SUPREME COURT OF CANADA

Citation: Canada (Attorney General) v. Hislop, [2007] 1 S.C.R. 429, 2007 SCC 10

Date: 20070301

Docket: 30755

Between:

Attorney General of Canada

Appellant/Respondent on cross-appeal

and

George Hislop, Brent E. Daum, Albert McNutt, Eric Brogaard and Gail Meredith

Respondents/Appellants on cross-appeal

‑ and ‑

Attorney General of Ontario, Attorney General of Quebec, Attorney General of Alberta and Egale Canada Inc.

Interveners

Coram: McLachlin C.J. and Bastarache, Binnie, LeBel, Deschamps, Abella and Rothstein JJ.

Reasons for Judgment: (paras. 1 to 136)

Concurring Reasons: (paras. 137 to 165)

LeBel and Rothstein JJ. (McLachlin C.J. and Binnie, Deschamps and Abella JJ. concurring)

Bastarache J.

Canada (Attorney General) v. Hislop, [2007] 1 S.C.R. 429, 2007 SCC 10

Attorney General of Canada Appellant/Respondent on cross‑appeal

v.

George Hislop, Brent E. Daum, Albert McNutt, Eric Brogaard and Gail Meredith Respondents/Appellants on cross‑appeal

and

Attorney General of Ontario, Attorney General of Quebec, Attorney General of Alberta and Egale Canada Inc. Interveners

Indexed as: Canada (Attorney General) v. Hislop

Neutral citation: 2007 SCC 10.

File No.: 30755.

2006: May 16; 2007: March 1.

Present: McLachlin C.J. and Bastarache, Binnie, LeBel, Deschamps, Abella and Rothstein JJ.

on appeal from the court of appeal for ontario

Constitutional law — Charter of Rights — Equality rights — Sexual orientation — Survivorship pensions — Whether provisions of Canada Pension Plan limiting eligibility for survivor benefits to same‑sex partners of deceased contributors who died on or after January 1, 1998 infringe right to equality — If so, whether infringement justified — Canadian Charter of Rights and Freedoms, ss. 1, 15(1) — Canada Pension Plan, R.S.C. 1985, c. C‑8, ss. 44(1.1), 72.

Constitutional law — Charter of Rights — Equality rights — Sexual orientation — Survivorship pensions — Whether estate has standing to advance equality rights claim on behalf of deceased survivor — Canadian Charter of Rights and Freedoms, s. 15(1) — Canada Pension Plan, R.S.C. 1985, c. C‑8, s. 60(2).

Constitutional law — Remedy — Retroactive remedy — Circumstances under which courts may limit retroactive remedy — Constitution Act, 1982, s. 52.

Following the 1999 decision of this Court in M. v. H., the federal government amended the Canada Pension Plan (“CPP”) in 2000 to extend survivor benefits to same‑sex partners by changing the definition of “spouse” to conform with the equality rights provisions of s. 15(1) of the Canadian Charter of Rights and Freedoms and by adding new provisions. However, under s. 44(1.1) of the CPP, eligibility was limited to same‑sex partners whose “spouse” died on or after January 1, 1998. Benefits were not retroactive to April 17, 1985, when s. 15(1) came into force, or the date of death of the “spouse”, whichever occurred later. Section 72(2) of the CPP precluded payments to same‑sex survivors for any month before July 2000, the month s. 72(2) came into force. The effect of s. 72(2) came to an end as of June 2001, because after that month, same‑sex and opposite‑sex survivors benefited from application of the general rule in s. 72(1), which limits survivors’ arrears benefits to not more than 12 months prior to the month in which the application is received. Finally, s. 60(2) of the CPP, a general provision not affected by the 2000 amendments, limits the right of estates of survivors from obtaining benefits if the application for the benefits is not made within 12 months after the death of the survivor. A class action was commenced by same‑sex survivors (“Hislop class”), challenging the constitutionality of the four sections. The Ontario Superior Court of Justice held that ss. 44(1.1) and 72(2) violated s. 15(1) and were not justified under s. 1 of the Charter. Same‑sex survivors were granted a constitutional exemption from ss. 60(2) and 72(1). Survivors were entitled to benefits retroactive to the later of the date of death of their partner and April 17, 1985, with interest. The Court of Appeal affirmed the finding that ss. 44(1.1) and 72(2) were unconstitutional, but held that ss. 60(2) and 72(1) did not infringe s. 15(1). Sections 44(1.1) and 72(2) were struck down, leaving a same‑sex surviving spouse entitled to survivor benefits subject only to the 12‑month cap on arrears and the limitation on estate claims (as were all benefits claimants).

Held: The appeal and cross‑appeal should be dismissed.

Per McLachlin C.J. and Binnie, LeBel, Deschamps, Abella and Rothstein JJ.: Essential to the question of differential treatment in the s. 15(1) analysis is the choice of comparator group. Under s. 44(1.1), the appropriate comparison is between same‑sex and opposite‑sex survivors whose partners died before January 1, 1998; under s. 72(2), it is between same‑sex and opposite‑sex survivors who applied for survivor’s pensions between July 2000 and June 2001. To frame, as suggested by the government, the comparator group in terms of the express distinction made in ss. 44(1.1) and 72(2) between two groups of survivors of same‑sex relationships, based on the date on which the relationship ended as a result of the death of one of the partners, would be to miss the fundamental reason for the 2000 CPP amendments — namely, to remedy unequal treatment between same‑sex and opposite‑sex couples. Here, the Court of Appeal was right to conclude that ss. 44(1.1) and 72(2) infringe s. 15(1) of the Charter. Both provisions treat same‑sex survivors differently from their comparator group; sexual orientation is recognized as an analogous ground of discrimination under s. 15(1), and the differential treatment discriminated in a substantive sense. [24] [37‑38] [60]

Section 44(1.1) is not saved under s. 1 of the Charter. In an appropriate case, the matching of benefits conferred under the CPP with obligations imposed on same‑sex partners under other legislation may be a pressing and substantial objective of limits on the extension of social benefit legislation as there are complex relationships between income tax credits, pensions, supplements and other entitlements, often based on conjugal relationships. It is not enough, however, to only make general reference to these relationships. There must be an explanation supported by relevant evidence as to what those relationships are, why they are relevant and why they justify the limit on the Charter right that has been found to be violated. In the absence of such evidence in this case, matching cannot be considered to be a pressing and substantial objective of s. 44(1.1). The absence of evidence also precludes a finding of rational connection between s. 44(1.1) and its objective and of minimal impairment of a Charter right. [47] [52‑54]

Likewise, the government has failed to establish a s. 1 justification for s. 72(2). Where the issue is solely whether Charter relief claimants should be entitled to the same retroactive benefits expressly available to their comparator group, absent cost considerations, it is difficult to see how denial of these benefits has a pressing and substantial objective. Here, there is an absence of evidence of cost justifying the provision. In the circumstances, there is no rational connection between s. 72(2) and its objective, nor does s. 72(2) minimally impair the Charter rights of the Hislop class. [65‑66]

With respect to s. 60(2) of the CPP, the estates of those survivors who died more than 12 months before the coming into force of the 2000 amendments do not have standing to claim a s. 15(1) Charter right on behalf of the deceased survivor. The use of the term “individual” in s. 15(1) was intentional and indicates that s. 15(1) applies to natural persons only. In the context in which the claim is made here, an estate is just a collection of assets and liabilities of a person who has died. It is not an individual and it has no dignity that may be infringed. Mr. Hislop’s situation, however, is different. Although he died between the time his notice of appeal was filed in this Court and the hearing of this appeal, he obtained judgment while he was still alive. When a judgment is obtained, the cause of action upon which the judgment is based is merged in the judgment. Therefore, where a party dies pending appeal, the appeal survives even if the original cause of action would not. Although s. 15(1) rights are personal, the constitutional issues raised here are issues of public importance. Given the public interest in ensuring that questions of law related to such rights be correctly decided, an appeal from a judgment raising such issues must be allowed to survive the party’s death pending the appeal. Moreover, it is a long‑standing principle of law that a litigant should not be prejudiced by an act of the court. Accordingly, the estate of any class member who was alive on the date that argument concluded in the Superior Court of Justice and who otherwise met the requirements under the CPP is entitled to the benefit of this judgment. [71‑77]

Although the Hislop class frames its s. 72(1) argument as an adverse effect discrimination claim, what it is seeking is a retroactive Charter relief. Since such a relief is unavailable in this case, it is not necessary to undertake a s. 15(1) analysis in respect of s. 72(1). Because courts are adjudicative bodies that, in the usual course of things, are called upon to decide the legal consequences of past happenings, they generally grant remedies that are retroactive to the extent necessary to ensure that successful litigants will have the benefit of the ruling. However, where the law changes through judicial intervention, it may be appropriate for a court to limit the retroactive effect of its judgment. While a substantial change in the law is necessary, it is not sufficient to justify purely prospective remedies. Once the “substantial change” threshold is met, factors such as good faith reliance by governments, fairness to the litigants and the need to respect the constitutional role of legislatures must be considered to determine whether it is appropriate to limit the retroactive effect of the remedy. Here, the M. v. H. decision marked a departure from pre‑existing jurisprudence on same‑sex equality rights, and all the other relevant factors also weigh in favour of limiting retroactive relief. First, given the state of the jurisprudence prior to M. v. H., the exclusion of same‑sex partners from the former CPP was based on a reasonable understanding of the state of s. 15(1) jurisprudence as it existed at the relevant time. Second, the government did not act in bad faith in failing to extend survivors’ benefits to same‑sex couples prior to M. v. H. Third, in seeking payment of arrears back as far as 1985, the Hislop class effectively overlooks the evolution in the jurisprudence of same‑sex equality rights. Its claim for a retroactive remedy is tantamount to a claim for compensatory damages flowing from the underinclusiveness of the former CPP. Imposing that sort of liability on the government, absent bad faith, unreasonable reliance or conduct that is clearly wrong, would undermine the important balance between the protection of constitutional rights and the need for effective government that is struck by the general rule of qualified immunity. [69] [86] [99‑100] [107‑117]

With respect to ss. 44(1.1) and 72(2), the appropriate remedy under s. 52 of the Constitution Act, 1982 is to strike down both provisions. Here, the extent of the inconsistency with the equality guarantee is co‑extensive with the scope of these two provisions. As such, the inconsistency can be cured without distorting or interfering with the rest of the legislative scheme. A declaration that ss. 44(1.1) and 72(2) are of no force and effect is in keeping with the scheme and obvious purpose of the 2000 amendments to extend the survivors’ benefit to same‑sex survivors. Finally, a temporary suspension of the declaration of invalidity is not appropriate in the present case. Suspensions should only be used where striking down the legislation without enacting something in its place would pose a danger to the public, threaten the rule of law or where it would result in the deprivation of benefits from deserving persons without benefiting the rights claimant. None of these factors are present here. As a result, a class member who was precluded by s. 44(1.1) or s. 72(2) from receiving the survivor’s benefit and who otherwise meets the eligibility requirements, will be entitled to payment of that benefit. The relevant date for the purpose of that payment is the date on which application was received or where no application was made because of the unconstitutional provisions, the date on which the statement of claim was filed. In no event are benefits payable in respect of a month prior to August 1999, which is the earliest month in respect of which a class member who applied for the survivor’s benefit on the day the 2000 amendments came into force could have been eligible. [121] [134]

Lastly, pre‑judgment interest is available in the instant case. The CPP is silent on the issue of pre‑judgment interest and cannot, as such, reasonably be interpreted as creating an exception to the entitlement created by s. 31 of the Crown Liability and Proceedings Act. [135]

Per Bastarache J.: Retroactivity of a constitutional remedy granted under s. 52(1) of the Constitution Act, 1982 is generally the norm in our constitutional jurisprudence, and a decision to deny such remedy — and the Hislop class’s challenge to s. 72(1) of the CPP in this case — should be based on the balancing of interests that must take place in any claim for relief from an unconstitutional law. Reliance on the existence of a substantial change of law is not an appropriate consideration in the context of Charter rights. Such an approach implies that these rights depend on judicial recognition of a new or newly recognized social environment for their genesis. By attaching importance to changing social conditions, it makes Charter rights dependent on how the majority of Canadians perceive the claimants’ rights. Society’s views of Charter claimants — especially in the context of vulnerable minorities — should not be a factor for determining whether a Charter right was part of the Constitution in 1985, or whether it sprung into existence later and thereby be a basis for denying retroactive relief. Our Constitution may change to correspond to new realities, but that does not mean that every time a new constitutional interpretation is adopted or a previous decision is overturned that the fundamental rights and freedoms guaranteed in our Charter have changed or that new ones have been created. There is a difference between changes in constitutional interpretation and actual constitutional change. Even if the existence of a substantial change in the law was an appropriate threshold criterion, no such change actually occurred in this case. A review of the case law after 1985 suggests that this Court did not upset established law when it handed down its decision in M. v. H. [137] [140] [143‑147] [157]

The normal retroactive effect of judgments, however, may need to be tempered in certain circumstances in order to protect other legitimate interests. Reasonable reliance, good faith, fairness to litigants and Parliament’s role are important considerations to be taken into account in deciding whether a retroactive constitutional remedy should be denied. Different considerations apply when deciding whether a suspended declaration of invalidity should be granted. A suspended declaration is a valid measure when a declaration of invalidity would pose a danger to the public, threaten the rule of law or deprive deserving persons of benefits without thereby benefiting the individual whose rights had been violated. A suspended declaration of invalidity is ultimately only a temporary limit on retroactivity; it does not determine whether governments are entitled to deny retroactive relief. It will not always be necessary to establish reasonable reliance in order to deny retroactive relief. In this case, given the state of the law pre‑M. v. H., it would be more accurate to emphasize the fact that the government was acting in good faith in the face of jurisprudential uncertainty. Finally, the nature of the constitutional violation is not relevant to deciding whether to deny a retroactive remedy. Subject to these clarifications, there is general agreement with how the majority has applied the other factors to this case. [159‑164]

Cases Cited

By LeBel and Rothstein JJ.

