Time for Buy-Back: Supreme Court Set to Hear Important Adverse Effects Discrimination Case

By: Jennifer Koshan and Jonnette Watson Hamilton

PDF Version: Time for Buy Back: Supreme Court Set to Hear Important Adverse Effects Discrimination Case

Case Commented On: Fraser v Canada (Attorney General), 2018 FCA 223 (CanLII), leave to appeal granted, 2019 CanLII 42345 (SCC)

In December, the Supreme Court of Canada will hear an appeal in an equality rights challenge under section 15(1) of the Canadian Charter of Rights and Freedoms. Several female members of the Royal Canadian Mounted Police argue that their employer’s pension rules – which denied pension buy-back rights to those who were job-sharing – discriminated against them based on their sex and family or parental status. The case is a classic example of adverse effects discrimination, involving a claim that a law or policy that is neutral on its face has an adverse impact on the basis of grounds protected under section 15(1). In this post we will review the Federal Court and Federal Court of Appeal decisions rejecting the women’s claim to set the stage for the upcoming appeal at the Supreme Court.

We have high hopes for this appeal in light of the Supreme Court’s May 2018 pay equity decisions, which were the Court’s first substantive rulings in favour of women’s equality under section 15(1) (see Quebec (Attorney General) v Alliance du personnel professionnel et technique de la santé et des services sociaux, 2018 SCC 17 (CanLII)(APP) and Centrale des syndicats du Québec v Quebec (Attorney General), 2018 SCC 18 (CanLII) (Centrale des syndicats du Québec) (CSQ) and our posts on those decisions here and here). However, the chances of victory for the claimants in Fraser are by no means certain because the Court has historically struggled with adverse effects discrimination claims based on sex (or other grounds). Indeed, the lower court judgments in this case display several of the typical problems that courts have had in previous adverse effects cases. As we wrote about in Adverse Impact: The Supreme Court’s Approach to Adverse Effects Discrimination under Section 15 of the Charter, these problems include higher evidentiary and causation requirements, assumptions about the claimants’ “choices”, formalistic comparator analyses, a focus on the “neutrality” of government policies, and a narrow definition of discrimination. In our view, it is time for the Court to buy back into the concept of adverse effects discrimination, something a majority of the Court has not done in the past 20 years – not since Eldridge v British Columbia (Attorney General), [1997] 3 SCR 624, 1997 CanLII 327 (SCC) and Vriend v Alberta, [1998] 1 SCR 493, 1998 CanLII 816 (SCC)).

Facts

The claimants were female members of the RCMP who participated in job-sharing under their employer’s policy in order to work reduced hours while their children were young. The RCMP Pension Plan did not deal specifically with job-sharing arrangements; rather, it treated the claimants as part-time workers. The Plan counted part-time and full-time years of service in an equivalent way for calculating years of pensionable service. However, the claimants’ pension contributions and benefits during their periods of job-sharing were based on the part-time hours they worked, and they were not entitled to contribute, or “buy-back”, the hours they would have worked if they had not been job-sharing (see Royal Canadian Mounted Police Superannuation Act, RSC 1985, c R-11 (RCMPSA) and Royal Canadian Mounted Police Superannuation Regulations, CRC, c 1393 (Regulation)). Their situation differed from that of RCMP members who took leave without pay (LWOP) for three months or longer, as that leave was fully pensionable based on the hours they regularly worked immediately prior to taking leave, if they returned to work and if they made additional contributions to their pension for the leave period. Had the claimants taken unpaid care and nurturing leave instead of job-sharing, they would have been eligible to buy back their pension benefits, rather than having those benefits and their ultimate retirement income reduced on account of their part-time hours.

The claimants argued that their inability to make additional contributions to their pensions based on the full-time rate for the time they were job-sharing deprived them of the equal benefit of the law based on their sex and family or parental status, contrary to section 15(1) of the Charter. They applied to the Federal Court, seeking declarations that the relevant portions of the RCMPSA and Regulation were invalid and to have their entitlement to pension buy-back rights read into the legislation.

The evidence at the Federal Court hearing (reported at 2017 FC 557 (CanLII)) consisted of affidavits as well as expert evidence. As explained by the Court of Appeal, the claimants’ affidavits described why they “chose to job-share” and “the affront they felt” from having their pensions reduced for the period of job-sharing, which showed “a lack of appreciation for female RCMP officers who chose to have children” (at para 15, emphasis added). The Court of Appeal noted that while there was some evidence of what a pension buy-back cost for one of the claimants who took a three year LWOP ($24,000) and evidence estimating the reduction in pension benefits for another job-sharing claimant (a 5% reduction), there was no evidence “comparing the pecuniary value of the job-sharing arrangement, inclusive of the reduced pension treatment, with the pecuniary value of an equivalent period of leave without pay” (at para 16). As Justice Catherine Kane put it at the Federal Court: “For those on LWOP who did not work elsewhere at all during that period, there would be economic disadvantages that must be considered, even if the member on LWOP has the option to buy-back their pension benefits” (2017 FC at para 129).

