COURT OF APPEAL FOR BRITISH COLUMBIA
Withler v. Canada (Attorney General),
2008 BCCA 539
Hazel Ruth Withler and Joan Helen Fitzsimonds
Attorney General of Canada
The Honourable Madam Justice Rowles
The Honourable Madam Justice Ryan
The Honourable Madam Justice Newbury
J. Arvay, Q.C.
Counsel for the Appellants
Counsel for the Respondent
Place and Date of Hearing:
Vancouver, British Columbia
6 & 7 September 2007
Place and Date of Judgment:
Vancouver, British Columbia
23 December 2008
Dissenting Reasons by: :
The Honourable Madam Justice Rowles
Written Reasons by:
The Honourable Madam Justice Ryan (P. 59, para. 145)
Concurred in by:
The Honourable Madam Justice Newbury
Reasons for Judgment of the Honourable Madam Justice Rowles:
 In class action proceedings, the appellants challenged the constitutionality of s. 47(1) of the Public Service Superannuation Act, R.S.C. 1985, c. P-36 [PSSA], and s. 66(1) of the Canadian Forces Superannuation Act, R.S.C. 1985, c. C-17, as amended [CFSA] on the ground that those sections and their companion regulations constitute age discrimination under s. 15 of the Canadian Charter of Rights and Freedoms.
 Section 15(1) of the Charter, which came into effect in 1985, provides:
15.(1) Every individual is equal before and under the law and has the right to the equal protection and equal benefit of the law without discrimination and, in particular, without discrimination based on race, national or ethnic origin, colour, religion, sex, age or mental or physical disability.
 Section 47(1) of the PSSA provides a supplementary death benefit (“SDB”) payment of twice the salary of the participant upon his or her death, “subject to a reduction of ten per cent, to be made as of the time that the regulations prescribe, for every year of age in excess of sixty-five attained by the participant”. A similar provision is included in s. 60(1) of the CFSA, except that the reduction commences at age 60. The provisions in the impugned legislation that stipulate a decreasing death benefit dependant upon the age of the participant at his or her death (the “Reduction Provisions”) are the subject of the constitutional challenge.
 The appellants, who are members of the plaintiff class, are the surviving spouses of those entitled to an SDB payment reduced on account of the age of their spouse at the time of death. The Withler action relates to the PSSA; the Fitzsimonds action to the CFSA. The plaintiffs in the class action were all in receipt of the SDB as well as the survivor’s pensions provided under the PSSA and the CFSA.
 The relief sought in the class action included a declaration that the Reduction Provisions are inconsistent with the Charter and are therefore of no force and effect, and judgment to the class of the amount by which the SDB payments to the claimants were reduced by operation of the Reduction Provisions.
 The action was certified as a class proceeding on 2 November 2001. On 16 May 2005, the following questions were certified as common issues:
(1) Do the plaintiff and class members have standing to allege a breach of Charter rights and to seek a remedy for Charter infringement?
(2) Are the plaintiff and class members to whom benefits were paid in respect of elective participants in the Plan precluded from bringing this action by virtue of the fact that the participants voluntarily maintained their coverage under the Plan or did not opt out of or discontinue their coverage under the Plan?
(3) Do the Reduction Provisions discriminate on the basis of age, contrary to s. 15(1) of the Charter, and, if so, are they saved by s. 1 of the Charter?
(4) If the Reduction Provisions discriminate as set out above and are not saved by s. 1, what is the appropriate remedy under the Charter?
(5) If the Reduction Provisions are constitutionally invalid, and if the plaintiff and the class members are not entitled to a remedy under s. 24(1) of the Charter, can they claim redress from the Crown for Charter infringement under common law or equitable principles?
(6) If the answer to the above question is “yes”, has the Crown been unjustly enriched through application of the Reduction Provisions?
(7) If so, are the Withheld Funds [funds to which the plaintiffs and the class members would have been entitled but for the Reduction Provisions] or an amount equivalent to the Withheld Funds, subject to an institutional or remedial constructive trust of an equitable lien in favour of the plaintiff and the class members?
(8) [abandoned at trial]
(9) Can the causes of action advanced and the remedies sought in these proceedings be barred by the Crown Liability and Proceedings Act, R.S.C. 1985, c.C-50, ss. 29 and 32, the provincial limitation periods or the doctrines of laches and acquiescence?
 Before the trial judge, the respondent argued, unsuccessfully, that the appellants did not have standing to seek a s. 15 Charter remedy. The judge’s reasons may be found at Withler v. Canada (Attorney General), 2006 BCSC 101, 51 C.C.P.B. 19. It was the respondent’s position that when the Reduction Provisions are triggered, not by the age of the claimant but by the age of the deceased plan participant, the allegation of age discrimination could only be advanced by participants in the statutory plan. The judge concluded that Benner v. Canada (Secretary of State),  1 S.C.R. 358, 143 D.L.R. (4th) 577, applied by analogy to the appellants’ claims. She held that it is the right of the surviving spouse that is targeted by the legislation and that there is “a connection between the plaintiffs’ rights and the differentiation made by the legislation between spouses whose husbands died before age 65 (or 60) and those whose husbands died after age 65 (or 60)” (at paras. 64-93).
 Having determined the first common issue, the judge then turned to the third common issue because she decided that the question of whether the Reduction Provisions constitute discrimination on the basis of age could be dispositive of the claim.
 The trial judge decided that the Reduction Provisions did not constitute age discrimination under s. 15 of the Charter. As a result, she dismissed the appellants’ action.
 On the appeal, the appellants take issue with the trial judge’s conclusion that the Reduction Provisions do not violate s.15 of the Charter. On its cross-appeal, the respondent argues that the judge erred in law in finding that the appellants had standing to assert a claim under s. 15 of the Charter.
 In my opinion, the appellants had standing to assert a claim under s. 15 of the Charter. In that regard, I am in substantial agreement with the reasons the trial judge gave on the standing issue and nothing further need be said about that question. I would therefore dismiss the cross-appeal.
 On the third common issue, it is my respectful view that in her analysis, the trial judge fell into error by not fully stating and consistently applying the comparator group the appellants had put forward and erred by failing to conclude that the challenged statutory provisions discriminated on the basis of age contrary to s. 15 of the Charter. It is also my view that the impugned provisions cannot be saved by the application of s. 1 of the Charter and that the Reduction Provisions must be amended so as to comply with the requirements of s. 15 of the Charter. My reasons for reaching those conclusions follow.
 Since preparing these reasons, I have had the privilege of reading the draft reasons for judgment of Madam Justice Ryan. In my respectful view, the Supreme Court of Canada’s decision in R. v. Kapp, 2008 SCC 41, 294 D.L.R. (4th) 1, released after this appeal was argued, does not lead to a different conclusion when determining whether the impugned provisions in the PSSA and the CFSA constitute age discrimination in this case.
II. History of the SDB provisions in the PSSA and the CFSA
 A summary of the legislative history of the Supplementary Death Benefit Plans (“SDB Plans”) in the context of the overall benefit scheme created by the PSSA and the CFSA is contained in the trial judge’s reasons at paras. 11-15 and as it is readily available, I need not repeat it here. The trial judge’s reasons also make reference to the evidence which shed light on the purpose of the SDB Plans (at paras. 111-154). What follows is a brief outline of the enactment of and amendments to the SDB Plans in the PPSA and the CFSA.
 The PSSA was first enacted on 1 January 1954 (S.C. 1952-53, c. 47). A year later, that is, on 1 January 1955, in the first set of amendments to the PSSA, the SDB Plan was added (S.C. 1953-54, c. 64, s. 2). The first SDB Plan provided: a benefit amount of the lesser of $5,000 or the participant’s annual salary (s. 39(1)(a)); a minimum benefit of 1/6 of the participant’s salary (s. 39(1)(a)); and an employee contribution of 40 cents/$1000 of coverage (s. 42). The beneficiary of the SDB Plan could be either the participant’s spouse or, in the absence of a spouse, the participant’s estate (s. 44). The benefit was subject to an annual reduction of 10% after the participant’s 61st birthday (s. 39(1)(a)) and the participant could elect to discontinue coverage after retirement (s. 40). There was a paid-up benefit of $500 at the participant’s 65th birthday. The minimum paid-up benefit was the minimum dollar amount to which life insurance could be reduced. At age 65, it became a “paid-up” benefit, which meant there was no longer a financial contribution required from the participant for this portion of the coverage.
 Over the years, there were a number of changes to the SDB Plan, as provided in Part II of the PSSA and the Supplementary Death Benefits Regulations, C.R.C., c. 1360. The major amendments to the SDB Plan were summarized by the trial judge as follows (at para. 13):
1960 Amendments [S.C. 1960, c. 38]
· Minimum Benefit: increased to the greater of 1/6 of the participant’s salary or $500, not less than $500 for elective participants [s. 21(1)];
· Beneficiary: the spouse of the deceased male participant, or estate if no spouse. For female participants, beneficiary is the estate [s. 26(1)];
1966 Amendments [S.C. 1966-67, c. 44]
· Benefit Amount: the maximum of $5,000 is removed from the PSSA [s. 22(1)].
1976 Amendments [S.C. 1974-75, c. 81]
· Beneficiary: widened to include participant’s estate, anyone over age of 18, charity or benevolent organization, religious or educational organization [ss. 23-24].
1992 Amendments [S.C. 1992, c. 46]
· Benefit Amount: doubled to twice the participant’s annual salary [s. 25(1)];
· Minimum Benefit: increased to not less than 1/3 participant’s salary or $5,000, whichever was greater [s. 25(1)];
· Employee Contribution: reduced to 20 cents/$1000 of coverage [s. 26];
· Paid-up Benefit: increased to $5000 at the participant’s 65th birthday [s. 26].
1999 Amendments [S.C. 1999, c. 34]
· Minimum Benefit: increased to not less than 1/3 participant’s salary or $10,000 whichever was greater [s. 98(1)];
· Employee Contribution: reduced to 15 cents/$1,000 coverage [s. 103];
· Reduction: 10% annual reduction delayed until after the participant's 66th birthday [s. 98(1)];
· Paid-up Benefit: increased to $10,000 at the participant’s 65th birthday [s. 98(1)].
 The CFSA was amended in 1966 (S.C. 1966, c. 44) to include SDB coverage for regular forces participants. Until then, regular forces participants had been covered by the SDB provisions in the PSSA. The introductory SDB provisions in the CFSA were similar to those in the PSSA, except the benefit amount was one year’s salary (s. 53); the employee contribution was 20 cents/$1,000 coverage (s. 53); and a 10% annual reduction commenced after the participant’s 60th birthday (s. 53).
 After 1966, the amendments made to the CFSA and the PSSA diverged. The major amendments to the CFSA and the corresponding regulations, the Canadian Forces Superannuation Regulations, C.R.C., c. 396, were set out in the trial judge’s reasons as follows (at para. 15):
1975 Amendments [S.C. 1975, c. 81]
· Beneficiary: the widow of the participant unless she predeceases him or he names his estate or another beneficiary [ss. 45-46].
1977 Amendments [S.C. 1976-77 s. 28, C.P.C. 1360, s. 26(1) (1978)]
· Beneficiary: widened to include the widow of the participant, the participant’s estate, anyone over [the] age of 18, charity or benevolent organization, religious or educational organization.
1992 Amendments [S.C. 1992, c. 46]
· Benefit Amount: increased to twice the participant’s salary [s. 52(1)];
· Minimum Benefit: increased to not less than $5,000 [s. 52(1)];
· Employee Contribution: reduced to 10 cents/$1,000 coverage [s. 52(1)];
· Paid-up Benefit: increased to $5000 at the participant’s 65th birthday [s. 54].
III. The provisions of the PSSA and the CFSA and their Regulations, which are the subject of the s. 15 Charter challenge
 The challenged provision of the PSSA provides:
47(1) … “basic benefit”, with respect to a participant, means an amount equal to twice the salary of the participant, if that amount is a multiple of one thousand dollars, or an amount equal to the nearest multiple of one thousand dollars above twice the salary of the participant, if the first-mentioned amount is not a multiple of one thousand dollars, subject to a reduction of ten per cent, to be made as of the time that the regulations prescribe, for every year of age in excess of sixty-five attained by the participant, except that ….