Distinguished: Kingstreet Investments Ltd. v. New Brunswick (Finance), [2007] 1 S.C.R. 3, 2007 SCC 1; considered: R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295; Schachter v. Canada, [1992] 2 S.C.R. 679; Miron v. Trudel, [1995] 2 S.C.R. 418; Nova Scotia (Workers’ Compensation Board) v. Martin, [2003] 2 S.C.R. 504, 2003 SCC 54; referred to: M. v. H., [1999] 2 S.C.R. 3; Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497; Egan v. Canada, [1995] 2 S.C.R. 513; Benner v. Canada (Secretary of State), [1997] 1 S.C.R. 358; R. v. Oakes, [1986] 1 S.C.R. 103; Stinson Estate v. British Columbia (1999), 70 B.C.L.R. (3d) 233, 1999 BCCA 761; Edmonton Journal v. Alberta (Attorney General), [1989] 2 S.C.R. 1326; Lew v. Lee, [1924] S.C.R. 612, aff’d [1925] A.C. 819; Reid v. Batty, [1933] O.W.N. 496, aff’d [1933] O.W.N. 817; Turner v. London and South‑Western Railway Co. (1874), L.R. 17 Eq. 561; Gunn v. Harper (1902), 3 O.L.R. 693; Hubert v. DeCamillis (1963), 41 D.L.R. (2d) 495; Monahan v. Nelson (2000), 186 D.L.R. (4th) 192, 2000 BCCA 297; In re Spectrum Plus Ltd. (in liquidation), [2005] 2 A.C. 680, [2005] UKHL 41; Reference re Remuneration of Judges of the Provincial Court of Prince Edward Island, [1998] 1 S.C.R. 3; R. v. Demers, [2004] 2 S.C.R. 489, 2004 SCC 46; Reference re Manitoba Language Rights, [1985] 1 S.C.R. 721; Edwards v. Attorney‑General for Canada, [1930] A.C. 124; Reference re Same‑Sex Marriage, [2004] 3 S.C.R. 698, 2004 SCC 79; Attorney General of Quebec v. Blaikie, [1979] 2 S.C.R. 1016; Re Residential Tenancies Act, 1979, [1981] 1 S.C.R. 714; Law Society of Upper Canada v. Skapinker, [1984] 1 S.C.R. 357; Hunter v. Southam Inc., [1984] 2 S.C.R. 145; United States v. Johnson, 457 U.S. 537 (1982); Chevron Oil Co. v. Huson, 404 U.S. 97 (1971); Mackin v. New Brunswick (Minister of Finance), [2002] 1 S.C.R. 405, 2002 SCC 13; Guimond v. Quebec (Attorney General), [1996] 3 S.C.R. 347; Tétreault‑Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22.

By Bastarache J.

Referred to: Nova Scotia (Workers’ Compensation Board) v. Martin, [2003] 2 S.C.R. 504, 2003 SCC 54; Jim Pattison Industries Ltd. v. The Queen, [1984] 2 F.C. 954; Davidson v. Davidson (1986), 33 D.L.R. (4th) 161, leave to appeal refused, [1987] 1 S.C.R. vii; Rath v. Kemp (1996), 1996 ABCA 378, 46 Alta. L.R. (3d) 1; Reference re Manitoba Language Rights, [1985] 1 S.C.R. 721; Miron v. Trudel, [1995] 2 S.C.R. 418; R. v. Hess, [1990] 2 S.C.R. 906; Reference re Same‑Sex Marriage, [2004] 3 S.C.R. 698, 2004 SCC 79; M. v. H., [1999] 2 S.C.R. 3, aff’g (1996), 96 O.A.C. 173; Andrews v. Ontario (Minister of Health) (1988), 64 O.R. (2d) 258; Vogel v. Manitoba (1992), 90 D.L.R. (4th) 84, rev’d (1995), 126 D.L.R. (4th) 72; Veysey v. Canada (Commissioner of the Correctional Service), [1990] 1 F.C. 321; Knodel v. British Columbia (Medical Services Commission) (1991), 58 B.C.L.R. (2d) 356; Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554; Egan v. Canada, [1995] 2 S.C.R. 513; McKinney v. University of Guelph, [1990] 3 S.C.R. 229; Kane v. Ontario (Attorney General) (1997), 152 D.L.R. (4th) 738; Vriend v. Alberta, [1998] 1 S.C.R. 493; Rosenberg v. Canada (Attorney General) (1998), 38 O.R. (3d) 577; Schachter v. Canada, [1992] 2 S.C.R. 679; Kingstreet Investments Ltd. v. New Brunswick (Finance), [2007] 1 S.C.R. 3, 2007 SCC 1.

Statutes and Regulations Cited

Amendments Because of the Supreme Court of Canada Decision in M. v. H. Act, 1999, S.O. 1999, c. 6, s. 68(2).

Canada Pension Plan, R.S.C. 1985, c. C‑8, ss. 2(1) “common-law partner” [ad. 2000, c. 12, s. 42(2)], “spouse” [ad. c. 30 (2nd Supp.), s. 1(3); rep. 2000, c. 12, s. 42(1)], 42(1) “survivor” [ad. 2000, c. 12, s. 44(3)], 44(1)(d) [repl. idem, s. 45(1)], 44(1.1) [ad. idem, s. 45(2)], 60(2), 72 [am. idem, s. 54].

Canadian Charter of Rights and Freedoms, ss. 1, 15(1), 24(1).

Constitution Act, 1982, s. 52(1).

Crown Liability and Proceedings Act, R.S.C. 1985, c. C‑50, s. 31.

Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.).

Modernization of Benefits and Obligations Act, S.C. 2000, c. 12, ss. 42, 45(2).

Authors Cited

Blackstone, William. Commentaries on the Laws of England, vol. 1. Oxford: Clarendon Press, 1765.

Canada. Parliament. Special Joint Committee on the Constitution of Canada. Minutes of Proceedings and Evidence of the Special Joint Committee of the Senate and of the House of Commons on the Constitution of Canada (1980-81), Issue No. 43, January 22, 1981, pp. 43:39‑43:44; Issue No. 44, January 23, 1981, pp. 44:6‑44:10; Issue No. 47, January 28, 1981, p. 47:88; Issue No. 48, January 29, 1981, pp. 48:4‑48:49.

Choudhry, Sujit, and Kent Roach. “Putting the Past Behind Us? Prospective Judicial and Legislative Constitutional Remedies” (2003), 21 S.C.L.R. (2d) 205.

Driedger, Elmer A. Construction of Statutes, 2nd ed. Toronto: Butterworths, 1983.

Driedger, Elmer A. “Statutes: Retroactive Retrospective Reflections” (1978), 56 Can. Bar Rev. 264.

Fisch, Jill E. “Retroactivity and Legal Change: An Equilibrium Approach” (1997), 110 Harv. L. Rev. 1055.

Hogg, Peter W. Constitutional Law of Canada, vol. 2, loose‑leaf ed. Scarborough, Ont.: Carswell, 1992 (updated 2005, release 1).

Lord Reid. “The Judge as Law Maker” (1972-1973), 12 J.S.P.T.L. 22.

Roach, Kent. Constitutional Remedies in Canada. Aurora, Ont.: Canada Law Book, 1994 (loose-leaf updated 2006, release 13).

Sampford, Charles. Retrospectivity and the Rule of Law. Oxford: Oxford University Press, 2006.

APPEAL and CROSS‑APPEAL from a judgment of the Ontario Court of Appeal (Charron, Feldman and Lang JJ.A.) (2004), 73 O.R. (3d) 641, 246 D.L.R. (4th) 644, 192 O.A.C. 331, 124 C.R.R. (2d) 1, 12 R.F.L. (6th) 71, [2004] O.J. No. 4815 (QL), setting aside in part a decision of Macdonald J. (2003), 234 D.L.R. (4th) 465, 114 C.R.R. (2d) 303, 50 R.F.L. (5th) 26, [2003] O.J. No. 5212 (QL). Appeal and cross‑appeal dismissed.

Roslyn J. Levine, Q.C., and Paul Vickery, for the appellant/respondent on cross‑appeal.

J. J. Camp, Q.C., R. Douglas Elliott, Sharon D. Matthews, Patricia A. LeFebour, R. Trent Morris and Sean M. Grayson, for the respondents/appellants on cross‑appeal.

Daniel Guttman and Janet E. Minor, for the intervener the Attorney General of Ontario.

Hugo Jean, for the intervener the Attorney General of Quebec.

Nick Parker, for the intervener the Attorney General of Alberta.

Written submissions only by Cynthia Petersen for the intervener Egale Canada Inc.

The judgment of McLachlin C.J. and Binnie, LeBel, Deschamps, Abella and Rothstein JJ. was delivered by

LeBel and Rothstein JJ. —

I. Introduction

[1] This is an appeal by the Attorney General of Canada (“government”) from a decision of the Ontario Court of Appeal upholding in part the decision of the Ontario Superior Court of Justice and a cross-appeal by the respondents (“Hislop class”). The government appeals the finding of unconstitutionality of ss. 44(1.1) and 72(2) of the Canada Pension Plan, R.S.C. 1985, c. C-8 (“CPP”), as enacted by the Modernization of Benefits and Obligations Act, S.C. 2000, c. 12 (“MBOA”). The Hislop class cross-appeals the finding of constitutionality of ss. 72(1) and 60(2) of the CPP and the denial of a remedy with respect thereto. We propose to dismiss both the appeal and the cross-appeal.

II. Overview

[2] Under the CPP, the spouse of a contributor was entitled to apply for a survivor’s pension after the death of the contributor. If the survivor’s pension was approved, it would be payable for each month following the death of the contributor.

[3] However, if the application of the survivor was not received by the government within 12 months of the death of the contributor, the arrears that could be claimed by the survivor were limited to a 12-month period preceding the receipt of the application. For example, if the contributor died in January 1995, and the survivor’s application was not received by the government until March 1998, the maximum monthly pension arrears that could be claimed would be for 12 months preceding March 1998.

[4] Until July 2000, for purposes of entitlement to a survivor’s pension under the CPP, the survivor had to have been married to the contributor or had to be a person of the opposite sex who was cohabiting with the contributor in a conjugal relationship at the time of the contributor’s death. Same-sex conjugal relationships were not recognized and the survivor of a same-sex conjugal relationship was not eligible to receive a survivorship pension under the CPP.

[5] In M. v. H., [1999] 2 S.C.R. 3, this Court struck down the opposite-sex definition of spouse in the Family Law Act, R.S.O. 1990, c. F.3, as contrary to s. 15(1) of the Canadian Charter of Rights and Freedoms. The declaration of invalidity was suspended for six months to give the Ontario government time to review the legislation. As a consequence of M. v. H., the federal government and the governments of other provinces also amended a number of their statutes to address the constitutional deficiency. The federal government enacted the MBOA, which amended 68 pieces of legislation. This appeal involves a constitutional challenge to remedial amendments made to the CPP by the MBOA to recognize same-sex conjugal relationships for the purpose of entitlement to survivor’s pensions under the CPP.

[6] The Hislop class commenced a class action claiming that the MBOA’s remedial amendments to the CPP were contrary to s. 15(1) of the Charter as being underinclusive. There are four issues to consider. First, the amendments restricted eligibility for survivor benefits under the CPP to survivors of same-sex conjugal relationships whose partners died on or after January 1, 1998. The Hislop class argued that survivors of contributing partners of same-sex conjugal relationships who died anytime after s. 15(1) of the Charter became effective (April 17, 1985) should be eligible to make an application for CPP survivorship benefits. The Ontario Court of Appeal agreed with the Hislop class and found that the eligibility restriction violated s. 15(1) of the Charter and that the restriction could not be justified under s. 1. In other words, survivors of same-sex conjugal relationships in which the contributors died any time after April 17, 1985 should be eligible to receive survivor’s pensions. We agree with that conclusion.

[7] Second, for eligible same-sex survivors, the MBOA remedial amendments granted monthly pension payments commencing in July 2000, when the amendments came into force. As such, the amendments did not grant benefits retroactively. Irrespective of when the same-sex survivor became eligible, the monthly pension payments would not commence before July 2000.

[8] The Hislop class argued that survivors of same-sex conjugal relationships should be entitled to retroactive benefits from the time of death of the same-sex conjugal contributing partner. The Ontario Court of Appeal dismissed the Hislop class’s claim for retroactive benefits.

[9] The third and fourth issues are of less pervasive significance. In respect of the third issue, as explained earlier, general provisions of the CPP unaffected by MBOA amendments allowed for monthly pension arrears for up to 12 months preceding the receipt by the government of an application for benefits. The Ontario Court of Appeal found that the MBOA amendments that granted benefits commencing in July 2000, without entitling a same-sex survivor the opportunity of up to 12 months of arrears as was available to opposite-sex survivors, were contrary to s. 15 of the Charter and could not be justified under s. 1.

[10] As to the fourth issue, the Ontario Court of Appeal held that estates of same-sex survivors have no standing to claim a remedy for breach of s. 15(1) of the Charter.

III. Statutory Provisions at Issue

[11] Prior to the MBOA, “spouse” was defined in the CPP as:

Canada Pension Plan, R.S.C. 1985, c. C-8

2. (1) . . .

“spouse”, in relation to a contributor, means,

(a) except in or in relation to section 55,

(i) if there is no person described in subparagraph (ii), a person who is married to the contributor at the relevant time, or

(ii) a person of the opposite sex who is cohabiting with the contributor in a conjugal relationship at the relevant time, having so cohabited with the contributor for a continuous period of at least one year, and

(b) in or in relation to section 55, a person who is married to the contributor at the relevant time,

and, in the case of a contributor’s death, the “relevant time”, for greater certainty, means the time of the contributor’s death;

[12] The MBOA came into force on July 31, 2000. Section 42(1) of the MBOA repealed the definition of spouse in the CPP. Section 42(2) of the MBOA substituted a definition of “common-law partner”. In that definition there is no express or implied reference to the gender of the partners. The definition of “common-law partner” reads:

2. (1) In this Act,

“common-law partner”, in relation to a contributor, means a person who is cohabiting with the contributor in a conjugal relationship at the relevant time, having so cohabited with the contributor for a continuous period of at least one year. For greater certainty, in the case of a contributor’s death, the “relevant time” means the time of the contributor’s death.

[13] By a series of other MBOA amendments to the CPP, survivors of same-sex conjugal relationships became entitled to receive benefits under the CPP. However, by s. 45(2) of the MBOA, s. 44(1.1) of the CPP was added whereby no survivorship pension would be payable to survivors of same-sex conjugal relationships unless they became survivors on or after January 1, 1998. In other words, if the partner of the same-sex survivor died before January 1, 1998, the survivor would not be eligible for a survivorship pension under the CPP. Section 44(1.1) provides:

44. . . .

(1.1) In the case of a common-law partner who was not, immediately before the coming into force of this subsection, a person described in subparagraph (a)(ii) of the definition “spouse” in subsection 2(1) as that definition read at that time, no survivor’s pension shall be paid under paragraph (1)(d) unless the common-law partner became a survivor on or after January 1, 1998.

[14] A second provision added by the MBOA was s. 72(2) of the CPP. Section 72(2) operated to preclude payments to same-sex survivors for any month before July 2000, the month s. 72(2) came into force. Section 72(2) provides:

72. . . .