The claimants’ expert provided evidence that working women in Canada bear a disproportionate burden of child-rearing and face “role overload” given the competing demands on their time (2018 FCA at para 19). He was of the view that these issues “may be particularly acute for women in policing, and most especially for those who work in rural and isolated areas with limited access to child care” (at para 19). However, the Court of Appeal noted that there was no evidence as to the overall percentage of RCMP officers who were women or mothers, nor of the sex or parental status of those working part-time or taking LWOP. Although a government witness provided evidence that in May 2010 and again in May 2014, all the RCMP members who were job-sharing were women, the Court of Appeal highlighted that “for many of them the reasons for job-sharing were reported to be unrelated to the need to care for young children” (at para 18). This differs from the Federal Court’s finding that “a significant majority” of the women in 2010 and 2014 cited child care as their reason for job-sharing (2017 FC at para 81).

In the Federal Court, Justice Kane dismissed the claimants’ application, holding that the challenged provisions did not violate section 15(1) of the Charter. The Federal Court of Appeal agreed with the conclusion reached by Justice Kane. However, in dismissing the claimants’ appeal, Justices Mary Gleason, Johanne Gauthier and Judith Woods disagreed with some aspects of Justice Kane’s reasoning. In particular, they noted that her judgment “tend[ed] to conflate the two steps in the section 15 analysis and [did] not squarely grapple with the requirements for a claim of adverse impact discrimination” (at para 37). We will summarise the Federal Court of Appeal decision first, as it only dealt with the first step of the test for section 15(1), followed by a summary of the Federal Court’s decision on the second step.

Federal Court of Appeal Decision

Writing for the Court of Appeal, Justice Gleason stated the test for discrimination by relying (amongst other cases) on the Supreme Court’s 2018 pay equity decisions, APP and CSQ. According to the Justice Gleason, the section 15(1) test requires:

… first, determination of whether the impugned law on its face or in its impact creates a distinction based on a ground enumerated in section 15 of the Charter or on analogous ground, and second, determination of whether such distinction imposes a burden or denies a benefit in a manner that has the effect of reinforcing or perpetuating prejudice or disadvantage (at para 39).

As mentioned above, the Court of Appeal decision dealt only with the first part of this test. Justice Gleason began by noting that, unlike the pay equity decisions which involved claims of direct discrimination, the case at hand concerned rules that were neutral on their face with respect to the grounds raised by the claimants. Any distinctions drawn by the RCMP Pension Plan were based on hours worked or whether the employee was on leave without pay – matters related to employment status, which is not a protected ground under section 15(1) (at para 41).

Turning to adverse effects discrimination, and relying on human rights case law, the Court of Appeal said that “two things are required” to establish this form of discrimination: “demonstration of adverse treatment as compared to others and demonstration that such treatment results from the particular characteristics that the protected group possesses” (at para 43, citing Ontario Human Rights Commission v Simpson-Sears, 1985 CanLII 18 (SCC), [1985] 2 SCR 536). The Court saw this test as consistent with that for adverse effects discrimination under the Charter, stating that Kahkewistahaw First Nation v Taypotat, 2015 SCC 30 (CanLII), [2015] 2 SCR 548– the Supreme Court’s most recent adverse effects discrimination decision – required prima facie proof of adverse or disproportionate impact of the law on the claimant group at step one of the test. Put in terms of the case at hand, the claimants “were required to show that [the pension rules] negatively impact them in a disproportionate way and that such impact [was] due to a protected or analogous ground” (at para 47).

Applying this test, the Court of Appeal noted the finding of Justice Kane that the claimants failed to prove the impugned provisions “had any negative impact at all” (at para 48). They agreed that it was important to consider the provisions in the overall context of the remuneration package offered to employees who were job-sharing and those who were on leave without pay. Comparing employees in these two categories, the Court remarked on the lack of evidence of the “relative value of the two packages” such that it was “impossible to conclude that job-sharing is adverse to being on a leave without pay” (at para 50).

Even if the complainants had proven adverse impact, they failed to show that it was based on sex and family or parental status. According to the Court of Appeal, “there was no evidence … to suggest that the option of a leave without pay was unavailable (either actually or practically) to female RCMP members who had young children. Nor was there any evidence to suggest that more men than women or more childless individuals than those with children had opted to take leaves without pay” (at para 52). In spite of the evidence that a significant majority of female RCMP officers cited child care as their reason for job-sharing, the claimants “were not denied buy-back rights based on their personal characteristics of being female RCMP members with young children, but rather because they elected to job-share as opposed to taking care and nurturing leave” (at para 53). Put another way, “[w]e must take care to distinguish between effects which are wholly caused, or are contributed to, by an impugned provision, and those social circumstances which exist independently of such a provision” (at para 54, quoting from Miceli-Riggins v Canada (Attorney General), 2013 FCA 158 (CanLII), which in turn relied on Symes v Canada, 1993 CanLII 55 (SCC), [1993] 4 SCR 695).