 The following provisions of the Supplementary Death Benefit Regulations are challenged:
15. The reduction referred to in the definition “basic benefit” in subsection 47(1) of the Act shall be made on April 1st and October 1st of each year, whichever date immediately follows the participant’s birthday. SOR/78-785, s. 1; SOR/92-716, ss. 7(F), 8(F), 9(F), 11; SOR/99-378, s. 6.
16. The reduction in the amount that certain participants are required to contribute pursuant to s. 53 of Act shall commence on April 1st or October 1st, whichever date immediately follows the anniversary of the birthday of the participant on which he became eligible for the reduction. SOR/92-716, s. 11.
 The provisions of the CFSA which are challenged are as follows:
60(1) … “basic benefit”, with respect to a participant, means an amount equal to twice the salary of the participant, if that amount is a multiple of two hundred and fifty dollars, or an amount equal to the nearest multiple of two hundred and fifty dollars above twice the salary of the participant, if the first-mentioned amount is not a multiple of two hundred and fifty dollars, subject to a reduction of ten per cent, to be made as of such time as the regulations prescribe, for every year of age in excess of sixty attained by the participant, except that …
 The challenged provisions of the Canadian Forces Superannuation Regulations are as follows:
52. The times when the reductions referred to in the definition “basic benefit” in subsection 60(1) of the Act shall be made are as follows:
(a) in the case of an elective participant who ceased to be a member of the regular force and to whom an annuity or pension is not payable under the Act or the Defence Services Pension Continuation Act, each reduction shall be made on each anniversary of the day (that is on or that follows the 61st birthday of the participant, whichever occurs first), on which an annual contribution under the Act is payable; and
(b) in any case, other than the case mentioned in paragraph (a), each reduction shall be made on the first day of April or the first day of October whichever date immediately follows each anniversary of the birthday of the participant commencing with his 61st birthday. SOR/92-717, ss. 9(F), 10.
IV. The trial judge’s reasons
 The trial judge took as her guide the analytical framework found in Law v. Canada (Minister of Employment and Immigration),  1 S.C.R. 497, 170 D.L.R. (4th) 1, to determine whether the impugned legislation violated s. 15 of the Charter (at paras. 94-103). The three broad inquiries the court must make, as set out in Law at para. 88, are as follows:
(A) Does the impugned law (a) draw a formal distinction between the claimant and others on the basis of one or more personal characteristics, or (b) fail to take into account the claimant’s already disadvantaged position within Canadian society resulting in substantively differential treatment between the claimant and others on the basis of one or more personal characteristics?
(B) Is the claimant subject to differential treatment based on one or more enumerated and analogous grounds?
(C) Does the differential treatment discriminate, by imposing a burden upon or withholding a benefit from the claimant in a manner which reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or a member of Canadian society, equally deserving of concern, respect, and consideration?
 As to the first two parts of the analysis, the trial judge concluded that the claimants were subject to differential treatment based on one of the enumerated grounds and that the Reduction Provisions drew a formal distinction based upon an analogous ground, namely, the age of the claimant’s spouse at death (at paras. 104‑05).
 On the third stage of the analysis, the judge first set out the four contextual factors that need to be considered in order to determine if the impugned law demeans the complainants’ dignity. The trial judge recognized at para. 102 that the test to be applied was that of a reasonable person, in circumstances similar to those of the claimants, who takes into account the following contextual factors relevant to the claim:
1. Pre-existing disadvantage, stereotyping, prejudices or vulnerability experienced by the individual or group at issue;
2. The correspondence, or lack thereof, between the ground or grounds on which the claim is based and the actual need, capacity, or circumstances of the claimant or others;
3. The ameliorative purpose or effects of the impugned law upon a more disadvantaged person or group in society than the complainants; and
4. The nature and scope of the interest affected by the impugned law.
 In considering whether the differential treatment amounted to discrimination, the trial judge purported to accept the comparator group advanced by the appellants, which she described as “all civil servants and members of the armed forces who received the full SDB, not reduced on the basis of age” (at paras. 106-10).
 In reaching her conclusion that the Reduction Provisions did not constitute age discrimination, the trial judge considered the entire package of benefits under the PSSA and the CFSA available to surviving spouses of any age, and found that the SDB was “not intended to be a long term stream of income for older surviving spouses because their long term income security is guaranteed” (at paras. 122-23). The judge concluded that the purpose of the SDB was to provide insurance, and that even if it corresponded “imperfectly” with the needs of its beneficiaries, there was still “a correspondence between needs and the payment, when examined in the context of the other pensions and benefits to which the plaintiffs are entitled” (at para. 156).
 In terms of the four contextual factors informing the third and final inquiry in the Law test, the trial judge said:
(i) … it can no longer be said that a majority of elderly Canadians are a disadvantaged group in Canada compared to all other age groups. I conclude that the plaintiffs have failed to prove that the elderly suffer from stereotyping, prejudice, or vulnerability based on their reduced income (at para. 158).
(ii) The reduction in the amount received by a surviving spouse is offset to some degree by the survivors’ pension which does assure the survivor spouse of a long term stream of income … when combined with the entire benefit package … I find the law does not fail to take into account the plaintiffs’ actual situation (at para. 159).
(iii) … the SBD is life insurance. The very essence of group life insurance is the egalitarian nature of the cost and the coverage…. This has an ameliorative purpose for the younger members of the employee group in terms of coverage, and an ameliorative purpose for the older member of the group in terms of cost…. The ameliorative purpose mentioned in Law v. Canada was probably not intended to capture the type of effects I mention here. But there is a parallel reasoning in that the effect of group life insurance benefits different people in different ways at different times in their life cycle. (para. 160); and
(iv) I would not characterize the disadvantage to the plaintiffs as severe. There was much testimony that they are more advantaged when considering the whole package of employment benefits than other seniors. When compared to the comparator group, although they receive a lesser SDB they receive a greater survivor pension (para. 161).
 On the third question under the Law test, the judge said at para 170:
The purpose of the SDB varies somewhat as the covered employee ages. At the younger ages it provides a limited stream of income for unexpected death where the surviving spouse is not protected by a pension. At older ages the purpose of the SDB is for the expenses associated with last illness and death. I conclude that the fact that the reduction means those costs may not be fully covered is not discrimination. It does not bear any of the hallmarks of discrimination as set out in the Law v. Canada analysis.
 In view of her conclusion that the Reduction Provisions did not constitute age discrimination under s. 15, the judge found it unnecessary to consider the remaining common issues and dismissed the appellants’ action.
V. The comparator group
 In my respectful view, the judge’s analysis of the third common issue was flawed because the comparator group the appellants had put forward was not fully stated and was not consistently applied. The comparator group was described as spouses of “all civil servants and members of the armed forces who received the full SDB, not reduced on the basis of age” (at para. 108). Based on the provisions in the legislation providing for benefits, it is plain that those who received the full SDB would also be eligible for a survivor’s pension. While eligibility for a survivor’s pension appears to me to be implicit in the comparator group the appellants advanced, the trial judge did not consistently include that component of the comparator group.
 In my opinion, the description of the comparator group ought properly to be stated as “all civil servants and members of the armed forces who received a full SDB, not reduced on the basis of age, and who are eligible for a survivor’s pension”.
 Before turning to the question of what effect adopting and applying the foregoing description of the comparator group would have in the Law analysis, I will try to explain how the comparator group used by the trial judge affected her analysis.
 The judge found that the legislation had a dual purpose: for the older employees, the benefit was intended to provide for the costs associated with last illness and death; whereas for the younger employees, who might not be in receipt of a survivor’s pension, the benefit was meant to provide temporary income replacement. Instead of restricting the definition of the comparator group so as to include only those members for whom the benefit provided was similar to the claimant group, the judge accepted the broad definition of comparator group as simply “all civil servants and members of the armed forces who received a full SDB, not reduced on the basis of age”. As a result, the judge encountered a number of analytical difficulties in attempting to compare the two groups of individuals.
 Considering that the purpose of the benefit for the claimant group is associated with last illness and death and not long-term income replacement, which is provided for by the members’ survivor’s pension, it appears to me that the comparator group ought to be refined to include “eligibility for survivor’s pension”. That refinement would ensure that the comparator group mirrors the claimant group, save for the distinction made on the analogous ground of age of the spouse. Narrowing the comparator group would allow the court to draw appropriate comparisons on the evidence and clarify the analysis.
 As noted above, the judge applied the analytical framework set out in Law and concluded that the first two inquiries were satisfied in that the Reduction Provisions drew a formal distinction based upon an analogous ground, namely the age of one’s spouse at death (paras. 104-05). However, she found that the differential treatment did not amount to discrimination under the third branch of the Law test.
 The judge stated that she accepted the appellants’ choice of comparator group as “all civil servants and members of the armed forces who received a full SDB, not reduced on the basis of age” on the basis that there seemed to be “no other more appropriate group” (at paras. 108, 110).
 On the appeal, the appropriateness of the comparator group used by the trial judge was not raised as a discrete issue. While the starting point in choosing the appropriate comparator group is the one advanced by the claimants, the court should nevertheless ensure that the comparator is appropriate and may refine the comparison within the scope of the grounds pleaded where that is warranted (Hodge v. Canada (Minister of Human Resources Development), 2004 SCC 65,  3 S.C.R. 357 at paras. 20-21; Law at para. 58).
 When considering the way in which the comparator group is defined, the court must take the purpose of the legislation and the nature of the benefit into account. In Canada (Attorney General) v. Hislop, 2007 SCC 10,  1 S.C.R. 429, aff’g (2004) 573 O.R. (3d) 685, 246 D.L.R. (4th) 644 (C.A.), the Supreme Court of Canada observed that “[i]t is the purpose of the [legislation] itself that determines the appropriate comparator group” (at para. 38). In particular, where the impugned legislation concerns access to a government benefit, “the initial focus is on what the legislature is attempting to accomplish” (Hodge at paras. 24, 26). The importance of aligning the comparator group with the intended purpose of the legislation is underscored by the fact that a number of cases before the Supreme Court of Canada have turned on that basis.
 In Lovelace v. Ontario, 2000 SCC 37,  1 S.C.R. 950, the claimants were non-status Indians who considered themselves discriminated against by a provincial government program favouring status Indians. However, the Court rejected the claimants' choice of comparator group on the basis that the program was undertaken to support a “government-to-government” relationship and therefore the potential universe of claimants was limited to “band and non-band aboriginal communities” (at paras. 7, 64). The purpose of the program was to provide benefits to aboriginal communities, not aboriginal individuals, and because this distinction was not in itself discriminatory, the Court would not second-guess what amounted to a policy decision made by the legislature.
 In Nova Scotia (Workers’ Compensation Board) v. Martin, 2003 SCC 54,  2 S.C.R. 504, chronic pain sufferers with work-related injuries claimed that the Nova Scotia Workers' Compensation Board discriminated against them on the basis of their disability. Under the relevant legislative scheme, while all injured workers were barred from bringing claims in tort against their employers, those not suffering from chronic pain were entitled to compensation for their injuries. Chronic pain sufferers, on the other hand, were largely excluded from the benefit scheme. Members of the claimant group attempted to compare themselves with chronic pain sufferers whose injuries were not employment-related, pointing to the fact that such persons were able to seek compensation in court. However, the Court rejected this comparison. The compensation benefits provided under the legislative scheme were by definition not available to members of the asserted comparator group; therefore, there was no proper alignment between the benefit sought and the ground of discrimination alleged (at para. 72). The appropriate comparator was defined as the group of workers subject to the legislation who did not have chronic pain and were eligible for compensation for their work-related injuries (at para. 71).
 In this case, the appellants assert that the “universe of people potentially entitled to the benefit” consists of the beneficiaries of public servants and members of the Canadian Forces. Therefore, the appropriate group to which the claimants should be compared are all individuals who received a full SDB, without age reduction. Although the trial judge accepted this comparator, it is evident from her reasons that she considered it problematic. The judge took the view that the heterogeneity within the claimant and comparator group made it impossible to draw any useful comparisons between them because individual members varied widely in terms of age and circumstance (at paras. 147‑48).
 Although the trial judge recognized that the purpose of the legislation had to be taken into account in defining the comparator group, she considered the purpose to vary according to age. Her approach was unusual and added a layer of complexity to the analysis. The judge found that for younger surviving spouses, the SDB provided a limited stream of income for unexpected death where the surviving spouse was not protected by a pension. By contrast, she found that for older surviving spouses, the SDB provided for expenses associated with last illness and death because their long term income security was guaranteed by pension (at para. 170).
