widow in her deceased husband’s estate, itself a form of maintenance obligation extending even beyond the death of the husband, was justified in part as a means to “prevent poorer widows from becoming a public liability”:  Chambers, supra, at p. 19.  Nevertheless, reducing expenditures on social assistance is not an overriding concern and cannot be pushed too far.  What legislative scheme is not extolled by its proponents as saving money for the public purse, or operating existing benefit schemes more efficiently?  The legislative history suggests that imposition of a support obligation upon unmarried spouses was seen primarily as a mechanism to encourage unmarried couples to marry:  Legislature of Ontario Debates, October 26, 1976, at p. 4103.  Fiscal considerations were decidedly secondary.  Indeed, that purpose has, to a considerable degree, been achieved:  Ontario Law Reform Commission, Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act, supra, at p. 6.

204                           In conclusion, Iacobucci J. denies that the purpose of the FLA more broadly, and Part III in particular, is to address the dynamic of dependence unique to cohabiting opposite-sex couples, and to remedy the disadvantages suffered by women in opposite-sex relationships (at para. 103) as well as to ensure the protection of children (at para. 104).  I cannot agree.  In my opinion, the text of the legislation, the preamble, and the legislative history inescapably lead to the conclusion that such is its very purpose.

205                           My colleague Bastarache J. describes the purpose of the FLA in various ways.  One purpose is a desire to impose support obligations upon “partners in relationships in which they have consciously signalled a desire to be so bound”, where those relationships possess a level of “permanence and seriousness”, where one of the partners assumes a degree of household responsibilities, or other sacrifices for the common benefit of the couple, and which cause or enhance an economic disparity between the partners (at para. 347).  A second purpose ascribed to the legislation is to ensure “a greater degree of autonomy and equality within the family unit” (para. 320).  A third purpose he attributes to the legislation is a “recognition of the significance of the procreative and socializing role of the opposite-sex family” (para. 318).  The final purpose Bastarache J. locates is the need to address “the widespread economic dependence faced by women, inside or outside the marital bond, upon the breakdown of the relationship” (para. 340).

206                           In addition to these various purposes, Bastarache J. correctly points out that the extension of this obligation to individuals in common-law relationships was to reflect “the subordinated position of women in non-marital relationships” (para. 339).  Further, he points out that the extension of this obligation to men was merely a reflection of the concern to “erase official gender bias” in legislation (para. 342). 

207                           In my view, the purposes as described by Bastarache J. are similar in many respects to the one I have identified.  As I discuss above, I am in agreement with his views on why the legislation was extended to include men in opposite-sex relationships.   I am further in agreement as to the focus of this legislation on women in opposite-sex relationships, which generally possess characteristics of economic disparity between the woman and her partner.  Where Bastarache J. and I part company is in our answers to the question “What is the purpose in excluding [same-sex couples]?” (para. 355).  In my opinion, Bastarache J. has already answered that question in his reasons.  He points out, at para. 298, that same-sex couples generally do not suffer the same economic and social inequalities in the relationship.  Both partners in a same-sex relationship are generally wage-earners (at para. 298).  There is not the same level of economic dependency between individuals in a same-sex relationship as there is in opposite-sex relationships (at para. 298).  Opposite-sex relationships perform a “social function” of raising children, and this function is “typically applicable” only to them (para. 319).  I wholly agree.  However, in my opinion, these conclusions, coupled with an analysis of the rest of the contextual factors, lead to a finding that the s. 15(1) claim must fail. 

208                           My colleague Bastarache J. concludes that notwithstanding the numerous contextual factors that support the legislature’s initiative in alleviating the position of women in opposite-sex relationships, there is no reason to exclude individuals in same-sex relationships. My colleague’s position would make it extremely difficult for this Court to deny the expansion of the scope of the spousal support regime to any relationship of dependency.  Making permanence the touchstone of the legislation will lead, in my view, to extending the FLA regime to any two (or more) people in relationships of permanence, so long as they can demonstrate exclusion on the basis of an enumerated or analogous ground.  The category of potential claimants may conceivably be quite large:  we simply do not know; the point is not addressed.  Such a consequence, however, flows inescapably from the premise underlying his reasons. 

209                           Bastarache J. asserts that underlying this concern for women rests the notion of dependency.  Further, he concludes that barring any evidence demonstrating that no homosexuals are dependent, they should be included.  In my view, this is too rigid and formalistic an application of the analysis set out in Law, supra.   As I explain below, the reason for excluding individuals in same-sex relationships is that there is an absence of reasons for including them.  The circumstances of opposite-sex couples are sufficiently unique to warrant granting a benefit (and imposing a burden).  There is, to my mind, no reason why the Legislative Assembly cannot take these realities into consideration, and indeed, make specific efforts to alleviate the underlying problems.  This is further developed in the evaluation of the contextual factor of “correspondence”, below.

C.  The Attorney General’s Concession

210                           I begin with an examination of whether s. 29 of the FLA violates s. 15(1)  of the Charter .  Before doing so, however, I must address a preliminary issue.  In both oral and written argument before us, the Attorney General conceded that s. 29 of the FLA violates s. 15(1)  of the Charter , but endeavoured to have the Court uphold the provision under s. 1.  This Court has traditionally taken a dim view of concessions in constitutional cases, given their potentially wide ramifications for persons or governments not parties to the particular case.  It is no different here.  Interveners may provide assistance to the Court, but their role is constrained, and in any event, their submissions are no substitute for full argument by the parties. 

211                           I would not wish to be taken as suggesting that parties before this Court should refrain from making concessions where appropriate.  A refusal to make appropriate concessions may undermine a litigant’s position.  That said, sometimes a concession is inappropriate, particularly where it deprives the Court of submissions on a central issue in the appeal.  As Cory J. rightly notes (at para. 45), we are not bound by the Attorney General’s concession, and speaking for myself, I think that it was the wrong position for the Attorney General to have adopted.

D.  Locating the Appropriate Comparator Group in the Present Appeal

212                           In Law, supra, Iacobucci J. pointed out at paras. 56-58, that it is essential to establish the appropriate comparator.  The equality guarantee contained in s. 15(1) does not guarantee equality in the abstract, and does not protect against violations of human dignity per se.  We must establish that there is differential treatment as between the claimant and another group.  The factors to consider include the purpose and effects of the legislation, and the biological, historical and sociological similarities or dissimilarites of the claimant and the comparator. To establish who the correct comparator is, we begin with the choice made by the claimant.  However, by analysing the various factors, it may be that the legislation in fact targets a different group, and the differential treatment is between the claimant and a comparator unselected by the claimant.   In my view, that is the case here.

213                           As I discuss below, many of the submissions made by the parties related to the similarities and dissimilarities between same-sex couples and opposite-sex couples.  Given the subject-matter of this legislation, many of these dissimilarities are highly relevant.  Again, it is my view that the subject-matter of this legislation relates to the unique social function played by opposite-sex couples, and to alleviate the economic disadvantages that flow from that social function.  In so doing, it targets primarily, but not exclusively, women who play a disproportionate child-rearing role in heterosexual relationships:  Moge, supra. 

E. Does the Legislation Create a Distinction?

214                           As indicated above, the first step in a s. 15(1) analysis is to determine whether the legislature has drawn a distinction between the claimant and others.  There can be little doubt in the present appeal that such a distinction exists.  Individuals who are part of a married couple or a cohabiting opposite-sex couple of a specified duration are entitled to invoke the spousal support provisions of Part III of the FLA.  By contrast, due to the definition of “spouse” in s. 29 of the FLA, individuals who are part of same-sex couples have no access to the spousal support provisions.  It must be emphasized, as this Court has held repeatedly, that not all distinctions drawn by the legislature will infringe s. 15(1).  There must be more.  I underscore the concern expressed by La Forest J. in Egan v. Canada, [1995] 2 S.C.R. 513, at para. 7:

... 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 inquiry in every case involving a protected group.

Nevertheless, the legislative provision under consideration in the present appeal clearly makes a distinction.  It must still be determined whether the differential treatment results in discrimination.  We must, therefore, proceed to the next step of the analysis. 

F.   Is the Differential Treatment Based on One or More Enumerated or Analogous  Grounds?

215            The respondent M. asserts that she has been discriminated against on the basis of her sexual orientation.  It is common ground, as this Court has held on several occasions, including Egan, supra, and more recently, Vriend, supra, that sexual orientation is analogous to those grounds of discrimination enumerated in s. 15(1)  of the Charter .  Facially, the legislation does not draw a distinction on the basis of sexual orientation.  The language of s. 29 of the FLA draws a distinction between relationships, not individuals.  As McLachlin J. pointed out in Miron v. Trudel, [1995] 2 S.C.R. 418, at para. 153:  “In theory, the individual is free to choose whether to marry or not to marry.   In practice, however, the reality may be otherwise.”  The same is arguably true for conjugal relationships with members of the opposite sex.  Although the choice to enter into conjugal relationships with members of the opposite sex is available in theory to homosexuals, in practice, it may be no choice at all.   Nonetheless, it is clear that s. 29 does not facially draw a distinction on the basis of sexual orientation.

216            What the legislature has done by its enactment of Part III of the FLA is to draw a distinction between certain -- though I note, not all -- opposite-sex relationships and all other relationships of financial dependence.  As La Forest J. underscored in Egan, supra, the legislature intended to make the facial distinction on the basis of spousal status, rather than sexual orientation.  However, although the distinction drawn by the legislation is made on the basis of spousal status, the effect of this legislation is to include a distinction on the basis of sexual orientation.  From the very beginning of our s. 15 jurisprudence, this Court has stated clearly that equality must be substantive as well as formal, and the principle continues to inform our law:  Vriend, supra, at para. 83.  Put another way, we must examine the effects of the impugned legislative distinction, and evaluate the statute’s consistency with the Charter ’s equality provisions on that basis.  Otherwise, we would create a victory of form over substance.  Even though s. 29 of the FLA does not draw a facial distinction on the basis of sexual orientation, the effect of s. 29’s definition of “spouse” is to do so.

217            Although the distinction drawn by the legislature has the practical effect of excluding same-sex couples who may live in relationships of financial dependence,  many other relationships of financial dependence are also excluded from the scope of Part III of the FLA.  All individuals who cohabit in non-conjugal relationships are excluded from its scope.  Indeed, even opposite-sex couples who have not cohabited continuously for a period of three years are excluded, unless they have children, despite the fact that they may exhibit financial dependence.  Nonetheless, the fact that others may also be excluded does not alter the fact that individuals in same-sex relationships are treated differently, and we must therefore ask whether this differential treatment is discriminatory on the basis of the analogous ground of sexual orientation.

G.  Does the Differential Treatment Discriminate?

1.  Does the Legislation Impose a Burden or Withhold a Benefit?

218            The third broad inquiry involves the determination of whether the distinction based on enumerated or analogous grounds is discriminatory.  The first aspect of this inquiry is to ask whether the legislature imposes a burden upon or withholds a benefit from the claimant.  The direct or indirect effect of the impugned legislation is considered.  Again, I agree with the majority that it cannot seriously be denied that the distinction drawn by the legislature in this case has resulted in disadvantage to the claimant.  Part III undoubtedly bestows a benefit -- access to a spousal support regime -- the denial of which results in a disadvantage to the respondent M.  Of course, as the majority rightly points out, the issue on this appeal is the threshold question of access to Part III of the FLA, not the viability or merit of any claim that the respondent M. might make under it.  To that extent, at least, this case is broadly similar to Vriend, supra, in which this Court considered the threshold question of access to Alberta’s human rights scheme rather than the merits of a claim made under that scheme.

219            I agree with the majority that access to the spousal support provisions of Part III of the FLA is a benefit in the way that this Court has defined that term for the purposes of s. 15(1)  of the Charter .  As Finlayson J.A.  observed in the Court of Appeal below ((1996), 31 O.R. (3d) 417, at p. 427), the spousal support regime set out in Part III of the FLA “is far more comprehensive than any known to the civil law in other areas”.  Aside from imposing the legal obligation to support one’s spouse, the FLA also provides a mechanism by which to apply for a support order (s. 33), outlines the powers of a court to make (s. 34) and to enforce (s. 35) such an order. Orders made pursuant to the FLA enjoy certain benefits:  they bind an estate (s. 34(4)) and may be indexed (s. 34(5)), and are accorded a degree of priority in the event that the obligee defaults or becomes insolvent (Bankruptcy and Insolvency Act , R.S.C., 1985, c. B-3, ss. 121(4) , 136(1) (d.1), 178(1) (c); Family Responsibility and Support Arrears Enforcement Act, 1996, S.O. 1996, c. 31). The support process is also facilitated and expedited by the requirement, contained in s. 41, that the parties must exchange financial statements.  In short, Part III of the FLA provides a relatively efficient and effective mechanism by which to resolve spousal support claims.  Part III of the FLA provides a benefit to those entitled to invoke it, and denial of access to that benefit is a disadvantage for the purposes of s. 15(1)  of the Charter .

220            At the same time, it is important to recognize, as my colleague Bastarache J. has done, that this legislation both provides a benefit to a class of people and imposes a burden on that same class.  He notes, at para. 300, that “[t]he cost of imposing the s. 29 regime is the reduction of autonomy of individuals affected.”  The legislation alters the fundamental right of individuals to structure their affairs freely, and, where the circumstances so warrant, imposes a support obligation where none previously existed.  Not only does the legislation impose a burden by reducing an individual’s autonomy, it also reduces some of the financial advantages of relationships that fall outside the definition of “spouse”.   This was expressly noted by the then Attorney General, Hon. Roy McMurtry, when discussing the rationale for extending the support obligation to common law spouses (Legislature of Ontario Debates, October 26, 1976, at p. 4103):

We believe it is in the community’s interest that some legal responsibilities flow from a common law relationship.  By imposing a support obligation on common law spouses in the same terms as the obligation on married persons, we will be removing at least some financial advantages of a common law union over a legal marriage.

2.  Stereotypical Application of Presumed Characteristics and the Contextual                Factor of Correspondence

221            Having established that the legislation confers a benefit, we must then turn to the question of whether the benefit is withheld “in a manner which reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration”:  Law, supra, at para. 88.  As I explain below, the concept of “stereotyping” often is linked to the contextual factor of “correspondence”.  Where the legislation takes into account the actual need, capacity or circumstances of the claimant and the comparator, then it is unlikely to rest on a stereotype.

222            The concept of stereotyping is well-rooted in the equality jurisprudence.  Although the central purpose of the s. 15(1) equality guarantee is to protect and promote human dignity, one of the principal mechanisms by which the Charter ’s equality guarantee achieves that purpose is by barring the use of stereotypes as a basis for legislative distinctions.  The jurisprudence recognizes that in the absence of reliance on a stereotype, discrimination will be difficult to establish:  Law, supra, at para. 64.  Wilson J., dissenting on other grounds in McKinney v. University of Guelph, [1990] 3 S.C.R. 229, at p. 387, confirmed that 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”.  McLachlin J. indicated in Miron, supra, at para. 128, that the unequal treatment at issue must be “based on the stereotypical application of presumed group or personal characteristics”.  It is true that in Law, this Court unanimously held that there may be, on occasion, other infringements of s. 15(1) which are not rooted in stereotype or prejudice.  However, such occasions will likely be rare.

223            Distinctions drawn on enumerated or analogous grounds usually rely upon the stereotypical application of presumed characteristics, rather than an accurate account of the true situation, or actual abilities, circumstances, or capacities.  The language of s. 15(1) leads us to suspect that distinctions made on the basis of enumerated or analogous grounds are discriminatory.  Indeed, that is precisely why they were selected as enumerated -- and later, analogous -- grounds, because our historical experience has shown that stereotypes and social disadvantage have often clung to them, with resulting infringements of human dignity:  Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, at p. 175.  Legislative distinctions drawn on those grounds will usually bear no relation to the “true situation” of the claimants.

224            As Iacobucci J. noted in Law, supra, at para. 64, generally speaking, a stereotype is an inaccurate generalization made with reference to a personal characteristic.  Specifically, it involves the attribution of a characteristic or set of characteristics to a group, which is then imputed to the individual members of that group because of their membership in it.  One takes a preconceived or fixed notion about a group of individuals identified by a personal characteristic, and assumes that all individuals identified by that personal characteristic fit that preconceived idea.  In invoking a stereotype, one either attributes characteristics which are not present, or fails to take into account characteristics which are present.  Either way, one is failing to treat an individual as he or she really is.