(2) In the case of a survivor who was the contributor’s common-law partner and was not, immediately before the coming into force of this subsection, a person described in subparagraph (a)(ii) of the definition “spouse” in subsection 2(1) as that definition read at that time, no survivor’s pension may be paid for any month before the month in which this subsection comes into force.

[15] The effect of s. 72(2) was to entirely preclude retroactive pension benefits to same-sex survivors. The Hislop class’s position is that same-sex survivors should be entitled to retroactive benefits to the month following the death of their same-sex conjugal partner.

[16] The incidental effect of s. 72(2) was to preclude for same-sex survivors the opportunity to seek up to 12 months of pension arrears prior to July 2000, an opportunity that was available to opposite-sex survivors under s. 72(1). Section 72(1) provides in relevant part:

72. (1) Subject to subsection (2) and section 62, where payment of a survivor’s pension is approved, the pension is payable for each month commencing with the month following

but in no case earlier than the twelfth month preceding the month following the month in which the application was received.

The effect of s. 72(2) came to an end as of June 2001, because after that month, same-sex and opposite-sex survivors benefited from the application of the general rule in s. 72(1).

[17] As indicated, s. 72(1) limits survivors’ arrears benefits to not more than 12 months prior to the month in which the application is received. The Hislop class says that s. 72(1) imposes adverse effect discrimination on survivors of same-sex conjugal relationships. While s. 72(1) is facially neutral, access to survivorship pensions for same-sex survivors was not recognized until the enactment of the MBOA. Thus, even if s. 72(2) was struck down, the effect of s. 72(1) is to limit same-sex survivors from claiming benefits for more than 12 months prior to when their application was made. The Hislop class argues that s. 72(1) should be suspended to enable same-sex survivors to obtain payment of pension benefits back to when their same-sex partner died.

[18] Finally, s. 60(2) of the CPP, a general provision not affected by the MBOA, limited the right of estates of survivors from obtaining benefits if the application for the benefits is not made within 12 months after the death of the survivor. Section 60(2) provides:

60. . . .

(2) Notwithstanding anything in this Act, but subject to subsections (2.1) and (2.2), an application for a benefit, other than a death benefit, that would have been payable in respect of a month to a deceased person who, prior to the person’s death, would have been entitled on approval of an application to payment of that benefit under this Act may be approved in respect of that month only if it is made within 12 months after the death of that person by the estate, the representative or heir of that person or by any person that may be prescribed by regulation.

[19] The Hislop class says that s. 60(2) should be suspended to enable estates of survivors of same-sex conjugal relationships to claim retroactive benefits for the period during which the survivor would have been entitled to benefits had same-sex survivors been recognized by the CPP as of the date of the death of the contributor. The suspension of s. 60(2) would only avail to the benefit of estates if ss. 44(1.1) and 72(2) are found to be unconstitutional.

IV. Judicial History

A. Ontario Superior Court of Justice (2003), 234 D.LR. (4th) 465

[20] Macdonald J. found that all the respondents in this case were in conjugal relationships with their same-sex partners for a number of years. She considered whether ss. 44(1.1) and 72(2) of the CPP breached s. 15(1) of the Charter, using the three‑part analysis set out in Law v. Canada (Minister of Employment and Immigration), [1999] 1 S.C.R. 497. The comparator group chosen by the Hislop class was married heterosexual couples. The trial judge agreed and found that the legislation treats same-sex survivors differently than survivors of married or opposite-sex common-law relationships. She observed that in Egan v. Canada, [1995] 2 S.C.R. 513, at para. 13, it was found that “‘sexual orientation’ is an analogous ground under s. 15(1)”. The judge also found that ss. 44(1.1) and 72(2) offended Hislop’s human dignity. She concluded that ss. 44(1.1) and 72(2) infringed s. 15(1) of the Charter.

[21] Macdonald J. considered the government’s argument that the Hislop class’s claim amounted to an impermissible claim for retroactive benefits for a breach of the Charter. She held that this Court’s judgment in Benner v. Canada (Secretary of State), [1997] 1 S.C.R. 358, provided “a complete answer to the arguments raised by the Crown” (para. 99). In her view, “. . . discrimination was experienced after the coming into force of s. 15 of the Charter. The plaintiffs have actually experienced discrimination since before the passage of the Charter, but they seek to apply the Charter prospectively, to discrimination that took place after April 17, 1985” (ibid.). She held that the government had failed to demonstrate that the exclusion of same‑sex survivors’ pensions was a reasonable limit on the Hislop class’s Charter rights under s. 1 of the Charter.

[22] Macdonald J. held that ss. 44(1.1) and 72(2) violated s. 15(1) of the Charter and were not justified under s. 1. Under s. 52(1) of the Constitution Act, 1982, she granted same‑sex survivors a constitutional exemption from the two general sections (ss. 60(2) and 72(1)) of the CPP. She also awarded interest under s. 31 of the Crown Liability and Proceedings Act, R.S.C. 1985, c. C-50, on the unpaid survivors’ pensions from February 1, 1992 or one month after the death of the contributor, whichever is later. She also awarded costs at approximately $1.3 million. She dismissed the Hislop class’s claims for symbolic damages of $20,000 for each class member under s. 24(1) of the Charter, and their claims for damages for breach of fiduciary duty and unjust enrichment.

B. Ontario Court of Appeal (Charron, Feldman and Lang JJ.A.) (2004), 73 O.R. (3d) 641

[23] The Court of Appeal affirmed Macdonald J.’s decision with regard to the unconstitutionality of ss. 44(1.1) and 72(2). However, the court queried why the trial judge considered ss. 60(2) and 72(1) in her analysis of remedy but did not review them in her s. 15(1) or s. 1 analysis. The court was informed that the trial judge’s remedy analysis followed the argument as presented at trial by the Hislop class’s counsel. The court invited written submissions as to the constitutional validity of ss. 60(2) and 72(1). The court then found that it had the necessary factual foundation and jurisdiction to consider the constitutional validity of these sections and that they must be subjected to a s. 15 and a s. 1 Charter analysis.

[24] The court concluded that ss. 44(1.1) and 72(2) offended s. 15(1) of the Charter. The court found that the appropriate comparator group is opposite‑sex surviving partners and that, although the impugned sections contain temporal distinctions as the government argued, those distinctions apply only to same‑sex surviving partners and not to opposite‑sex survivors. Both ss. 44(1.1) and 72(2) treat same‑sex survivors differently from their comparator group. The court said that sexual orientation is now recognized as an analogous ground under s. 15(1) of the Charter. The differential treatment discriminated in a substantive sense, and the discrimination could not be justified under s. 1.

[25] In dealing with s. 60(2), the Court of Appeal held that estates could not make a s. 15(1) Charter application because s. 15(1) only applies to “[e]very individual”. Survivors’ estates enjoy no s. 15(1) rights and there was no basis to assess whether such rights were breached by s. 60(2).

[26] With respect to s. 72(1), the court held that it did not limit same-sex survivors’ rights. The court observed, at para. 106, that “[i]t is only once s. 44(1.1) and s. 72(2) are declared unconstitutional that s. 72(1) may have an adverse effect” on the Hislop class, by limiting pension entitlement to 12 months of arrears, no matter when the partner died. Most same-sex survivors did not apply for the survivor’s pension when their partners died, because they were not eligible to apply under the CPP. The court held that s. 72(1) did not violate s. 15(1) of the Charter because, “in the context of the MBOA, it had no adverse effect on the claimants, and at this stage, it is not clear that it will have an adverse effect on the class” (para. 111). Therefore, there was no need to consider s. 1.

[27] The Court of Appeal upheld the declaration of invalidity in respect of ss. 44(1.1) and 72(2). The court held that the two general sections, ss. 60(2) and 72(1) of the CPP, do not breach s. 15(1) of the Charter. Consequently, there was no remedy to be granted in respect of those sections. The court set aside the constitutional exemption which was ordered by the trial judge with respect to ss. 60(2) and 72(1).

[28] The Court of Appeal held that the trial judge properly exercised her discretion on pre‑judgment interest. She was entitled to award pre‑judgment interest and there is no basis to interfere with that disposition. Costs were awarded on a partial indemnity basis.

C. Appeal and Cross-Appeal to This Court

[29] The government now appeals the declaration of invalidity of ss. 44(1.1) and 72(2) and the awarding of pre-judgment interest on pension arrears. The Hislop class cross-appeals the finding of the Court of Appeal which set aside the constitutional exemptions ordered by the trial judge with respect to ss. 60(2) and 72(1).

V. The Issues

[30] On October 6, 2005, the Chief Justice stated the following constitutional questions:

1. Does s. 44(1.1) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

2. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

3. Does s. 72(2) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

4. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

5. Does s. 60(2) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

6. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

7. Does s. 72(1) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

8. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

[31] By way of cross-appeal, the Hislop class submitted three questions:

(1) Can estates advance s. 15(1) rights?

(2) Is there an entitlement to interest on survivor’s pension arrears?

(3) Did the Court of Appeal err in setting aside the constitutional exemption in respect of ss. 60(2) and 72(1)?

[32] In essence, the issues are:

(1) Should survivors of same-sex conjugal relationships in which the contributing partner died prior to January 1, 1998 be eligible to receive CPP survivorship benefits?

(2) Should survivors of same-sex conjugal relationships in which the contributing partner died at any time after April 17, 1985 be entitled to retroactive CPP benefits from the month following the death of the contributing partner?

(3) Should estates of deceased survivors be entitled to claim Charter relief which the survivors would have been entitled to claim?

VI. Analysis

[33] The MBOA was enacted in response to this Court’s decision in M. v. H. The Act seeks to eliminate from federal legislation distinctions based on sexual orientation. In the context of survivorship benefits under the CPP, the MBOA amendments entitle survivors of conjugal relationships, whether they be of the same or opposite sex, access to the same benefit regime. In other words, the gender of the parties to a conjugal relationship is now irrelevant in the survivorship provisions of the CPP. Even though the MBOA is remedial legislation, the complaint is that the relief granted by the MBOA does not go far enough, in that it fails to confer eligibility on survivors whose same-sex partner died prior to January 1, 1998 and fails to grant retroactive relief.

A. Section 44(1.1)

[34] It will be convenient to first deal with the eligibility restriction. The MBOA recognized that the former CPP regime violated s. 15(1) of the Charter because it failed to recognize survivors of same-sex conjugal relationships. Section 44(1.1) continues to fail to recognize survivors of same-sex conjugal relationships whose contributing partner died before January 1, 1998.

[35] Section 15(1) of the Charter provides:

Equality Rights

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.

[36] The test applicable to s. 15(1) of the Charter was articulated by this Court in Law. In short, there must be (1) differential treatment on the basis of a personal characteristic, (2) that is an enumerated or analogous ground, (3) which is discriminatory in purpose or effect.

(1) Differential Treatment

[37] As the Court of Appeal observed, essential to the question of differential treatment is the choice of comparator group. Throughout this litigation, the government has argued that s. 44(1.1) draws a temporal distinction only. The government’s position is that the provisions of the MBOA do not differentiate between same-sex couples and opposite-sex couples, but rather, between two groups of survivors of same-sex relationships, based on the date their relationships ended as a result of one partner’s death. It cannot, therefore, violate s. 15(1) because a temporal basis for a distinction is not an enumerated or recognized analogous ground of discrimination. In our opinion, the courts below were correct in rejecting this argument.

[38] To frame the comparator group in terms of the express distinction made in s. 44(1.1) between survivors whose partners died before January 1, 1998 and those whose partners died on or after that date would be to miss the fundamental reason for the enactment of the MBOA. In M. v. H., this Court held that the distinction in the spousal support regime between same-sex and opposite-sex couples was unconstitutional and that it could not be saved under s. 1. The MBOA was expressly intended to extend equal treatment to same-sex partners in a wide range of statutes. It is the purpose of the MBOA itself that determines the appropriate comparator group. What must be compared is the subset of same-sex survivors that remains excluded from the CPP survivor’s benefits, i.e., those whose partners died before January 1, 1998, and similarly situated opposite-sex survivors. The appropriate comparator group in respect of the s. 44(1.1) analysis is survivors of opposite-sex conjugal relationships whose partners died before January 1, 1998.

[39] If the government was correct, remedial legislation intended to address the constitutional infirmity of existing legislation, but which limited eligibility for relief on a temporal basis, could never be the subject of a successful s. 15(1) Charter challenge. That is because a temporal basis of distinction is not one based upon grounds enumerated in s. 15(1) or grounds analogous thereto. When the government enacts remedial legislation, that legislation may still violate s. 15(1) requirements. The fact that it is remedial legislation does not immunize it from Charter review.

[40] In this Court, the government’s primary challenge to the s. 15(1) decision of the Court of Appeal was based on the choice of comparator group. We find the Court of Appeal did not err in the selection of comparator group.

[41] The government cursorily argued that the Charter should be interpreted in a manner recognizing the evolution of societal opinions and the incremental recognition of analogous grounds. The government says the MBOA amendments restricting eligibility to survivors whose partners died on or after January 1, 1998 are consistent with that evolution. With respect, we do not see how the evolution of societal opinions and the incremental recognition of analogous grounds bear on whether s. 44(1.1) continues the discrimination which the MBOA amendments are intended to remedy. The question is not whether there was recognition of same-sex discrimination prior to 1998. Rather, the question is whether the prior discrimination which was recognized by the MBOA is or is not now being remedied. If survivors of same-sex conjugal relationships whose partners died before 1998 continue to be ineligible for CPP survivorship pensions, the prior discrimination that has been recognized by Parliament in enacting the MBOA continues for such survivors. For these reasons, we do not accept the government’s evolution argument as a valid response to the s. 15(1) claim in respect of s. 44(1.1).

[42] This disposes of the government’s s. 15(1) arguments. The government did not argue any other error in the Charter analysis of the Court of Appeal. We agree with that analysis and need not duplicate it here.

(2) Section 1

[43] Section 1 of the Charter provides:

Guarantee 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.

Under s. 1, the government has the burden to demonstrate that a discriminatory provision is a reasonable limit on a s. 15(1) Charter right. If it meets this burden, the law will be saved as being a demonstrably justified reasonable limit on that right.

[44] The framework for a s. 1 analysis is the well-known Oakes test (see R. v. Oakes, [1986] 1 S.C.R. 103, at pp. 138-40). The Oakes test may be formulated as two main tests with subtests under the second branch, but it may be easier to think of it in terms of four independent tests. If the legislation fails under any one test, it cannot be justified. The four tests ask the following questions:

(1) Is the objective of the legislation pressing and substantial?

(2) Is there a rational connection between the government’s legislation and its objective?

(3) Does the government’s legislation minimally impair the Charter right or freedom at stake?