 As the Supreme Court of Canada has stated, the comparator group should mirror the characteristics of the claimant group relevant to the benefit or advantage sought, except for the personal characteristic related to the enumerated or analogous ground raised as the basis for the discrimination (Hodge at para. 23), which in this case is age.
 It is therefore my view that the appropriate comparator group should be restated as “all civil servants and members of the armed forces who received a full SDB, not reduced on the basis of age, and who are eligible for survivor’s pension”. Restating the comparator group in that way brings clarity to the analysis and affects the way in which the evidence is considered in relation to the four contextual factors that inform the third branch of the Law test.
 The way in which the trial judge described the comparator group led to a number of analytical problems that appear throughout her analysis and, in particular, when she drew comparisons between the claimants and broad categories of people unrelated to the SDB plan. By way of illustration, I have set out below the instances in which the comparator group used by the trial judge was improperly applied.
 In her discussion of whether the claimants failed to establish that they suffered from a pre-existing disadvantage owing to their low level of income and wealth, the trial judge noted that the claimants had particular advantages in comparison to the “universe of elderly persons”, “all other age groups”, the “statistical average” and “the general population” (at paras. 140, 158). She also referred to the fact that federal government employees had a seven to nine percent salary advantage over employees in the private sector (at para. 140).
 The trial judge referred to the comparator group as “the entire under age 65 population” when considering the evidence related to the second contextual factor (at para. 149).
 The trial judge found that the disadvantage to the plaintiffs was not severe and, in reaching this conclusion, gave weight to the breadth of testimony showing that the claimant group was more advantaged than “other seniors” when considering the whole package of employment benefits (at para. 161).
 The inconsistent application of the chosen comparator group stems from the fact that the comparator group as stated by the trial judge was too broadly defined. Refining the comparator group in the manner I have suggested would permit appropriate comparisons to be drawn on the evidence because the purpose of the legislation and the nature of the benefit provided would remain constant. This would, in turn, affect the analysis most significantly in relation to the consideration of whether the distinction drawn on the basis of age corresponds to the actual need, capacity, or circumstances of the claimants.
 The trial judge accepted as a fact that the “costs of last illness and death increase quite dramatically in each decade over 65” (para. 154). Nevertheless, she concluded that the legislation did not fail to take into account the claimants’ needs because of the coverage provided under the government of Canada benefit plans, which offset the reduction in the SDB amount to a certain degree (at para. 154). Further, the judge noted that the legislation also included “consideration of the younger working public servants … [who] may not have a pension to rely upon” (at para. 139).
 An application of the revised comparator group would render the fact that members of the claimant group receive survivor’s pensions and other government benefits under the Plans irrelevant. The resulting comparison would reveal two groups of surviving spouses receiving similar pensions and benefits, with one group subject to the Reduction Provisions and the other group not, solely on the basis of the age of their spouse. In light of the fact that the trial judge accepted that the purpose of the legislation is to provide for the costs associated with last illness and death, and that the costs of last illness increase dramatically with age in each decade over 65, the logic of reducing the benefit provided on the basis of age becomes questionable.
 Ensuring that the comparator and claimant group mirror one another in terms of benefits and pensions, save for the Reduction Provisions, does not mean that the overall context of benefits covered by other government schemes should be completely disregarded. Indeed, members of the claimant group may have a higher income than members of the comparator group as a result of larger survivor’s pensions. The respondent asserts that the Supreme Court of Canada has recognized that it is appropriate in the discrimination inquiry to consider whether a benefit which is denied in one act is replaced by compensation in another act. While this may be true in some instances, the case law indicates that this inquiry should be postponed to the s. 1 analysis (Tétreault-Gadoury v. Canada (Employment and Immigration Commission),  2 S.C.R. 22 at 42, 81 D.L.R. (4th) 358; Egan v. Canada,  2 S.C.R. 513 at para. 156, 124 D.L.R. (4th) 609, Cory J., dissenting). As Cory J. explained in his dissenting opinion in Egan:
Postponing this inquiry to s. 1 is appropriate because, if a claimant has established that the challenged legislation has denied an equal benefit of the law, then the government would, under s. 1 of the Charter, bear the onus of demonstrating that the denial was offset and justified by benefits provided under other provincial legislation.
 In my opinion, the respondent’s arguments, as well as the trial judge’s comments, relating to the egalitarian nature of group life insurance and the constraints associated with the cost of providing these types of benefits should properly be considered as justification arguments under the s. 1 Charter analysis.
VI. Analysis: Do the Reduction Provisions discriminate on the basis of age, contrary to s. 15(1) of the Charter?
 In Law, the Supreme Court of Canada established the analytical framework for determining whether an impugned law violates s. 15(1) of the Charter. The discrimination analysis is carried out in three steps. First, the court must determine whether the impugned law draws a distinction between the claimant and others on the basis of a personal characteristic. Second, the court must examine whether the basis for the differential treatment is one of the enumerated or analogous grounds. Third, the court must determine whether this differential treatment is discriminatory in purpose or effect (Hislop at para. 36).
 In this case, the trial judge held that the Reduction Provisions drew a formal distinction based upon an analogous ground, namely, the age of one’s spouse at death. However, she concluded that the distinction did not amount to substantive discrimination. Thus, the third stage of the discrimination analysis is the focus of the appeal.
(a) Substantive discrimination
 Not all distinctions drawn on the basis of an enumerated or analogous ground amount to substantive discrimination (Lavoie v. Canada, 2002 SCC 23,  1 S.C.R. 769 at paras. 76, 87; Canadian Foundation for Children, Youth and the Law v. Canada (Attorney General), 2004 SCC 4,  1 S.C.R. 76 at paras. 86-87 [Canadian Foundation]). A law that imposes differential treatment will be considered discriminatory if it conflicts with the purpose of s. 15(1) (Law at para. 41; Lovelace at para. 55). Justice Iacobucci described the purpose of s. 15(1) in Law at para. 51 as follows:
It may be said that the purpose of s. 15(1) is to prevent the violation of essential human dignity and freedom through the imposition of disadvantage, stereotyping, or political or social prejudice, and to promote a society in which all persons enjoy equal recognition at law as human beings or as members of Canadian society, equally capable and equally deserving of concern, respect and consideration.
 To have a complete understanding of how an impugned law impacts human dignity, a full contextual analysis is required, taking into account the surrounding legislative, historical, and social circumstances relevant to the claim (Lovelace at para. 55). However, for the purposes of this appeal, it is important to note that the emphasis on context does not invite a broad, generalized examination of the facts in evidence, which was the approach taken at trial. As was stated in Lovelace at para. 55, “this contextual analysis is a directed inquiry; it is focused through the application of contextual factors which have been identified as being particularly sensitive to the potential existence of substantive discrimination” (emphasis added).
 The framework derived from Law mandates consideration of the following contextual factors: (1) the presence of pre-existing disadvantage, vulnerability, stereotyping or prejudice directed at the person or group at issue; (2) the correspondence, or lack thereof, between the ground upon which the differential treatment is based and the actual needs, characteristics and circumstances of the affected person or group; (3) any ameliorative purpose or effect of the legislation upon a more disadvantaged group; and (4) the nature and scope of the interest affected by the legislation (Lavoie at para. 38).
 At each stage of the analysis, the central focus must remain on that of a reasonable person in circumstances similar to the claimant (Law at paras. 59-61; Lovelace at para. 55).
(b) Pre-existing disadvantage, stereotyping, prejudice, or vulnerability experienced by the individual or group at issue
 The first contextual factor examines whether members of the claimant group suffer from pre-existing disadvantage, stereotype, prejudice or vulnerability. Differential treatment of individuals belonging to a group that has experienced historic disadvantage is more likely to have a discriminatory impact because it often exacerbates that disadvantage (Law at para. 63; Corbiere v. Canada (Minister of Indian and Northern Affairs),  2 S.C.R. 203 at para. 70, 173 D.L.R. (4th) 1 at para. 70). As McLachlin C.J. explained in Gosselin v. Quebec (Attorney General), 2002 SCC 84,  4 S.C.R. 429 at para 30:
Historic patterns of discrimination against people in a group often indicate the presence of stereotypical or prejudicial views that have marginalized its members and prevented them from participating fully in society. This, in turn, raises the strong possibility that current differential treatment of the group may be motivated by or may perpetuate the same discriminatory views. The contextual factor of pre-existing disadvantage invites us to scrutinize group-based distinctions carefully to ensure that they are not based, either intentionally or unconsciously, on these kinds of unfounded generalizations and stereotypes.
 Furthermore, in Law, Iacobucci J. emphasized that pre-existing vulnerability and disadvantage is a relevant indicator of substantive discrimination, but not a prerequisite to establishing a Charter violation (at para. 65). The claimants are not required to prove that they are more disadvantaged than the comparator group. At this stage of the inquiry, the Supreme Court of Canada has expressly rejected adopting a “relative disadvantage” approach, as this would involve a “race to the bottom” (Lovelace at paras. 58-59, 69; Martin at para. 88).
 Another crucial point to note is that the focus of the inquiry must align with the alleged ground of discrimination (Hodge at para. 37). In Gosselin, the claimant brought a class action challenging a social welfare regulation that subjected persons under the age of 30 to more onerous requirements than persons over 30. In an attempt to establish pre-existing disadvantage, the claimant emphasized the disadvantages experienced by all welfare recipients as a class, rather than focusing on disadvantages experienced by persons under 30. McLachlin C.J., writing for the majority, rejected this argument, explaining at para. 35:
The ground of discrimination upon which she founds her claim is age. The question with respect to this contextual factor is therefore whether the targeted age-group, comprising young adults aged 18 to 30, has suffered from historic disadvantage as a result of stereotyping on the basis of age.
[Emphasis in original]
 Accordingly, the focus of the inquiry in this case should be directed towards examining whether older surviving spouses experience pre-existing disadvantage on the basis of their age. Therefore, it does not matter whether the claimants have advantages over seniors in the private sector by virtue of their status as recipients of government benefits because this description does not properly align with the alleged ground of discrimination.
 The appellants claim that older people are disadvantaged in terms of overall economic well-being as a result of increasing health care costs and declining income. Beyond financial insecurity, the appellants also point to the loneliness and despair that often results from the death of a spouse during the later years of life. Furthermore, the appellants allege that seniors are subject to stereotypes that portray them as either physically or mentally incapable, or conversely, living carefree without problems worthy of attention.
 The respondent, in reply, points to the evidence of the individual claimants that indicates they are not vulnerable: none of them receive a Guaranteed Income Supplement (“GIS”); they are financially secure; and they are even able to travel.
 There is, however, ample evidence in the record to establish that seniors as a group suffer from pre-existing disadvantage and vulnerability based on their economic well-being. Any assessment of overall economic well-being must take into account wealth, income and costs. The evidence on all three of these elements points to the vulnerability of seniors.
 First, a table from a Statistics Canada report dated February 2002, entitled “Wealth Inequality” by Rene Morrisette, Xuelin Zhang and Marie Drolet, shows that in 1984 the average wealth of the major income recipients in a family steadily rose from the age of 35 to 64 and then declined from the age of 65 onwards to a level below the average wealth of persons in the age 45 to 54 age category (see Withler at para. 148). The trial judge rejected this report as evidence of seniors’ disadvantage on the basis that it failed to show that persons 65 and older were less economically well-off than the entire under age 65 population (at para. 149). However, as I have indicated earlier, the claimants are not required to prove that they are the most disadvantaged. Furthermore, even if the claimants were required to show relative disadvantage, the relevant comparison would be with the comparator group, which is surviving spouses eligible for the full SDB and pension benefits. This would more than likely exclude persons in the lower age categories, i.e., the only sector of society that had lower average wealth.
 Second, the data on average wealth must be considered in light of the fact that older people have increasing difficulty finding and maintaining employment (see Law at para. 101; Gosselin at para 60). Fewer income-generating opportunities available at older ages increase the financial insecurity of older people.
 Third, the evidence of Professor Forget shows that the health care costs incurred in the last six months of life increase dramatically with age and are particularly high for seniors over the age of 60 to 65 (paras. 151-154 of the trial judge’s reasons). That evidence will be discussed fully under the second contextual factor. Although medical costs are only one part of overall living expenses, the totality of the evidence on all three elements of economic well-being supports the conclusion that seniors are economically vulnerable.