225            The concept of “stereotyping” is often linked to the contextual factor of “correspondence”.  At para. 70 of Law, supra, Iacobucci J. states that “legislation which takes into account the actual needs, capacity, or circumstances of the claimant and others with similar traits in a manner that respects their value as human beings and members of Canadian society will be less likely to have a negative effect on human dignity”.  Where the distinction is so directly related to the objective reality of the ground under consideration as to be, in effect, a simple description or reflection of the circumstances, merit, or actual situation of the claimant, any differences that exist in such cases may be due to social, biological, or constitutional facts, rather than discrimination or prejudice, and exist apart from any legislative distinctions based upon them.  They are beyond the scope of the ground of discrimination.  Indeed, failure to treat those differences as differences may itself constitute a breach of the requirements of substantive equality.  A description is unlikely to be a stereotype when it is an accurate account of the characteristic being described.

226            Where the legislation “takes into account the claimant’s actual situation” in a manner that reflects his or her human dignity, it will be less likely to infringe s. 15(1).  With this understanding in mind, I now turn to consider these factors in light of the legislation under challenge in this case.

227            The contextual factor of correspondence is critical for determining the central issue in this appeal, whether the analogous ground of sexual orientation nourishes a challenge to the definition of “spouse” and renders the exclusion of gay and lesbian persons from that definition discriminatory under s. 15(1)  of the Charter .  The definition of “spouse”, as I have already suggested, is an extension of marriage.  To be a spouse is, in essence, to be as if married, whether or not one is actually married.  Spousehood is a social and cultural institution, not merely an instrument of economic policy.  The concept of “spouse”, while a social construct, is one with deep roots in our history.  It informs our legal system:  the status of “spouse” is defined, recognized, and regulated by the legislature.  Under the Constitution Act, 1867 , the federal government regulates capacity to marry (s. 91(26)), and the provincial governments regulate solemnization of marriage, that is, matters of form and ceremony (s. 92(12)), along with property and civil rights (s. 92(13)).  It is rooted in Western history, in which the concept of “spouse” has always referred to a member of a cohabiting opposite-sex couple.  That is what it means to be a “spouse”:  Andrews v. Ontario (Minister of Health) (1988), 64 O.R. (2d) 258 (H.C.J.), at p. 261; Dean v. District of Columbia, 653 A.2d 307 (D.C. 1995), at p. 362.  That well-recognized definition does not discriminate on the basis of sexual orientation, any more than the status of “child” or “adult” discriminates on the basis of age, or “male” or “female” discriminates on the basis of sex. 

228            Our system of family law is, to a great degree, based upon the legal rights and duties flowing from marriage.  The law recognizes marriage as a status voluntarily acquired through contract, and endows that status with both rights and duties.   Individuals who do not live within the institution of marriage are, for the most part, not subject to the rights and obligations that attach to it.  So, for example, the FLA’s sections on family property (Part I) and the matrimonial home (Part II) apply only to married couples, not unmarried, cohabiting opposite-sex couples, or anyone else.  In general, unmarried individuals who cohabit outside marriage do not bear the full range of legal responsibilities of a married couple, and neither the benefits nor the burdens of marriage extend to them.  As I explained in Miron, supra (and as La Forest J. outlined in greater detail in Egan, supra), marriage is a fundamental social institution because it is the crucible of human procreation and the usual forum for the raising of children.  That is the primary, though not sole, purpose of the institution of marriage: Layland v. Ontario (Minister of Consumer and Commercial Relations) (1993), 104 D.L.R. (4th) 214 (Ont. Div. Ct.), at pp. 222-23.  The very importance of such an institution can render description of it difficult, as Dickson C.J. recognized in Brooks v. Canada Safeway Ltd., [1989] 1 S.C.R. 1219, at p. 1237.  Distinguishing description from justification may be similarly difficult.

229            Mere appeal to history or tradition, of course, is insufficient to establish the existence of the actual need, capacity, or circumstances of the claimant or others: Edwards v. Attorney-General for Canada, [1930] A.C. 124 (P.C.), at p. 134.  The proper evaluation of an equality claim requires the Court to subject our social practices to critical evaluation, as Wilson J. explained in R. v. Turpin, [1989] 1 S.C.R. 1296, at p. 1332.  Nonetheless, the recognition that social values change and evolve does not alter the reality that certain biological and social realities endure.  In Big M Drug Mart, supra, at p. 344, and on several occasions since, this Court has observed that the Charter  was not enacted in a vacuum. The Charter requires the Court to subject our social institutions to a critical eye, but it does not mandate that we should sweep them away.  Legislative distinctions based on an accurate appreciation of biological and social realities may not amount to discrimination, where the legislation corresponds to the characteristics in a manner that respects the claimant’s human dignity.

230            Legislation such as Part III of the FLA, which provides differing economic and legal treatment for married couples and opposite-sex common law couples is premised on certain assumptions about the nature of the economic, social, and legal dynamic of the relationships those individuals have.  That economic, social, and legal dynamic, I should emphasize, is a multi-faceted one, and some of its features will be shared by other relationships, including some long-term same-sex relationships.  One of the central questions, therefore, is whether those underlying assumptions, embodied in the distinction drawn in s. 29, are based on stereotypes.  As La Forest J. recognized in Egan, supra, at para. 25, an opposite-sex couple 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. . . . [T]his is the unit in society that fundamentally anchors other social relationships and other aspects of society.  [Emphasis in original.]

231            I pause to address the question of whether the present appeal can be distinguished from the decision of this Court in Egan.  The initial response to this claim is that Egan and this case deal with different legislation, and as such, I agree with Iacobucci J. that the case “must be evaluated on its own merits” (para. 75).  Further, in my view, the present appeal is an easier case than was Egan.  The present appeal is concerned with the support obligation, which, though extended to cohabiting opposite-sex couples, is an essential feature of marriage itself.  That cannot be said (at least to the same degree) about the payment made by the state to opposite-sex couples which was at issue in Egan.  The spousal support obligation is unquestionably a core feature of the institution of marriage itself.  True, that obligation has been extended to unmarried cohabiting opposite-sex couples by legislative action.  Yet that should not obscure the fact that the extension was carefully tailored for a specific purpose, and that the nature of the obligation was established in the marriage context before it was ever extended to unmarried cohabiting opposite-sex couples.  I thus find Cory J.’s statement that “this appeal has nothing to do with marriage per se” (para. 52) entirely unconvincing.

232            The Legislative Assembly has restricted the meaning of “spouse” in Part III of the FLA.  However, it is critical to see that the restriction has not been made on the basis of stereotypical assumptions regarding group or personal characteristics, as set out in para. 64 of Iacobucci J.’s reasons in Law, and as I have set out above.  To the contrary, s. 29’s definition of “spouse” corresponds with an accurate account of the actual needs, capacity and circumstances of opposite-sex couples as compared to others, including  same-sex couples.  Those differences stem, in part, from the biological reality that only opposite-sex couples can bear children.  That biological reality means that opposite-sex couples play a unique social role.  That social role often leads to the well-established economic vulnerability of women in long-term opposite-sex relationships, often (though not always) stemming from the decision to bear and raise children.

233            Cohabiting opposite-sex couples are the natural and most likely site for the procreation and raising of children.  This is their specific, unique role.  Section 29 of the FLA is specially structured to meet this reality.  This can be seen from the fact that the necessary cohabitation period to come within the scope of the section is reduced where there is a child of the relationship.  As well, a child may make a claim under Part III of the FLA against his or her parents, regardless of the parents’ relationship when the child was conceived, raised, or thereafter (s. 31).

234            This unique social role of opposite-sex couples has two related features.  First, it is notorious, as my colleague Iacobucci J. noted in Symes v. Canada, [1993] 4 S.C.R. 695, at p. 762, that “women bear a disproportionate share of the child care burden in Canada”, and that this burden is borne both by working mothers and mothers who stay at home with their children.  The second feature is that one partner (most often the woman) tends to become economically dependent upon the other. 

235            The evidence is uncontroverted that women in long-term opposite-sex relationships tend to become economically dependent upon their spouses.  This dependence arises for several related reasons.  First, women as a group are economically disadvantaged by comparison to men.  Women, on average, earn less than men do.  A woman cohabiting in an opposite-sex relationship is thus likely to earn less than her male partner.  This relative economic disadvantage is both the cause and the effect of gender roles.  Second, women in long-term opposite-sex relationships commonly -- though, I emphasize, not inevitably -- waive educational and employment opportunities or prospects for economic advancement in order to bear and raise children and to shoulder a greater share of domestic responsibilities.  They are also more likely to choose employment that facilitates temporary exit from and subsequent re-entry into the work force, employment that tends to be more insecure and to pay less.

236            Of course, not all procreation takes place within marriage.  Indeed, recognition of this growing reality was an important impetus to the legislature’s decision to extend certain rights and obligations to unmarried common law opposite-sex couples in the 1970s.  There is, obviously, no requirement that married couples bear children, or have the capacity to do so.  Some married couples are unable to have children.  Some choose not to have children.  Some married couples adopt children.  Conversely, it is possible for same-sex couples to have children, either from previous opposite-sex relationships, through adoption, or through artificial insemination.  So too, of course, can some single individuals.  These circumstances are, however, as La Forest J. observed in Egan, supra, at para. 26, “exceptional”.  To acknowledge that they exist does not alter the demographic, social, and biological reality that the overwhelming majority of children are born to, and raised by, married or cohabiting couples of the opposite sex, and that they are the only couples capable of procreation.  Indeed, by definition, no child can be born of a same-sex union:  a third party must be involved.

237            Even in the absence of children, women in cohabiting opposite-sex relationships often take on increased domestic responsibilities which limit their prospects for outside employment, precisely because their lower average earnings make this an efficient division of labour for the couple.  Again, gender roles are both a cause and effect of this division of labour.  If the relationship breaks down, the woman is usually left in a worse situation, probably with impaired earning capacity and limited employment opportunities.  The economic disadvantages faced by women upon the breakdown of opposite-sex relationships, as indicated by reduced earning capacity, more fragile employment prospects, and underinvestment in education and training,  occur for the very reason that the woman did not anticipate that she would have to support herself.  She engaged in a division of labour with her former partner in the expectation that such an arrangement would yield a higher joint economic position.

238            Moreover, if there are children from the relationship, the woman is far more likely to have custody of them upon breakdown. Although the cost of supporting the children is addressed by the FLA’s recognition, in s. 31, of an obligation of parents to support their children, custody of children makes it more difficult for a woman to earn a living, because it reduces her opportunities to work.  These features, of course, are not present in all long-term opposite-sex relationships, and such evidence as there is suggests that as women’s legal equality is increasingly reflected in social and economic fact, women in long-term opposite-sex relationships as a group are likely to become less susceptible to economic dependence upon their male partners.  Nonetheless, the dynamic I have described continues to be manifested in a great many such relationships and has a well-established empirical basis and deep historical roots. 

239            Although in most cases, the woman in an opposite-sex relationship will be the one to suffer from the systemic dynamic of dependence, this will not always be the case.  As I discussed above, as women’s economic situation improves, more claims will likely be brought by men in the future.  However, I emphasize that while women are the primary group who suffer from this dynamic of dependence, they are not the exclusive group.  The evidence before this Court demonstrates that when the female partner is not suffering from this dynamic of dependence, the male partner often is.  In 1992, for example, 25.2 percent of all married couples were characterized by the wife being the full-time wage-earner, and the husband either working part-time or not at all:  Statistics Canada, Family Expenditure in Canada 1992 (1994), at p. 160.  Studies presented to this Court demonstrate that when men are in such a position, they expect less from their partner in housework:  P. Blumstein and P. Schwartz, American Couples (1983), at p. 151.  In other words, the dynamic of dependence also exists for men in opposite-sex relationships, as a type of “division of labour” is created, where the man will often assume additional duties at home while his partner is at work:  M. S. Schneider, “The Relationships of Cohabiting Lesbian and Heterosexual Couples:  A Comparison”, Psychology of Women Quarterly, 10 (1986),  at p. 234.  Other evidence submitted to this Court demonstrate other forms of dependency that are similarly unique to individuals in opposite-sex relationships:  J. M. Lynch and M. E. Reilly, “Role Relationships:  Lesbian Perspectives”, Journal of Homosexuality, 12(2) (Winter 1985/86), at p. 53.  Although the dynamic of dependence unique to opposite-sex relationships plays out differently for men, it flows from similar factors:  in essence, the dynamic of dependence reduces autonomy and increases attachment in heterosexual relationships:  ibid., at p. 56.

240            These realities are captured in the status of “spouse”.  To my mind, they are realities that the legislature may address by extending some (though not all) of the rights and obligations that attach to marriage to cohabiting opposite-sex couples (again, it must not be forgotten that not all opposite-sex couples are included within the scope of the definition) without transgressing constitutional boundaries on legislative action if it does not also extend them to same-sex couples.  It is both legitimate and reasonable for the Legislative Assembly to extend special treatment to an important social institution.  The position I advance also finds support in foreign jurisprudence, including the recent decision of the Court of Justice of the European Communities in Grant v. South-West Trains Ltd., [1998] I.C.R. 449, at paras. 35-36.

241            It is this dynamic of dependence that the legislature has sought to address by way of Part III of the FLA.  The question before this Court is whether the Charter  compels the extension of the legislature’s efforts to address this problem to long-term same-sex couples.  In arguing that the Charter  does just that, the respondent M. and several of the interveners contend that long-term same-sex relationships manifest many of the features of long-term opposite-sex relationships.  This may well be true.  But to my mind, this argument is inadequate.  It fails to demonstrate that the specific feature of long-term opposite-sex relationships that the legislature has sought to address by way of Part III of the FLA, what I have called the dynamic of dependence, is also present in long-term same-sex relationships.  Indeed, it is almost certain that it could not be, because the dependency of women in long-term opposite-sex relationships arises precisely because they are opposite-sex relationships.  There is simply no evidence that same-sex couples in long-term relationships exhibit this type of dependency in any significant numbers. 

242            Indeed, the evidence before us is to the contrary.  That evidence indicates that lesbian relationships are characterized by a more even distribution of labour, a rejection of stereotypical gender roles, and a lower degree of financial interdependence than is prevalent in opposite-sex relationships:  Schneider, supra, at p. 237.  Same-sex couples are much less likely to adopt traditional sex roles than are opposite-sex couples:  M. Cardell, S. Finn and J. Marecek, “Sex-Role Identity, Sex-Role Behavior, and Satisfaction in Heterosexual, Lesbian, and Gay Male Couples”, Psychology of Women Quarterly, 5(3) (Spring 1981), at pp. 492-93.  Indeed, “research shows that most lesbians and gay men actively reject traditional husband-wife or masculine-feminine roles as a model for enduring relationships”:  L. A. Peplau, “Lesbian and Gay Relationships”, in J. C. Gonsiorek and J. D. Weinrich, eds., Homosexuality:  Research Implications for Public Policy (1991), 177, at p. 183.

243            The evidence before us also indicates that partners in a lesbian couple are more likely to each pursue a career and to work outside the home than are partners in an opposite-sex couple:  ibid., at pp. 183-84; N. S. Eldridge and L. A. Gilbert, “Correlates of Relationship Satisfaction in Lesbian Couples”, Psychology of Women Quarterly, 14 (1990), at p. 44.  As members of same-sex couples are, obviously, of the same sex, they are more likely than members of opposite-sex couples to earn similar incomes, because no male-female income differential is present.  For the same reason, the gendered division of domestic and child-care responsibilities that continues to characterize opposite-sex relationships simply has no purchase in same-sex relationships. 

244            Undoubtedly, in some same-sex relationships, one partner may become financially dependent on the other.  This may happen for any number of reasons, including explicit or implicit agreement, differences in age, health, or education, and so on.  However, no pattern of dependence emerges.  Put another way, dependence in same-sex relationships is not systemic:  it does not exhibit the gendered dependency characteristic of many cohabiting opposite-sex relationships.  Due to the high degree of equality observed in lesbian relationships, very few women were dependent on their same-sex partners for financial support, and even differences in income between same-sex partners did not affect women’s perception of their financial dependence on one another in same-sex relationships:  Lynch and Reilly, supra, at p. 66.