(4) Is the deleterious effect of the Charter breach outweighed by the salutary effect of the legislation?

[45] As the MBOA was the legislative response to this Court’s decision in M. v. H., the Act obviously has a pressing and substantial purpose. However, in M. v. H., the Court also held that “when dealing with underinclusive legislation it is important also to consider the impugned omission when construing the objective” (para. 100).

[46] It is therefore necessary to consider whether the objective of s. 44(1.1) is pressing and substantial. The effect of s. 44(1.1) is to limit eligibility for CPP survivorship pensions in the case of survivors of same-sex conjugal relationships. No survivorship pension is payable where the partner died prior to January 1, 1998.

[47] The government’s pressing and substantial objective argument in this Court is different than in the Ontario courts. In the Ontario courts, the government’s pressing and substantial argument was based on the evolutionary recognition of Charter rights and that, prior to M. v. H., there would have been no rights to same-sex partners. In this Court, the government argues that the pressing and substantial objective of s. 44(1.1) is the matching of benefits conferred under the CPP with obligations imposed on same-sex partners under other legislation.

[48] In the government’s factum and in its oral argument there was no reference to evidence before the trial judge supporting the matching of benefits and obligations argument. The reasons of the trial judge and the Court of Appeal make no reference to the matching argument. It appears that this argument was raised for the first time in this Court.

[49] We acknowledge that there may be cases in which the pressing and substantial objective of legislation and the impugned provisions are obvious and may be deduced from the legislation itself. However, in the majority of cases, in order to satisfy the pressing and substantial objective test, the government must adduce some evidence to support its argument.

[50] In the case of s. 44(1.1), the issue is one of eligibility only. It is not a provision that purports to make benefits retroactive or that would confer a lump-sum windfall on a survivor. Therefore, it is not obvious why the matching argument supports the limitation imposed by s. 44(1.1).

[51] The government seems to relate the January 1, 1998 eligibility date in the CPP to provisions of the Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.), allowing for a three-year limitation period for the government to re-assess or the individual to refile individual tax returns under that Act. Again, it is not clear why re-assessment or refiling under the Income Tax Act would have any relation to eligibility under the CPP. The eligibility date of January 1, 1998 in the CPP has no bearing on income of years prior to 2000 which is the first year in which same-sex survivors would actually receive CPP benefits. Therefore the matching argument relying on limitation periods in the Income Tax Act is perplexing.

[52] If there was something to the matching argument in respect of eligibility, one would think that s. 44(1.1) would not have made the eligibility date January 1, 1998, but rather would have related it to the date the legislation came into force, July 2000. The fact that the government saw no difficulty in granting eligibility as of January 1, 1998 suggests that matching cannot be considered to be a pressing and substantial objective of s. 44(1.1).

[53] We do not say that in an appropriate case the matching of benefits and obligations may not be a pressing and substantial objective of limits on the extension of social benefit legislation. We also acknowledge the government’s argument that there are complex relationships between income tax credits, pensions, supplements and other entitlements, often based on conjugal relationships. But it is not enough to make general reference to these relationships. There must be an explanation supported by relevant evidence as to what those relationships are, why they are relevant and why they justify the limit on the Charter right that has been found to be violated.

[54] The absence of evidence justifying the matching argument also precludes a finding of rational connection between s. 44(1.1) and its objective. Without evidence, it is impossible to understand why s. 44(1.1) is rationally connected to the matching of benefits and obligations. For the same reason there is no evidence as to why s. 44(1.1) is minimally impairing of the s. 15 right to which same-sex survivors have been found to be entitled.

[55] More than a vague argument raised for the first time in this Court by the government will be necessary if a s. 1 argument is to succeed. It cannot succeed in this case. Because the government has failed to establish the pressing and substantial objective of the provision, its rational connection to its objective and that it minimally impairs the Hislop class’s Charter rights, its argument fails in respect of s. 44(1.1) of the CPP.

B. Section 72(2)

[56] Section 72(2) became effective in July 2000. The earliest month for which survivorship benefits may be payable to a survivor of a same-sex relationship is therefore July 2000.

[57] The Hislop class says that s. 72(2) is to be contrasted with s. 72(1) which is the provision which applies generally to all survivors eligible for CPP survivorship pensions. Section 72(1) provides that, in the general case, survivorship pension arrears are payable for a period of up to 12 months preceding the month following the month in which the application for the pension was received.

[58] After June 2001, survivors of same-sex and opposite-sex relationships are equally entitled to up to 12 months of arrears benefits under s. 72(1). However, s. 72(2) provides that, in the transitional period from July 2000 to June 2001, same-sex survivors could not access arrears prior to July 2000. By reason of s. 72(2), same-sex and opposite-sex survivors are treated differently in that transitional period.

[59] The government makes the same “comparator group” argument in respect of s. 72(2) as it did in respect of s. 44(1.1). It says the MBOA amendments did not differentiate between same-sex and opposite-sex couples but rather between two groups of survivors of same-sex relationships based on the date on which the relationship ended as a result of the death of one of the partners.

[60] The government is correct that s. 72(2) creates a temporal distinction. But that does not make its choice of comparator group correct. The appropriate comparison is between same-sex and opposite-sex survivors who applied for the survivor’s pensions between July 2000 and June 2001. As with s. 44(1.1), the government’s challenge to the s. 15(1) decision of the Court of Appeal is based solely on the choice of comparator group and the evolutionary recognition of analogous grounds. For the reasons we have expressed in our analysis of the s. 15(1) arguments in respect of s. 44(1.1), we do not find merit in the government’s submissions in respect of s. 72(2).

[61] With respect to s. 1, the government seems to make two arguments in support of s. 72(2). One is that the general principle is that legislation operates from the time of enactment forward and that this approach is consistent with prior amendments to the CPP. The other is that the focus was to fix the liability of the CPP as of July 2000.

[62] We agree that remedial legislation generally operates prospectively. However, the only effect of striking down s. 72(2) is to render s. 72(1) applicable to survivors of same-sex conjugal relationships during the July 2000 to June 2001 period. For those same-sex survivors who applied for survivors’ pensions during that time, they, like opposite-sex survivors, could be entitled to up to 12 months of arrears depending upon when the contributor died.

[63] The effect of s. 72(1) applying to same-sex survivors during the period from July 2000 to June 2001 means that survivors who applied during that period could be entitled to arrears, at most, back to August 1999. While the striking down of s. 72(2) will grant access of up to 12 months of pension arrears, this is a result of retroactive benefits which Parliament itself granted by enacting s. 72(1).

[64] We recognize that costs may be a factor in a s. 1 analysis. The government says that July 2000 was selected in order to fix the liability of the CPP as of that date. However, as of July 2000, the CPP had to anticipate up to 12 months of arrears liability arising from opposite-sex relationships. The government did not refer to evidence to suggest that cost was the reason to deny same-sex survivors up to 12 months of benefit arrears during the transitional period.

[65] We are not persuaded that, in the circumstances of this case, s. 72(2) is justified under s. 1 of the Charter. Access of up to 12 months of arrears payments prior to July 2000 is provided by existing legislation, namely s. 72(1). Parliament itself has provided for limited retroactive benefits. Where the issue is solely whether Charter relief claimants should be entitled to the same retroactive benefits expressly available to their comparator group, absent cost considerations, it is difficult to see how denial of these benefits has a pressing and substantial objective. Here, there is an absence of evidence of cost justifying the provision. In the circumstances, we cannot find there is a rational connection between s. 72(2) and its objective, or indeed, that s. 72(2) minimally impairs the Charter rights of the Hislop class.

[66] The government has failed to establish a s. 1 justification for s. 72(2).

C. Section 72(1)

[67] Because we have found that s. 72(2) violates s. 15(1) and is not saved by s. 1 of the Charter, same-sex survivors will be entitled to up to 12 months of pension arrears for the period from August 1999 to July 2000, pursuant to s. 72(1).

[68] However, the Hislop class takes issue with s. 72(1) itself. The Hislop class argues that although s. 72(1) is facially neutral, its 12-month limitation on pension arrears has an adverse effect on same-sex survivors. It says that same-sex survivors were unable, prior to July 2000, to make a claim for survivorship benefits. Based upon s. 44(1.1) and s. 72(2) being struck down, it is argued that same-sex survivors should be entitled to claim retroactive benefits from the time they became survivors after April 17, 1985 when s. 15(1) of the Charter came into force. For example, if an individual became a survivor of a same-sex relationship in 1995, the survivor should be entitled to a lump sum of benefits for the period from 1995 to the time the application was received. The Hislop class argues that the operation of s. 72(1) should be suspended for same-sex survivors to enable them to obtain those retroactive benefits.

[69] Although the Hislop class frames the s. 72(1) argument as an adverse effect discrimination claim, the issue which the argument raises is, in fact, one of remedy. What the Hislop class is seeking is retroactive Charter relief. Their request for a constitutional exemption from the limitation on arrears in s. 72(1) is, in effect, a request for a remedy in respect of their exclusion from the survivors’ benefits by the pre-MBOA CPP between 1985 and 2000. As will be explained hereafter, this Court has been explicit in restricting entitlement to retroactive Charter relief of this nature. Because the remedy sought by the Hislop class is unavailable in any event, it is not necessary to undertake a s. 15(1) analysis in respect of s. 72(1).

D. Section 60(2)

[70] Where a survivor entitled to a survivor’s CPP pension dies without making application for that pension, the survivor’s estate may apply and obtain the pension benefit to which the survivor would have been entitled, provided the estate makes application within 12 months after the death of the survivor. The Hislop class submits that, because some same-sex survivors had been deceased for over 12 months when the MBOA amendments to the CPP came into effect, their estates should be able to apply for the benefits to which the survivors would have been entitled and that the 12-month limitation in s. 60(2) should be suspended so as not to bar such estate claims.

[71] The threshold issue is whether the estates of those survivors who died more than 12 months before the coming into force of the MBOA amendments to the CPP may have standing to claim a s. 15(1) Charter right on behalf of the deceased survivor. Only if they have such standing may the Court even entertain an argument that s. 60(2) should not apply to such estates. The Hislop class relies on R. v. Big M Drug Mart Ltd., [1985] 1 S.C.R. 295, at pp. 322-23. Big M Drug Mart dealt with s. 2 of the Charter which uses the term “[e]veryone”. The term used in s. 15(1) is more precise and narrower, as it allows rights to “[e]very individual”.

[72] The government submits, on the basis of the British Columbia Court of Appeal judgment in Stinson Estate v. British Columbia (1999), 70 B.C.L.R. (3d) 233, 1999 BCCA 761, that s. 15(1) rights cannot be enforced by an estate because those rights are personal and terminate with the death of the affected individual. The government also submits that estates are not individuals but artificial entities incapable of having their human dignity infringed. In addition, the government relies on the Special Joint Committee on the Constitution (see Minutes of Proceedings and Evidence of the Special Joint Committee of the Senate and of the House of Commons on the Constitution of Canada (1980-81), Issue No. 43, January 22, 1981, at pp. 43:39-43:44; see also Issue No. 44, January 23, 1981, at pp. 44:6-44:10; Issue No. 47, January 28, 1981, at p. 47:88; and Issue No. 48, January 29, 1981, at pp. 48:4-48:49), which substituted the words “every individual” for “everyone” in s. 15(1) in response to the Minister of Justice’s desire “to make it clear that this right would apply to natural persons only” (p. 43:41). The government further argues that this Court has held that s. 15(1) rights could not be claimed by other entities such as corporations (see Edmonton Journal v. Alberta (Attorney General), [1989] 2 S.C.R. 1326, at p. 1382, per La Forest J.).

[73] In our opinion, the government’s submissions have merit. In the context in which the claim is made here, an estate is just a collection of assets and liabilities of a person who has died. It is not an individual and it has no dignity that may be infringed. The use of the term “individual” in s. 15(1) was intentional. For these reasons, we conclude that estates do not have standing to commence s. 15(1) Charter claims. In this sense, it may be said that s. 15 rights die with the individual.

[74] Mr. Hislop’s individual situation, however, is different. Although he died between the time his notice of appeal was filed in this Court and the hearing of this appeal, he obtained judgment while he was still alive.

[75] When a judgment is obtained, the cause of action upon which the judgment is based is merged in the judgment: Lew v. Lee, [1924] S.C.R. 612, aff’d on this point [1925] A.C. 819 (P.C.); Reid v. Batty, [1933] O.W.N. 496 (H.C.J.), aff’d [1933] O.W.N. 817 (C.A.). In Lew, Anglin C.J. explained that, because of the doctrine of merger, the issue in an appeal is not the original cause of action but rather the legality and validity of the judgment. As such, where a party dies pending appeal, the appeal survives even if the original cause of action would not.

[76] It should be noted that Anglin C.J. relied in part on a provision in the British Columbia Supreme Court Rules (currently r. 15(2)), which provided that whether a cause of action survives or not, the death of either party between verdict or finding of the issues of fact and judgment will not give rise to abatement and that judgment may be entered notwithstanding death. He reasoned that, a fortiori, the right to enforce a judgment or defend it on appeal must also survive. In our view, his analysis is applicable in the instant case irrespective of any legislative provision. Although s. 15(1) rights are personal, the constitutional issues raised here are issues of public importance. Given the public interest in ensuring that questions of law related to such rights be correctly decided, an appeal from a judgment raising such issues must be allowed to survive the party’s death pending the appeal.

[77] Although the preceding comments are sufficient to dispose of the issue in relation to Mr. Hislop himself, because this is a class action, it is appropriate to clarify with more precision the time at which s. 15(1) rights crystallize. Merger, as we have explained, occurs when judgment is entered. Nevertheless, it is a long-standing principle of law that a litigant should not be prejudiced by an act of the court (actus curiae neminem gravabit): Turner v. London and South-Western Railway Co. (1874), L.R. 17 Eq. 561. Based on this principle, in cases where a plaintiff has died after the conclusion of argument but before judgment was entered, courts have entered judgment nunc pro tunc as of the date that argument concluded: see Gunn v. Harper (1902), 3 O.L.R. 693 (C.A.); Hubert v. DeCamillis (1963), 41 D.L.R. (2d) 495 (B.C.S.C.); Monahan v. Nelson (2000), 186 D.L.R. (4th) 192, 2000 BCCA 297. We affirm the correctness of this approach and conclude that the estate of any class member who was alive on the date that argument concluded in the Ontario Superior Court, and who otherwise met the requirements under the CPP, is entitled to the benefit of this judgment.