 The respondent maintains that the claimants have not shown that they have personally experienced disadvantage because they are better off than average Canadian seniors. Nevertheless, it must be remembered that regardless of the particular circumstances of the individual claimants, the issue is whether pre-existing disadvantage exists among elderly persons as a group (see Corbiere at para. 19; Martin at para. 76). Even if all of the claimants are to be considered personally ‘advantaged’ in some respects, this does not serve to negate pre-existing disadvantage experienced by seniors as a class, which is the relevant focus of this inquiry. This point was illustrated in Lavoie in relation to historic disadvantage of non-citizens as a class. The court recognized that while the particular claimants in the case were all relatively educated, it was well-established that non-citizens as a group suffered from political marginalization, stereotyping and historical disadvantage (at para. 45).
 In addition, it is of little significance that the percentage of single elderly Canadians considered to be living on a low income has reduced considerably or that the median wealth for Canadians over age 65 is rising. The fact that circumstances may have improved over time for a particular class of individuals does not mean that members of that class are no longer vulnerable or subject to negative stereotyping. If that were the case, the same argument could be made about virtually every enumerated or analogous ground. “Less disadvantage” does not necessarily equal “no disadvantage”.
(c) The correspondence, or the lack thereof, between the ground or grounds on which the claim is based and the actual need, capacity or circumstances of the claimant or others
 The second contextual factor examines the relationship between the enumerated or analogous ground (i.e., the age of one’s spouse at death) and the nature of the differential treatment (Granovsky v. Canada (Minister of Employment and Immigration, 2000 SCC 28,  1 S.C.R. 703 at para. 61). A law that takes into account the actual and particular needs or abilities of individuals or groups, rather than relying on stereotypes or generalized assumptions, is less likely to be found discriminatory (Law at paras. 70-71). The focus of this inquiry is on the extent to which the impugned law considers the claimants’ situation in a manner that respects their value as full members of Canadian society (Law at para. 53; Martin at para. 92).
 In this case, the trial judge found that the Reduction Provisions do not fail to take into account the actual needs and circumstances of the claimants. While the judge accepted that the costs of last illness and death increase with age, particularly in each decade over the age of 65, she found that the Reduction Provisions are offset to a certain degree by the full package of other benefits to which the claimants are entitled, including higher pensions (at para. 154).
 The appellants argue that the reason why the claimants might have higher survivor’s pensions than the younger group is both because their spouses earned it and because they have a greater need for long-term income replacement. Accordingly, it should not be used as justification for reducing assistance with covering the costs of last illness and death.
 In reply, the respondent argues that the egalitarian nature of the legislative scheme as a form of group life insurance necessarily involved the allocation of scarce resources in a way that obtained the maximum benefit for the greatest number of employees. The respondent points to the reality that distinctions are commonly made on the basis of age in group life insurance plans and reductions at some stage are necessary in order to ensure the financial viability of the plan for everyone. Relying on Gosselin, the respondent contends that Parliament's decision to give greater support to younger people accords with “logic and common sense” because they might not have access to the same level of pension benefits as the claimants. The respondent objects to the appellants’ sole focus on the elderly and urges this Court to consider the needs of others covered by the scheme as well.
 When considering this contextual factor, it is crucial to remember that the perspective from which the discrimination analysis must be conducted is that of a reasonable person in circumstances similar to the claimants (Law at para. 61; M. v. H.,  2 S.C.R. 3 at para. 67, 171 D.L.R. (4th) 577 at para. 67). From this point of view, it can hardly be said that the Reduction Provisions correspond to the needs of the claimants, namely, surviving spouses over the age of 60 or 65. The law is intended to provide older surviving spouses with assistance covering the costs of last illness and death. The evidence before the trial judge demonstrated that older people bear a higher burden than younger people in terms of health care expenses because their needs are greater and yet the coverage is the same. If Parliament had properly considered the actual needs and circumstances of the claimants in this regard, the law would have provided them with more or at least equal assistance than younger surviving spouses, rather than less.
 The respondent argues that the claimants’ higher pension levels compensate for the reduced provisions, which presumes that the claimants have less need for their pension than those in the comparator group. In fact, the opposite is true. As the Court noted in Law, older people have more difficulty replacing a deceased spouse’s income over the longer term and this economic vulnerability requires they receive a higher pension than younger people (Law at paras. 103-04). The Reduction Provisions do not just fail to consider the needs of surviving spouses over the age of 60 or 65 – they directly conflict with the claimants’ actual needs related to the costs of last illness and death. By forcing the claimants to fall back on their pensions, the Reduction Provisions undermine the purpose for which pensions were provided in the first place, i.e., to counteract their economic vulnerability with long-term income replacement.
 The trial judge accepted that some of the claimants “may have incurred some costs not covered by the SDB or other coverage available to them”, however, she nevertheless found that the legislation does not fail to meet its purpose (at paras. 154‑55). In coming to that conclusion, the judge relied on the fact that the law is a “balancing exercise” designed to take into account not only the interests of the claimants, but also “the interests of the public to ensure that the civil service is treated equitably but not over generously” (at para. 155).
 Had the trial judge considered the impact of the Reduction Provisions from the perspective of a reasonable person in circumstances similar to the claimants, she most likely would have come to a different conclusion. The equality guarantees in s. 15 of the Charter are concerned with the protection of essential human dignity, not with whether certain violations of human dignity are in the public interest. On this point, Bastarache J.’s comments in Lavoie are instructive:
47 … Section 15(1) requires the claimant to show that her human dignity and/or freedom is adversely affected. The concepts of dignity and freedom are not amorphous and, in my view, do not invite the kind of balancing of individual against state interest that is required under s. 1 of the Charter....
48 … [T]he exigencies of public policy do not undermine the prima facie legitimacy of an equality claim. A law is not “non-discriminatory” simply because it pursues a pressing objective or impairs equality rights as little as possible.... While these are highly relevant considerations at the s. 1 stage, the suggestion that governments should be encouraged if not required to counter the claimant's s. 15(1) argument with public policy arguments is highly misplaced. Section 15(1) requires us to define the scope of the individual right to equality, not to balance that right against societal values and interests or other Charter rights.
 The overarching remedial purpose of s. 15 must also be borne in mind at this stage of the analysis. As the Ontario Court of Appeal stated in Hislop at para. 54, “the court must not be sidetracked by considerations that more appropriately belong in the s. 1 or remedy analysis.” Within the s. 15 analysis, the purpose of the legislation should not be construed in a way that mirrors the “legitimate objective” analysis under s. 1. Here, the purpose of the program as it applies to the claimants is to provide for the costs of last illness and death; it is not to provide an egalitarian group life insurance scheme. Arguments referring to the cost constraints of running such a program relate to the government’s possible justification for creating a discriminatory law and should not be considered when seeking to answer the question of whether the law is discriminatory in the first place. As Binnie J. explained in Canadian Foundation at para. 97 (dissenting in part):
Care must be taken, however, to ensure that the “correspondence” factor is kept to its original purpose as a marker for discrimination and not allowed to become a sort of Trojan horse to bring into s. 15(1) matters that are more properly regarded as “reasonable limits ... demonstrably justified in a free and democratic society” (s. 1).
 By providing less to older spouses than younger spouses when it is clear that their needs are higher in this regard, the Reduction Provisions do not take into account the claimants’ needs. A reasonable person in circumstances similar to the claimants would not be comforted by the fact that the reductions in her benefits enabled surviving spouses in similar circumstances under the age of 60 or 65 to receive full benefits, nor would she feel less of an infringement upon her human dignity in knowing that the Reduction Provisions satisfied “cost-effectiveness” concerns of the overall program.
(d) The ameliorative purpose or effect of the impugned law upon a more disadvantaged person or group in society
 The third contextual factor examines whether the legislation has an ameliorative purpose or effect upon a more disadvantaged person or group in society (Law at para. 72). The case law distinguishes between two types of legislation: targeted ameliorative legislation and comprehensive ameliorative legislation. Laws that act to remedy a disadvantage of a particular group of individuals in society are less likely to be considered discriminatory when those who are excluded from the scope of legislation are more advantaged with respect to the circumstances addressed by the legislation (Lovelace at paras. 85-86; Law at para. 72; M. v. H. at para. 71). Conversely, laws that are designed to benefit the population in general, yet exclude historically disadvantaged claimants, will “rarely escape the charge of discrimination.” (Law at para. 26, quoting Andrews v. Law Society of British Columbia,  1 S.C.R. 143 at 174-75, 56 D.L.R. (4th) 1; Vriend v. Alberta,  1 S.C.R. 493 at paras. 94-104, 156 D.L.R. (4th) 385 at paras. 94-104).
 The appellants maintained that the Reduction Provisions do not have an ameliorative purpose acting to benefit a more disadvantaged group. The respondent, in reply, acknowledges that the law was not drafted with an ameliorative purpose “as contemplated by this contextual factor”. Nevertheless, the respondent emphasizes that provision of full benefits to younger spouses, which is only made possible by reducing benefits for older spouses, was designed to address the “increased vulnerability of individuals, families and partners who are in unusual and difficult circumstances.”
 It is undisputable that the loss of a spouse at any age will result in increased vulnerability for the surviving partner. Neither is it disputed that the provision of full death benefits to younger spouses would help ameliorate this vulnerability. However, as stated in Law at para. 70:
The fact that the impugned legislation may achieve a valid social purpose for one group of individuals cannot function to deny an equality claim where the effects of the legislation upon another person or group conflict with the purpose of the s. 15 guarantee.
 In this case, Parliament’s decision to provide older surviving spouses with less assistance covering the costs of last illness and death than younger spouses does not have an ameliorative purpose or effect within the meaning of this contextual factor such as to point away from substantive discrimination. For the reasons stated above, those who are subject to the Reduction Provisions (i.e., the claimants) are more disadvantaged in terms of their ability to cover the costs associated with last illness and death than those who are not subject to the reductions (i.e., the comparator group). While the overall benefit scheme can be considered generally ameliorative in the sense that some amount of benefits are provided to all surviving spouses, it is the challenged distinction within the scheme, the Reduction Provisions, with which we are concerned (see Falkiner v. Ontario (Director, Income Maintenance Branch, Ministry of Community and Social Services) (2002), 59 O.R. (3d) 481, 212 D.L.R. (4th) 633 (C.A.) at para. 99). The decision to provide older surviving spouses with less assistance covering the costs of last illness and death than younger surviving spouses was not motivated by a desire to ameliorate a more disadvantaged sector in society. Rather, the evidence indicates that the Reduction Provisions were motivated primarily by financial considerations.
(e) The nature and scope of the interest affected by the impugned law
 The final factor evaluates the economic, constitutional and societal significance of the interest adversely affected by the law (Law at para. 74). This involves both a qualitative and quantitative analysis. It is relevant to consider whether the distinction restricts access to a “fundamental social institution” or affects a “basic aspect of full membership in Canadian society”. Furthermore, distinctions that result in a complete non-recognition or exclusion of a group’s interests will more likely point towards substantive discrimination than legislation that at least partially recognizes or accommodates the group (Egan, L’Heureux-Dubé J. at paras. 63-64).
 The respondent contends that the interest affected is purely economic; I must disagree. In M. v. H., Cory J. characterized as fundamental “the ability to meet basic financial needs following the breakdown of a relationship characterized by intimacy and economic dependence” (at para. 72). Similarly, I find that the nature of the interest affected in the present case is fundamental. The ability to cover the financial costs associated with last illness and death of a spouse is intimately linked to the degree and quality of care that can be provided which, in turn, can have a direct impact on the health and well-being of the ill or dying spouse. For the surviving spouse, the Reduction Provisions increase not only her financial burden, but also her psychological burden in terms of having to assume a larger role in caring for her partner informally at home. Furthermore, to the extent that the Reduction Provisions force the surviving spouse to use part of her pension, the ability to meet her own day-to-day expenses and needs subsequent to the loss of her partner is also adversely affected. As such, I find that the economic nature of the interest in this case is highly associated with physical and mental integrity interests as well.
 While the nature of the interest affected is at the higher end of the scale in terms of importance, the scope of the interest affected is more towards the middle end of the scale in terms of severity. In Egan, L’Heureux-Dubé J. explained that the more severe and localized the economic consequences, the more likely that the impugned law is discriminatory (at paras. 63-64). Rather than completely denying surviving spouses over the age of 60 or 65 any death benefits, the law reduces the level of assistance provided. Yet, depending on the number of years one is subject to the reductions, the ultimate benefit received could be reduced quite significantly. With this in mind, I find that the overall effect of the law under this contextual factor points towards substantive discrimination.