245            Mere need in an individual case, unrelated to systemic factors, is, in my view, insufficient to render the FLA’s scheme constitutionally underinclusive.  This is especially true because the scheme involves, in counterpart to an access to support, a restriction on freedom and a burden.  Taken as a group, same-sex relationships simply do not resemble opposite-sex relationships on this fundamental point.  Consequently, I see no reason why the Charter  requires the legislature to treat them identically with regard to it.  The FLA scheme was intended to address need of a particular kind, and does so.  There is no evidence that that particular need exists to any significant degree outside of long-term opposite-sex relationships.  Consequently, although the distinction drawn by s. 29 of the FLA undoubtedly denies a benefit to individuals in same-sex couples, and the distinction may be seen, in its effects, to be drawn on the basis of sexual orientation, no discrimination arises, because no stereotypical assumptions motivate the distinction.  On the contrary, the legislation takes into account the claimant’s actual need, capacity and circumstances as compared with individuals in opposite-sex couples and by doing so it does not violate human dignity.

246            Indeed, the then Chair of the Ontario Law Reform Commission, discussing the Commission’s 1993 Report, made just such an observation (J. D. McCamus, “Family Law Reform in Ontario”, in Special Lectures of the Law Society of Upper Canada 1993: Family Law:  Roles, Fairness and Equality (1994), 451, at pp. 470-71):

The provisions of the Family Law Act appear to be designed to give effect to the sorts of reasonable expectations and to compensate for the kinds of uncompensated contributions that are likely to occur in marital relationships and in unmarried heterosexual relationships that follow a similar pattern.  The kinds of expectations and contributions that arise in these contexts rest, to some extent at least, on gender-based assumptions concerning such questions as the division of labour within the home and  the making of decisions concerning career choices and other economic matters....  The general pattern of such relationships, and, moreover the confusion generated by partial coverage, was sufficient justification, in the Commission’s view, to warrant inclusion of unmarried heterosexual couples on an ascriptive basis.  It was also the Commission’s view, however, that we have, at the present time, no sound basis for assuming that similar considerations would apply to significant numbers of gay and lesbian couples.  It may be that gay and lesbian relationships do not, as a general proposition, correspond to the same model and accordingly, that equality considerations do not demand identical treatment. [Emphasis added.]

247            My colleague Iacobucci J. concedes this point in stating that “there is evidence to suggest that same-sex relationships are not typically characterized by the same economic and other inequalities which affect opposite-sex relationships” (para. 110).  In his view, however, this only goes to show that allowing members of same-sex couples to bring applications under Part III of the FLA might not, as a practical matter, result in the award of many support orders, because such individuals would only infrequently exhibit the dependence or need necessary to obtain an award.  This leads Iacobucci J. to the conclusion that the mere prediction that few claims by members of same-sex couples would succeed if such claims could be brought cannot be used to support the position that such claims should be barred from the beginning.  As I discussed earlier, this same line of thought pervades the reasons of my colleague Bastarache J.

248            However, this contention also proves too much.  Indeed, as Professor (now Dean) Hogg indicates (Constitutional Law of Canada (loose-leaf ed.), vol. 2, at s. 52.7(b)), carried to its logical conclusion, few limitations upon the category of individuals eligible to seek a support order under Part III of the FLA could withstand constitutional scrutiny.  Friends who share an apartment do not typically exhibit a relationship of financial dependence.  Yet there may be particular cases in which they do, and if so, according to my colleague’s reasoning, at least so long as an enumerated or analogous ground can be located, be it race, age, physical handicap or other, one of them should be able to seek a spousal support order under Part III of the FLA, because the legislation would have failed to take into account their already disadvantaged position within Canadian society:  Law, supra, at para. 88.  I find this reasoning unattractive.  As I indicated in Thibaudeau v. Canada, [1995] 2 S.C.R. 627, at para. 143, “legislation must be assessed in terms of the majority of cases to which it applies.  The fact that it may create a disadvantage in certain exceptional cases while benefiting a legitimate group as a whole does not justify the conclusion that it is prejudicial.”  Indeed, such was the case in Law, where it was shown that, in the exceptional case, some younger women may be less capable of providing for their own long-term need.  In this case, many couples (and indeed, groups) live in relationships not comprehended by s. 29’s definition of “spouse”.  Thus, the distinction drawn by the FLA is not between (1) opposite-sex couples and (2) same-sex couples; it is between (1) certain opposite-sex couples and (2) all non-spousal relationships. 

249            I pause to underline that nothing in my reasons should be taken as suggesting that same-sex couples are incapable of forming enduring relationships of love and support, nor do I wish to imply that individuals living in same-sex relationships are less deserving of respect.  To this end, I reiterate the position that the Court recently adopted in Vriend, supra, where I concurred with the majority reasons of Cory and Iacobucci JJ., that discrimination on the basis of sexual orientation is abhorrent and corrosive of our values.  However, the difference between this case and Vriend is that in this case the Legislative Assembly has not discriminated on the basis of arbitrary distinctions or stereotypes.  In Vriend, the stated purpose of the legislation was to address comprehensively discrimination in several contexts such as employment and housing.  By failing to include sexual orientation as one of the protected grounds, the legislation was thus found to be “underinclusive”, having regard to its stated purpose.  Unlike those listed in the legislation, homosexuals were denied access to the remedial procedures specifically designed to redress discrimination.  Whereas in Vriend, the target of the legislation was those individuals who suffered discrimination in these contexts, the legislation here is entirely different.  Here, we are asked whether the legislature violates the Charter  by imposing a special support regime on individuals who are in a particular type of relationship that fulfills a special social function, and has special needs, without extending that support regime to other types of relationships which do not, as a group, fulfil a similar role or exhibit those needs.  Considering all of the contextual factors, I believe that the question must be answered in the negative.  While discrimination on the basis of sexual orientation is abhorrent, mere distinction that takes into account the actual circumstances of the claimant and comparison group in a manner which does not violate the claimant’s human dignity is not.

250            Individuals must be treated with equal respect.  They must not be discriminated against on the basis of the stereotypical application of irrelevant personal characteristics.  Yet the state is not barred from recognizing that some relationships fulfil different social roles and have specific needs, and responding to this reality with positive measures to address those differences.  The state may, in certain circumstances, draw distinctions on the basis of accurate personal characteristics, where those characteristics are germane to the legislative goal.  As Iacobucci J. stated in Law, supra, in some cases “such a difference in treatment [will be] appropriate in light of the historical, biological and sociological differences” between the claimant and comparison group (para. 71).  In my view, the principles adopted by the Legislative Assembly to govern spousal support under Part III of the FLA do not infringe any constitutional requirements, and as a consequence, are within the scope of legislative competence.

251            It is plain to any observer that same-sex couples may, and do, form relationships which are similar in many ways to those formed by opposite-sex couples.  Much of the respondent M.’s argument and evidence sought to demonstrate the similarities between same-sex relationships and opposite-sex relationships.  Same-sex couples may exhibit companionship, love, loyalty, and economic intermingling.  The underlying assumption of the respondent M.’s argument is that s. 29 relies on a stereotype in suggesting that only a couple made up of a “man and woman” exhibit these features.

252            This argument, in my view, misses the mark.  The distinction drawn by s. 29 does not discriminate because it does not involve the stereotypical application of presumed group or personal characteristics, and it does not otherwise have the effect of perpetuating or promoting the view that individuals in same-sex relationships are less deserving of concern, respect, and consideration.   The evidence bears out the contention that the legislative distinction is drawn on the basis of a true appreciation of the facts.  As I see it, the key question is whether there is a correspondence between the ground on which the claim is based and the actual need, capacity, or circumstances of the claimant or others.  I am convinced that there is.  The one feature that distinguishes the two types of relationship is that which is addressed by the FLA:  the necessarily gendered nature of the relationship, which in a great many cases leads to economic dependency based on gender, often (though not always) due to the presence of children.  Moreover, the economic disparity between men and women in relationships only occurs in opposite-sex relationships.  By definition, a same-sex relationship cannot exhibit these features.  Dependency may arise in some same-sex relationships, to be sure, but it must by necessity stem from a different cause, and as I have discussed, it is much less likely to occur.  Where a legislative distinction is drawn on the basis of an accurate picture of capacity and circumstance (no one suggests that merit is at issue here), there is no stereotype, and discrimination is unlikely:  Law, supra; Miron, supra, at para. 132 (per McLachlin J.).

3.  Contextual Factors of Pre-Existing Disadvantage, Stereotyping, Prejudice     or Vulnerability, and Ameliorative Purpose

253            Having established that the legislation does not rest on a stereotype, but rather on the correspondence between the ground of the distinction and the claimant’s actual needs, capacity or circumstances, we must consider whether any of the remaining contextual factors establish that the legislation perpetuates or promotes “the view that the individual is less capable or worthy of recognition or value as a human being or as a member of Canadian society, equally deserving of concern, respect, and consideration” (Law, at para. 88).  One of the contextual factors to consider is whether the claimant experiences “[p]re-existing disadvantage, stereotyping, prejudice, or vulnerability”.  Iacobucci J. explained in Law that an analysis of these factors is relevant because “the perpetuation or promotion of their unfair social characterization . . . will have a more severe impact upon them, since they are already vulnerable” (para. 63).  Additionally, where legislation has an ameliorative purpose, we should consider whether it excludes a historically disadvantaged group (at para. 72).  If it does, it will be an important indicator of discrimination.

254            My colleague Cory J. appears to take judicial notice of the fact that “there is significant pre-existing disadvantage and vulnerability” experienced by homosexuals, which is exacerbated by this legislation (para. 69).  With respect, a more detailed understanding of the pre-existing disadvantage and prejudice experienced by individuals in same-sex relationships is called for, in order to properly assess whether this legislation perpetuates the view that the claimant is less deserving of concern or respect.  By simply labelling same-sex couples as a “disadvantaged group”, Cory J. appears to fall into the trap of creating a “strict dichotomy of advantaged and disadvantaged groups, within which each claimant must be classified” (Law, at para. 68 ).  In my view, it would be incongruous with the purpose of s. 15(1) that a member of a (“generally”) disadvantaged group should have a higher entitlement to the protection of the Charter , even if they are relatively advantaged in relation to the subject-matter of the legislation.  This is why, in Andrews v. Law Society of British Columbia, supra, Wilson J. believed it to be important that “[r]elative to citizens, non-citizens are a group lacking in political power and as such vulnerable to having their interests overlooked and their rights to equal concern and respect violated” (emphasis added) (p. 152). As Iacobucci J. pointed out in Law, supra, at para. 67, “[t]here is no principle or evidentiary presumption that differential treatment for historically disadvantaged persons is discriminatory.”  We must identify what disadvantages the group faces, because no group will be uniformly disadvantaged in all respects.  Simply labelling a group as “disadvantaged” results in a mechanistic and rigid application of s. 15(1), which this Court repeatedly has rejected.

255            The types of disadvantage suffered by homosexual individuals, and by extension, same-sex couples in Canadian society, were canvassed by Cory J. in  dissent in Egan, supra.  He explained that the disadvantage experienced by homosexuals include being the victims of public harassment and verbal abuse, violent crime, and exclusion from some employment and services.  All of these forms of discrimination are abhorrent, and any legislation tending to aggravate or failing to offer equal protection against these pre-existing disadvantages would likely infringe s. 15(1).  However, there are two questions which still must be answered:  are same-sex couples disadvantaged in the area covered by the legislation, and does the FLA aggravate this pre-existing disadvantage? 

256            Are individuals in same-sex relationships disadvantaged in relation to the subject-matter of the legislation?  As I have described in the preceding section, they are not.  Individuals in same-sex relationships do not carry the same burden of fulfilling the social role that those in opposite-sex relationships do.  They do not exhibit the same degree of systemic dependence.  They do not experience a structural wage differential between the individuals in the relationship.  In this sense, individuals in same-sex relationships are an advantaged group as compared to individuals in opposite-sex relationships.  As such, there is no need to consider whether the legislation aggravates or exacerbates any pre-existing disadvantage. 

257            Nor can it be said that the ameliorative legislation excludes a group which is disadvantaged in relation to the subject-matter of the legislation. The main targets of the ameliorative legislation, partners in opposite-sex relationships (particularly women), suffer a structural disadvantage which is unknown to individuals in same-sex relationships.  Just because one form of disadvantage is addressed by this legislation does not mean that all forms of disadvantage must be addressed by it.  In essence, this legislation is the type of ameliorative legislation expressly contemplated by L’Heureux‑Dubé J. in McKinney, supra, and by Cory J. in Vriend, supra, at para. 96:

It is not as if the Legislature had merely chosen to deal with one type of discrimination.  In such a case it might be permissible to target only that specific type of discrimination and not another.  This is, I believe, the type of case to which L’Heureux‑Dubé J. was referring in the comments she made in obiter in her dissenting reasons in McKinney (at p. 436):  “in my view, if the provinces chose to enact human rights legislation which only prohibited discrimination on the basis of sex, and not age, this legislation could not be held to violate the Charter ”.

4.  Contextual Factor of the Nature and Scope of the Affected Interest

258            The fourth factor to consider is the nature and scope of the affected interest.  Here, we are to consider how “severe and localized the . . . consequences on the affected group” are:  Egan, supra, per L’Heureux-Dubé J., as adopted by Iacobucci J. in Law, supra, at para. 74.  The economic, constitutional, and societal significance of the interest is to be considered at this stage.  If the legislation restricts access to a fundamental social institution, or unnecessarily promotes non-recognition of a particular group, then it is a factor which supports the conclusion that the legislation is discriminatory.

259            In this case, it is clear that there is an impact on the claimant’s group.  Of course, the same could be true of any legislation which targets a specific group to receive a benefit (and a corresponding burden).  However, it must be emphasized that the claimants are not deprived of access to support, but rather, they are not forced to participate in a mandatory support regime.  The cost of receiving this benefit is the imposition of a burden on a class of people.  The burden is not only a reduction in autonomy, but also is an imposition of financial obligations.  It is a departure from the general principle in Western societies that individuals enjoy freedom and are expected to provide for themselves.  Fully aware of the historical and systemic disadvantage faced by individuals in opposite-sex relationships (mainly women), the legislature saw fit to depart from this fundamental principle of self-sufficiency and to restrict their freedom by imposing this burden on individuals in opposite-sex relationships.

260            While the legislature does not force individuals in same-sex relationships to provide support, it also does not prevent them from doing so.  Individuals in same-sex relationships are free to formulate contracts which impose support obligations upon themselves, just as the FLA does for some opposite-sex couples.  I do not understand how it could be said that individuals in same-sex relationships are rendered “invisible” by non-inclusion in a regime merely because they have the same rights and obligations as all persons other than certain opposite-sex couples, particularly as they can impose equivalent support obligations by way of contract.  It is clear that this may result in some additional expenses relating to the contract, but it is difficult to see how this possible expense results in discriminatory non-recognition of the group, and how it results in “severe and localized consequences”.  Clearly, when compared with the consequences that flowed from arbitrary non-recognition in the Alberta Individual’s  Rights Protection Act in Vriend, supra (firing from jobs, denial of housing, etc.), these consequences do not appear nearly as “severe”.  More importantly, as I explain in the following section, neither the non-recognition of same-sex couples in s. 29 nor these consequences constitute a violation of human dignity.

5.  Human Dignity

261            The central purpose of the equality guarantee in s. 15(1)  of the Charter  is the protection and promotion of human dignity.  The concept of human dignity is concerned with the autonomy, self-worth, and self-respect of individuals.  As Iacobucci J. points out in Law, supra, “[h]uman dignity is harmed by unfair treatment premised upon personal traits or circumstances which do not relate to individual needs, capacities, or merits” (para. 53).  Human dignity is the lifeblood of the equality guarantee, and forms the basis of each of the contextual factors I have already identified.  However, s. 15(1) is not a guarantee of human dignity per se.  It is a comparative equality guarantee which focuses on discrimination as between groups or persons that leads to a denial of human dignity.