E. Remedies

[78] We now turn to the question of remedy. In challenging s. 72(1) of the CPP, the appellants seek a fully retroactive remedy. They argue that a declaration of invalidity under s. 52 of the Constitution Act, 1982 necessarily operates back to the coming into force of s. 15 of the Charter. In order to determine the validity of this claim, we must first consider the nature of constitutional remedies and the circumstances under which courts may limit a retroactive remedy and craft a prospective remedy. In the present appeal, we conclude that a prospective remedy would be appropriate. A retroactive remedy would be unwarranted in respect of s. 72(1) CPP. We will also determine the appropriate remedies in respect of s. 44(1.1) and s. 72(2) CPP.

[79] In substance, the position of the appellants is predicated on the traditional — often called Blackstonian — view that judges never make law, but merely discover it. In this perspective, the courts are said to apply the law as it really was or has been rediscovered. As a consequence of the declaration of nullity, the appellants claim that they are entitled to the full benefits of the law, in conformity with an understanding of the Constitution, which is deemed to have never changed.

[80] The supremacy clause, now enshrined at s. 52, is silent about the remedies which may flow from a declaration of nullity. Does it mean that such a declaration is always both prospective and retroactive? This does not appear to have been the position of our Court throughout the incremental development of the law of constitutional remedies after the adoption of the Charter. A body of jurisprudence now accepts the legitimacy of limiting the retroactive effect of a declaration of nullity and of fashioning prospective remedies in appropriate circumstances.

(1) Retroactive and Prospective Remedies Under the [Charter](https://www.canlii.org/en/ca/laws/stat/schedule-b-to-the-canada-act-1982-uk-1982-c-11/latest/schedule-b-to-the-canada-act-1982-uk-1982-c-11.html)

[81] The Constitution empowers courts to issue constitutional remedies with both retroactive and prospective effects: see, e.g., Schachter v. Canada, [1992] 2 S.C.R. 679, at p. 719. Section 24(1) of the Charter enables individuals who have had their Charter rights violated to seek redress for those past wrongs and “obtain such remedy as the court considers appropriate and just”. Section 24(1) may also, in some situations, enable the claimant to recover damages, which are necessarily retroactive: Schachter, at pp. 725-26.

[82] Section 52(1) instructs courts to declare unconstitutional legislation of no force or effect. When a court issues a declaration of invalidity, it declares that, henceforth, the unconstitutional law cannot be enforced. The nullification of a law is thus prospective. However, s. 52(1) may also operate retroactively so far as the parties are concerned, reaching into the past to annul the effects of the unconstitutional law: see, e.g., Miron v. Trudel, [1995] 2 S.C.R. 418.

[83] This Court has applied in many cases the “declaratory approach” to constitutional remedies, which implies that s. 52(1) remedies are often given retroactive effect. See, for example, Nova Scotia (Workers’ Compensation Board) v. Martin, [2003] 2 S.C.R. 504, 2003 SCC 54, at para. 28, Gonthier J. On this view, s. 52(1) remedies are deemed to be fully retroactive because the legislature never had the authority to enact an unconstitutional law. In the words of Professor Hogg, a declaration of constitutional invalidity “involves the nullification of the law from the outset” (P. W. Hogg, Constitutional Law of Canada (loose-leaf ed.), vol. 2, at p. 55-2 (emphasis added)). If the law was invalid from the outset, then any government action taken pursuant to that law is also invalid, and consequently, those affected by it have a right to redress which reaches back into the past.

[84] As mentioned above, the declaratory approach is derived from Blackstone’s famous aphorism that judges do not create law but merely discover it: W. Blackstone, Commentaries on the Laws of England (1765), vol. 1, at pp. 69-70. It reflects a traditional and widespread understanding of the role of the judiciary in a democratic state governed by strong principles of separation of powers between courts, legislatures and executives. In this perspective, courts grant retroactive relief applying existing law or rediscovered rules which are deemed to have always existed. On the other hand, legislators fashion new laws for the future.

[85] Blackstone’s declaratory approach has not remained unchallenged in modern law. Commentators and courts have pointed out that judges fulfill a legitimate law-making function. Judges do not merely declare law; they also make law. These critics argue that Blackstone’s view is a fiction as judges make law, especially in the common law world. See, e.g., Lord Reid, “The Judge as Law Maker” (1972-1973), 12 J.S.P.T.L. 22. They say such a fiction should not be turned into an ironclad principle.

[86] However, this acknowledgement does not require abandoning Blackstone’s declaratory approach altogether. The critique of the Blackstonian approach applies only to situations in which judges are fashioning new legal rules or principles and not when they are applying the existing law. In instances where courts apply pre-existing legal doctrine to a new set of facts, Blackstone’s declaratory approach remains appropriate and remedies are necessarily retroactive. Because courts are adjudicative bodies that, in the usual course of things, are called upon to decide the legal consequences of past happenings, they generally grant remedies that are retroactive to the extent necessary to ensure that successful litigants will have the benefit of the ruling: see S. Choudhry and K. Roach, “Putting the Past Behind Us? Prospective Judicial and Legislative Constitutional Remedies” (2003), 21 S.C.L.R. (2d) 205, at pp. 211 and 218. There is, however, an important difference between saying that judicial decisions are generally retroactive and that they are necessarily retroactive. When the law changes through judicial intervention, courts operate outside of the Blackstonian paradigm. In those situations, it may be appropriate for the court to issue a prospective rather than a retroactive remedy. The question then becomes what kind of change and which conditions will justify the crafting of judicial prospective remedies.

[87] The House of Lords recently adopted this view in the course of its discussion of prospective overruling: In re Spectrum Plus Ltd. (in liquidation), [2005] 2 A.C. 680, [2005] UHKL 41. The words of Lord Nicholls at para. 34 are particularly apt:

[Blackstone’s declaratory] theory is still valid when applied to cases where a previous decision is overruled as wrong when given. Most overruling occurs on this basis. These cases are to be contrasted with [those] where the later decision represents a response to changes in social conditions and expectations. Then, on any view, the declaratory approach is inapt. In this context [this] approach has long been discarded. It is at odds with reality.

[88] Although Lord Nicholls’ statement arose in the common law context, this Court has implicitly adopted a similar line of reasoning in constitutional law. Despite this Court’s endorsement of the Blackstonian declaratory approach, in its development of the law of constitutional remedies, it has frequently seen fit to temper the retroactive effect of s. 52(1) remedies and adopt a position similar to that of Lord Nicholls. On occasion, this Court has expressly stated that the s. 52(1) remedy would “apply prospectively” only: Reference re Remuneration of Judges of the Provincial Court of Prince Edward Island, [1998] 1 S.C.R. 3, at para. 18. In that reference, the Court held that the provinces’ remuneration scheme for provincial judges violated the right to a trial before an independent and impartial tribunal. In fashioning its remedy, the Court side-stepped Blackstonian doctrine and provided for a “transition period of one year before th[e] requirement [took] effect” (para. 18).

[89] The use of the “transition period” after a finding of unconstitutionality is inconsistent with the declaratory approach. It keeps in force by judicial decision a law that was deemed never to have been of force or effect. Thus, the fact that this Court has occasionally prescribed transition periods, intended to give a s. 52(1) remedy prospective effect only, suggests that this Court is not wedded to the declaratory approach in all Charter cases.

[90] In another type of situation, which arises more frequently, the Court has held that providing immediate and retroactive judicial remedies may be “inappropriate” when “doing so would create a lacuna in the regime before Parliament would have a chance to act”: R. v. Demers, [2004] 2 S.C.R. 489, 2004 SCC 46, at para. 57. In such cases, the Court has temporarily suspended the declaration of invalidity of the unconstitutional legislation to avoid creating a “legal vacuum” or “legal chaos” before Parliament or the legislature has the opportunity to enact something in place of the unconstitutional legislation: Reference re Manitoba Language Rights, [1985] 1 S.C.R. 721 (“Manitoba Language Rights Reference”), at p. 747; Schachter. In Schachter, this Court held that the suspended declaration of invalidity was appropriate when giving immediate retroactive effect to the Court’s declaration of invalidity would (a) “pose a danger to the public”; (b) “threaten the rule of law”; or (c) “result in the deprivation of benefits from deserving persons”, such as when the legislation was “deemed unconstitutional because of underinclusiveness rather than overbreadth”: Schachter, at p. 719.

[91] Like transition periods and other purely prospective remedies, the suspended declaration of invalidity is not fully consistent with the declaratory approach. By suspending the declaration of invalidity, the Court allows the constitutional infirmity to continue temporarily so that the legislature can fix the problem. In other words, the Court extends the life of a law which, on the Blackstonian view, never existed.

[92] Although if the legislature fails to comply with the Court’s order within the period of suspension, the Court’s declaration would apply retroactively, the purpose of a suspended declaration of invalidity can be to facilitate the legislature’s function in crafting a prospective remedy. The temporal delay in striking down the law also has the effect of extending the life of an unconstitutional law. In such cases, to allow the claimants to recover concurrent retroactive relief would be at cross-purposes with the Court’s decision to grant a suspended declaration of invalidity: Schachter, at p. 720.

[93] The determination of whether to limit the retroactive effect of a s. 52(1) remedy and grant a purely prospective remedy will be largely determined by whether the Court is operating inside or outside the Blackstonian paradigm. When the Court is declaring the law as it has existed, then the Blackstonian approach is appropriate and retroactive relief should be granted. On the other hand, when a court is developing new law within the broad confines of the Constitution, it may be appropriate to limit the retroactive effect of its judgment.

[94] The approach which our Court has adopted in respect of the crafting of constitutional remedies also flows from its understanding of the process of constitutional interpretation, which the “living tree” metaphor neatly describes. From the time Lord Sankey L.C. used these words to characterize the nature of the Canadian Constitution, courts have relied on this expression to emphasize the ability of the Constitution to develop with our country (Edwards v. Attorney-General for Canada, [1930] A.C. 124 (P.C.), at p. 136). This Court has often stated that the Canadian Constitution should not be viewed as a static document but as an instrument capable of adapting with the times by way of a process of evolutionary interpretation, within the natural limits of the text, which “accommodates and addresses the realities of modern life”: Reference re Same-Sex Marriage, [2004] 3 S.C.R. 698, 2004 SCC 79, at para. 22; see also Attorney General of Quebec v. Blaikie, [1979] 2 S.C.R. 1016, at p. 1029; Re Residential Tenancies Act, 1979, [1981] 1 S.C.R. 714, at p. 723; Law Society of Upper Canada v. Skapinker, [1984] 1 S.C.R. 357, at p. 365; Hunter v. Southam Inc., [1984] 2 S.C.R. 145, at p. 155.

[95] It is true that the “living tree” doctrine is not wedded to a particular model of the judicial function. At times, its application may reflect the fact that, in a case, the Court is merely declaring the law of the country as it has stood and that a retroactive remedy is then generally appropriate. In other circumstances, its use recognizes that the law has changed, that the change must be acknowledged and that, from a given point in time, the new law or the new understanding of some legal principle will prevail.

[96] The question is no longer the legitimacy of prospective remedies, but rather when, why and how judges may rule prospectively or restrict the retroactive effect of their decisions in constitutional matters. The key question becomes the nature and effect of the legal change at issue in order to determine whether a prospective remedy is appropriate. The legitimacy of its use turns on the answer to this question.

[97] There must be change, but what kind of change will be enough? Given the often incremental nature of changes in judge-made law in a common law system, the question is bound to raise difficulties. Various formulas have been suggested. For example, at a time when it appeared to be moving towards a broad recognition of prospective overruling, the U.S. Supreme Court relied on a notion of a “clear break with the past”. The limitation on proactive remedies could be considered after litigants had established such a change (see United States v. Johnson, 457 U.S. 537 (1982), at p. 549).

[98] On other occasions, the same court crafted different formulas. In an earlier case, in 1971, it relied on a test which tended to look at both the nature of the change and of some of its consequences (Chevron Oil Co. v. Huson, 404 U.S. 97 (1971), at pp. 106‑7). It identified three relevant considerations:

First, the decision to be applied nonretroactively must establish a new principle of law, either by overruling clear past precedent on which litigants may have relied . . . or by deciding an issue of first impression whose resolution was not clearly foreshadowed. . . . Second, it has been stressed that “we must . . . weigh the merits and demerits in each case by looking to the prior history of the rule in question, its purpose and effect, and whether retrospective operation will further or retard its operation.” . . . Finally, we have weighed the inequity imposed by retroactive application, for “[w]here a decision of this Court could produce substantial inequitable results if applied retroactively, there is ample basis in our cases for avoiding the ‘injustice or hardship’ by a holding of nonretroactivity.”

See also: J. E. Fisch, “Retroactivity and Legal Change: An Equilibrium Approach” (1997), 110 Harv. L. Rev. 1055, at pp. 1060-63; C. Sampford, Retrospectivity and the Rule of Law (2006), at pp. 211-12.

[99] Change in the law occurs in many ways. “Clear break with the past” catches some of its diversity. It can be best identified with those situations where, in Canadian law, the Supreme Court departs from its own jurisprudence by expressly overruling or implicitly repudiating a prior decision. Such clear situations would justify recourse to prospective remedies in a proper context. But other forms of substantial change may be as relevant, especially in constitutional adjudication, where courts must give content to broad, but previously undefined, rights, principles or norms. The definition of a yet undetermined standard or the recognition that a situation is now covered by a constitutional guarantee also often expresses a substantial change in the law. The right may have been there, but it finds an expression in a new or newly recognized technological or social environment. Such a legal response to these developments properly grounds the use of prospective remedies, when the appropriate circumstances are met. A substantial change in the law is necessary, not sufficient, to justify purely prospective remedies. Hence, we must now turn to what else must be considered once legal change has been established.

[100] Although the list of such factors should not be considered as closed, some of them appear more clearly compelling. They may include reasonable or in good faith reliance by governments (Miron, at para. 173; Mackin v. New Brunswick (Minister of Finance), [2002] 1 S.C.R. 405, 2002 SCC 13, at para. 78), or the fairness of the limitation of the retroactivity of the remedy to the litigants. Courts ought also consider whether a retroactive remedy would unduly interfere with the constitutional role of legislatures and democratic governments in the allocation of public resources (Benner, at para. 103; Schachter, at p. 710).

[101] A careful consideration of reliance interests is critical to this analytical process. Although legal mechanisms, such as the de facto doctrine, res judicata or the law of limitations, may mitigate the consequences of declaratory rulings in certain circumstances, they do not address every situation. Fully retroactive remedies might prove highly disruptive in respect of government action, which, on the basis of settled or broadly held views of the law as it stood, framed budgets or attempted to design social programs. Persons and public authorities could then become liable under a new legal norm. Neither governments nor citizens could be reasonably assured of the legal consequences of their actions at the time they are taken.