The claim therefore failed at the first step of the test for discrimination, as it had before Justice Kane in the lower court. The Court of Appeal closed by recognizing “the very real and significant challenges working mothers face, especially in male-dominated workplaces.” However, they went on to state that “this social reality does not give rise to a constitutional right to increased pension benefits in the absence of discrimination” (at para 61). Instead, it was up to Parliament to decide whether to provide the claimants with pension buy-back rights.

Federal Court Decision

Because the Court of Appeal did not consider the second step of the test for discrimination, we will summarise the Federal Court decision on the second step. However, caution is necessary in discussing the Federal Court judgment for two reasons. First, the Court of Appeal articulated the test differently than did the lower court, in part because the Supreme Court’s pay equity decisions were not rendered until after the lower court’s decision, and it was the pay equity decisions that clarified that the Supreme Court had adopted a new test for section 15 in its decisions subsequent to R v Kapp, 2008 SCC 41 (CanLII), [2008] 2 SCR 483 at para 17 and Withler v Canada (Attorney General), 2011 SCC 12 (CanLII), [2011] 1 SCR 396, the two decisions that Justice Kane relied upon. Second, and as the Court of Appeal noted, Justice Kane had conflated the two steps of the section 15 analysis (2018 FCA at para 37).

Justice Kane noted the parties’ agreement that the applicable test was articulated by the Supreme Court of Canada in Kapp, at para 17 and at para 30:

(1) Does the law create a distinction based on an enumerated or analogous ground? (2) Does the distinction create a disadvantage by perpetuating prejudice or stereotyping? (FC 2017 at para 62)

She also acknowledged the Supreme Court’s more recent section 15(1) decisions in Quebec (Attorney General) v A, 2013 SCC 5 (CanLII), [2013] 1 SCR 61 and Kahkewistahaw First Nation v Taypotat, 2015 SCC 30 (CanLII), [2015] 2 SCR 548, quoting (at para 78) from Justice Rosalie Abella’s judgment in Quebec v Athat:

The root of s. 15 is our awareness that certain groups have been historically discriminated against, and that the perpetuation of such discrimination should be curtailed. If the state conduct widens the gap between the historically disadvantaged group and the rest of society rather than narrowing it, then it is discriminatory. (at para 332)

In addition, Justice Kane relied upon two Federal Court of Appeal section 15(1) decisions (at paras 101-106). In Grenon v Canada, 2016 FCA 4 (CanLII), the Court considered the claim of a male taxpayer who was denied a deduction for legal expenses he had incurred for the determination of child support payments. The Court held that although 92.8% of child support payors were men, this was insufficient to establish adverse effects discrimination. Rather, “[t]he law, when applied to men as opposed to women, must have a qualitatively different impact on men. A mere numerical imbalance will not suffice” (Grenon at para 41). Thompson v Canada (Attorney General), 2016 FCA 253 (CanLII) involved the claim of a civilian employee who was severely disabled in a crash of a Canadian Forces aircraft, and who was entitled to less favourable treatment than the military personnel who were also injured in the crash. The Court found that the differential treatment was based on the claimant’s employment status rather than his disability and denied his claim because employment status is not an enumerated or analogous ground under section 15(1).

Based on this case law, Justice Kane held the claimants had not proved that the relevant provisions of the RCMPSA and Regulation created a distinction based on enumerated or analogous grounds. According to the Court of Appeal, Justice Kane found that adverse impact had not been proven and that any impact was due to the claimants working part-time hours (2018 FCA at para 24). She also found that the claimants’ argument “confound[ed] the underlying social circumstances with the consequences of the law” (2017 FC at para 139, quoting Grenon at para 43).

Although it was unnecessary, Justice Kane went on to find that the claimants had also failed to meet the requirements of the second step of the Kapp/Withler test for section 15(1): Does the distinction create a disadvantage by perpetuating prejudice or stereotyping? Justice Kane acknowledged the evidence of historic disadvantage to women in the workforce due to their primary responsibility for child care, one result of which was their disproportionate representation in the part-time labour force and consequent reduced pension income (2017 FC at paras 168-169). She also acknowledged the challenge posed to women with children by the impact of policing’s patrol duties, shift work and isolated postings.

Nevertheless, she held that the impugned provisions did not perpetuate disadvantage because there was “no evidence that the RCMPSA was or is a disincentive to recruitment of women to the RCMP” (2017 FC at para 170). Further, she found there was “no evidence of any historic disadvantage to women or women with parental status arising from [the RCMP] Pension Plan” (2017 FC at para 171). Instead, she noted that the RCMPSA treated all members the same, using a mathematical formula that did not include any input for sex or any other prohibited ground (2017 FC at para 172).