262            As I have laboured to point out at every step in this analysis, in no way does this legislation violate the claimant’s human dignity.  It must be recalled that the vantage point used in making this conclusion is “the reasonable person, in circumstances similar to those of the claimant, who takes into account the contextual factors relevant to the claim”:  Law, supra, per Iacobucci J., at para. 88.  As L’Heureux‑Dubé J. points out in Corbière v. Canada (Minister of Indian and Northern Affairs), [1999] 2 S.C.R. 203, “the ‘reasonable person’ considered by the subjective-objective perspective understands and recognizes not only the circumstances of those like him or her, but also appreciates the situation of others” (emphasis in original) (para. 65).  It is important to note that it is not a denial of human dignity to recognize difference; to the contrary, acknowledging individual personal traits is a means of fostering human dignity.  By recognizing individuality, and rejecting forced uniformity, the law celebrates differences, fostering the autonomy and integrity of the individual.

263            Therefore, we must ask, having considered all of the contextual factors discussed above, should a reasonable person in circumstances similar to the claimant feel her dignity is demeaned?  In my view, I cannot see how she could.  The legislation in question is ameliorative legislation that seeks to better the position of a group of people who are uniquely disadvantaged.  Although the claimant is a member of a group that suffers pre-existing, historical disadvantage, the claimant’s group is relatively advantaged in relation to the subject-matter at hand.  The evidence does not demonstrate that there is the same level of dependency in same-sex relationships that warrants granting the benefit (and imposing the burden) of access to the mandatory spousal obligation regime.  The evidence does not suggest that same-sex relationships fulfil the same social role, nor that they suffer the consequences of that unique social role.  In sum, the legislation takes into account accurate differences between that group and the claimant in a manner which respects the claimant’s human dignity by not relying on a stereotype.

6.  Competing Views on Whether Section 29 of the FLA is Discriminatory

264            I wish to address briefly the opinion of my colleagues on this third broad inquiry about whether the legislation is discriminatory.  Bastarache J. acknowledges that Law, supra, has set out a number of contextual factors we are to consider, but instead of analysing them, he appears content to simply conclude that “[i]n this case, discrimination exists because of the exclusion of persons from the regime on the basis of an arbitrary distinction, sexual orientation.  Having regard to these contextual factors, I am convinced that this distinction is one that puts into question the dignity of the person affected” (para. 291).  With respect, I do not agree with arriving at conclusions on the contextual factors set out in Law so quickly.  In particular, as I discussed above, it appears to me that much of Bastarache J.’s points clearly demonstrate that the legislation corresponds with the reality of the claimant. 

265            My colleagues Cory and Iacobucci JJ. similarly contend that extending the spousal support regime in Part III of the FLA to cohabiting same-sex couples would not undermine the objectives of the legislation, and would enhance the dignity of the claimant.  However, this proposition is the understandable consequence of defining the legislative purpose as Cory and Iacobucci JJ. have done.  It is not responsive to the argument that what the Legislative Assembly has attempted to achieve by way of the FLA is to impose a mandatory support regime on individuals who are in a certain social condition or set of circumstances.  As I have outlined above, excluding cohabiting same-sex couples from the scope of Part III of the FLA does not mean that individuals who are cohabiting in same-sex relationships are any less worthy of respect, nor does it have any objective impact on their dignity as individuals.  To the contrary, the FLA does not impose spousal support obligations upon cohabiting same-sex couples because, in the legislature’s view (and, I might add, in the view of some gays and lesbians, including the respondent H.), same-sex couples do not have the same structural needs as opposite-sex couples, and same-sex couples do not, as a group, fulfil the same social function.  Those needs arise out of the very nature of opposite-sex relationships, and stem from a biological reality.

266            This differing view of the purpose and target of the legislation results in a different analysis throughout each of the contextual factors.  For example, Cory J. explains that because, in his view, the legislation targets conjugal, permanent relationships (which he previously referred to as “intimate” relationships), the legislation does not correspond to the reality of the claimant, who is in a conjugal, permanent (“intimate”) relationship (para. 70).  At para. 69, he explains that the legislation’s ameliorative purpose excludes individuals such as the claimant, who are equally disadvantaged as the target group.

267            In the s. 1 analysis, Iacobucci J. gets to the heart of the matter with his observation that the spousal support provisions contained in Part III of the FLA are “simultaneously underinclusive and overinclusive” (para. 113).  I agree, although in my view, having advanced this proposition, he draws the wrong lesson from it.  In and of itself, the proposition is not remarkable.  It describes almost all legislation.  To legislate is to draw distinctions between people on specified grounds.  Given the capacity of a legislature in a modern regulatory state to govern a province populated by millions of people, those distinctions can never entirely correspond with the particular circumstances of individuals, as was the case in Law, supra.  This point was made well by La Forest J. in Egan, supra, at para. 25:

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....

268            The Charter cannot possibly require the Legislative Assembly to revise the FLA to exclude non-procreative opposite-sex couples from its scope.  As La Forest J. indicated, the legislative and administrative scheme necessary to do so would be highly intrusive and would likely violate Charter  privacy guarantees.  By contrast, exclusion of same-sex couples, who are inherently, rather than situationally, non-procreative, from the FLA support regime raises none of these concerns. 

269            My colleagues, relying upon the statement of Charron J.A. in the Court of Appeal below, suggest that it was the extension of s. 29 of the FLA from married couples alone to cohabiting opposite-sex couples that grounds the respondent M.’s claim in the present appeal.  As my colleagues observe, many of the provisions of the FLA are limited in their application to married couples alone.  The argument, as I understand it, is that although limiting support obligations to married couples alone might be a defensible legislative goal, once the legislature took the step of extending support obligations to unmarried opposite-sex couples, the entire FLA spousal support scheme was imperiled, because the legislature could not fashion a constitutionally sound distinction between unmarried opposite-sex couples and unmarried same-sex couples.  Although not without its initial attractions, this argument ultimately founders. 

270            The entire issue in this appeal is whether the distinction drawn by the Legislative Assembly between cohabiting opposite-sex couples and all other relationships is maintainable.  My colleagues’ implicit suggestion is that the simplest way to render s. 29 constitutionally viable would be to restrict its application to married couples alone.  I see no reason to conclude that the Charter  operates so restrictively.  Here, the Legislative Assembly has made a distinction on the basis of a fundamental biological and social reality to address a unique form of disadvantage specific to opposite-sex couples.  On the other hand, if the Legislative Assembly is obliged to address all of the manifestations of economic interdependence between individuals, it is difficult to see how drawing the line to include same-sex couples, but to exclude those relationships which are not conjugal (and which can be similarly distinguished on enumerated or analogous grounds), can itself withstand constitutional challenge. 

271            It may be, of course, that extending the definition of “spouse” to include same-sex couples or other relationships or otherwise providing for support obligations based on dependency would be a prudent or reasonable policy decision.  My colleague Iacobucci J.’s view, for example, is that to make this extension would advance what he sees as the purpose of the legislation.  However, the wisdom or desirability of such an extension is not itself a matter properly within the consideration of this Court.  The legislature itself considered the desirability of extending the scope of s. 29 to same-sex couples, but in the end, decided to the contrary.  We must take seriously the contention that the legislation violates the Charter .  Yet we must be careful not to jump from the assertion that a legislative change would be prudent to the contention that such a change is constitutionally mandated.  In my view, my colleagues make that jump.  For this reason, I respectfully disagree.

H.  Section 1  of the Charter 

272            As it is my view that the respondent M. has not demonstrated an infringement of s. 15(1)  of the Charter  in the present appeal, it is not necessary to engage in a s. 1 analysis. 

III. Conclusion

273            One of the fundamental principles in Canadian society is that individuals enjoy freedom and are expected to provide for their own needs.  This rule is obviously not absolute.  Canadians take pride in our social programs which lend a hand to those who, for differing reasons, cannot provide for themselves.  At the same time, these social programs have specific policy goals, and in pursuit of those goals, they target specific groups of people.  When the State provides assistance, only those who need the assistance should receive it.  In this appeal, the impugned legislation sought to redress a historical fact that individuals in opposite-sex relationships suffer a systemic dynamic of dependence, which manifests in a support obligation that exists not only while the two individuals are in the relationship, but also after the relationship breaks down.  Usually, it is the female partner who suffers the greatest burden upon marriage or common-law relationship breakdown:  Moge, supra.  The legislature has, since 1859, used a variety of legislative tools to alleviate this systemic suffering, which is unique to opposite-sex relationships.  The statutory language, the preamble, and the legislative debates reveal that this legislation is one of those tools.

274            In my view, the s. 15(1) claim in this case fails because s. 29 of the FLA seeks to ameliorate a historical and structural disadvantage specific to individuals in certain types of opposite-sex relationships, and in so doing, accurately corresponds with the needs, capacity, and circumstances of the claimant and these opposite-sex couples.  Although individuals in same-sex relationships suffer pre-existing disadvantage in many areas of life,  it has not been shown that this is one of them.  In fact, the contrary has been shown: individuals in same-sex relationships generally exhibit less dependency in the relationship; they do not have a structural wage differential between the partners in the relationship; and they do not exhibit the same gendered division of domestic and child-care responsibilities. Although any of these elements may be present in a same-sex relationship, none will have been created by the structural dynamic of dependence which the legislature has seen fit to address, but rather will be attributable to the individual idiosyncrasies of the claimant.

275            In this case, the respondent M. claimed that a benefit granted opposite-sex couples has been denied her on a distinction based on the analogous ground of sexual orientation.  However, the contextual factors demonstrate that there has been no discrimination and the claimant falls outside the scope of protection accorded by that ground.  This distinction does not rest on the stereotypical application of presumed group or individual characteristics; to the contrary, it rests on the correspondence between the ground of distinction and the actual needs, capacity, and circumstances of the claimant and those of the group the legislation targets.  The right to equality in s. 15(1) does not guarantee equality in the abstract; it rests on a comparison with others.  This requires us to examine whether the claimant group suffers pre-existing disadvantage, stereotyping, prejudice or vulnerability as compared with the selected comparison group, and as related to the subject-matter of the legislation.  In this case, individuals in same-sex relationships are not disadvantaged in relation to the dynamic of dependence which the legislation seeks to address.  As such, the ameliorative purpose of the FLA is not underinclusive of a group which is disadvantaged in relation to that purpose.  Moreover, although the claimant is affected by her exclusion from the mandatory support regime, this regime both confers a benefit and imposes a burden.  Mandatory support restricts personal choice and reduces the concomitant financial advantages.   The legislation does not render individuals in same-sex relationships “invisible”.  They are fully entitled to impose support obligations upon themselves, if they so choose.  However, the circumstances unique to individuals in opposite-sex relationships which warrant the reduction of that group’s autonomy do not similarly exist in same-sex relationships.

276            By analysing all of the contextual factors, it is apparent that the claimant’s human dignity is not violated by s. 29 of the FLA.  A reasonable person in the position of the claimant, having taken into account all of the contextual factors relevant to the claim, would not find their human dignity violated by a provision which appropriately takes into account their actual needs, capacity, and circumstances as compared to those of opposite-sex couples subject to the legislation.  For these reasons, it is my view that the s. 15(1) claim must fail.

IV - Disposition           

277            For all of the above reasons, I would allow the appeal, set aside the order of the Court of Appeal, and answer the constitutional questions as follows: 

1.  Does the definition of “spouse” in s. 29 of the Family Law Act, R.S.O. 1990, c. F.3, infringe or deny s. 15(1)  of the Canadian Charter of Rights and Freedoms ?

Answer:  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 ?

Answer:  Given my response to question 1, the question is not applicable.

278            I concur with my colleagues on the disposition as to costs.

The following are the reasons delivered by

279            Major J. -- Although I agree with much of the reasoning of my colleagues and their result, I reach the same result on a narrow basis.

280            I agree with Justices Cory and Iacobucci that the respondent M. in this case was excluded from applying for a benefit as a result of her relationship on the basis of sexual orientation contrary to s. 15(1)  of the Canadian Charter of Rights and Freedoms .  As my colleagues conclude, that exclusion served no purpose and unnecessarily burdened the public purse; it was, therefore, unjustified.

281            The purpose of s. 29 of the Family Law Act, R.S.O. 1990, c. F.3, as stated in para. 53 of Justices Cory and Iacobucci’s reasons is to allow persons who become financially dependent on one another in the course of a lengthy “conjugal” relationship some relief from financial hardship resulting from the breakdown of that relationship.  The relationship at issue in this case meets those requirements.  The Ontario Court of Appeal ((1996), 31 O.R. (3d) 417) found a lengthy conjugal relationship that resulted in financial dependence.  The exclusion of same-sex couples from the scheme to determine and redress this dependence on the basis of their sexual orientation denies them equal benefit of the law contrary to s. 15(1)  of the Charter .  

282            In order to dispose of this appeal it is unnecessary to consider whether other types of long-term relationship may also give rise to dependency and relief.  The  respondent M. asks nothing more than the government recognize that her relationship with H. existed, and that it might have given rise to the sort of financial dependency contemplated in the FLA.  The statute’s categorical exclusion of an individual whose situation falls squarely within its mandate, and who might be entitled to its benefits and protection, denies that person the equal concern and respect to which every Canadian is entitled, and constitutes discrimination.

283            The exclusion at issue in this case could not be shown to further any purpose.  In fact, it undermines the intention of the legislation which was designed in part to reduce the demands on the public welfare system, a factor which I find compelling.  By leaving potentially dependent individuals without a means of obtaining support from their former partners, s. 29 burdens the public purse with their care.  Therefore, the offending provisions are not rationally connected to the valid aims of the legislation and cannot be justified under s. 1  of the Charter .

284            Accordingly, I would dispose of the appeal in the manner proposed by my colleagues.

The following are the reasons delivered by

285            Bastarache J. ‑‑ On June 9, 1994, the Ontario legislature defeated Bill 167, which sought to extend the definition of “spouse” in various Acts to include same-sex couples.  This has been interpreted as a deliberate exclusion of same-sex couples from the definition of “spouse” in s. 29 of the Family Law Act, R.S.O. 1990, c. F.3 (“FLA”), which is impugned in the case before this Court.  In that sense, it is true that this case is about the status of same-sex couples in the family law regime of Ontario.  Things are, however, not as simple as they may appear.  It has been argued before us that this case is essentially about the degree of deference to be given to legislatures in designing public policy and that the central legal issue here is to determine whether there is a constitutional obligation on the legislature to afford same-sex couples the same status under the family law regime as that afforded to opposite-sex couples.  These are very contentious questions.  The scope of the legal issues raised and the implications of our decision may in fact be greater than expected when the action was first initiated.  This explains why we have heard forceful presentations by interveners, and why emotions run high.  It is easy to understand, in this context, why the Court has been invited by some parties to take sides; but this is not the role of a court.  A court’s role is to give a generous and liberal interpretation to s. 15(1)  of the Canadian Charter of Rights and Freedoms , and to apply s. 1  of the Charter  in a fair and reasonable way in order to determine, in legal terms, whether the legislature has breached its obligations under the Charter .

286            This case, like all Charter  challenges to legislation, represents another episode in the continuing dialogue between the branches of government.  As this Court recently outlined in Vriend v. Alberta, [1998] 1 S.C.R. 493, when Canadians collectively chose to adopt the Charter , Canada changed from a system of parliamentary supremacy to a system of constitutional supremacy.  In doing so, Canadians assigned the role of judicial review to the courts so that the rights given to individuals under the Charter  could not be unjustifiably infringed by any legislature or government.  This Court, in Vriend, set out the proper role of the courts on judicial review of government action.  At para. 136, Cory and Iacobucci JJ., for the majority, state:

Because the courts are independent from the executive and legislature, litigants and citizens generally can rely on the courts to make reasoned and principled decisions according to the dictates of the constitution even though specific decisions may not be universally acclaimed.  In carrying out their duties, courts are not to second-guess legislatures and the executives; they are not to make value judgments on what they regard as the proper policy choice; this is for the other branches.  Rather, the courts are to uphold the Constitution and have been expressly invited to perform that role by the Constitution itself.  But respect by the courts for the legislature and executive role is as important as ensuring that the other branches respect each others’ role and the role of the courts.