VII. Summary and conclusion on the discrimination issue
 A contextual analysis of the differential treatment imposed by the Reduction Provisions leads to the conclusion that this legislative distinction violates the essential human dignity of surviving spouses whose SDB has been reduced on the basis of the age of their spouses at death. This amounts to substantive discrimination. Justice Iacobucci described human dignity in the following terms in Law at para. 53:
Human dignity means that an individual or group feels self-respect and self-worth. It is concerned with the physical and psychological integrity and empowerment. Human dignity is harmed by unfair treatment premised upon personal traits or circumstances which do not relate to individual needs, capacities, or merits. It is enhanced by laws which are sensitive to the needs, capacities, and merits of different individuals, taking into account the context underlying their differences. Human dignity is harmed when individuals and groups are marginalized, ignored, or devalued, and is enhanced when laws recognize the full place of all individuals and groups within Canadian society. Human dignity within the meaning of the equality guarantee does not relate to the status or position of an individual in society per se, but rather concerns the manner in which a person legitimately feels when confronted with a particular law.
 For the reasons outlined above, a reasonable person in circumstances similar to the claimants would feel ignored and devalued by the Reduction Provisions. While the government is not required to provide a benefit, if it chooses to do so it must be provided in a non-discriminatory manner. The government cannot escape the claimants’ assertion of discrimination by pointing to the fact that private employers incorporate reduction provisions into their life insurance schemes, for it is government action that must withstand Charter scrutiny. The sense of infringement to the claimants’ dignity that is caused by the Reduction Provisions is not alleviated by the fact that the distinction is commonplace.
 What is perhaps of most concern about the Reduction Provisions in this case is that they are premised on the assumption that older surviving spouses can readily draw on their pensions with little or no consequence. Yet by requiring surviving spouses to use their pensions to compensate for receiving a reduced death benefit, the law exacerbates their income vulnerability, which is the very harm against which survivor’s pensions are meant to protect. The effect of the Reduction Provisions is that older couples are essentially forced to choose between the medical needs of the ill spouse preceding death and the needs of the surviving spouse thereafter.
 As noted by the court below, no complaints were filed about the Reduction Provisions prior to Margolis v. Canada, 2001 FCT 85, 201 F.T.R. 55. The trial judge concluded at paras. 126 and 164 of her reasons that this was pertinent evidence indicating that very few beneficiaries felt their dignity was being violated by the Reduction Provisions. In my opinion, it is doubtful if anything could reasonably be inferred from the claimants’ silence. Moreover, it is important to remain sensitive to the potential impact of discriminatory laws on vulnerable segments of society whose problems may go unnoticed by the majority of the population. The very purpose of the equality guarantees in the Charter is to promote the view that every individual is worthy of recognition or value as members of Canadian society (Granovsky at para. 58, quoting L’Heureux-Dubé J. in Egan at para. 39). To penalize older surviving spouses on the basis of their previous silence would render the s. 15 Charter protection meaningless.
 For the reasons stated, I am of the view that the Reduction Provisions draw a distinction on the basis of age that violates essential human dignity. As a consequence of that conclusion, the burden must shift to the respondent to show that the discrimination is justifiable in a free and democratic society under s. 1 of the Charter.
 There was evidence before the trial court which would permit a s. 1 analysis, but in light of her conclusion that the impugned legislation does not infringe s. 15 of the Charter, the trial judge did not engage in such an analysis. As a result, we do not have the benefit of any reasons on that issue.
 The appellants have urged us to consider the issue of whether the impugned provisions of the PSSA and the CFSA and their Regulations can be saved under s. 1 of the Charter, in light of the extensive evidence presented at trial. I agree that the evidence presented permits a section 1 analysis to be undertaken and it is to that issue I now turn.
VIII. Can the impugned legislation be saved under s. 1?
(a) The framework for the s. 1 analysis
 Section 1 of the Charter provides:
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.
 In R. v. Oakes,  1 S.C.R. 103, 26 D.L.R. (4th) 200 (cited to S.C.R.), the Supreme Court of Canada established the framework for conducting a s. 1 analysis: Oakes at 138-140; Dagenais v. Canadian Broadcasting Corp.,  3 S.C.R. 835 at 889, 120 D.L.R. (4th) 12. The analysis requires consideration of the following questions:
1. Is the objective of the law pressing and substantial?
2. Is there a rational connection between the objective of the infringing measure and the means chosen to achieve the objective?
3. Does the impugned law impair rights as little as possible?
4. Are the deleterious effects of the law proportional to both the salutary effects and the importance of the objective?
 The burden of proof on a s.1 analysis shifts to the party who wishes to uphold the law. The standard of proof requires proof on a balance of probabilities that the infringing measure is justifiable (Oakes at 136-137). If my analysis of the discrimination issue is correct, the respondent bears the burden of proof.
 Oakes also makes clear that at each stage of the analysis, the proper focus of enquiry is the limit on the Charter right itself: Martin at para. 107. For that reason, the focus must remain throughout on the question of whether the Reduction Provisions can be saved under s. 1, not whether the overall legislative scheme of “group life insurance” is justifiable.
 Finally, the Supreme Court of Canada has made it clear that the evidence led in support of the government’s claim that the impugned law is a reasonable limit under s. 1 is “very important to the outcome” (Newfoundland (Treasury Board) v. N.A.P.E., 2004 SCC 66,  3 S.C.R. 381 at para. 55). As McLachlin J. stated in RJR-MacDonald Inc. v. Canada (Attorney General),  3 S.C.R. 199, 127 D.L.R. (4th) 1 [RJR], writing for the majority at para. 128: “The process is not one of mere intuition, nor is it one of deference to Parliament's choice. It is a process of demonstration.” (emphasis in original)
 Despite the jurisprudence on the subject of the importance of an evidentiary record, in this case the respondent takes the position that the s. 1 analysis “can be conducted on the face of the legislation itself” and that parts of the Oakes test are “self-evident”.
 In RJR, McLachlin J. stated at para. 133:
The s. 1 inquiry is by its very nature a fact-specific inquiry. In determining whether the objective of the law is sufficiently important to be capable of overriding a guaranteed right, the court must examine the actual objective of the law. In determining proportionality, it must determine the actual connection between the objective and what the law will in fact achieve; the actual degree to which it impairs the right; and whether the actual benefit which the law is calculated to achieve outweighs the actual seriousness of the limitation of the right. In short, s. 1 is an exercise based on the facts of the law at issue and the proof offered of its justification, not on abstractions.
 Supreme Court of Canada jurisprudence has also made it clear that in order to properly consider the evidence led in support of the infringement on Charter rights, close attention must be paid to context. The social and economic context in which the impugned law is situated is important to the analysis because it informs the type of proof that the courts may demand of the government in attempting to discharge its burden and the degree of deference which the court should accorded to Parliament’s decision: Thomson Newspapers Co. v. Canada (Attorney General),  1 S.C.R. 877 at paras. 87-88, 159 D.L.R. (4th) 385 [Thomson]; M. v. H. at paras. 78-80). As well, the Supreme Court of Canada has indicated that the appropriate level of deference varies according to the nature of the problem that Parliament is attempting to address and the context in which the limitation on rights is imposed (RJR at para. 135, per McLachlin J.). For example, the courts have given greater deference to the legislatures or Parliament where the law is designed to protect vulnerable groups, where the law is based on complex social-science evidence, where the law is concerned with mediating between competing sectors of society, or where the law is involved in the allocation of scarce resources (see for example R. v. Butler,  1 S.C.R. 452, 89 D.L.R. (4th) 449; RJR at para. 135; Irwin Toy Ltd. v. Quebec (Attorney General),  1 S.C.R. 927 at 994-95, 58 D.L.R. (4th) 577). In such situations, the answer to the problem which the law seeks to address may not be clear and for that reason the courts are reluctant to substitute their view for that of the government. However, it must be remembered that “[d]eference must not be carried to the point of … accepting Parliament’s view simply on the basis that the problem is serious and the solution difficult”. (RJR at para. 136; see also Vriend at para. 54, cited with approval in M. v. H. at para. 78). To do so would be to “diminish the role of the courts in the constitutional process and to weaken the structure of rights upon which our constitution and our nation is founded” (RJR at para. 136).
 In this case, the question of deference must be considered but given that the Supreme Court of Canada has said “the question of deference is not an issue that can be determined … at the outset of the [s.1] inquiry”, I will defer discussion of the appropriate level of deference to later in the analysis (see M. v. H. at para. 80).
(b) The application of the Oakes test
 The first step of the Oakes test requires the respondent to establish that the objective of the law is sufficiently important to justify the infringement on a guaranteed Charter right: RJR at para. 60; R. v. Big M Drug Mart Ltd.,  1 S.C.R. 295 at 352, 18 D.L.R. (4th) 321. At a minimum, the government’s purpose in enacting the law must relate to concerns that are pressing and substantial in a free and democratic society.
 For the purpose of the justification analysis, the Supreme Court of Canada has emphasized that the objective of the law should be stated “as precisely and specifically as possible”: Thomson at para. 98, citing RJR at para. 144; Vriend at para. 110. However, the Supreme Court of Canada has more recently stated that that it is permissible for the government to state its objective in terms of one main broad objective pursued by way of several sub-objectives, so long as the more precise aims of the legislature are made clear: Health Services and Support – Facilities Subsector Bargaining Assn. v. British Columbia, 2007 SCC 27,  2 S.C.R. 391 at para. 146 [Health Services].
 In this case, the respondent submits that the purpose of insurance, and the supplementary death benefit in particular, is “to provide insurance coverage to participants at a reasonable cost” which entails “providing full coverage to those active employees when they need it most – when their wealth is minimal – and still providing some benefit to those retired employees when their wealth is highest.” The respondent contends that ensuring the financial viability of the plans and the competitiveness of the federal public service is “undoubtedly an important and pressing objective.” The objective is restated by the respondent as “providing a competitive benefit package at a reasonable cost to employees while ensuring that the benefit package is sustainable”.
 The appellants argue that cost savings rarely justify a Charter infringement. In support of that submission, the appellants referred to Martin, Reference re Remuneration of Judges of the Provincial Court of Prince Edward Island,  3 S.C.R. 3, 150 D.L.R. (4th) 577 [Reference re Remuneration of Judges], Singh v. Minister of Employment and Immigration,  1 S.C.R. 177, 17 D.L.R. (4th) 422, and N.A.P.E.
 Despite the respondent’s stated objectives, the appellants submit that the “real reason” underlying the Reduction Provisions is to make sure that there is “continued political and public support for compensation of the public service”, a matter alluded to by the respondent. In the appellants’ submission, such an objective is not only “not a compelling” objective, it is an impermissible one.
 What is the actual objective of the law? As Peter Hogg notes, “[t]he identification of the objective of a challenged law is a task of considerable practical and theoretical difficulty”: Constitutional Law of Canada, 5th ed. (Scarborough: Carswell, 2006) at §38.9(a).
 In referring to the purpose of life insurance, the respondent attempts to describe the objective in terms of the overall goal of the legislative scheme. However, the “objective” with which we are concerned here is the objective of the infringing measure, not the purpose of legislation in general, “since it is the infringing measure and nothing else which is sought to be justified.” (RJR at para. 144, per McLachlin J.) For example, the challenged legislation in RJR, the Tobacco Products Control Act, S.C. 1988, c. 20, was enacted as part of a broad legislative and policy scheme to protect Canadians from the health risks associated with tobacco use. However, as McLachlin J. noted, the objectives of the offending provisions, which imposed an advertising ban and restricted trade mark usage, were much more narrow, that is, “to prevent people in Canada from being persuaded by advertising and promotion to use tobacco products” (at para. 144). As stated by one author, “[the court’s] primary focus must remain upon the characterization which most directly relates to the reason for violating the constitutional right.” (C. Sherrin, Objectionable Objectives?: An Analysis of the First Branch of the Oakes Test (D.Jur. Thesis, Osgoode Hall Law School, York University, 1994) at 137, cited in Hogg at §38.9(a))
 The overall goal of the legislation in this case may be to provide a group life insurance scheme, but the objective of the Reduction Provisions is clearly related to cost concerns. This is supported by the expert opinion of Gordon Argue, who stated in his report that “the decision to restrict benefits after normal retirement age is primarily influenced by financial considerations”. Essentially, the purpose of the Reduction Provisions is to save money in providing less benefit to spouses beyond the age of 60 or 65 so that more funds are available to spouses under the age of 60 or 65.