[102] The strict declaratory approach also hardly appears reconcilable with the well-established doctrine of qualified immunity in respect of the adoption of unconstitutional statutes which our Court applied, for example, in cases such as Mackin and Guimond v. Quebec (Attorney General), [1996] 3 S.C.R. 347. Where legislation is found to be invalid as a result of a judicial shift in the law, it will not generally be appropriate to impose liability on the government. As Gonthier J. wrote in Mackin, it is a general rule of public law that “absent conduct that is clearly wrong, in bad faith or an abuse of power, the courts will not award damages for the harm suffered as a result of the mere enactment or application of a law that is subsequently declared to be unconstitutional” (para. 78). The rationale for this qualified immunity, which applies equally to actions for damages based on the general law of civil liability and to claims for damages under s. 24(1) of the Charter, was aptly expressed by Gonthier J.:

Thus, the government and its representatives are required to exercise their powers in good faith and to respect the “established and indisputable” laws that define the constitutional rights of individuals. However, if they act in good faith and without abusing their power under prevailing law and only subsequently are their acts found to be unconstitutional, they will not be liable. Otherwise, the effectiveness and efficiency of government action would be excessively constrained. Laws must be given their full force and effect as long as they are not declared invalid. Thus it is only in the event of conduct that is clearly wrong, in bad faith or an abuse of power that damages may be awarded. [para. 79]

The same principles will apply in respect of claims for retroactive benefits under s. 15 of the Charter. Whether framed as a remedy under s. 52 or s. 24(1), it may be tantamount to a claim for compensatory damages flowing from the underinclusiveness of the legislation.

[103] People generally conduct their affairs based on their understanding of what the law requires. Governments in this country are no different. Every law they pass or administrative action they take must be performed with an eye to what the Constitution requires. Just as ignorance of the law is no excuse for an individual who breaks the law, ignorance of the Constitution is no excuse for governments. But where a judicial ruling changes the existing law or creates new law, it may, under certain conditions, be inappropriate to hold the government retroactively liable. An approach to constitutional interpretation that makes it possible to identify, in appropriate cases, a point in time when the law changed, makes it easier to ensure that persons and legislatures who relied on the former legal rule while it prevailed will be protected. In this way, a balance is struck between the legitimate reliance interests of actors who make decisions based on a reasonable assessment of the state of the law at the relevant time on one hand and the need to allow constitutional jurisprudence to evolve over time on the other.

[104] Having regard to the above-mentioned criteria, it is possible to distinguish this case from some cases where fully retroactive remedies were granted. Miron provides an example of when it would not be appropriate for courts to limit the retroactive effect of a s. 52(1) remedy. In Miron, the appellant was injured while a passenger in a vehicle driven by an uninsured driver. He made a claim for accident benefits against the insurance policy of his unmarried partner but his claim was denied on the basis that the policy covered only legally married spouses. Writing for the majority, McLachlin J. (as she then was) held that the distinction based on marital status was discriminatory under s. 15(1) of the Charter. She concluded that retroactive reading-up of the legislation was an appropriate remedy, which entitled the appellant to the retroactive benefit of his partner’s insurance policy.

[105] In Miron, it would not have been appropriate for the Court to limit the retroactive effect of the remedy and grant a purely prospective remedy. First, the government did not meet the threshold factor of showing a substantial change from the existing law. As early as 1980, the Ontario Legislature was able to agree on a formula to extend death benefits to certain unmarried persons. And in 1981, in the context of the Ontario Human Rights Code, the Legislature agreed on a definition of “spouse” as the person to whom a person of the opposite sex is married or with whom the person is living in a conjugal relationship outside marriage. In other words, Ontario’s vehicle insurance legislation was out of step with the evolving understanding of “spouse” as it existed in other Ontario statutes. Therefore, the Court’s holding in Miron — that the vehicle insurance legislation’s definition of “spouse” violated s. 15 — was not a substantial change from the existing law. To the contrary, it reflected an understanding of s. 15 of the Charter that was already understood by the Ontario Legislature in the context of the Ontario Human Rights Code and other provincial legislation. Because the finding of a s. 15 infringement in Miron did not represent a substantial change in the law, it would have been inappropriate to limit the retroactive effect of its decision.

[106] However, even if the government had succeeded in meeting the substantial change requirement, other factors militated against limiting the retroactive effect of the remedy. In reaching her conclusion, McLachlin J. drew support from three observations. First, she observed that the legislature had, since the accident occurred, amended the applicable legislation to include unmarried partners, thus allaying any concerns about interfering unduly with legislative objectives. The amended legislation provided “the best possible evidence of what the Legislature would have done had it been forced to face the problem the appellants raise[d]” (para. 180). Second, considerations of fairness to the successful litigant also weighed in favour of retroactivity, as providing a retroactive remedy in this case was the only means of “cur[ing] an injustice which might otherwise go unremedied”: ibid. Third, McLachlin J. noted that the distinction based on marital status was unreasonable, even at the time the impugned legislation was enacted (para. 173). Because the Legislature ought to have known that the vehicle insurance legislation was out of step with a modern understanding of “spouse”, it could not reasonably exclude common law spouses from insurance coverage.

[107] It should be noted that, in Miron, all of the factors discussed above — good faith reliance by governments, fairness to the litigants and the need to respect the constitutional role of legislatures — favoured a retroactive remedy. In a number of cases, however, these factors may pull in different directions, with some factors favouring a retroactive remedy and others favouring a purely prospective remedy. In such cases, once the “substantial change” threshold criterion is met, it may be appropriate to limit the retroactive effect of the remedy based on a balancing of these other factors. This balance must be struck on a case-by-case basis.

[108] A second situation that must be distinguished was considered by this Court in its recent judgment in Kingstreet Investments Ltd. v. New Brunswick (Finance), [2007] 1 S.C.R. 3, 2007 SCC 1, wherein it held that taxes collected pursuant to an ultra vires regulation are recoverable by the taxpayer. The difference between the result in Kingstreet and the type of situation in the present case may be understood in terms of a basic distinction between cases involving moneys collected by the government and benefits cases. Where the government has collected taxes in violation of the Constitution, there can be only one possible remedy: restitution to the taxpayer. In contrast, where a scheme for benefits falls foul of the s. 15 guarantee of equal benefit under the law, we normally do not know what the legislature would have done had it known that its benefits scheme failed to comply with the Charter. In benefits cases, a range of options is open to government. The excluded group could simply be included in the existing benefit scheme as was the result in Tétreault-Gadoury v. Canada (Employment and Immigration Commission), [1991] 2 S.C.R. 22. It could also be included in a modified benefit scheme, adopted by legislative amendments, as occurred in Schachter. Also, in Schachter, the Court alluded to the possibility of an elimination of the benefit (p. 702). In our political system, choosing between those options remains the domain of governments. This principle points towards limiting the retroactive effect of remedies in s. 15 benefits cases in which the other above-mentioned criteria are met.

(2) The Appropriate Remedy in This Case

(a) Limits on the Retroactive Effect of the Remedy in the Context of This Case

[109] Same-sex equality jurisprudence since 1985 is illustrative of the sort of legal shift that gives rise to new law and justifies consideration of prospective remedies. The factors mentioned above also weigh in favour of limiting the retroactive effect of the remedy in the context of this case.

(i) The Substantial Change in the Law

[110] This Court’s decision in M. v. H. marked a departure from pre-existing jurisprudence on same-sex equality rights. In 1995, a majority of this Court upheld the exclusion of same-sex partners from old age security legislation in Egan, with four judges finding no s. 15(1) violation, and one judge concluding that the scheme was contrary to s. 15(1) but that it could be justified under s. 1. Four years later in M. v. H., eight members of this Court held that the exclusion of same-sex partners from the spousal support provisions under the Family Law Act was contrary to s. 15(1) and could not be saved under s. 1. M. v. H. thus marks a clear shift in the jurisprudence of the Court, where it moved away from the plurality’s holding in Egan and came to a new understanding of the scope of equality rights.

[111] Bastarache J. disagrees with our conclusion on the nature of the change brought about by M. v. H. He cites lower court decisions rendered prior to Egan and before M. v. H. to show that the law on same-sex equality rights remained unsettled until M. v. H. However, in our system, the Supreme Court has the final word on the interpretation of the Constitution: Manitoba Language Rights Reference, at p. 745. A majority of the Court in Egan rejected the appellants’ claim for equal benefits under the law. It was a fact that this Court held in Egan that the Constitution did not require equal benefits for same-sex couples. This fact changed only after M. v. H. when this Court held that it was unconstitutional to exclude same-sex couples from the definition of spouse in the Family Law Act. The threshold requirement for limiting the retroactive effect of the remedy has been satisfied. The Court must now consider other relevant factors. In this case, reliance interests, fairness concerns, the government’s good faith, and the need to respect Parliament’s legislative role all weigh in favour of limiting retroactive relief.

(ii) Reasonable Reliance

[112] Given the state of the jurisprudence prior to M. v. H., the exclusion of same-sex partners from the former CPP was based on a reasonable understanding of the state of s. 15(1) jurisprudence as it existed after Egan and before M. v. H. Admittedly, the Court in Egan was divided over whether to extend old age security benefits to same-sex couples, with four judges dissenting. After M. v. H., it became apparent that the Egan dissent had prevailed. However, the benefit of hindsight does not undermine the government’s reasonableness in relying on Egan.

[113] In holding that the government reasonably relied on the pre-M. v. H. jurisprudence, we do not seek to justify the slowness of legislatures and courts alike in recognizing Charter rights. Rather, we acknowledge the fact that although the Constitution embodies the supreme law and the enduring values of this country, it is up to the courts to interpret and apply those provisions. In Manitoba Language Rights Reference, this Court held:

The Constitution of a country is a statement of the will of the people to be governed in accordance with certain principles held as fundamental and certain prescriptions restrictive of the powers of the legislature and government. It is, as s. 52 of the Constitution Act, 1982 declares, the “supreme law” of the nation, unalterable by the normal legislative process, and unsuffering of laws inconsistent with it. The duty of the judiciary is to interpret and apply the laws of Canada and each of the provinces, and it is thus our duty to ensure that the constitutional law prevails. [p. 745]

[114] The text of the Constitution establishes the broad confines of the supreme law, but it is up to the courts to interpret and apply the Constitution in any given context. The inviolability of the Constitution ensures that our nation’s most cherished values are preserved, while the role of the courts in applying the Constitution ensures that the law is sufficiently flexible to change over time to reflect advances in human understanding. But it also means that the Constitution, at any snapshot in time, is only as robust as the court interpreting it. If the judiciary errs or is slow to recognize that previous interpretations of the Constitution no longer correspond to social realities, it must change the law. However, in breaking with the past, the Court does not create an automatic right to redress for the Court’s prior ruling. Where the government’s reliance on the unconstitutional law was reasonable because it was relying on this Court’s jurisprudence, it will be less likely that a right to retroactive relief will flow from a subsequent declaration of invalidity of the unconstitutional law.

(iii) Good Faith

[115] Our comments above indicate that the government did not act in bad faith in failing to extend survivors’ benefits to same-sex couples prior to M. v. H. It is significant that the survivors’ benefit scheme under the former CPP was never struck down by a court of competent jurisdiction. Rather, recognizing the likely implications of this Court’s ruling in M. v. H. for that scheme, Parliament endeavoured to pre-emptively correct the constitutional deficiencies therein by enacting remedial legislation. Because the government acted in good faith by attempting pre-emptively to correct a constitutional infirmity soon after it was discovered, it would be inappropriate to reach back further in time and impose a retroactive remedy.

(iv) Fairness to Litigants

[116] In seeking payment of arrears back as far as 1985, the Hislop class effectively asks this Court to overlook the evolution in the jurisprudence of same-sex equality rights that has taken place and to declare that the understanding to which we have come over that period of time was in fact the law in 1985. This position cannot be sustained. Although M. v. H. declares what the Constitution requires, it does not give rise to an automatic right to every government benefit that might have been paid out had the Court always interpreted the Constitution in accordance with its present-day understanding of it. M. v. H. was not a case like Miron where limiting the retroactive effect of the s. 52(1) remedy would have granted the “successful” claimant a hollow victory. In contrast, a purely prospective remedy in M. v. H. was not meaningless. M. v. H. resulted in wide-scale amendments to federal and provincial legislation across the country to extend government benefits to same-sex couples. Equally important, M. v. H. helped usher in a new era of understanding of the equal human dignity of same-sex couples. One could not say that M. v. H. granted those litigants only a Pyrrhic victory.

(v) Respecting Parliament’s Role

[117] Achieving an appropriate balance between fairness to individual litigants and respecting the legislative role of Parliament may mean that Charter remedies will be directed more toward government action in the future and less toward the correction of past wrongs. In the present case, the Hislop class’s claim for a retroactive remedy is tantamount to a claim for compensatory damages flowing from the underinclusiveness of the former CPP. Imposing that sort of liability on the government, absent bad faith, unreasonable reliance or conduct that is clearly wrong, would undermine the important balance between the protection of constitutional rights and the need for effective government that is struck by the general rule of qualified immunity. A retroactive remedy in the instant case would encroach unduly on the inherently legislative domain of the distribution of government resources and of policy making in respect of this process.

(vi) Conclusion

[118] For the foregoing reasons, the retroactive relief sought by the Hislop class is unavailable under the law applicable to constitutional remedies. It is not therefore necessary to carry out a s. 15(1) analysis in respect of s. 72(1).

(b) Remedies Arising From the Specific Provisions: Sections 44(1.1) and 72(2)

[119] We turn now to a consideration of the appropriate remedy for the specific constitutional violations that have been identified in this case.

[120] Writing for a majority of the Court in Schachter, Lamer C.J. explained that three questions must be answered when s. 52 of the Constitution Act, 1982 is engaged: (1) what is the extent of the inconsistency between the impugned provision and the Charter; (2) can that inconsistency be dealt with alone, by way of severance or reading in, or is it too inextricably linked to other parts of the legislation; and (3) should a declaration of invalidity be temporarily suspended?

[121] In the present case, ss. 44(1.1) and 72(2), although found within remedial legislation, restrict the availability of that legislation to a marginalized group. The extent of the inconsistency with the equality guarantee is co-extensive with the scope of these two provisions. As such, the inconsistency can be cured without distorting or interfering with the rest of the legislative scheme. A declaration that ss. 44(1.1) and 72(2) are of no force and effect is in keeping with the scheme and obvious purpose of the MBOA to extend the survivors’ benefit to same-sex survivors. Finally, a temporary suspension of the declaration of invalidity is not appropriate in the present case. As Lamer C.J. noted in Schachter, at p. 716, such suspensions are “serious matter[s] from the point of view of the enforcement of the Charter” because they allow an unconstitutional state of affairs to persist. Suspensions should only be used where striking down the legislation without enacting something in its place would pose a danger to the public, threaten the rule of law or where it would result in the deprivation of benefits from deserving persons without benefiting the rights claimant (p. 719). None of these factors are present in the case at bar.