287            Although I agree with Cory and Iacobucci JJ. on the outcome and applicable remedies in this case, I have come to a different conclusion regarding the objectives of the legislation under review.  While I find that the purpose of the legislation is constitutional, I am of the opinion that the effect of the legislation is to discriminate contrary to s. 15(1).  My analysis is therefore at variance with that proposed by my colleagues.  In regard to the description of the facts, the account of the decisions at trial and in the Court of Appeal, I simply adopt the reasons of Cory J.

Analysis

288            This Court’s approach to the s. 15 inquiry has been recently revisited in Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497.  According to this unanimous decision, the analysis should focus on three central issues:  (a) whether a law imposes differential treatment between the claimant and others, in purpose or effect; (b) whether one or more enumerated or analogous grounds of discrimination are the basis for the differential treatment; and (c) whether the law in question has a purpose or effect that is discriminatory within the meaning of the equality guarantee.

289            I agree with Charron J.A. that s. 29 of the FLA has drawn a distinction between opposite-sex partners and same-sex partners in relationships of permanence.  The comparison is best made, not with married couples, whose status was consensually acquired, but with unmarried cohabiting couples.  In essence it is the inclusion of unmarried couples in the support scheme and the failure to deal with same-sex couples that creates the impugned distinction.

290            Section 29 of the FLA creates a distinction because M. is denied the right to make an application for support under the FLA against her partner H.  Same-sex couples are capable of meeting all of the statutory prerequisites set out in ss. 29 and 1(1) of the FLA, but for the requirement that they be a man and a woman.  Charron J.A. was correct in holding that the main distinction drawn by the legislation between M. as a person in a same-sex relationship and a person in an opposite-sex relationship is based on sexual orientation.  Same-sex couples who cohabit for the requisite time would qualify as spouses within the meaning of s. 29 of the FLA were it not for their sexual orientation.  It is now well established that sexual orientation is a personal characteristic analogous to those found in s. 15(1).

291            In order to determine whether the distinction is discriminatory within the meaning of s. 15, Law suggests that the court consider the contextual factors relevant to the claim such as pre-existing disadvantage, correspondence with actual need, the ameliorative purpose of the law and the nature and scope of the interest protected.  In this case, discrimination exists because of the exclusion of persons from the regime on the basis of an arbitrary distinction, sexual orientation.  Having regard to these contextual factors, I am convinced that this distinction is one that puts into question the dignity of the person affected.  Section 29 of the FLA, in effect, blocks gay men and lesbians from accessing the legal procedure by which they might seek support from their partner after a relationship has ended.  This exclusion suggests that their union is not worthy of recognition or protection.  There is therefore a denial of equality within the meaning of s. 15.

Justification Under Section 1 of the Charter 

292            Infringements on the broad guarantees in the Charter  may be legally justified according to the standard set out in s. 1:

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.

293            This Court has developed a consistent approach to whether legislation is a reasonable limit demonstrably justified in a free and democratic society, as required by s. 1.  There are two stages to this analysis.  At the first stage, the objective or purpose behind the limit on the Charter  guarantee is evaluated to determine if it is of sufficient importance; the second stage considers whether the legislative means chosen are rationally connected to the legislative objective, whether those means minimally impair the Charter  guarantee that has been infringed, and, finally, whether the infringement of the Charter  right is nevertheless too severe relative to the benefits arising from the measure.  In short, the first stage evaluates legislative ends, while the second stage evaluates legislative means.  Both evaluations are made in light of the underlying values of the Charter , which inform the application of s. 1 at every stage.

Contextual Approach to Analysis Under Section 1

294            A number of contextual factors must be considered in order to determine the degree of deference owed to the legislature in applying the various steps inherent in the s. 1 analysis.  As I explained in Thomson Newspapers Co. v. Canada (Attorney General), [1998] 1 S.C.R. 877, at para. 87:

The analysis under s. 1  of the Charter  must be undertaken with a close attention to context.  This is inevitable as the test devised in R. v. Oakes, [1986] 1 S.C.R. 103, requires a court to establish the objective of the impugned provision, which can only be accomplished by canvassing the nature of the social problem which it addresses.  Similarly, the proportionality of the means used to fulfil the pressing and substantial objective can only be evaluated through a close attention to detail and factual setting.  In essence, context is the indispensable handmaiden to the proper characterization of the objective of the impugned provision, to determining whether that objective is justified, and to weighing whether the means used are sufficiently closely related to the valid objective so as to justify an infringement of a Charter  right.

295            Background is particularly important in determining the deference to be afforded to the legislature.  The degree of deference cannot be determined by a crude distinction between legislation that pits the state against the individual and legislation that mediates between different groups within society.  A court should consider a variety of factors when assessing whether a limit has been demonstrably justified in accordance with s. 1.  In Thomson Newspapers, I considered the following factors to be of importance:  the vulnerability of the group which the legislator seeks to protect (as in Irwin Toy Ltd. v. Quebec (Attorney General), [1989] 1 S.C.R. 927; Ross v. New Brunswick School District No. 15, [1996] 1 S.C.R. 825, at para. 88); that group’s own subjective fears and apprehension of harm (as in R. v. Keegstra, [1990] 3 S.C.R. 697, per McLachlin J., at p. 857); and the inability to measure scientifically a particular harm in question, or the efficaciousness of a particular remedy (as in R. v. Butler, [1992] 1 S.C.R. 452, at p. 502).  These are not rigid categories of justification, but rather instances of relevant contextual factors.

Context of the Exclusion:  The Social Science Evidence

296            I have been greatly aided in my consideration of the existing social science evidence by the voluminous Brandeis briefs and articles submitted by the respondent.  While this evidence is an important source of information for this Court, I must stress that care should be taken with social science data.  When dealing with studies exploring the general characteristics of a socially disadvantaged group, a court should be cautious not to adopt conclusions that may in fact be based on, or influenced by, the very discrimination that the courts are bound to eradicate.  Judges, in fact, should be diligent in examining all social science material for experimental, systemic or political bias of any kind.  With this caution in mind, I will briefly consider the material before this Court.

297            The 1993 Ontario Law Reform Commission Report on the Rights and Responsibilities of Cohabitants Under the Family Law Act discusses the possible rationales for the exclusion of same-sex couples from the spousal support provisions of the FLA.  It states (at pp. 45-46):

A second possible rationale for restricting the definition of spouse to couples of the opposite sex is the argument that the rights granted to spouses in the Family Law Act respond to the particular needs of women.  Traditionally, heterosexual relationships adhere to a pattern in which there is a sexual division of labour.  Men have primary responsibility for earning an income and women have primary responsibility for unwaged labour in the home.  Even in relationships in which women work outside of the home, a greater responsibility for child care and household services may force them to limit their involvement in the workforce or may hinder their career development.  As well, because of systemic discrimination against women in the workforce, they may earn less than their male spouses.  The fact that a woman earns a lower wage creates an economic incentive for the couple to further the male spouse’s career.  These economic and social pressures frequently result in women suffering disproportionately from the breakdown of a heterosexual relationship. . . .

The relationships of same-sex couples may not correspond to the traditional heterosexual model.  In addition, if both partners are of the same sex, any wage differential between the two cannot be ascribed to systemic discrimination in the workplace.  Of course, the relationships of some same-sex couples may mimic the heterosexual pattern, with one partner performing unwaged labour in the home or restricting her involvement in the paid workforce, but we do not have sufficient information to know how many same-sex relationships adopt this pattern.  We can be certain, however, that systemic discrimination against women in the workplace will not influence the economic relationship of same-sex couples in the same way as it does heterosexual couples.  This difference might appear to justify restricting the Family Law Act protections to heterosexual couples.  Yet the Family Law Act is not drafted to respond to this objective; the terms of the statute apply generally, granting rights and responsibilities to both men and women, whether or not their relationship conforms to the traditional pattern.  As a result, the restrictive definition of spouse does not appear rationally connected to this objective.  [Emphasis added.]

298            The statements of fact made by the Commission’s Report crystallize the key issue:  that same-sex couples may not suffer from the same economic and social inequalities which tend to create a likelihood of dependence in heterosexual unions.  This simply means that the dynamic of gender inequality does not propel same-sex unions in that direction in the way that it does for opposite-sex couples.  In discussing evidence of actual characteristics of dependence in same-sex relationships, the Report states (at p. 47):

This justification for excluding same-sex couples from the definitions of spouse used in the Family Law Act is difficult to evaluate without more evidence than is available.  A recent Canadian study of the relationships of lesbian cohabitants identified points of similarity and distinction.  While the lesbian relationships studied were found to have similar qualities of intimacy, these couples had relatively lesser feelings of stability than heterosexual relationships.  This may reflect the absence of societal support for lesbian relationships.  The lesbian couples were also less economically interdependent than heterosexual couples and shared unwaged household work more equitably.  Other commentators have observed that ‘(m)ost lesbians and gay men are in ‘dual-worker’ relationships, so that neither partner is the exclusive ‘breadwinner’ and each partner has some measure of economic independence.  Further, examinations of the division of household tasks, sexual behaviour, and decision making in same-sex couples find that clear-cut and consistent husband-wife roles are uncommon’.  While this material is suggestive, it is not sufficient to confirm the anti-assimilationist position.  [Emphasis added.]

This description is directly relevant to the question at hand and I have found from a survey of the social science evidence that this view is widely supported, especially with respect to economic power relations within same-sex relationships.  Many sociological studies indicate that both partners are wage-earners in the vast majority of same-sex relationships.  A study which provides an overview of the social science research states:

Same-sex relationships may be less likely than heterosexual relationships to involve gender-typed roles.  Available data indicate that most lesbians and gay men do not play rigid “husband/wife” roles in such areas as decision-making, sexual behavior, and the division of household tasks; although task specialization often occurs, it typically is based on individual skills and preferences, with neither partner assuming exclusively “masculine” or “feminine” tasks.  [Emphasis in original.]

(G. M. Herek, “Myths About Sexual Orientation:  A Lawyer’s Guide to Social Science Research” (1991), 1 Law & Sexuality 133, at p. 163.)

In a study of various aspects of patterns in lesbian relationships, one researcher makes the following remarks (S. Slater, The Lesbian Family Life Cycle (1995), at pp. 50-51):

Lesbians tend to work especially hard to equalize relational power in the financial arrangements between the partners.  As women, many lesbians have personally experienced financial obstacles resulting from the culture’s reserving the majority of its resources for white men.  Already schooled in the link between money and power, lesbians commonly grapple with how to allocate the family’s money, with both partners being aware of these decisions’ symbolic importance with regard to the balance of power between the partners.

                                                                   . . .

Frequently lesbians struggle to have financial agreements demonstrate their simultaneous commitments to the partners’ independence, equality, and interdependence.  While lesbian couples do typically expect both partners to be employed (unless there are small children at home or there are other extenuating circumstances), they may not bring home equal paychecks.  Nevertheless, lesbians commonly strive to break the traditional linkage between income size and relational power seen in heterosexual couples and to reject the concept of there being a head of the household. As Blumstein and Schwartz report, “Lesbians do not expect to support another adult or be head of a family in the same way a husband expects to take on the position of breadwinner.  A lesbian sees herself as a worker, not a provider or a dependent.”  These researchers found sufficient distinction among lesbians to report that “money establishes the balance of power in relationships, except among lesbians.”  [Emphasis added.]

299            The preponderance of this social science evidence indicates that same-sex, particularly lesbian, relationships do not generally share the imbalance in power that is characteristic of opposite-sex couples and which causes economic dependency in the course of an intimate relationship.  Moreover, the likelihood of household responsibilities becoming a strong source of interdependence and division of labour is lessened with the smaller number of children in same-sex households.  Although this is not determinative on its own, it does indicate that a catalyst for dependence and division of labour within the relationship is significantly more common in opposite-sex couples than in same-sex relationships, particularly gay male relationships.

300            The legislative structure at issue in this case involves mandatory ascription of status, obligations and entitlements.  This case is not about the right to choose to be bound by support obligations -- that choice already exists within the law of contract.  The cost of imposing the s. 29 regime is the reduction of the autonomy of individuals affected.  It is argued that while the imposition of such a regime makes sense in the context of a situation of social inequality which itself interferes with the exercise of autonomy, the situation of gay and lesbian couples generally does not suggest that their autonomy is impaired.

301            On the other hand, it is important to consider that same-sex couples do not have the benefit of consensual access to the family law regime through marriage and that the obligations incurred through the importation of the regime will impose no actual costs on those same-sex partners who are in a situation of equality.  The FLA’s support provisions affect only those relationships in which there is actual economic dependence of one partner on the other.  The issue is whether relationships of dependence in one category can be differentiated from relationships of dependence in the others.

302            The context in which the present s. 1 analysis must be pursued is one in which, although there are fundamental differences in the general economic relationships between heterosexual and same-sex partners, there is also recognition of the ascription of family status to same-sex relationships in society in general, as well as in some legislation and government policies, i.e., the proposed amendments to the Immigration Act; see Building on a Strong Foundation for the 21st Century:  New Directions for Immigration and Refugee Policy and Legislation (1998); Adoption Act, R.S.B.C. 1996, c. 5, s. 5(1); Family Relations Act, R.S.B.C. 1996, c. 128, s. 1(1).

303            The recognition of the prevalence of discrimination against non-heterosexuals and of the need to eradicate this form of discrimination is another part of the social context to be considered in this case.  As Cory J. considered in Egan v. Canada, [1995] 2 S.C.R. 513, at paras. 175-78, many legislators in Canada and other jurisdictions have recognized sexual orientation as a prohibited ground of discrimination.  Indeed, the Canadian Human Rights Act , R.S.C., 1985, c. H-6 , and the human rights acts of all of the provinces prohibit discrimination based on sexual orientation.  This is further reflected in s. 718.2 (a)(i) of the Criminal Code , R.S.C., 1985, c. C-46 , which establishes heightened penalties for crimes motivated by bias, prejudice or hate based on, inter alia, sexual orientation.

Does This Context Suggest that More or Less Deference Should Be Accorded to Legislative Choices in This Area?

304            The s. 1 analysis in the present case must evaluate what degree of underinclusion in the accomplishment of the purposes behind the definition of  “spouse” is permitted when these purposes, and those of the Act as a whole, are balanced against Charter  values.  In other words, even if most individual partners in same-sex relationships are not in a position more typical of a woman in an opposite-sex relationship, some are.  If the Court can perceive this to be the case, can it rewrite the boundary in order to include that smaller number of individuals who are in such a position, or must it defer to legislative determination of the issue?  In reviewing the factors relevant to determining the level of deference to be accorded to the legislature, I have concluded that deference should not be given a dominant role in the present instance.

305            There are several criteria for determining the degree of deference which ought to be shown legislative classifications which are challenged as a violation of s. 15.  First, there is the nature of the interest actually affected by the inclusion or exclusion from the class.  The more fundamental the interest affected, the less deference a court should be prepared to accord to the legislature.  In this case, although the respondent is not left in a total legal vacuum, she is denied the assistance of counselling and mediation of the family law regime, as well as the benefit of access to the specific remedy of spousal support.  In particularly egregious cases, remedies are available against a long-time partner with whom one has laboured to build up common wealth through the equitable doctrine of the constructive trust (Pettkus v. Becker, [1980] 2 S.C.R. 834; Peter v. Beblow, [1993] 1 S.C.R. 980), but this does not alter the fact that the consequences of the failure to include the respondent in s. 29 are significant.

306            The respondent M. submits that the harm that arises due to the inapplicability of s. 29 goes beyond the inability to recover a share of what has been invested in the relationship.  She argues that the true significance of the inapplicability of s. 29 is that it renders gay and lesbian couples invisible to the law, and it treats them as less worthy of respect and dignity because their intimate relationships are not accorded any legal recognition.  There is force to this argument.

307            The respondent H. replies to this argument by noting that the source of this invisibility is not the failure to stretch s. 29 to cover same-sex couples; rather, it is the government’s failure to provide a unique legislative framework to legally recognize the significance and legal ramifications of same-sex unions.  The Law Reform Commission itself, supra, at p. 45, stated that “we do not have sufficient information to know how many same-sex relationships adopt [the opposite-sex] pattern” with respect to division of labour in the household.  It proposed a number of policy alternatives that would specifically address the position of same-sex couples and perhaps others.