 The issue of whether cost-related concerns can constitute a legitimate objective for the purposes of the justification analysis under s. 1 is contentious. Prior to the Supreme Court of Canada’s decision in N.A.P.E., there seemed to be little room for treating financial considerations as pressing and substantial concerns. In Singh, the court found that the procedures for determining refugee status claims under the Immigration Act 1976, 1976-77 (Can.), c. 52, which denied the claimant an oral hearing, were inconsistent with the principles of fundamental justice and thus incompatible with s. 7 of the Charter. On the matter of s. 1, the government argued that the Immigration Appeal Board was “already subjected to a considerable strain in terms of the volume of cases which it was required to hear” and that if the Board was required to give an oral hearing for every application for redetermination of a claim, it would constitute an unreasonable burden on the Board’s resources (at para. 68).
 In considering the government’s cost-related arguments, Wilson J. speaking for the three members of the Supreme Court of Canada who applied the Charter, said as follows (at para. 70):
…I have considerable doubt that the type of utilitarian consideration brought forward by [the government] can constitute a justification for a limitation on the rights set out in the Charter. Certainly the guarantees of the Charter would be illusory if they could be ignored because it was administratively convenient to do so. No doubt considerable time and money can be saved by adopting administrative procedures which ignore the principles of fundamental justice but such an argument, in my view, misses the point of the exercise under s. 1 … it seems to me that the basis of the justification for the limitation of rights under s. 7 must be more compelling than any advanced in these appeals.
 Further, even if the cost of compliance with the Charter was relevant, Wilson J. was not satisfied that the government has discharged its burden of showing that the cost was “so prohibitive as to constitute a justification within the meaning of s. 1.” (at para. 73).
 Justice Wilson’s reasons in Singh were cited with approval by Lamer C.J. for the Court in Reference re Remuneration of Judges at para. 281. Referring to his earlier decision in Schachter v. Canada,  2 S.C.R. 679 at 709, 93 D.L.R. (4th) 1, that “budgetary considerations cannot be used to justify a violation under s. 1”, Lamer C.J. reiterated that “a measure whose sole purpose is financial, and which infringes Charter rights, can never be justified under s. 1” (at paras. 282-284) (emphasis added).
 In Martin, four principle objectives had been submitted in support of s. 10B of the Workers’ Compensation Act, S.N.S. 1994-95, c. 10, which excluded chronic pain sufferers from receiving the same benefits as other injured workers. The first concern put forth was to “maintain the viability of the Accident Fund set up by the Act to compensate injured workers”, which the Court noted suffered from significant unfunded liability (at para. 108). In considering this objective, Gonthier, J. for the Court stated that while budgetary considerations could not normally be “invoked as free-standing pressing and substantial objective[s]”, they may be relevant at the minimal impairment stage of the analysis if the law was supported by some other non-financial goal. Further, Gonthier J. noted that in some circumstances, controlling expenditures might constitute a legitimate purpose under s. 1, but he found it unnecessary to decide this point (at para. 109, citing Eldridge v. British Columbia (Attorney General),  3 S.C.R. 624 at para. 84, 151 D.L.R. (4th) 577).
 N.A.P.E. is presently the only case in which the Supreme Court of Canada has concluded that under unique circumstances, financial concerns, when combined with other public policy considerations, can form a legitimate objective capable of overriding Charter rights. The infringing measure at issue in N.A.P.E. was a piece of legislation which cancelled arrears and delayed commencement of a pay equity increase that the government of Newfoundland and Labrador had promised, by way of a signed agreement, in favour of female employees in the health care sector. The Court found that at the time the legislation was enacted, it was indisputable that the provincial government was experiencing a “severe fiscal crisis” (at para. 59). The government put forward evidence that showed that in addition to the cuts to pay equity, the government had taken other drastic steps to control its expenditures, including the closing of hundreds of acute care hospital beds, freezing of student grants, termination of medicare coverage for certain items, and the laying off of 2000 employees (at para. 61). With this in mind, Binnie J. stated (at para. 72):
… [The] courts will continue to look with strong scepticism at attempts to justify infringements of Charter rights on the basis of budgetary constraints. To do otherwise would devalue the Charter because there are always budgetary constraints and there are always other pressing government priorities. Nevertheless, the courts cannot close their eyes to the periodic occurrence of financial emergencies when measures must be taken to juggle priorities to see a government through the crisis.
[Italic emphasis in original; underline emphasis added]
 Justice Binnie determined that the purpose behind the legislation was not solely financial. He observed at para. 75:
The government in 1991 was not just debating rights versus dollars but rights versus hospital beds, rights versus layoffs, rights versus jobs, rights versus education and rights versus social welfare. The requirement to reduce expenditures, and the allocation of the necessary cuts, was undertaken to promote other values of a free and democratic society.
[Italic emphasis in original; internal citations omitted]
 Justice Binnie ultimately concluded that the government’s financial concerns were “pressing and substantial”. The fact that the government took other drastic, wide-ranging measures in order to reduce its spending, gave merit to the argument that the cost-related objectives were legitimate.
 Since N.A.P.E., the Supreme Court of Canada has confirmed that objectives related to budgetary constraints will continue to be viewed with “strong scepticism”, but courts should consider whether the government has established other pressing and substantial objectives (Health Services at para. 147).
 It appears to me that N.A.P.E. carved out a very limited exception to the general rule that cost-related concerns cannot form a pressing and substantive objective for the purposes of a s. 1 analysis. In my view, this case does not fall within the exception. This point was addressed succinctly by the appellants:
In NAPE, the Court described the kind of cost savings scenarios that might justify an infringement of rights. The language used there – “serious financial situation,” “drastic circumstances,” “exceptional financial crisis” – shows just how far this case is from that type of situation. In contrast to the serious deficit that the Newfoundland government was experiencing in NAPE, the federal SDB Plans have seen growing surpluses. Indeed, it was against this backdrop that the Reduction Provisions were first challenged in Federal Court where the government did not even raise a s. 1 defence.
[Internal citations omitted; emphasis added.]
 Putting the purpose of the legislation in the best possible light, it may be said that the goal is to provide a sustainable group life insurance scheme at reasonable cost by offering full coverage to those employees who need it most and still providing some coverage to those employees who have a lesser need. If such a cost-related objective were accepted by this Court, it could be considered pressing and substantial. If that were the case, it would then be necessary to consider whether the manner in which the government chose to pursue this objective, by reducing the amount of death benefits beyond the age of 60 or 65, was proportional. However, the problem with stating the objective of the law in this manner is that it does not “supply a reason for infringing the Charter right.” (Hogg at §38.9(a)) In other words, it does not address the objective behind reducing benefits for those over 60 or 65.
 The only other objective stated by the respondent is to “ensure that the Government remains a competitive employer.” In my opinion, that goal is impermissibly vague and broad. Even if it were to be accepted, it would inevitably fail the minimal impairment test.
 With the foregoing in mind, I have concluded that the respondent has not met its burden of establishing that there is a pressing and substantial goal underlying the Reduction Provisions, sufficient to warrant overriding the constitutional right to equality.
 Nevertheless, I will go on to address the other stages for the sake of completeness.
(ii) Rational connection
 At the first stage of the proportionality analysis, the respondent must demonstrate that the infringement on the right to equality guaranteed in the Charter is rationally connected to the legislative goal. So long as the law is not arbitrary, unfair, or based on irrational considerations, it will meet the rational connection requirement (Oakes at 139, 141-142). However, although the threshold is low, it must be remembered that the Oakes case itself failed this stage of the analysis (see also Benner).
 Is reducing benefits to spouses beyond the age of 60 or 65 rationally connected to the goal? If the purpose of the legislation is related to cost-savings, then a reduction in benefits is clearly rationally connected to that goal. However, if the goal is to allocate benefits according to need, then the rational connection is much more tenuous. For the reasons I have given in the course of my s. 15 analysis, I am of the view that it has not been established that younger employees have a greater need for the death benefit than older people.
(iii) Least drastic means
 At the second stage of the proportionality analysis, the respondent must show that the infringing measures impair equality rights as little as reasonably possible in order to achieve the legislative objective. The law need not stand up to a test of perfection. Provided the law falls within a “range of reasonable alternatives” it will not be found to be overbroad (RJR at para. 160; Thomson at para. 118). However, as McLachlin J. cautioned in RJR at para. 160, “if the government fails to explain why a significantly less intrusive and equally effective measure was not chosen, the law may fail” (cited in Thomson at paras. 44 and 118).
 It is true that greater deference to government is given where the law is concerned with the provision of social benefits. In this case, the respondent contends that the Reduction Provisions mediate between conflicting claims of different groups and that deference should be accorded to Parliament’s choice on this basis. In support of that argument, the respondent relies on the following passage from McKinney v. University of Guelph,  3 S.C.R. 229 at 285, 76 D.L.R. (4th) 545:
In assessing proportionality and particularly the issue whether there has been a minimal impairment to a constitutionally guaranteed right, it must be remembered that we are concerned here with measures that attempt to strike a balance between the claims of legitimate but competing social values. In the case of broadly based social measure like these, where government seeks to mediate between competing groups, it is by no means easy to determine with precision where the balance is to be struck.
 The respondent also refers to Sopinka J.’s comments in Egan at para. 64, that the “government must be accorded some flexibility in extending social benefits”.
 In my view, however, it must be remembered that the government “benefit” provided in this case is promised under plans to which the employees themselves contribute. In addition, the evidence shows that the government’s own employer-paid portion of the benefit is relatively low and the employee-paid portion is relatively high when viewed in the context of other benefit plans. With this in mind, the argument that the courts should defer to the government in deciding how to allocate these funds is considerably less convincing.
 Further, in Thomson, the Supreme Court of Canada noted that it is often difficult to distinguish between cases in which the state is acting as mediator between different groups, for which deference would be appropriate, and cases in which the state acts as the singular antagonist, for which deference would not be appropriate (at paras. 88 and 90, referring to RJR at para. 135). In RJR, for example, the law restricted the advertising and promotion of tobacco products as a means of mediating between the tobacco companies’ commercial interests and the public’s interest in reducing advertising-related use of harmful tobacco products. In this case, by contrast, there was no conflict between the two relevant groups, younger employees versus older employees, until the Reduction Provisions take effect. Unlike RJR, the Reduction Provisions were not brought into effect in order to solve a pre-existing problem, which would demand greater deference. In this case, the law itself created the competition between the two groups for limited resources. This point was raised in case comment by Alyn James Johnson, “Abdicating Responsibility: The Unprincipled use of Deference In Lavoie v. Canada” (2004) 42 Alta. L. Rev. 561.
 The degree to which the government’s resources are “limited” is also in question. The appellants argue that as a result of high employee contributions, the plans have accumulated a “significant surplus”. In that regard, Mr. John Christie gave the following testimony:
Well, as of 2004, the amount in the account is $2.1 billion. The reserve that would need to be kept there for future paid-up benefit is of the order of $400 million. So you could say that the surplus was of the order of $1.6, $1.7 billion.
 The existence of a surplus – and one as substantial as this – casts doubt on the contention that the Reduction Provisions impair rights as little as is reasonably possible. The appellants suggest that if employer contributions were increased and the excessive surplus reduced by a change in benefit levels, the Reduction Provisions could be eliminated without threatening the financial viability of the overall plan. The only difference between the evidence of the parties, the appellants claim, is in the amount of time for which the plans would be sustainable, with the appellants’ expert saying 20 years and the respondent’s expert saying 65 years.
 This case is not the first time that alternatives to the Reduction Provisions have been raised. In 1992, when Parliament was considering Bill C-55, to amend the PSSA, the Public Service Alliance of Canada submitted its views to a legislative committee mandated to study that Bill and recommended the reductions should be eliminated “given that the reserves in the supplementary death benefit accounts are substantial”. The recommendations were not accepted. However, on cross-examination, Ms. Joan Arnold, a witness called by the respondent, admitted that at the time the amendments were proposed in 1992, there was no discussion amongst Treasury Board staff for the rationale of the Reduction Provisions beginning at age 60. No studies were done as to the comparative needs of spouses of contributors who died over the age of 60 versus under the age of 60. Finally, no one examined the needs of spouses whose partners died over the age of 60 and were in receipt of government programs (such as the pension programs, old age security, and CPP) as compared to the needs of spouses whose partners died under the age of 60 and were not in receipt.