[122] Throughout these proceedings, the Crown has taken the position that the specific provisions are constitutionally unassailable because they merely ensure that the remedial provisions in the MBOA apply prospectively (i.e., not retroactively) from the time of the enactment. Further, the Crown has relied on the argument that same-sex equality rights, as they are understood today, are not what they were in 1985. In effect, it is the Crown’s position that striking down the specific provisions would be tantamount to applying today’s legal understanding of equality rights to past situations.

[123] As we have explained above, the specific provisions violate s. 15(1) and cannot be justified under s. 1. The constitutional analysis related to these provisions does not depend on any particular conception of the equality rights of same-sex survivors prior to the enactment of the MBOA. The specific provisions, as applied to same-sex survivors today, are discriminatory.

[124] Any concerns about retroactivity in respect of s. 44(1.1) are misplaced because they mistakenly construe the survivor’s benefit as a payment in respect of a past event, namely the death of the survivor’s spouse or common law partner, rather than the ongoing status of being a survivor. In our view, the principles articulated by this Court in Benner are a complete answer to the Crown’s s. 44(1.1) argument.

[125] Benner involved a challenge to provisions of the Canadian Citizenship Act that afforded different treatment to applicants for Canadian citizenship, where those applicants were born before February 14, 1977, based on whether they were born of Canadian mothers or Canadian fathers. Children of Canadian fathers were automatically entitled to Canadian citizenship upon registration of their birth, while children of Canadian mothers were required to apply for citizenship, swear an oath and pass a security check. Benner challenged the legislative distinction on s. 15(1) grounds, while the Crown resisted the claim in part on the basis that Benner’s claim involved an impermissible retroactive application of the Charter to events that had taken place before the Charter came into force. Writing for a unanimous court, Iacobucci J. rejected the Crown’s argument and, in so doing, provided a useful clarification of the concepts of retroactivity, retrospectivity and prospectivity.

[126] It should be noted that Benner engaged questions related to the retroactive application of the Charter, while the instant case raises questions related to the retroactive application of ordinary legislation and to the availability of retroactive remedies for unconstitutional legislation. Although retroactivity raises different concerns in the two contexts, the basic conceptual distinctions between retroactivity, retrospectivity and prospectivity remain. As such, Iacobucci J.’s elucidation of these concepts is directly relevant to the present case.

[127] First, Iacobucci J. addressed the difference between retroactivity and retrospectivity, citing, at para. 39, E. A. Driedger’s explanation from “Statutes: Retroactive Retrospective Reflections” (1978), 56 Can. Bar Rev. 264, at pp. 268‑69:

A retroactive statute is one that operates as of a time prior to its enactment. A retrospective statute is one that operates for the future only. It is prospective, but it imposes new results in respect of a past event. A retroactive statute operates backwards. A retrospective statute operates forwards, but it looks backwards in that it attaches new consequences for the future to an event that took place before the statute was enacted. A retroactive statute changes the law from what it was; a retrospective statute changes the law from what it otherwise would be with respect to a prior event. [Emphasis in original.]

Next, Iacobucci J. turned to the difference between retrospectivity and prospectivity, again citing Professor Driedger. He reproduced the following passage from Driedger’s Construction of Statutes (2nd ed. 1983), at p. 192, at para. 42 of his reasons:

These past facts may describe a status or characteristic, or they may describe an event. It is submitted that where the fact‑situation is a status or characteristic (the being something), the enactment is not given retrospective effect when it is applied to persons or things that acquired that status or characteristic before the enactment, if they have it when the enactment comes into force; but where the fact‑situation is an event (the happening of or the becoming something), then the enactment would be given retrospective effect if it is applied so as to attach a new duty, penalty or disability to an event that took place before the enactment. [Emphasis added.]

It is true, as Iacobucci J. observed, that the distinction between situations involving discrete events and those involving ongoing conditions is not always clear. For example, while the death of a spouse or common law partner is an event, the fact of “being a survivor” is an ongoing condition. The challenge lies in determining whether, in all the circumstances, a particular legislative scheme relates primarily to the past event or the current condition resulting from it (para. 46).

[128] On the facts before the Court in Benner, Iacobucci J. concluded that the appellant’s claim did not involve a retrospective application of the Charter because it was a claim flowing from Benner’s ongoing status as the child born outside Canada of a Canadian mother before the specified date. The date on which Benner first acquired that status was, in Iacobucci J.’s view, immaterial.

[129] The same reasoning is controlling in the present case. As the Court of Appeal found, the purpose of the CPP is “to provide a secure government-controlled pension for retired persons in their senior years, and thereafter to provide for their surviving spouses and partners” (para. 55). The basis for the survivor’s benefit is the continuing status of being a survivor. As Iacobucci J. observed in Benner, at para. 56, “the important point is not the moment at which the individual acquires the status in question, it is the moment at which that status is held against him or disentitles him to a benefit”. There is, as such, no merit to the government’s argument that striking down s. 44(1.1) would be tantamount to requiring retroactive remedies for Charter violations. The remedy for the class members in relation to s. 44(1.1) is prospective in that it entails granting equal benefit of the law, prospectively, to people who are survivors of same‑sex partners.

[130] The government raises the additional argument that the Court should not strike down s. 44(1.1) because doing so would have the unintended effect of excluding same‑sex survivors from eligibility for the survivor’s benefit. The Crown’s argument is predicated on the view that the survivor’s benefit is paid in respect of a past event such that any entitlement to it crystallizes at the time that the survivor’s partner dies and that, but for s. 44(1.1), no benefit would be payable to any survivor whose same‑sex common‑law partner died before July 2000. As we have stated, this is an untenable view of what the survivor’s benefit is and how entitlement to it works. A close look at the July 2000 amendments to the CPP reveals that s. 44(1.1) is in fact a limiting provision, such that simply striking it down leaves a coherent provision that provides for equal treatment, in terms of eligibility, of opposite‑sex survivors and same‑sex survivors.

[131] As part of the 2000 amendments, the opposite-sex definition of “spouse” in s. 2(1) of the CPP was repealed. In its place, a definition for “survivor” that includes both opposite-sex and same-sex spouses was added to Part II, “Pensions and Supplementary Benefits”, in s. 42(1). Finally, s. 44(1)(d), which sets out the basic entitlement to the survivor’s benefit, was amended by replacing reference to the statutory definition of “spouse” with reference to the new statutory definition of “survivor”. The amended provision provides that “subject to subsection (1.1), a survivor’s pension shall be paid to the survivor of a deceased contributor . . .”.

[132] Once eligibility is understood in terms of the current status of being a survivor, it becomes clear that the effect of the amendments listed above is to extend entitlement to the survivor’s benefit under s. 44(1)(d) to same‑sex survivors. The function of s. 44(1.1) is purely restrictive: it limits that entitlement based on the date on which the applicant became a survivor. As such, striking down s. 44(1.1) has no unintended or undesirable effect.

[133] The analysis in relation to s. 72(2) is somewhat different. Prior to the enactment of the MBOA, it is true that same-sex survivors had no right to the survivor’s benefit. Striking down s. 72(2) admittedly alters the legal consequences of having been a survivor (a past situation) in the 12 months preceding July 2000, when the MBOA came into force. To the extent that this involves a retroactive change in the law, it flows necessarily from the fact that Parliament has included in the CPP a right to up to 12 months of arrears. It is clearly open to Parliament to legislate retroactively, which it has done in s. 72(1). The Charter simply requires that same-sex survivors receive equal treatment as their opposite-sex counterparts. To the extent that s. 72(2) withholds from same-sex survivors a right to arrears to which similarly situated opposite-sex survivors are entitled, it is of no force and effect.

[134] In conclusion, class members who were precluded by s. 44(1.1) or s. 72(2) from receiving the survivor’s benefit, and who otherwise meet the eligibility requirements, will be entitled to payment of that benefit. In the circumstances, the relevant date for the purpose of that payment is the date on which application was received or where no application was made because of the unconstitutional provisions, the date on which the statement of claim was filed. In no event are benefits payable in respect of a month prior to August 1999, which is the earliest month in respect of which a class member who applied for the survivor’s benefit on the day the MBOA came into force could have been eligible.

VII. Interest

[135] For the reasons given by the Court of Appeal, we reject the government’s contention that pre-judgment interest is not available in the instant case. In short, s. 31 of the Crown Liability and Proceedings Act provides that, subject to an express provision in another enactment, successful litigants are entitled to pre-judgment interest from the Crown in the same way that they would be so entitled as against any other litigant. The CPP is silent on the issue of pre-judgment interest and cannot, as such, reasonably be interpreted as creating an exception to the entitlement created by s. 31 of the Crown Liability and Proceedings Act.

VIII. Disposition

[136] The appeal and cross-appeal are dismissed without costs. The constitutional questions are answered as follows:

1. Does s. 44(1.1) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

Yes.

2. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

No.

3. Does s. 72(2) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

Yes.

4. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

No.

5. Does s. 60(2) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

No.

6. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

It is not necessary to answer this question.

7. Does s. 72(1) of the Canada Pension Plan, R.S.C. 1985, c. C‑8, infringe s. 15(1) of the Canadian Charter of Rights and Freedoms?

It is not necessary to answer this question.

8. If so, is the infringement a reasonable limit prescribed by law as can be demonstrably justified in a free and democratic society under s. 1 of the Canadian Charter of Rights and Freedoms?

It is not necessary to answer this question.

The following are the reasons delivered by

[137] Bastarache J. — I have read the joint reasons of my colleagues Justices LeBel and Rothstein, and, while I agree with their disposition of this appeal, I cannot fully agree with their approach to the question of retroactive constitutional remedies. In my view, their reliance on the existence of a substantial change of law is an inappropriate consideration in the context of rights guaranteed by the Canadian Charter of Rights and Freedoms and is, in any event, inapplicable to this appeal. The decision to deny retroactive relief, and the appellants’ challenge to s. 72(1) of the Canadian Pension Plan, R.S.C. 1985, c. C-8 (“CPP”), specifically, should be based purely on the balancing of interests that must take place in any claim for relief from an unconstitutional law.

The Retroactivity of Constitutional Remedies

[138] It is important to be clear about the nature and reasons for retroactive constitutional remedies. The general and well-established rule is that remedies for constitutional violations apply retroactively (S. Choudhry and K. Roach, “Putting the Past Behind Us? Prospective Judicial and Legislative Constitutional Remedies” (2003), 21 S.C.L.R. (2d) 205, at p. 211). The basis for general retroactivity is not Blackstone’s declaratory theory, but the Constitution itself. Section 52(1) of the Constitution Act, 1982 establishes that any law which is inconsistent with the Constitution of Canada “is, to the extent of the inconsistency, of no force or effect”. The Constitution exists independently of judicial decisions and, as such, any law which is inconsistent with it is invalidated from the moment the law came into effect. Gonthier J. explained this principle clearly in Nova Scotia (Workers’ Compensation Board) v. Martin, [2003] 2 S.C.R. 504, 2003 SCC 54, at para. 28:

The invalidity of a legislative provision inconsistent with the Charter does not arise from the fact of its being declared unconstitutional by a court, but from the operation of s. 52(1). Thus, in principle, such a provision is invalid from the moment it is enacted, and a judicial declaration to this effect is but one remedy amongst others to protect those whom it adversely affects. In that sense, by virtue of s. 52(1), the question of constitutional validity inheres in every legislative enactment. [Emphasis added.]

(See also P. W. Hogg, Constitutional Law of Canada (loose-leaf ed.), vol. 2, at p. 55-2; K. Roach, Constitutional Remedies in Canada (loose-leaf), at para. 14.920.)

[139] This rule applies equally in the context of the Charter. Laws which violate Charter guarantees are invalid from the date that the provisions of the Charter came into force: Jim Pattison Industries Ltd. v. The Queen, [1984] 2 F.C. 954 (T.D.), at p. 957; Davidson v. Davidson (1986), 33 D.L.R. (4th) 161 (B.C.C.A.), at p. 170, leave to appeal refused, [1987] 1 S.C.R. vii; Rath v. Kemp (1996), 1996 ABCA 378, 46 Alta. L.R. (3d) 1 (C.A.), at para. 25. In the case of a violation of s. 15, this would mean that, in theory, a law which violates the Charter would be invalid as of April 17, 1985 or the date of its adoption, if later.

[140] The general norm of retroactivity has been reaffirmed many times by this Court. In Reference re Manitoba Language Rights, [1985] 1 S.C.R. 721, the finding that the Province of Manitoba has been constitutionally required since 1890 to enact its laws both in English and in French applied not only for the future, but also retroactively to all of its laws that had been enacted in English only since 1890. The declaration of invalidity was suspended in order to preserve the rule of law, but there was no doubt that a finding of unconstitutionality applied retroactively. In Miron v. Trudel, [1995] 2 S.C.R. 418, a violation of s. 15(1) of the Charter was cured by reading the excluded group into the legislation, a remedy that applied retroactively. In R. v. Hess, [1990] 2 S.C.R. 906, the Court severed an offending portion of the Criminal Code, R.S.C. 1985, c. C-46, and ordered a new trial on the basis of the provision as amended even though, strictly speaking, it was not the law of the land when the alleged crime had been committed. These cases are only examples, but they confirm that retroactivity of a constitutional remedy granted under s. 52(1) is the norm in our constitutional jurisprudence, not the exception.

[141] With respect, my colleagues ignore the fact that Blackstone’s comments on the nature of judicial law making were made in the context of the common law and find only limited application in the constitutional context. As it pertains to the common law, Blackstone’s declaratory theory is necessarily a fiction because there is no independent source for common law rules. The common law is by definition judge-made. It does not exist in some jurisprudential ether for judges to discover. Therefore when judges overturn precedents or establish a new common law rule, they are undoubtedly making new law.

[142] The same cannot be said for judicial decisions which interpret and apply the Constitution. The Constitution exists independently of judicial decisions. Judges do not “make” the Constitution every time they interpret its provisions. Interpretations of what the Constitution requires may change, but the underlying rights and freedoms endure. Charter rights are not created every time a court expressly overrules or implicitly repudiates a prior decision or gives “content to broad, but previously undefined, rights, principles or norms” (LeBel and Rothstein JJ., at para. 99). The rights and freedoms in the Charter were guaranteed to all Canadians from the moment the Charter came into force.