308            I agree that the failure to provide same-sex couples with any consensual avenue for mutual and public recognition perpetuates a legal invisibility which is inconsistent with the moral obligation of inclusion that informs the spirit of our Charter .  The best means of eliminating this invisibility may be for the government to conduct consultation and studies in order to create legislation that accords with the expectations and circumstances of same-sex couples.  In my view, there is nothing preventing the legislature from taking this initiative even if the only present means of addressing the discrimination is to declare that same-sex partners shall have access to the existing family law regime.

309            The vulnerability of the group that is excluded by the definition in question is also relevant in assessing the quality of the interest affected by the exclusion, and the degree of deference that is appropriate with respect to other Charter  guarantees (Irwin Toy, supra, at p. 995; Ross, supra, at para. 88).  This can be an ambiguous factor in cases involving social legislation since the vulnerability of the included group will often bespeak a higher degree of deference to the government’s program, while the vulnerability of the excluded group will support the opposite approach.  This case, however, does not involve a balancing of interests, although there is clear evidence of adverse impact on a vulnerable group.

310            Another factor militating in favour of deference is complexity.  In deciding the standard of review of administrative decisions, one of the criteria to be considered is the level of expertise required of the decision-maker in settling the question in dispute.  The animating principle is not that a court should shy away from difficult decisions, but rather that, with regard to certain types of questions, a greater degree of deference might be owed to non-judicial decision-makers.  In my view, the polycentricity of the issues relating to definitions of marriage and common-law spousal status warrant judicial caution in overruling legislative classifications.  Section 29 is part of an elaborate network of obligations and entitlements contained not only in other provisions of the FLA, but also in many other statutes.  Many of these obligations and entitlements are directly related to the central role of the traditional family in social legislation.  We must recognize the value of establishing integrated social programs and norms that reflect clear policy objectives.  When the Court decides to redraw the classificatory boundary, it removes a single piece from that interlocking scheme.  Consequently, a situation arises in which some legislative provisions which apply to opposite-sex couples will apply to same-sex couples, while others will not, depending upon whether a particular section has been challenged, and the judicial outcome of the challenge.  Courts are simply ill-suited to manage holistic policy reform.  If a court must intervene, it must therefore circumscribe that intervention as much as possible.  As this Court acknowledged in Winnipeg Child and Family Services (Northwest Area) v. G. (D.F.), [1997] 3 S.C.R. 925, at para. 24:

This change to the law of tort is fraught with complexities and ramifications, the consequences of which cannot be precisely foretold. At what stage would a fetus acquire rights? Could women who choose to terminate a pregnancy face injunctive relief prohibiting termination, relief which this Court rejected in Tremblay v. Daigle? Alternatively, could they face an action for damages brought on behalf of the fetus for its lost life? If a pregnant woman is killed as a consequence of negligence on the highway, may a family sue not only for her death, but for that of the unborn child? If it is established that a fetus can feel discomfort, can it sue its mother (or perhaps her doctor) and claim damages for the discomfort? If the unborn child is a legal person with legal rights, arguments can be made in favour of all these propositions. Some might endorse such changes, others deplore them. The point is that they are major changes attracting an array of consequences that would place the courts at the heart of a web of thorny moral and social issues which are better dealt with by elected legislators than by the courts. Having broken the time‑honoured rule that legal rights accrue only upon live birth, the courts would find it difficult to limit application of the new principle to particular cases. By contrast, the legislature, should it choose to introduce a law permitting action to protect unborn children against substance abuse, could limit the law to that precise case.  [Emphasis added.]

311            Of course, complexity alone cannot be a justification for a breach of a Charter  right.  But a court should, as in cases of administrative judicial review, be aware of the limits to its own institutional competence in deciding at what point it should replace legislative judgment with its own judgment as to acceptable classificatory distinctions.  As Dickson C.J. remarked in R. v. Edwards Books and Art Ltd., [1986] 2 S.C.R. 713, at pp. 781-82:  “A ‘reasonable limit’ is one which, having regard to the principles enunciated in Oakes, it was reasonable for the legislature to impose.  The courts are not called upon to substitute judicial opinions for legislative ones as to the place at which to draw a precise line.”  These words are even more apposite in the context of legislation implicating a chain of interconnecting interests of which a court can be only dimly aware in any one case brought before it.

312            In the particular circumstances of this case, I believe it is possible to isolate the feature of the family law regime which has been challenged because the advocated extension of its application has no impact on the integrity of the regime or its purposes.  Obviously, this may not be the case with regard to other features of this same regime.  Accordingly, a challenge to other provisions of the FLA would require separate judicial analysis.

313            I do not agree with the proposition that the novelty of the basis of a challenge to a legislative classification leaves the Court powerless; nor do I accept that a process of incremental legislative change necessarily provides justification in the face of a classificatory scheme which is found to be substantially underinclusive or underinclusive with respect to fundamental interests (per Sopinka J. in Egan, supra, at para. 111).  Alone, these considerations do not provide any rationale for granting deference to the legislature’s choice.  The only relevance of incrementalism is that social realities may evolve over time and affect the justification under s. 1.  If social realities emerge which demonstrate that a legislative classification fails to recognize a significant number of individuals who fall within the purpose of the definition, and that this failure indicates that a Charter  right has been infringed, then the marker must be altered.  Subject to that consideration, the legislature is perfectly entitled to address the situations that it considers most severe or most deserving of immediate legislative attention.  It need not identify all the perceived harm in society and address it in every context and in every degree (McKinney v. University of Guelph, [1990] 3 S.C.R. 229, and Edwards, supra).  This is because the cost-benefit balance to the state and to others affected by the legislation changes as different social situations are addressed.

314            In this case, with regard to the specific issue of support obligations between spouses, the issue is neither administrative difficulty nor a strain on financial resources.  The legislature defined a classification which it perceived to present the most unequal of these domestic relationships whose legal status had not been consensually determined, and where it perceived the cost-benefit analysis to be most favourable.  There is no reason to give special deference to legislative choices in this case on the basis that the arbitration of social needs had to be made and priorities established.

315            Another helpful criterion which is used in determining the proper attitude of deference is the source of the rule.  Although I would be reluctant to place significant weight on this factor alone, it can be used as a helpful indicator of the quality of the decision.  Rules that are the product of common law development, or which are made by unelected decision-makers, ought to be accorded less deference in the absence of other factors.  Delegated decision-makers are presumptively less likely to have ensured that their decisions have taken into account the legitimate concerns of the excluded group, while a legislative expression of will presumptively indicates that all interests have been adequately weighted (see M. Jackman, “Protecting Rights and Promoting Democracy:  Judicial Review Under Section 1 of the Charter ” (1996), 34 Osgoode Hall L.J. 661, at pp. 668-69).  If, as Professor J. H. Ely (Democracy and Distrust (1980)) and Professor R. Dworkin (Freedom’s Law (1996)) suggest, one of our principal preoccupations in the equality guarantee is to ensure that the rights of all have been taken into account in the decision-making process, then processes which are more procedurally careful and open deserve greater deference.  That presumption will certainly not immunize legislation from review.  The specific refusal by the Alberta legislature to include sexual orientation as a prohibited ground of discrimination of the Individual’s Rights Protection Act did not prevent this Court from finding that distinction to be a violation of the equality guarantee (Vriend, supra, at para. 115; see also Romer v. Evans, 116 S.Ct. 1620 (1996), where even an amendment by plebiscite was struck down as a patent infringement on the right to equality).  In those cases, despite the democratic nature of the processes, there was no significant justification for the distinction given in the course of the deliberations.  Rather than a guarantee that equal consideration has been given, a democratic procedure merely gives greater weight to the facts, and the interpretation of facts, upon which the legislator has relied and that are open to reasonable disagreement.

316            In reviewing the debates leading to the rejection of Bill 167, the Equality Rights Statute Law Amendment Act, 1994, I noted two objections voiced by the Justice Critic for the Conservative Party:  first, that the institution of marriage was being altered by these amendments; and second, that there were other groups to whom some or all of the benefits and burdens of legal recognition of interdependence might also be extended should the foundation for the present classification be changed (Legislative Assembly of Ontario, Official Report of Debates, June 1, 1994, at p. 6583).  These objections are neither patently fallacious nor do they indicate an animus or prejudice against same-sex couples.  The problem is that they do not address the Charter  right to equal concern and respect and that they were voiced in the context of a wide-ranging amendment that would have affected a variety of statutes with different foundations and purposes.

317            A final factor with regard to deference which must be examined in this case is the role of moral judgments in setting social policy.  In this respect, the comments of Sopinka J. in Butler, supra, at pp. 492-93, are instructive:

The obscenity legislation and jurisprudence prior to the enactment of s. 163 were evidently concerned with prohibiting the “immoral influences” of obscene publications and safeguarding the morals of individuals into whose hands such works could fall.  The Hicklin philosophy posits that explicit sexual depictions, particularly outside the sanctioned contexts of marriage and procreation, threatened the morals or the fabric of society. . . .

I agree with Twaddle J.A. of the Court of Appeal that this particular objective is no longer defensible in view of the Charter . . . .

On the other hand, I cannot agree with the suggestion of the appellant that Parliament does not have the right to legislate on the basis of some fundamental conception of morality for the purposes of safeguarding the values which are integral to a free and democratic society.

The Court came to the conclusion in Butler that although the obscenity provisions of the Criminal Code  had their origins in moral considerations, there was nevertheless an objective harm to society which, in contemporary society, was the principal objective of the section.  The social science evidence demonstrated a reasonable apprehension of harm to women arising from these depictions that justified infringing on the freedom of expression in the Charter .

318            I am struck by the similarity between the evolving moral values ascribed to the family unit by society and these observations by Sopinka J.  Society has an interest in the traditional family.  The vast majority of children born in our society are born and raised in this environment, notwithstanding the development of reproductive technologies which arguably make this family form biologically unnecessary.  In truth, this opposite-sex family form is a product of socialization.  In recognition of the significance of the procreative and socializing role of the opposite-sex family, the modern state has created a host of inducements for this family form, in addition to the obligations between the parties which are intended to mitigate the insecurities created by traditional patterns of gender inequality and specialization.

319            Both the inducements, and the rights and obligations within the couple, confer an objective benefit to society by creating a regime in which opposite-sex partners will suffer the least harm by virtue of engaging in the sometimes risky enterprise of a family.  Even though the institution of marriage is imbued with moral significance for many people, which is the source of their objection to the extension of any marital or quasi-marital status to same-sex couples, there is a social function performed by that legal status which grants a benefit on society, and which is typically applicable to male-female unions, given the current social context of gender inequality.  To the extent that moral factors play a role in supporting an important social institution, I do not believe it is wrong for the Court to be aware of the special sensitivities of those judgments in society.  Like all factors, they must necessarily be assessed in light of Charter  values.

320            I am satisfied, however, that the government’s legitimate interest in setting social policies designed to encourage family formation can be met without imposing through exclusion a hardship on non-traditional families.  There is no evidence that the social purpose of s. 29 would be endangered by the extension in its application.  In fact, the extension sought is consistent with the legislative purpose of ensuring a greater degree of autonomy and equality within the family unit.

321            A review of the foregoing factors indicates that there is no need to be deferential to the legislative choice in this case.  The nature of the interest affected by the exclusion is fundamental, the group affected is vulnerable, it is possible to isolate the challenged provision from the complex legislative scheme, there is no evidence of the government establishing priorities or arbitrating social needs, the legislative history indicates that there was no consideration given to the Charter  right to equal concern and respect, and the government’s interest in setting social policy can be met without imposing a burden on non-traditional families.  Thus, as there is no basis on which to defer to the legislature, I will proceed to a strict application of the traditional test set out in R. v. Oakes, [1986] 1 S.C.R. 103.

Legislative Purpose

322            Determining legislative purpose is theoretically and practically a difficult task.  As Professor Hogg has remarked (Constitutional Law of Canada (loose-leaf ed.), vol. 2, at p. 35-17):

At the practical level, the objective of the legislators in enacting the challenged law may be unknown.  To be sure, the courts will now willingly receive the legislative history of the law, but this is often silent or unclear with respect to the provision under attack.  Courts have not been troubled by this difficulty as much as one might expect.  They usually assume that the statute itself reveals its objective, and they may pronounce confidently on the point even if there is no supporting evidence.

Despite these obstacles, the search for legislative intention has been laid as the cornerstone of the s. 1 analysis.  It has even been suggested that “how the Court characterizes the objective of the impugned legislation essentially determines whether legislation should be struck down or upheld” (E. P. Mendes, “The Crucible of the Charter :  Judicial Principles v. Judicial Deference in the Context of Section 1”, in G.-A. Beaudoin and E. Mendes, The Canadian Charter of Rights and Freedoms  (3rd ed. 1996), at p. 3-14).  Given the particular difficulties surrounding the determination of the legislative purpose in this case, it may be necessary at this point to sound some of the theoretical underpinnings of this approach, and to define precisely the nature of the task.

323            The search for legislative purpose as a method of statutory interpretation is not a novel concept.  In Heydon’s Case (1584), 3 Co. Rep. 7a, 76 E.R. 637, Lord Coke reports, at p. 638:

And it was resolved by them, that for the sure and true interpretation of all statutes in general . . . four things are to be discerned and considered:  —

1st.  What was the common law before the making of the Act.

2nd.  What was the mischief and defect for which the common law did not provide.

3rd.  What remedy the Parliament hath resolved and appointed to cure the disease of the commonwealth.

And, 4th.  The true reason of the remedy; and then the office of all the Judges is always to make such construction as shall suppress the mischief . . . according to the true intent of the makers of the Act. . . .  [Emphasis added.]

The “mischief and defect” formulation does not refer to the actual impact of the legislation, but rather to the category of harm at which the legislation was addressed.  In this passage, purpose is portrayed as an objective fact which can be determined independently of any statement of intention by the legislature, rather than an elaborate search for a fleeting legislative intent.  This may be explained by the fact that, until the landmark decision by the House of Lords in 1993, English courts were strictly prohibited from examining legislative debates as material aids to construction (Pepper v. Hart, [1993] A.C. 593).  However, Canadian courts may consider a wide range of intrinsic and extrinsic sources when searching for legislative purpose (R. Sullivan, Driedger on the Construction of Statutes (3rd ed. 1994), at pp. 51-59).  Although legislative history will often be helpful in determining the precise harm sought to be remedied by law-makers, the ultimate standard for determining the category of harm is the provisions of the legislation itself and the social facts to which it is addressed.  There is, thus, an objective and subjective component to the search for legislative intent and their relative importance will often depend upon the evidence before the court.

324            Usually, the objective and subjective aspects of the search for legislative purpose work hand in hand.  On the one hand, the court must examine the social context of the terms of the legislation, and on the other, it should heed unambiguous indications in the legislative history of what the legislature believed it was addressing.  Only where the two are manifestly inconsistent should legislative history be given little weight.  Such an approach not only accords with general principles of statutory interpretation, but more precisely accords with the role of s. 1, which is to demand of the government a justification for its interference with a Charter  guarantee.  Since that onus of justification is placed on the government by virtue of s. 1, it is appropriate that the legislature’s own views and characterization of a social problem be taken seriously.

325            This is particularly true in evaluating challenges based on s. 15.  Although a social problem could change so drastically as to render a subjective Parliamentary purpose anachronistic, there will be many cases where the evaluation of social relationships is more subtle and open to disagreement by reasonable people, not only because the interpretation of facts may be debatable, but also because normative judgments often play a significant role in setting policy.  In those cases, legislative determinations must be given some scope.  The Court should not simply substitute its own opinion for that of the legislature where the nature of the legislative classification involves disputable social phenomena.

326            There are various theoretical justifications for giving careful consideration to legislative history when considering the legislative purpose of an equality claim. The legislative history may reveal that the legislature has misapprehended in some fundamental way, whether through prejudice or ignorance, the circumstances of a particular group that is affected by a classification.  Rooting out this kind of unequal treatment based on a serious misunderstanding of the characteristics of a group, or based on sheer antipathy, emphasizes the importance of subjective legislative purpose, however obscure it may be.  It is simply impossible to analyse whether the legislature has failed to take into account, on an equal footing, the concerns and characteristics of a particular group without, to some degree, examining the terms of their deliberations.