 Other alternatives have also been suggested, namely, that the Reduction Provisions should begin at retirement rather than at some arbitrarily-defined age. The appellants also put forth the suggestion that instead of having the Reduction Provisions triggered at retirement, they could begin at the later of retirement or a certain age, thus “tying the reduction primarily to employment status, and only using age to extend the benefit.” This suggestion was recommended in 1982 by Mr. Hart Clark and proposed by senior Treasury Board staff in 1998, yet no studies were ever done to calculate the cost of changing the Reduction Provisions to begin at retirement, rather than age. I note that the respondent appears implicitly to acknowledge that a reduction in death benefits makes more sense when linked to employment status rather than age, as it bases a number of its arguments on a distinction between “active” versus “retired” employees.
 In my view, the respondent has failed to meet its burden of showing that the Reduction Provisions are minimally impairing.
(iv) Proportionate effect
 At the final stage, this Court must consider the salutary effects compared to the deleterious effects of the law. The respondent argues that the deleterious effect of the Reduction Provisions is that the plan participants have “less comfort or assurance” than without the reduction, but that this is outweighed by the benefit of providing life insurance to all employees, particularly the newer employees. The appellants, on the other hand, point to the fact that the Reduction Provisions deny to those over the age of 60 or 65 the full benefit “that would have helped them cope, with dignity, with … the most stressful and painful events of their lives – the last illness and death of their spouse.”
 Considering the surplus, the actual benefit of the Reduction Provisions is questionable. The harm that the Reduction Provisions have caused, however, is much more tangible and extends beyond mere “discomfort”.
(c) Conclusion on the s. 1 analysis
 For the foregoing reasons, I am of the view that the s. 15 violation cannot be upheld under s. 1 in a free and democratic society. A reasonable explanation for the infringement will not suffice; there must be a legitimate justification.
 For the reasons stated I would allow the appeal and set aside the order dismissing the appellants’ class action.
 In my view, this is a case in which the government must be afforded the opportunity to consider how the impugned legislation ought to be amended so that it does not infringe s. 15 of the Charter. The monetary remedy sought by the appellants is one that requires careful analysis and would best be undertaken after the amendments have been made. I consider that eight months would be an adequate time in which to consider the amendments and the monetary compensation to be afforded to the appellants as a result.
“The Honourable Madam Justice Rowles”
Reasons for Judgment of the Honourable Madam Justice Ryan:
 I have had the privilege of reading in draft the reasons of my colleague Madam Justice Rowles. I regret that I am unable to agree with her conclusion that this appeal should be allowed. I have concluded that the trial judge did not err in law in her definition or use of the comparator group put forward by the appellants. In my view the resolution of this appeal turns on the finding of facts and the characterization of those facts by the trial judge. Those findings were based in the evidence presented at trial and cannot be said to be unreasonable. The assessment of those facts by the trial judge is not a matter for appellate intervention. In the result I would dismiss the appeal.
 My colleague has set out the essential facts and legislation involved in the appeal. I will only repeat the facts and law necessary to these reasons.
 This is a class action brought to challenge s. 47(1) of the Public Service Superannuation Act, R.S.C. 1985, c. P-36 (“PSSA”), s. 60(1) of the Canadian Forces Superannuation Act, R.S.C. 1985, c. C-17 (“CFSA”) and their attendant regulations. As set out in the reasons of Madam Justice Rowles, these provisions create a “supplementary death benefit” which is paid on the death of a plan member to a named beneficiary of an amount equal to twice the salary of the plan member. However, once the CFSA plan member attains the age of 60, and in the case of a PSSA plan member the age of 65, the benefit payable begins to reduce by ten percent per year to the date of death. The parties have referred to these provisions compendiously as the “Reduction Provisions”.
 The appellants are all surviving spouses or partners of plan members to whom the Reduction Provisions applied. In these reasons I will refer to the appellants interchangeably as “the appellants”, “the claimants”, or “the claimant group”.
 The appellants alleged that the Reduction Provisions target age as a determinate, not for any legitimate reason, but on the basis of age stereotyping – specifically because the elderly either do not need the benefit or are not worthy of receiving it. Their position is that in reality they are in need of the death benefit as much if not more than the partners of younger deceased plan members who receive the full benefit and that its denial to them perpetuates the discrimination forbidden by the equality provisions of the Charter. Focusing on the affront to human dignity that results from such discrimination, Campbell J. summed up their position in the case of Margolis v. Canada, 2001 FCT 85,  F.C.J. No. 402, in this way, at para. 31:
… The effect of the message being communicated to the participant in such a plan over the age of 60 is that, compared to younger participants, your death, and the financial problems to be experienced by your surviving spouse on your death, are not as worthy of concern and support. …
 The respondent argued that the Reduction Provisions, while contingent on age, differentiated, but could not be said to discriminate (within the meaning of the Charter) because they are a part of a statutory group insurance and pension package designed to meet the changing financial requirements of employees as each progresses through the workforce from young worker to retiree. The trial judge agreed in essence with the respondent.
 The record shows that the parties to this litigation governed themselves by the analysis set out by the Supreme Court of Canada in Law v. Canada (Minister of Employment and Immigration),  1 S.C.R. 497 (“Law”). The action was argued using the Law criteria and the trial judge faithfully followed it in organizing her reasons for judgment.
 The Law case sets out four contextual factors which should assist a court in determining whether the law or benefit in question has a discriminatory impact. On this appeal the appellants submit that the trial judge misapplied each of those four contextual factors. While not resiling from any of his other arguments, Mr. Arvay, counsel for the appellants, emphasized in oral argument that the second contextual factor – the question of the correspondence between the ground on which the claim is based and the actual need, capacity or circumstances of the claimants – was really the crux of the appeal. It was also the crux of the claimants’ case before the trial judge. Through the testimony of both experts and individual claimants the appellants attempted to show that the cost of final illness and death increased with age, that financial well-being declines with age, that employment is more difficult to find, and that to decrease a death benefit simply on the basis of age, at a time when it is most needed, is unreasonable and discriminatory. The trial judge either did not accept these assertions or found that they were offset by other benefits. The appellants say that the trial judge reached the wrong conclusions on these points because she failed to properly analyze the contextual factors, and consistently applied the wrong comparator group.
 As I will develop in these reasons, I am of the view that the conclusions reached by the trial judge were not unreasonable and were not the result of faulty reasoning or use of the wrong comparator group.
The Test for Determining a Section 15 Violation
 After this claim was heard, and judgment on the appeal was reserved in this Court, the Supreme Court of Canada released its decision in R. v. Kapp, 2008 SCC 41 (“Kapp”). In that case the majority commented, in dicta, on the Court’s earlier decision in Law.
 The majority reasons in Kapp (McLachlin C.J.C. and Abella J.) began with a return to the language of that court’s seminal decision in Andrews v. Law Society of British Columbia,  1 S.C.R. 143 (“Andrews”), which established that s. 15 of the Charter is designed to promote substantive, rather than formal, equality and that, as stated in Andrews, “the main consideration must be the impact of the law on the individual or group concerned”.
 The majority in Kapp went on to observe that in Andrews, McIntyre J. writing for the Court viewed the notion of discriminatory impact through the lens of two concepts: (1) the perpetuation of prejudice and disadvantage to members of a group on the basis of personal characteristics identified in the enumerated and analogous grounds; and (2) stereotyping on the basis of these grounds that results in a decision that does not correspond to a claimant’s or group’s actual circumstances and characteristics. The majority noted that Andrews was decided on the basis of the second concept. In my view the second concept is what is largely, though not solely, at issue in the case at bar – did the government deny the plan members a benefit (here the full supplementary death benefit) by creating a plan that does not correspond to the claimants’ actual circumstances and characteristics?
 The majority in Kapp also made reference to the conclusion in Andrews that a finding of discrimination might be grounded in the fact that a particular law or program perpetuated the disadvantage of a group defined by enumerated or analogous s. 15 grounds. In that context McIntyre J. (at 174) had defined “discrimination” as:
… [D]iscrimination may be described as a distinction, whether intentional or not but based on grounds relating to personal characteristics of the individual or group, which has the effect of imposing burdens, obligations, or disadvantages on such individual or group not imposed upon others, or which withholds or limits access to opportunities, benefits, and advantages available to other members of society.
 Subsequently, as the majority observed in Kapp, the Court in Law shifted the focus of the discriminatory impact inquiry somewhat. The Court in Law suggested that discrimination should be defined in terms of the impact of the law or program on the “human dignity” of members of the claimant group having regard to four contextual factors. That said, Law did not retreat from the proposition, implicit in Andrews, that the aspect of human dignity protected by s. 15 of the Charter consisted of protection from the imposition of burdens or the withholding of benefits based on irrelevant or attributed personal characteristics. Indeed in his discussion of the meaning of “human dignity” in Law, Iacobucci J. said this at para. 53:
… Human dignity means that an individual or group feels self-respect and self-worth. It is concerned with physical and psychological integrity and empowerment. Human dignity is harmed by unfair treatment premised upon personal traits or circumstances which do not relate to individual needs, capacities, or merits. …
 But in shifting the discussion to the more abstract concept of human dignity and the suggestion that its diminution, if any, might be measured by the four contextual factors, the inquiry to determine discriminatory impact became more subjective and theoretically difficult.
 The majority in Kapp recognized the several criticisms made of Law in the academic literature and in some of the subsequent case law, particularly the conceptual difficulties that have arisen from the attempt in Law to employ human dignity as a legal test. The majority began by noting that human dignity is an essential value underlying not only s. 15 but all of the rights guaranteed by the Charter. They observed that even with the guidance of the four contextual factors set out in Law, the abstract and subjective notion of human dignity can not only become confusing and difficult to apply, but also becomes an additional burden on equality claimants, “rather than the philosophical enhancement it was intended to be” (para. 22).
 The majority in Kapp went on to say that “[c]riticism has also accrued for the way Law has allowed the formalism of some of the Court’s post-Andrews jurisprudence to resurface in the form of an artificial comparator analysis focussed on treating likes alike” (at para. 22). They finished their remarks with this:
 The analysis in a particular case, as Law itself recognizes, more usefully focuses on the factors that identify impact amounting to discrimination. The four factors cited in Law are based on and relate to the identification in Andrews of perpetuation of disadvantage and stereotyping as the primary indicators of discrimination. Pre-existing disadvantage and the nature of the interest affected (factors one and four in Law) go to perpetuation of disadvantage and prejudice, while the second factor goes to stereotyping. The ameliorative purpose or effect of a law or program (the third factor in Law) goes to whether the purpose is remedial within the meaning of s. 15(2). (We would suggest, without deciding here, that the third Law factor might also be relevant to the question under s. 15(1) as to whether the effect of the law or program is to perpetuate disadvantage.)
 Viewed in this way, Law does not impose a new and distinctive test for discrimination, but rather affirms the approach to substantive equality under s. 15 set out in Andrews and developed in numerous subsequent decisions. The factors cited in Law should not be read literally as if they were legislative dispositions, but as a way of focussing on the central concern of s. 15 identified in Andrews – combatting discrimination, defined in terms of perpetuating disadvantage and stereotyping.
 In the result I think it can be said that although Law broadened the focus of the discrimination inquiry, to its larger abstract meaning, it also proposed a very specific four-part test centred on the primary indicators of discrimination (the perpetuation of disadvantage and stereotyping) to assist in determining whether the government action has a discriminatory impact. Kapp has reminded the courts that the essential question under s. 15(1) is whether the distinction created by the impugned legislation is based on personal characteristics of an individual or group that has the effect of imposing burdens, obligations, or disadvantages on such individual or group not imposed on others, or which withholds or limits access to opportunities, benefits, and advantages available to other members of society. In making this inquiry, the legislation must be placed in context which the four factors described in Law may help to delineate.
The Test Applied by the Trial Judge and the Grounds of Appeal
The Application of the Test in Law by the Trial Judge
 As I mentioned earlier, the trial judge closely followed the test in Law. She concluded that the first two steps in the test had been met, namely (in the language of Law), (a) that the impugned benefit drew a distinction between the claimant(s) and others on the basis of one or more personal characteristics, here, age of their partners, and (b) that the differential treatment was based on an analogous or enumerated ground. This left the third inquiry, that is, whether the differential treatment discriminated by “withholding a benefit from the claimant[s] in a manner which reflects the stereotypical application of presumed group or personal characteristics, or which otherwise has the effect of perpetuating or promoting the view that the individual is less capable or worthy of recognition or value as a human being or member of Canadian society, equally deserving of concern, respect and consideration” (Law, para. 88). To determine this question, the trial judge looked at the four contextual factors. It is in her analysis of the factors that the appellants say she erred.