[143] By justifying the denial of retroactive relief in part on the existence of a “substantial change in the law”, my colleagues give Charter rights an uncertain status. I cannot accept an approach that, for remedial purposes, implies that Charter rights can be here one day and gone the next or, conversely, that they depend on judicial recognition of “a new or newly recognized technological or social environment” (para. 99) for their genesis. Such reasoning represents a watering down of the promise made to all Canadians when the Charter was enacted. By attaching importance to changing social conditions, it makes Charter rights dependent on how the majority of Canadians perceive the claimants’ rights. With respect, I cannot see why society’s views of Charter claimants — especially in the context of vulnerable minorities — should be a factor for determining whether a Charter right was part of the Constitution in 1985, or whether it sprung into existence later and thereby be a basis for denying retroactive relief.

[144] I should note that I am not advocating for a view of the Constitution that says that it is frozen in time. The “living tree” metaphor aptly describes how through time our Constitution may change to correspond to new realities. But that does not mean that every time a new constitutional interpretation is adopted or a previous decision is overturned that the fundamental rights and freedoms guaranteed in our Charter have changed or that new ones have been created. There is a difference between changes in constitutional interpretation and actual constitutional change. Furthermore, the “living tree” doctrine is a doctrine of “progressive interpretation” (Reference re Same‑Sex Marriage, [2004] 3 S.C.R. 698, 2004 SCC 79, at paras. 22-23 (emphasis added)), necessary to ensure that our Constitution does not become rigid and unresponsive to Canadian society. It should not be used as a justification for denying relief to a particular group of Charter claimants.

[145] For these reasons, I cannot accept my colleagues’ critique of the declaratory approach as a basis for denying retroactive relief. As will be explained below, there are important reasons for denying retroactive relief in certain circumstances. I am largely in agreement with my colleagues on what they are. But I cannot agree that they have anything to do with the success or failure of the “Blackstonian paradigm” in the context of constitutional law.

[146] The dangers of my colleagues’ approach are adequately evidenced when applied to the claimants in this appeal. The starting point of their analysis is that there was a substantial change in the law between 1985, when s. 15(1) of the Charter came into force, and this Court’s decision in M. v. H., [1999] 2 S.C.R. 3. The implication is that the right of same-sex spouses not to be excluded from survivor benefits did not form part of the Constitution until 1999. To put it bluntly, s. 15(1) of the Charter did not extend to same-sex couples until this Court said it did. I note that my colleagues are not simply saying that this Court’s interpretation of the Constitution had changed between 1985 and 1999. If that were the case, it would be sufficient to base their denial of retroactive relief solely on the good faith reliance of the government. Instead, by relying on a critique of the declaratory theory and the “living tree” doctrine, my colleagues assert, in essence, that the Constitution actually changed between 1985 and 1999 and that the claimants, unlike other Canadians, were not entitled to its protection in 1985. Such an approach runs counter to the spirit of the Charter and should not be countenanced.

[147] Furthermore, it is not at all clear that a substantial change in the law actually occurred sufficient to trigger a departure from the norm of retroactivity under my colleagues’ threshold approach. A review of the case law before and after Egan suggests that it was not the “final word” on s. 15(1) and discrimination of same-sex couples.

[148] Courts did not begin pronouncing on the extent of protection that s. 15 afforded to same-sex couples until the late 1980s, and then with conflicting results. In Andrews v. Ontario (Minister of Health) (1988), 64 O.R. (2d) 258, the Ontario High Court of Justice held that it was not discriminatory to deny provincial health insurance benefits to same-sex dependants based on the definition of “spouse” in the applicable legislation. In Vogel v. Manitoba (1992), 90 D.L.R. (4th) 84 (Man. Q.B.), exclusion of a same-sex partner from an employer’s spousal benefit plan was held not to be discriminatory.

[149] In other cases, the opposite result was achieved. In Veysey v. Canada (Commissioner of the Correctional Service), [1990] 1 F.C. 321, the Federal Court—Trial Division found that it was an infringement of s. 15(1) to deny visitation rights to the same-sex partner of an inmate and that the infringement could not be saved under s. 1. In Knodel v. British Colombia (Medical Services Commission) (1991), 58 B.C.L.R. (2d) 356 (S.C.), in direct opposition to Andrews, regulations excluding same-sex dependants from claiming provincial health insurance benefits were found to be an unjustified infringement of s. 15(1) and a declaration that same-sex couples be read into the definition of “spouse” was made.

[150] This Court first came close to the issue in Canada (Attorney General) v. Mossop, [1993] 1 S.C.R. 554, which involved a provision of a collective agreement that restricted bereavement leave to members of an employee’s “immediate family”. The complainant had sought to have the day he took off work to attend the funeral of his same-sex partner’s father count as bereavement leave. The majority (in a 4-3 decision) decided the issue on the narrow ground that the Canadian Human Rights Act, R.S.C. 1985, c. H-6, did not include sexual orientation as a ground of discrimination. Lamer C.J. noted, however, that the case could have been argued on the basis that the Charter required sexual orientation to be read into the Canadian Human Rights Act. Because the Charter argument had not been made, he felt the Court had to apply the Act as it stood. He made it clear that he would have perhaps decided the case differently if sexual orientation had been a prohibited ground of discrimination (p. 582).

[151] The next time this Court dealt with the issue was in Egan v. Canada, [1995] 2 S.C.R. 513. In a narrow 5-4 decision, this Court held that the definition of “spouse” under the Old Age Security Act, R.S.C. 1985, c. O-9, for the purposes of spousal allowances under the Act, was an infringement of s. 15(1), but was saved under s. 1. Sopinka J. cast the deciding vote. Central to his reasoning was that governments should be accorded some leeway in extending social benefits. In this regard, he agreed with La Forest J.’s suggestion in McKinney v. University of Guelph, [1990] 3 S.C.R. 229, that governments should be allowed time to take incremental measures to deal with discrimination (para. 105).

[152] It was soon clear that Egan was not the final word on the matter. The Manitoba Court of Appeal rendered its judgment in Vogel v. Manitoba (1995), 126 D.L.R. (4th) 72, a few months after Egan. It reversed the trial judge’s decision and held that it was discriminatory under provincial human rights legislation to exclude same-sex partners from employee benefit plans on the basis that same-sex partners were not included in the term “spouse”. Helper J.A. and Scott C.J.M. relied on Egan to find that the eligibility criteria of the scheme were discriminatory. A little more than a year later, the Ontario Court of Appeal released its decision in M. v. H. (1996), 96 O.A.C. 173. Charron J.A. (as she then was) found that the definition of “spouse” contained in the Family Law Act, R.S.O. 1990, c. F.3, and which excluded same-sex couples, was a violation of s. 15(1) and could not be saved under s. 1. Although she took notice of this Court’s decision in Egan, she was of the view that it was not determinative since the objectives of the legislation in each case were different (para. 84). The Court of Appeal’s decision in M. v. H. was relied upon in Kane v. Ontario (Attorney General) (1997), 152 D.L.R. (4th) 738 (Ont. Ct. (Gen. Div.)), to hold that the exclusion of same-sex partners from the definition of “spouse” in Ontario’s Insurance Act, R.S.O. 1990, c. I.8, was also unconstitutional.

[153] In Vriend v. Alberta, [1998] 1 S.C.R. 493, this Court rejected the incrementalist approach that was relied upon in part by Sopinka J. in Egan to find that the constitutional violation in that case was saved under s. 1.

[154] Relying on its own jurisprudence and this Court’s decision in Vriend, the Ontario Court of Appeal held in Rosenberg v. Canada (Attorney General) (1998), 38 O.R. (3d) 577, that a provision of the Income Tax Act, R.S.C. 1985, c. 1 (5th Supp.), that limited the definition of “spouse” to opposite sex couples was unconstitutional.

[155] Finally, in M. v. H., this Court upheld the Ontario Court of Appeal’s decision that the exclusion of same-sex couples from the definition of spouse in the Family Law Act was unconstitutional. Iacobucci J. distinguished this Court’s holding in Egan on the same grounds as the Court of Appeal had, namely that the legislative objectives at issue were entirely different (para. 75). It was in response to this decision that the federal government decided to enact the amendments to federal legislation, including the CPP, contained in the Modernization of Benefits and Obligations Act, S.C. 2000, c. 12.

[156] As can be seen, the majority of cases both before and after Egan indicated that differential treatment of same-sex couples was discriminatory and could not be justified under s. 1. It is true that this Court has the final word on the interpretation of the Constitution. However, I find it difficult to believe that Egan was the final word — even for a time — on the application of s. 15(1) to same-sex couples. Egan was decided on the basis of s. 1 arguments that were tailored to the specific legislative objectives at issue in that case. After Egan, it was far from clear whether its reasoning would apply outside the realm of social assistance schemes such as Old Age Security. Indeed, a number of decisions, including this Court’s decision in M. v. H., distinguished Egan on this basis. It did not stand for the application of s. 15(1) to all instances of legislative exclusions of same-sex couples. Given the CPP’s status as an insurance-based scheme, where contributions are paid by employers and employees, it is difficult to see how Egan definitely settled the issue with regard to the exclusion of same-sex couples from survivor’s benefits under the CPP.

[157] Furthermore, given the contradictory decisions both before and after Egan, the closeness of the decision in that case, and the difficult nature of the issues at stake, it is difficult to see Egan as definitively establishing what the Constitution required. The reality is that it was for a time unclear exactly how s. 15(1) would apply to same-sex couples. The judicial process can be slow. It took time for this Court and others to articulate the correct constitutional principles to be applied to legislative exclusions of same-sex couples. That does not mean that this Court was upsetting established law when it handed down its decision in M. v. H. In short, even if the existence of a substantial change in the law was an appropriate threshold criterion, no such change actually occurred in this case. We must therefore look to other reasons for denying retroactive relief.

The Correct Approach to Retroactive Remedies in This Case

[158] It is well established in Canadian constitutional law that courts will seek to balance competing interests when devising remedies for constitutional violations (Roach, at paras. 3.680 to 3.780). The starting point is that past constitutional violations should be corrected. This recognizes the grave nature of constitutional infringements:

Constitutional remedies are matters of considerable importance. Constitutional law expresses the most fundamental restraints and obligations of governments. Violations of constitutional rights are a serious matter both for those who suffer from the violation and the public in general.

(Roach, at para. 1.10)

Retroactive constitutional remedies ensure that the claimants who have brought the successful action, as well as similarly situated claimants, can benefit from the judgment (Choudhry and Roach, at p. 210).

[159] However, the normal retroactive effect of judgments may need to be tempered in certain circumstances in order to protect other legitimate interests (see Choudhry and Roach, at pp. 209-11). The use of transition periods and suspended declarations of invalidity are accepted ways of temporarily limiting the retroactive effect of constitutional remedies in order to prevent legal vacuums and introduce new procedural requirements. They are not evidence of the Court operating outside of the Blackstonian paradigm, but rather a recognition that in certain circumstances other legitimate interests may require retroactivity to be limited.

[160] The question is what legitimate interests should be taken into account in deciding to deny a retroactive constitutional remedy. For the most part, I agree that reasonable reliance, good faith, fairness to litigants and Parliament’s role are important considerations. I would simply like to add what I see as a few points of clarification.

[161] First, the question of whether to deny a retroactive remedy is different from deciding whether to grant a suspended declaration of invalidity; the same considerations do not apply to both. The latter will be a valid measure when a declaration of invalidity would pose a danger to the public, threaten the rule of law or deprive deserving persons of benefits without thereby benefiting the individual whose rights had been violated (Schachter v. Canada, [1992] 2 S.C.R. 679, at p. 719). A suspended declaration of invalidity is ultimately only a temporary limit on retroactivity; it does not determine whether governments are entitled to deny retroactive relief to the claimants in acting to cure the constitutional defect. Reasonable reliance, good faith, fairness to litigants and Parliament’s role will be appropriate to consider when courts are confronted with the question of what remedy the claimant is entitled to. This will occur, for example, when deciding whether a remedy of “reading in” should apply retroactively or whether a legislative provision enacted in response to a declaration of invalidity which limits retroactive relief should be read down or severed.

[162] Second, to my mind, establishing reasonable reliance will not always be necessary in order to deny retroactive relief. In this case, even on my colleagues’ analysis, the government’s reliance on Egan could not be a justification for denying relief before Egan was decided. Indeed, although it was not entirely clear, most of the case law before Egan suggested that governments could not justifiably exclude same-sex couples. Given the state of the law pre-M. v. H., it would be more accurate to emphasize the fact that the government was acting in good faith in the face of jurisprudential uncertainty. Although not determinative, this should be taken into account in deciding whether to depart from the norm of retroactive relief.

[163] Finally, my colleagues seem to suggest in their discussion of this Court’s recent decision of Kingstreet Investments Ltd. v. New Brunswick (Finance), [2007] 1 S.C.R. 3, 2007 SCC 1, that the nature of the constitutional violation is relevant to deciding whether to deny a retroactive remedy. I do not agree. A government has no more right to discriminate in the provision of benefits than it does to collect unconstitutionally levied taxes. In Kingstreet, there were no legitimate concerns with applying the general rule of retroactivity. The legislature retained the ability to enact remedial legislation to cure any adverse effects (para. 25), and a purely prospective remedy would have left the claimant empty-handed.

[164] These clarifications made, I am in general agreement with how my colleagues have applied the other factors to this appeal. Particularly relevant, it seems to me, is the fact that the Modernization of Benefits and Obligations Act was enacted in response to this Court’s decision in M. v. H. In that case, a suspension of the declaration of invalidity was ordered so as to allow the Ontario government flexibility to cure the constitutional defect. That flexibility implicitly included the ability to limit the retroactive effect of any remedial legislation. Indeed, this is what the Ontario legislature chose to do. The remedial legislation was made prospective from November 20, 1999 (Amendments Because of the Supreme Court of Canada Decision in M. v. H. Act, 1999, S.O. 1999, c. 6, s. 68(2)). Similar flexibility should be accorded to the Canadian government in this case. The legislative branch is better able to deal with distributional concerns than are courts, and its choices should be respected so long as they fall within the limits of the Constitution.

Conclusion

[165] For these reasons, I agree that the appeal and cross-appeal should be dismissed without costs.

Appeal and cross‑appeal dismissed.

Solicitor for the appellant/respondent on cross‑appeal: Attorney General of Canada, Ottawa.

Solicitors for the respondents/appellants on cross‑appeal: Roy Elliott Kim O’Connor, Toronto.

Solicitor for the intervener the Attorney General of Ontario: Attorney General of Ontario, Toronto.

Solicitor for the intervener the Attorney General of Quebec: Department of Justice, Sainte‑Foy.

Solicitor for the intervener the Attorney General of Alberta: Alberta Justice, Edmonton.

Solicitors for the intervener Egale Canada Inc.: Sack Goldblatt Mitchell, Toronto.