327            Also, an examination of the legislative history may demonstrate that the legislature failed to accord equal concern to the welfare of some disadvantaged groups.  Again, only a review of the processes of evaluation by a legislature can unveil whether this is what has taken place.  It is, however, necessary to consider whether the inclusiveness requirement is offended at the time of the claim, and not only at the time of the adoption of the legislation under review.  Many statutes adopted before the Charter  were unassailable when first enacted, but were held to offend the Charter  at a later date.  Examples of this phenomenon include Oakes, supra, which struck down s. 8 of the Narcotic Control Act, R.S.C. 1970, c. N-1; R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295, which struck down the federal Lord’s Day Act, R.S.C. 1970, c. L-13; R. v. Morgentaler, [1988] 1 S.C.R. 30, which struck down s. 251 of the Criminal Code, R.S.C. 1970, c. C-34; and Andrews v. Law Society of British Columbia, [1989] 1 S.C.R. 143, which struck down s. 42 of the Barristers and Solicitors Act, R.S.B.C. 1979, c. 26.  It is also important to note here, as in Vriend, supra, the deliberate decision of the legislature to exclude same-sex couples from the regime.

328            From a theoretical standpoint, it makes sense that legislative history should play a particularly important role in the s. 1 analysis.  Within the broad guarantees of the Charter , there can be no doubt that judges share with the legislature the task of making the law in many areas.  For some, this is a regrettable situation involving an irreducible struggle between the legislative and judicial branches of government.  Others have conceived of the relationship differently.  I pointed out in my introductory remarks that a court’s true task is cooperation and dialogue with the legislature in order to ensure that the democratic will is given the clearest expression possible, within the limits imposed by the Charter  (P. W. Hogg and A. A. Bushell, “The Charter Dialogue Between Courts and Legislatures (Or Perhaps The Charter of Rights Isn’t Such A Bad Thing After All)” (1997), 35 Osgoode Hall L.J. 75, at p. 105):

Judicial review is not “a veto over the politics of the nation,” but rather the beginning of a dialogue as to how best to reconcile the individualistic values of the Charter  with the accomplishment of social and economic policies for the benefit of the community as a whole.

See also P. L. Strauss, “The Courts and The Congress:  Should Judges Disdain Political History?” (1998), 98 Colum. L. Rev. 242, at pp. 262-64.  Engaging in that dialogue requires the courts and the legislatures to be speaking the same language.  Rather than being an impermissible incursion into the legislative domain, judicial consideration of legislative intent is a way of respecting the voice of the government of the community and, hence, of the community itself.  The s. 1 analysis places a burden on the government to justify the legislative incursion on the Charter  right.  It is therefore appropriate that governmental intention should, where possible, be considered and evaluated on its own terms to explain why the restriction on a Charter  right is justifiable.

329            Determining the legislative purpose behind the creation of a category that grants benefits to some, and withholds them from others, may sometimes be a difficult task.  This is because social legislation often explains explicitly why the category is being created, not why it is being limited to certain persons and not others.  Yet both of these facts are fundamental to the s. 1 analysis since the limitation on the right to equality necessarily involves a claim by an individual that he or she is not being treated equally relative to others.  Determining whether differential treatment is justified involves an analysis of the legislature’s reasons for the creation of the benefit or burden, but also its reasons for limiting it to a certain class.  The reasons for limitation do not always flow logically from the reasons for inclusion.  For example, the scope of many acts granting financial benefits are circumscribed by a government’s need to operate within fiscal constraints.  Such a concern is usually totally separate and distinct from the reasons for granting a benefit in the first place.  Thus, as Driedger, supra, states in describing the interpretation of statutes in general:  “To appreciate the purpose of the provision the court must ask not only why the credit [the benefit] was created but also why it was available for some investments but not for others.  Until this second question has been answered, the purpose of the provision is not fully known” (p. 63).  The danger with identifying only the broad purposes for inclusion in an act, without any appreciation of the reasons for limiting that definition to a particular group, is that it makes any exclusion that is not inherent in the reasons for inclusion appear not rationally connected to the purpose.  This approach was recently affirmed in Vriend, supra, at para. 110:

Section 1  of the Charter  states that it is the limits on Charter  rights and freedoms that must be demonstrably justified in a free and democratic society.  It follows that under the first part of the Oakes test, the analysis must focus upon the objective of the impugned limitation, or in this case, the omission.  Indeed, in Oakes, supra, at p. 138, Dickson C.J. noted that it was the objective “which the measures responsible for a limit on a Charter  right or freedom are designed to serve” (emphasis added) that must be pressing and substantial.  [Emphasis in original.]

330            The “limit” on the equality right in this case is the failure to treat individuals in same-sex relationships that otherwise meet the criteria of Part III of the FLA in the same way as individuals in opposite-sex relationships exhibiting similar characteristics.  The relevant reasons for that limit are not only the reasons for having created a regime of spousal support, but also the reasons for limiting the scope of that regime.  Of course, the legislature may not have explicitly considered the position of the claimant’s group.  Legislative intent is seldom so transparent or individuated; nor does it always address future circumstances or developments, or provide articulated rebuttals of future constitutional challenges.  In such circumstances, the proper approach is to consider first, the legislation’s objectives in creating the category, and second, any evidence as to the general reasons for limiting the category, even if it is impossible to say with precision why the group in question was excluded.  In considering the broader reasons for inclusion and for limitation, the court may undertake the s. 1 analysis with a full view of the legislature’s objectives.

331            This Court has been very careful in its jurisprudence to avoid the error of looking only at the reasons for inclusion, even if this has sometimes required great sensitivity in discerning the legislative reasons for limiting the defined class in a particular way.  In Eldridge v. British Columbia (Attorney General), [1997] 3 S.C.R. 624, the failure to provide interpretation services to the deaf in the provincial act providing universal medical care was challenged.  The analysis for s. 1 purposes clearly focused on the specific objectives of the exclusionary decision in question, not on the overall purposes of the Act (at para. 84):

Assuming without deciding that the decision not to fund medical interpretation services for the deaf constitutes a limit “prescribed by law”, that the objective of this decision -- controlling health care expenditures -- is “pressing and substantial”, and that the decision is rationally connected to the objective, I find that it does not constitute a minimum impairment of s. 15(1).  [Emphasis added.]

In Benner v. Canada (Secretary of State), [1997] 1 S.C.R. 358, the provision of the Citizenship Act challenged under s. 15(1) allowed children of a Canadian father to apply for citizenship without conditions, while children of Canadian mothers were required to undergo a security check and swear allegiance.  Iacobucci J. articulated the objective relevant for s. 1 purposes as follows (at para. 94):

The appellant accepted that the objectives of the impugned provisions -- to provide access to citizenship while establishing a commitment to Canada and safeguarding the security of its citizens -- were sufficiently pressing and substantial to warrant limiting a Charter  right.  I believe he was correct to do so.  Ensuring that potential citizens are committed to Canada and do not pose a risk to the country are pressing and substantial objectives.  [Emphasis added.]

332            In both of these cases, the Court was careful to consider the reasons for exclusion in conjunction with the overall aims of the legislation.  Indeed, it is impossible to define the reasons for excluding the class claiming a violation of the right to equality without understanding the reasons for inclusion.  As this Court emphasized in Vriend, supra, at para. 111:

. . . the objective of the omission cannot be fully understood in isolation.  It seems to me that some consideration must also be given to both the purposes of the Act as a whole and the specific impugned provisions so as to give the objective of the omission the context that is necessary for a more complete understanding of its operation in the broader scheme of the legislation.

333            The necessity of this approach also emerges from the particular nature of most social legislation conferring benefits.  The reason for this approach is clear.  The Oakes analysis presupposes a tension between the objectives of the Charter  right and the objectives of the particular legislative provision that is alleged to constitute a violation of the Charter  guarantee.  Only after that tension is identified does it make sense to ask whether that legislative objective is pressing and substantial, and whether the precise means adopted are so closely related and narrowly tailored to that objective as to warrant derogation from the Charter  guarantee.  If the tension of objectives is removed, then almost any exclusion that detracts from the ambit of the broad legislative goal will fail the s. 1 test, because, simply by virtue of being an exclusion, it cannot be rationally connected with the goal.  Only when the specific purpose or objective of the exclusion is articulated are the tests under Oakes, supra, properly engaged.  This is particularly true in cases involving the guarantee of equality.  Unlike most legislation which infringes ss. 2( a ) , (b), (d) and 7  to 14  of the Charter , the broad purposes of entitlement-granting legislation will seldom come into conflict with s. 15 .  Usually, the purposes are perfectly congruent and it is necessary to articulate the purpose of the limitation in order to identify the underlying tension between the legislative purpose and the Charter .

334            Another danger in not rigorously following this approach is the tendency to suggest that nothing is taken away from the included class by making others eligible for the benefit.  Such a formulation misses the point.  The issue under s. 1 is whether the government acted in a reasonable fashion in limiting the class based on one of the prohibited enumerated or analogous characteristics described in s. 15.  That requires an analysis of its reasons for limiting the class as it did.  If the government had a valid reason for limiting the class, and it used means which were proportional to this objective, then the limitation on the equality rights of those excluded is justified.  Whether anything is taken away from the included class is entirely irrelevant.  The inquiry is not into a possible detriment to the included class.

335            In this case, the majority of the Court of Appeal defined the relevant purpose for the s. 1 analysis in the following manner, at p. 450:  “Part III of the FLA on Support Obligations is clearly part of a legislative scheme aimed at providing ‘for the equitable resolution of economic disputes that arise when intimate relationships between individuals who have been financially interdependent break down’” (emphasis added).  A second purpose was found, at p. 450, to be that “the legislature intended, where applicable, to alleviate the burden on the public purse by shifting the obligation to provide support for needy persons to those parents and spouses who have the capacity to provide support to these individuals”.  In considering the specific extension to common law spouses by amendment in 1978, Charron J.A. states, at p. 451:

I am unable to identify any additional purposes distinct from those already raised.  In my view, the legislature must simply be taken to have recognized that, in so far as spousal support obligations were concerned, it was neither fair nor effective to choose marriage as the exclusive marker for the identification of intimate relationships giving rise to economic interdependence which might require, upon breakdown, some access to the equitable dispute-resolution scheme created by the legislation.

336            The limit on the right to equality that is challenged in this case is a definition which includes a particular class, but excludes all other classes.  The relevant purpose for the s. 1 analysis, then, is the rationale behind the definition itself.  The only rationale for the definition adopted, according to Charron J.A., is that it sought to recognize “intimate” relationships going beyond marriage, and to reduce dependency on government assistance.

337            In my view, this is an unwarranted recharacterization of the legislative purpose.  Rather than being a justification for the definition adopted by the government, Charron J.A. offers little more than a description of one characteristic of the class which falls within s. 29 and a reference to a mere incident of the failure to enforce the obligations of equal sharing in the conjugal partnership.  In fact, there is much more to the story that must be examined before the rationale behind the definition in s. 29 can be fully appreciated.

338            The respondent M. relied heavily on the extension of spousal support benefits to opposite-sex common law relationships and much debate on the Family Law Reform Act, 1978, S.O. 1978, c. 2, focused on this particular aspect of the definition.  In introducing the bill to the legislature, the Attorney General of the day, the Honourable Roy McMurtry, sought to justify the extension of benefits beyond the relationship of marriage (Legislature of Ontario Debates, October 26, 1976, at p. 4103):

This part also creates a limited obligation of one common law spouse to support the other.  Where two persons have lived together as if married, their relationship often takes on the same financial characteristics as a marriage.  One person frequently becomes dependent on the other, especially if there is a child of the union.  If one of these two people is no longer self-sufficient, it seems reasonable to look to the other to assist in restoring him or her to financial independence.  Certainly it is more desirable to place a support obligation on common law spouses than to have a large number of persons who are living common law looking to public welfare for support instead.

On a later day, the Attorney General responded to criticism that support obligations ought not to be imposed on relationships outside of marriage, and made the purpose behind that part of the definition even clearer (Legislature of Ontario Debates, November 18, 1976, at p. 4793):

If there is no financial dependency, there will be no need for support by the other party and there will not be a successful claim for support.

By contrast, however, there are many people living together in such relationships who are being exploited by their partner.  They have been induced to enter into the relationship and to stay home and raise the children arising from the union, or children of another union, and have thus been put in a position of total dependency on the person as a result of being out of the labour market for a lengthy period of time.  Many of these people are later abandoned and, under the present law, they have nowhere to turn but to the welfare authorities for support.

This is not a small problem.  For example, in September of this year, the government of Ontario has paid out family benefits to over 13,000 unmarried mothers and their 26,000 dependent children, totalling over $3.5 million for that month alone.  [Emphasis added.]

339            The primary legislative purpose in extending support obligations outside the marriage bond was to address the subordinated position of women in non-marital relationships.  But for this social problem, no legislation would have been passed to impose support obligations on unmarried cohabitees.  Support for this interpretation of the purpose of the Act is also to be found in previous judgments of this Court.  In Miron v. Trudel, [1995] 2 S.C.R. 418, L’Heureux-Dubé J. stated, at para. 97:

Both the courts and the legislatures have, in recent years, acknowledged and responded to the injustices that often flow from power imbalances of this type and have thereby given increasing recognition to non-traditional forms of relationships.  Why else did the Ontario legislature in 1986 extend benefits from married persons to cohabiting partners in over 30 Ontario statutes, several of which raised issues of financial interdependence that are analogous to the impugned provisions of the Insurance Act?  Why else has the Ontario Family Law Act, R.S.O. 1990, c. F.3, imposed an obligation of mutual support on common law spouses since 1978?  . . .  In all of these cases, although the language has generally employed gender-neutral references to “spouses”, it is indisputable that much of the impetus for these changes stemmed from courts’ and legislatures’ increasing recognition of the disadvantage endured by dependent spouses, most often women, within the context of those relationships.  [Emphasis added.]

340            The legislative history suggests unequivocally that the social harm which this definition was seeking to address was the widespread economic dependence faced by women, inside or outside the marital bond, upon the breakdown of the relationship.  The Attorney General’s second statement and reference to 13,000 mothers on social assistance makes it clear that this is the social problem which his government has in mind, first in order to ameliorate their economic position, and second to reduce reliance on public resources.  Moreover, his strong references to exploitation and inducement to enter into common law relationships reveals that the government was responding to a wider context of gender inequality which pressures women into remaining in unsatisfactory legal situations vis à vis their partner because the man refuses to undertake any of the legal obligations of marriage.  Nor is this purpose disjunctive with that of the support obligation in the family law regime as it existed prior to 1975.  As the Law Reform Commission noted in Part VI of its Report on Family Law (1975), dealing with support obligations between married spouses, at pp. 6-7:

There are historical reasons why the law still adopts a protective posture towards married women. . . .

The force of the conventional view that married women should occupy an exclusively domestic role was the most substantial contributing factor to the practical disabilities faced by those who were inclined to strive for economic self-sufficiency. . . .  In short, although married women secured the right to acquire separate property in 1884, if the law had not imposed a strict support obligation on husbands, it might have been said then of most married women that they merely acquired the legal capacity for destitution in their own right.

341            Ultimately, the Commission recommended that spousal obligations should be extended to wives because “[i]f a spouse’s need or dependency is to assume increased importance as a basis upon which support obligations are determined [as distinguished from marital fault], then the law must recognize that a husband’s need may also be such as to cast a positive obligation upon the wife for its alleviation” (p. 10).

342            The extension of the legislative remedy to men represents a separate strand of legislative intent.  The Attorney General was concerned, particularly at the introduction of the legislation, to avoid using gender-specific language.  In my view, this simply reflects a wider concern to erase official gender bias from the language of the FLA, rather than a concern to place “intimacy” and “interdependence” as the sole rationale for granting spousal support.  The Attorney General himself describes as one of the “basic themes” of the reform bill that it “recognizes the equality of the sexes.  It confers no privileges and imposes no disability on either men or women as a group” (Legislature of Ontario Debates, October 26, 1976, at p. 4102).  This is confirmed by Part VI of the Ontario Law Reform Commission’s Report on Family Law, supra.  It states, at p. 10:

If a support obligation is to be imposed upon wives under provincial law, it is reasonably clear that the justification for so doing cannot be founded upon evidence of widespread present need on the part of dependent husbands.  If it were otherwise, there would have been many more reported instances of claims by husbands for support under section 11  of the Divorce Act  which has given the court power, since July, 1968, to award support to a husband on very general criteria.  Clearly, this has not occurred in the reported cases.  [Emphasis added.]