The Grounds of Appeal
 The appellants articulate only one generalized ground of appeal in their factum – “that the trial judge erred in holding that the Reduction Provisions do not violate s. 15 of the Charter.” In their argument, however, the appellants took issue with the way in which the trial judge approached each of the four contextual factors described in Law. As best as I can extract them, the points are:
a. The learned trial judge erred in finding that the claimant group was not the subject of pre-existing disadvantage because:
i. her decision was based on the circumstances of the claimant group whereas at this stage of the analysis the appellants needed only to show, which they did, that the elderly suffer from real disadvantage and stereotyping; and
ii. the trial judge was inconsistent in comparing the claimant group to all persons over 65 years of age, and/or the comparator group.
b. The trial judge erred in finding that there was a lack of correspondence between the enumerated ground on which the claim is based and the nature of the differential treatment because:
i. her finding that the reducing nature of the death benefit was offset by the pension and other government benefits was wrong. The evidence showed in fact that there was little difference in the benefits the claimant group and the comparator groups received, and
ii. the evidence showed that the medical and financial needs of the elderly are greater than those of younger people.
c. The trial judge erred in her assessment of the ameliorative purpose of the benefits in question, in particular that she erred in characterizing the supplemental death benefit as “group life insurance”.
d. The trial judge erred in analyzing the nature of the benefit in question. In particular she erred in concluding that the disadvantage of the claimants was not severe since she concluded that they were better off than most seniors; and
e. The trial judge erred in failing to analyze the issue of human dignity because she did not discuss how a reasonable person would legitimately feel or act if placed in the same position as the claimants.
 For reasons which will be explained presently, I do not find it necessary to deal with all of the appellants’ arguments.
Critical Findings of Fact
 As I mentioned earlier, the trial judge began her analysis of the appellants’ claim under s. 15 with a finding that the provisions in question drew a distinction between surviving spouses receiving the supplementary death benefit on the basis of a personal characteristic of their deceased spouse – age. She then found that the resulting differential treatment was based on an analogous ground. This took the trial judge to the third stage of the Law analysis – determining whether this differential treatment on the basis of age constituted discrimination within the meaning of s. 15. Before turning specifically to the four contextual factors, the trial judge addressed the composition of the comparator group and then the legislative context and purpose of the legislation in question. It was in this portion of her reasons, dealing with the context and purpose of the legislation, that the trial judge made her critical findings of fact. I will turn to those findings before returning to the question of the comparator group and the four contextual factors.
 In my view the critical findings of fact are these:
1. The supplementary death benefit was created in the early 1950s when employer-sponsored group life insurance had become popular. Group life insurance was popular because of its broad coverage without medical examination at a modest and stable cost from the first day of employment.
2. One of the problems that group life insurance addressed was the need of junior civil servants for term insurance before their superannuation credit was built up.
3. For younger employees the supplementary death benefit is designed to act as group life insurance to insure unexpected death. Younger civil servants who die unexpectedly may not have built up their superannuation benefits.
4. For older, retired employees, the reduced supplementary death benefit is designed to cover, to some extent, the costs of last illness and death.
5. The supplementary death benefit is a part of a package of employment benefits provided in the PSSA and the CFSA. For example:
a. The PSSA provides a survivor’s pension upon the plan member’s death, generally one-half of the plan member’s unreduced pension.
(Survivors’ pensions are defined benefit plans paying 50% of the average salary of the plan participants in the five years prior to retirement. They are indexed, adjusted annually and backed by the solvency of the government of Canada).
b. The PSSA provides the Public Service Health Care Plan, which covers plan members and their dependants. It supplements provincial and territorial health care plans and provides 80% reimbursement for extended health expenses.
c. The PSSA provides the Public Service Dental Care Plan, which covers plan members as well as their spouses and dependants.
(Coverage under these two health plans continues as long as the plan member is employed or receiving a pension benefit. Survivor spouses are also covered if they are receiving survivor benefits).
6. Considering the entire package of benefits available to surviving spouses of any age, the supplemental death benefit is not intended to be a long-term stream of income for older surviving spouses because their long-term income security is guaranteed by the surviving spouse portion of their defined benefit pension plan.
The Comparator Group
 As the Chief Justice reiterated in Auton (Guardian ad litem of) v. British Columbia (Attorney General),  3 S.C.R. 657 at para. 51, the choice of the correct comparator is crucial, since the comparison between the claimants and this group permeates every stage of the analysis. At para. 53 the Chief Justice said:
… the comparator group should mirror the characteristics of the claimant or claimant group relevant to the benefit or advantage sought, except for the personal characteristic related to the enumerated or analogous ground raised as the basis for the discrimination: Hodge, supra, at para. 23. The comparator must align with both the benefit and “the universe of people potentially entitled” to it and the alleged ground of discrimination: [citation omitted].
 The comparator group proposed by the claimants in this case was “all civil servants and members of the armed forces who received the full supplementary death benefit, not reduced on the basis of age.” The appellants now say, for the first time on appeal (although as Madam Justice Rowles has pointed out, not as a discrete ground of appeal), that the trial judge should have refined the comparator group to consist only of those who were in receipt of both the full supplementary death benefit and a survivor’s pension.
 I would not give effect to this ground of appeal. In the case at bar, the universe of those potentially entitled to the benefit are the named beneficiaries, who were in a spousal or spousal equivalent relationship with a plan member at the time of the plan member’s death. The appellants assert their claim of discrimination solely on the basis of age, that is, that they say it was discriminatory that their supplementary death benefit was reduced on the basis of the age of their partners at death. While the personal circumstances of each survivor are obviously not the same, they all share (or “mirror”) the one characteristic relevant to the benefit sought. That characteristic is their entitlement to survivor benefits as the widows or widowers of plan members. In other words, they have only one difference relevant to the entitlement to the benefit – the age of their deceased partners. To narrow the comparator group to a subset of the larger group would deprive the court of the ability to fully analyze whether it can be said that the distinction made by the legislation resulted in discrimination within the meaning of s. 15 of the Charter.
The Other Grounds of Appeal
 The other grounds of appeal relate to the way in which the trial judge analyzed the four contextual factors proposed by the court in Law. Again, those factors are:
1. Pre-existing disadvantage, stereotyping, prejudice, or vulnerability experienced by the individual or group at issue.
2. The correspondence or lack thereof, between the ground or grounds on which the claim is based and the actual need, capacity, or circumstances of the claimant or others.
3. The ameliorative purpose or effects of the impugned law upon a more disadvantaged person or group in society.
4. The nature and scope of the interest affected by the impugned law. The more severe and localized the consequences of the legislation for the affected group, the more likely that the differential treatment responsible for these consequences is discriminatory within the meaning of s. 15(1).
The First Contextual Factor
 Under the first contextual factor the appellants attempted to show that the elderly suffer from the disadvantage of poverty and stereotyping. As to economic disadvantage the trial judge concluded that:
 Much of the evidence I heard from Professors Forget and Chaykowski about poverty of the elderly was largely irrelevant to the plaintiff group because they, as a group, are better off than the statistical average. But even using the average income and wealth for elderly Canadians it can no longer be said that a majority of elderly Canadians are a disadvantaged group in Canada compared to all other age groups. I conclude that the plaintiffs have failed to prove that the elderly suffer from stereotyping, prejudice, or vulnerability based on their reduced income. The plaintiffs have failed to prove that, as a group, they suffer from pre-existing disadvantage, stereotyping, prejudice or vulnerability based on their economic well-being.
 The appellants argue that in reaching her conclusion, the trial judge was inconsistent in comparing the claimant group to all persons over 65 and/or the comparator group. In fact the reasons mentioned all of these groups. In the end, however, the trial judge accepted the evidence adduced by the Attorney General that it cannot be said that the majority of elderly Canadians are a disadvantaged group compared to other age groups. This finding was based on the evidence called at trial which I need not canvass here. The finding, based on the evidence, cannot be said to be unreasonable.
 The appellants complain that the trial judge erred in analyzing the first contextual factor by taking into account the circumstances of the claimant group. They say at this stage it should have been enough to show that the elderly suffer from real disadvantage and stereotyping. However, as the reasons for judgment demonstrate in the paragraph I have just quoted, the trial judge did make findings relating to the majority of elderly Canadians.
 I would not accede to the grounds relating to the first contextual factor.
The Second Contextual Factor
 As I mentioned early in these reasons, the correspondence between the analogous ground and the nature of the differential treatment was the crux of the appellants’ case. Here, the appellants say that the trial judge was wrong to take into account the pension and other government benefits the claimant group received because there was little difference between what the claimants received and what the other survivors received by way of benefits. In my view this misses the point the trial judge was making. The argument was that the reduction in the supplementary death benefit did not take into account the fact that while the claimant group might be in receipt of a pension, the members of the group would still need the money provided by the supplementary death benefit to pay for such things as the increased cost of health care. The trial judge accepted that the costs of health care were higher for the elderly but, since other aspects of the employment package were intended to assist with a percentage of such costs, the supplementary death benefit did not need to take that fact into account.
 I would not accede to this ground of appeal.
The Third Contextual Factor
 Under this ground, the appellants submit that the trial judge erred in finding that the supplemental death benefit was a form of group life insurance. I will not rehearse the evidence on this point. It is fully set out in the reasons for judgment of the trial judge. There was evidence on which the trial judge could reach this conclusion. It was not unreasonable or unsupported by the evidence. I would dismiss this ground of appeal.
The Fourth Contextual Factor
 The appellants say that the trial judge erred in analyzing the nature of the interest affected by the impugned law, and in particular, erred in finding that the disadvantage was not severe since the appellants were better off than most seniors. Again, it was not wrong for the trial judge to reach this conclusion on the basis of the evidence before her. I would not accede to this ground of appeal.
The Failure to Address the Question of Human Dignity
 The appellants say that the trial judge erred in failing to consider how a reasonable person would legitimately feel or act if he/she were placed in the same position as a claimant. Although the trial judge did not directly address this issue, it is clear that implicit in her reasons she concluded that a similarly situated person in the position of the appellants would understand that the package of benefits received by his or her partner was designed to benefit them in different ways at different times in their life cycle. The reasonable person would also understand that crafting the correct combination of benefits to meet the needs of current and retired employees across the broad sweep of age, sex, family status and residence is a daunting task and that such a scheme could not be expected to meet the needs of everyone in all circumstances. In context, they would not reasonably consider the Reduction Provisions to be discriminatory.
 This case demonstrates the difficulty that arises when one attempts to isolate for criticism a single aspect of a comprehensive insurance and pension package designed to benefit an employee’s different needs over the course of his or her working life. The trial judge concluded that, viewed in context, the supplemental death benefit was the part of a larger scheme comprised of group insurance and pensions designed to look after the changing needs of an employee as he or she remained in the workforce and then retired. At the younger ages, the supplementary death benefit provided a limited stream of income for unexpected death where the surviving spouse is not protected by a pension. At older ages, the purpose of the supplementary death benefit is for expenses associated with last illness and death. The comprehensive plan, while not a perfect fit for each individual, did not meet the hallmarks of discrimination given that it was a broad-based scheme meant to cover the competing interests of the various age groups covered by the plan.
 I cannot say that in reaching the conclusions that she did, the trial judge erred in law or in fact that would warrant appellate intervention.
 I would dismiss the appeal.
The Cross Appeal
 At trial and in this Court the respondent maintained that the appellants did not have standing to challenge the benefits they received. The trial judge rejected that position. The respondents have cross-appealed on that issue.
 In my view a cross appeal was unnecessary in this case. The respondent was entitled to support the order made by the trial judge (that the action be dismissed) on any basis founded in the evidence. (Perka v. The Queen,  2 S.C.R. 232 at 240; Kehler v. Surrey (District) (1992), 70 B.C.L.R. (2d) 381 (C.A.) at para. 4.)
 In any event, given my conclusions as to the substantive issues on the appeal, it is not necessary to consider the issues raised on the cross-appeal. I would dismiss it.
“The Honourable Madam Justice Ryan”
“The Honourable Madam Justice Newbury”