343            The major controversy with respect to the definition of those entitled to support at the time of these debates and reports was the ascription of legal obligations to the parties in the absence of a formal, legal recognition by both parties that this is what they desired (Legislature of Ontario Debates, November 18, 1976, at p. 4801).  The FLA fundamentally altered the nature of those obligations by extending them beyond marriage and therefore beyond the realm of consensual undertaking.  It did so because of a pressing social concern:  that many women found themselves in a legal vacuum at the end of a relationship; that their economic dependence was worsened as a result of those relationships; and that the economic position of women generally in society relative to men placed them in a position where they might be induced to enter or remain in those relationships without the legal safeguards of marriage even when they might otherwise want the mutual rights and obligations which would thereby be imposed.  In that context, the government chose to extend spousal support beyond marriage and into the realm of an ascribed status based upon objective criteria of the length of cohabitation, or the presence of children from the relationship.  The fact that the language of the Act gives men the same access to spousal support reflects the government’s expressed desire to ensure gender equality in the language of the Act, and the Law Commission’s concern to refute the notion that women are legally dependent on men but not vice versa.  To my mind, in determining the purpose of the definition, and in particular, the extension of support based on ascribed status, it would be absurd for the purposes of the s. 1 analysis to ignore the pressing social harm to which the government was consciously responding.

344            It is also important to take into account the legislative context in which the impugned section is being considered.  This requires that I examine the general nature of the FLA, the interrelationship between Part III, where the impugned definition is found, and the other parts of the FLA, and the interrelationship between this Act and related social legislation.

345            The general nature of the Act can be determined in part by reference to its preamble and nomenclature.  The original title of the 1978 Act was An Act to reform the Law respecting Property Rights and Support Obligations between Married Persons and in other Family Relationships.  This title has now been shortened to the Family Law Act.  The Act concerns the rights and obligations of men and women who are married or living in common law relationships akin to marriage.  The Act also creates other familial support obligations including the obligations of parents to support children and obligations of children to support parents.  The preamble states:  “Whereas it is desirable to encourage and strengthen the role of the family; and whereas for that purpose it is necessary to recognize the equal position of spouses as individuals within marriage and to recognize marriage as a form of partnership . . . and to provide for other mutual obligations in family relationships, including the equitable sharing by parents of responsibility for their children”.  The FLA deals with family property, the matrimonial home, support obligations, domestic contracts, a dependant’s claim for damages, and amendments to the common law.  In essence, it provides for a regime that is required to ensure equality of spouses and the security of children and parents in the traditional family.  It addresses on various points the imbalance resulting from the situation of women with regard to men in conjugal relationships.

346            The FLA cannot be dissociated from other Acts dealing with family matters.  Without giving a complete list, I would name the Child and Family Services Act, R.S.O. 1990, c. C.11, regarding spousal consent for child placement or adoption, the Change of Name Act, R.S.O. 1990, c. C.7, Consent to Treatment Act, 1992, S.O. 1992, c. 31, Absentees Act, R.S.O. 1990, c. A.3, Coroners Act, R.S.O. 1990, c. C.37, Estates Act, R.S.O. 1990, c. E.21, Succession Law Reform Act, R.S.O. 1990, c. S.26, Execution Act, R.S.O. 1990, c. E.24, Family Support Plan Act, R.S.O. 1990, c. S.28, Workers’ Compensation Act, R.S.O. 1990, c. W.11, Mental Health Act, R.S.O. 1990, c. M.7, Pension Benefits Act, R.S.O. 1990, c. P.8, Municipal Act, R.S.O. 1990, c. M.45, Partnerships Act, R.S.O. 1990, c. P.5, Income Tax Act, R.S.O. 1990, c. I.2, Ontario Guaranteed Annual Income Act, R.S.O. 1990, c. O.17, Land Transfer Tax Act, R.S.O. 1990, c. L.6, Retail Sales Tax Act, R.S.O. 1990, c. R.31, Business Corporations Act, R.S.O. 1990, c. B.16, Election Act, R.S.O. 1990, c. E.6, Evidence Act, R.S.O. 1990, c. E.23, Juries Act, R.S.O. 1990, c. J.3, Members’ Conflict of Interest Act, R.S.O. 1990, c. M.6, Municipal Elections Act, R.S.O. 1990, c. M.53, Securities Act, R.S.O. 1990, c. S.5, Small Business Development Corporations Act, R.S.O. 1990, c. S.12, Nursing Homes Act, R.S.O. 1990, c. N.7, Rent Control Act, 1992, S.O. 1992, c. 11, Trustee Act, R.S.O. 1990, c. T.23, and the Marriage Act, R.S.O. 1990, c. M.3.  This legislation is further evidence of the fact that Ontario has recognized the special nature of the traditional family and its predominant social form in our society, involving well-recognized and complex rules, obligations and rights.

347            A scrupulous examination of the provisions of the FLA, including s. 29, and of the legislative debates surrounding its passage, does not suggest that there was any specific purpose behind defining the category so as to exclude same-sex couples.  Rather, those debates and the text of the law indicate that the general purpose of the legislator was to confine the non-voluntarily assumed support obligations as narrowly as possible to those who actually needed the intervention of a mandatory scheme.  In particular, the legislative debates focus on the disadvantaged economic position of women relative to men, the household division of labour as between women and men, and the hardships that women frequently suffer upon the breakdown of non-marital relationships of significant interdependence.  I would define the legislative purpose of the definition in Part III of the FLA as follows:  to impose support obligations upon partners in relationships in which they have consciously signalled a desire to be so bound (i.e., through marriage); and upon those partners in relationships of sufficient duration to indicate permanence and seriousness, and which involve the assumption of household responsibilities, or other career or financial sacrifices, by one partner for the common benefit of the couple, and which cause or enhance an economic disparity between the partners.  I would add that there is nothing upon which, in coming to this purpose, one can rely to determine that “intimacy” is in part, or substantially, related to the purpose in question.  I cannot understand why, on one hand, the Court of Appeal would extract “intimacy” as the sole criterion on which the government relied for its definition of the eligible category while, on the other, excluding the glaring reality of gender inequality which the legislative history establishes was clearly uppermost in the legislator’s mind in extending benefits into the non-consensual domain.

348            In coming to this purpose, I have examined the text of the FLA, the social context, and the legislative history of the provision, with an emphasis on the latter two.  I have come to the conclusion that the inclusion or exclusion of same-sex couples was not contemplated in 1975-1976.  The inference with regard to excluded categories is that they were not recognized as forming families in the traditional sense.  This analysis is however complicated by subsequent legislative history, in particular the rejection of the Equality Rights Statute Law Amendment Act, 1994 by the Ontario legislature.  The question is, to what extent could an amendment -- whether failed or successful -- alter the original purpose underlying a limitation on a Charter  right?  The basic principle spelled out by this Court in Big M Drug Mart Ltd., supra, at p. 335, is that the purpose of a law does not shift over time.  While there may be a shift in emphasis in a purpose, the purpose itself may not be fundamentally re-articulated by a court in light of subsequent societal changes.  To be sure, whether a purpose is pressing and substantial may well depend on the state of society, and on the effects of the legislation; but the objective which is tested remains the same.  A new intervention by the legislature may however constitute the reaffirmation of a legislative purpose, or provide evidence of a new or redefined purpose.  In this case, the failed legislative amendment provides evidence relative to the purpose of the exclusion created by the definition in s. 29.

349            The proposed amendments failed and therefore failed to alter the text of the Act in any way.  But leaving the legislative text intact in light of a changed social context is itself indicative of legislative purpose.  Although the legislative debates indicate much discussion irrelevant to the impact of the amendment on s. 29 of the FLA, what is relevant in the 1994 debates is the clear indication that the legislature desired to reaffirm its recognition of the categories of persons then subject to the family law regime and not put in question its longstanding support for the traditional family.  It can therefore be inferred that the legislature’s purpose was also to exclude all types of relationships not typically characterized by the state of economic dependency apparent in traditional family relationships.

350            Constitutional review of the validity of legislation is not an exercise of ordinary statutory interpretation.  The object of considering the legislative purpose is not to determine the scope of the legislation or of the impugned provision.  In the present instance, the object of the constitutional review is, rather, to determine whether there is evidence that the legislature intended to limit the benefit of the FLA and s. 29 for a purpose that is inconsistent with its Charter  obligations.  In the absence of such evidence, the object is to determine whether the criteria chosen to limit the categories of beneficiaries under the Act are consistent with Charter  values, in light of the fact that the application of the criteria is detrimental to an analogous group as defined by s. 15(1)  of the Charter .

Is the Objective of Section 29 Pressing and Substantial?

351            In the analysis above, I have looked closely at legislative history and social context in order to determine the purpose of the impugned definition.  This is an important part of taking seriously the rationale urged by the government for distinguishing between ascribed status for same-sex relationships and gendered relationships.  In determining whether the objective is pressing and substantial, it is necessary first to examine whether the creation of a class according to the government’s purposes fulfills some important social function.

352            In my view, in taking male-female relationships as the basic group with which the legislature was concerned, it is all too obvious that there is recognition of the fact that many women are still in a position of vulnerability in their long-term relationships with men.  This arises from the general social context of women in our society relative to men.  In Moge v. Moge, [1992] 3 S.C.R. 813, at pp. 853-54 and 861-62, this Court made numerous observations on the feminization of poverty and disproportionate impact of divorce on women and children.

353            These statements make clear that the problem perceived in 1976 by the government of Ontario is still of great importance.  Indeed, the potential incidence of this problem with respect to unmarried relationships has increased substantially in the last few years.  The 1996 Census indicates that the number of children living in a family of married relationships remained static, while the number of children living in a family of common-law relationships increased by 52 percent (Statistics Canada, “More lone-parent and common-law families” in Infomat: A Weekly Review (October 17, 1997), at p. 4).  The census statistics also show that 73.7 percent of all families are headed by a married couple, and 61 percent of those families have children in the household.  11.7 percent of families were headed by a common-law couple, 47 percent of whom have children living in the household.  Another Statistics Canada study reports that among breakdowns of marital relationships that had children in the household, women suffer a median loss in adjusted family income of 23 percent, while men’s adjusted income rises by 10 percent in the year after separation (Statistics Canada, Family Income After Separation (1997), at p. 17).  Given the findings in Symes v. Canada, [1993] 4 S.C.R. 695, and Moge, there can be no doubt that the presence of these children enhances the likelihood of the economic vulnerability of women upon the breakdown of a conjugal and permanent relationship.

354            I conclude from these statistical figures and judicial statements that the need for imposition of support payments by one party to a relationship to the other, in the case of traditional family relationship breakdown, is a pressing and substantial objective in Canadian society.  It is clear that with respect to at least one category captured by the government’s legislative purposes, gendered relationships of some permanence, there is a high likelihood of serious economic detriment to women on the breakdown of those relationships.  The justification for legislative intervention affecting the autonomy of heterosexual couples does not however explain the pressing need to exclude all other family relationships from the governmental regime.  There is no evidence that their inclusion would cause any particular difficulty.  The only arguments given to justify exclusion are that inclusion would undermine the traditional family and that there is evidence that same-sex relationships are not typically characterized by the economic imbalance observed in traditional conjugal relationships.  When evaluating the legislative objective of s. 29, it is necessary to consider whether the objective is consistent with Charter  values.  We cannot limit the inquiry to the positive purpose of the legislation, and not take into account its effect on excluded groups.  In order to be consistent with Charter  values, the purpose of the definition in s. 29 must be respectful of the equality of status and opportunity of all persons.  It would be consistent with Charter  values of equality and inclusion to treat all members in a family relationship equally and all types of family relationships equally.  It is, however, inconsistent with s. 15  to deny equal treatment to a member of a family relationship on the basis of an analogous ground.  Where an enumerated or analogous ground forms the basis of the discrimination, as in this case, then s. 15 is triggered.  However, where the distinction is drawn along other lines, the onus is on the claimant to demonstrate that it involves a new analogous ground.  For example, two sisters living together in an economically dependent relationship will not a priori satisfy this requirement.

355            The appellant Attorney General insists that the exclusion is serving a valid purpose by not imposing on same-sex couples a reduction in freedom and autonomy that is mandated by economic imperatives largely irrelevant to same-sex couples.  Even if one were to accept that the government’s true purpose in adopting the limitation is justified for the above mentioned reasons, there would be no rational connection between that purpose (excluding classes not generally experiencing economic imbalance because there is no need of special protection in their case and because there is no justification for limiting their freedom and autonomy) and the total exclusion dictated by s. 29.  Can it be said that this exclusion assists in achieving the objective of eradicating economic dependency within families?  No.  The gravamen of the respondent M.’s case is that her relationship was one of dependency resulting in serious economic detriment, and that her situation falls four square within the purposes of the legislation defined by the government.  In essence, the respondent M. is claiming that her equality rights have been impaired because of an underinclusive marker to define dependent family relationships, which is the government’s purpose.  Even though most same-sex couples do not experience economic imbalance, some do.  What is the purpose in excluding them?

356            This exclusion is not a valid means of achieving the positive purpose of s. 29, economic equality within the family.  By defining restrictively the scope of the family concept, s. 29 in effect is restricting the reach of equality.  When, as here, the exclusion specifically detracts from the general legislative purpose, the objective of the restriction cannot be considered pressing and substantial.  This follows from the reasons of Cory and Iacobucci JJ. in Vriend, supra, where they state, at para. 116:

In my view, where, as here, a legislative omission is on its face the very antithesis of the principles embodied in the legislation as a whole, the Act itself cannot be said to indicate any discernible objective for the omission that might be described as pressing and substantial so as to justify overriding constitutionally protected rights.

Even if the primary purpose of s. 29 was simply to recognize and promote the traditional family, and not to secure economic equality within the couple, which could be considered simply a means to an end, the exclusion of same-sex partners could not be demonstrably justified.  Denial of status and benefits to same-sex partners does not a priori enhance respect for the traditional family, nor does it reinforce the commitment of the legislature to the values in the Charter .  No evidence was adduced showing any beneficial impact of the exclusion on society, or what Charter  values would be served by the exclusion; but the detrimental effects clearly exist, both for the individual without recourse to the family law regime, and for society, faced with the prospect of giving social assistance to that aggrieved individual in need.  As for the protection of the freedom and autonomy of persons engaged in a same-sex relationships, s. 29 will only affect those who are in fact in situations of economic imbalance analogous to that which more commonly occurs in the case of heterosexual relationships.  The entitlement resulting from a wider definition of “spouse” does not create an absolute right to support.  The justification for interference with personal autonomy is therefore the same for same-sex partners and opposite-sex partners.

Conclusion

357            Accordingly, I find that the definition of “spouse” in s. 29 of the FLA is an infringement of s. 15(1)  of the Charter , and that it is not demonstrably justified in a free and democratic society according to s. 1.  I agree with Iacobucci J.’s treatment of remedy and costs and would, like him, dismiss the appeal and the cross-appeal with solicitor-client costs to both respondents M. and H. in the proceedings before this Court.

Appeal and cross-appeal dismissed with costs, Gonthier J. dissenting on the appeal.

Solicitors for the appellant:  Robert E. Charney and Peter C. Landmann, Toronto.

Solicitors for the respondent M.:  McMillan Binch, Toronto.

Solicitors for the respondent H.:  Borden & Elliot, Toronto.

Solicitors for the intervener the Foundation for Equal Families:  Elliott & Kim, Toronto.

Solicitor for the intervener the Women’s Legal Education and Action Fund (LEAF):  Carol A. Allen, Toronto.

Solicitors for the intervener Equality for Gays and Lesbians Everywhere (EGALE):  Sack Goldblatt Mitchell, Toronto.

Solicitor for the intervener the Ontario Human Rights Commission:  Joanne D. Rosen, Toronto.

Solicitors for the intervener the United Church of Canada:  Weir & Foulds, Toronto.

Solicitors for the interveners the Evangelical Fellowship of Canada, the Ontario Council of Sikhs, the Islamic Society of North America and Focus on the Family:  Lerner & Associates, Toronto.

Solicitors for the intervener REAL Women of Canada:  Stikeman, Elliott, Toronto.