In addition, because the RCMPSA was a “stable and reliable pension plan” (2017 FC at para 172) and thus a social benefit, “the ameliorative effect of the law on others and the multiplicity of interests it attempts to balance” (2017 FC at para 160, citing Withler at para 67) had to be taken into account, as well as the policy goals of those who designed the pension plan, the appropriateness of the lines drawn by the plan given its intent and purpose, and “resource implications” (2017 FC at para 173). For Justice Kane, drawing the line at employment status was the obvious place to do so (2017 FC at paras 175-176). The RCMPSA had an “overall ameliorative effect on all members” and the formula based on employment status ensured “overall fairness” (2017 FC at para 177).

Justice Kane also found that the claimants had made an informed and free choice. The RCMPSA did not interfere with their choices or prevent them from buying RRSPs or deny them the freedom to choose how to balance the competing demands of work and child care (2017 FC at paras 179-180).

Nor did Justice Kane find that there was evidence of stereotyping. Women who combined the roles of parent and RCMP member were not seen as any less worthy of recognition or respect than women who were full-time caregivers or full-time RCMP members (2017 FC at para 181-184). The reason the claimants were not able to make contributions at the full-time rate in order to receive full pension benefits on retirement was “because “technical qualification requirements were not met” (2017 FC at para 185, quoting Miceli-Riggins, supra at para 84). There was no need for the pension plan to “perfectly correspond to [their] needs” (2017 FC at para 186).

Commentary

As noted in our introduction, the judgments in this case display many of the problems that courts have had with adverse effects discrimination cases. In our commentary, we map these problems onto the two steps of the test for section 15(1).

Step One Problems

Recall that the Federal Court of Appeal articulated the first step of the test for discrimination as “whether the impugned law on its face or in its impact creates a distinction based on a ground enumerated in section 15 of the Charter or on analogous ground” (at para 39). We do not take issue with this formulation generally, but rather with its application by the Court of Appeal.

One of the problems we have noted with adverse effects discrimination cases is that they often take a formalistic approach to the comparative analysis inherent in finding a “distinction”. The claimants appear to have selected employees on LWOP as their comparator group (see para 2, which sets out the claimants’ argument that “they were treated less favourably than those who are absent from work on leave without pay of more than three months’ duration”). It is true that if one focuses on pension buy-back rights, it is useful to compare job-sharing and LWOP employees – the latter group can buy back pension benefits, while the former cannot. However, Withler– another case involving pension benefits that we wrote about here– tells us that there may be more than one relevant comparator in any given case (at para 72). Another approach in Fraser would have been to compare part-time job-sharing employees with full-time employees, as it is the latter group who receives full pension benefits without any need to buy back. Seen this way, pension buy-back rights are a way of remedying the reduced pension benefits that part-time employees receive as compared to full-time employees, rather than the benefit itself.

A related problem is that the Court of Appeal considered the pension benefits in the overall context of the remuneration package, following the approach in Withler. One of the problems with taking this approach is that it casts too wide a net when it comes to the comparative analysis – we might assume that anyadverse distinctions between groups could be dissipated if the net was cast widely enough. It was on the basis of considering the overall employment context that the Federal Court of Appeal found a lack of evidence of the “relative value of the two packages” received by LWOP and part-time employees.

Comparing part-time and full-time employees with respect to their pension benefits compares favourably to the Supreme Court’s approach in human rights cases. For example, in Moore v British Columbia (Education), [2012] 3 SCR 360, 2012 SCC 61 (CanLII), the Court examined a claim by a student with learning disabilities that he was being denied access to education in a discriminatory way. The lower courts in that case had compared Jeffrey Moore to students with other disabilities, but Justice Abella rejected that approach, finding that:

Comparing Jeffrey only with other special needs students would mean that the District could cut all special needs programs and yet be immune from a claim of discrimination. It is not a question of who else is or is not experiencing similar barriers. This formalism was one of the potential dangers of comparator groups identified in Withler (at para 30, citation omitted).

We can see this formalism problem at play in the Court of Appeal’s finding that it is “impossible to conclude that job-sharing is adverse to being on a leave without pay” (at para 50). Why compare part-time employees to another group that does not get full benefits when their overall employment context is examined? If pension buy-back rights for LWOP employees were eliminated by the RCMP, would that mean that neither group was entitled to full pension benefits? From the perspective of substantive equality, it makes more sense to compare part-time employees without full pension benefits to full-time employees, the group with the clear advantage in this context. It is open to the Supreme Court to revise the comparator analysis on appeal to use multiple comparators, and we urge them to do so.

When comparing part-time and full-time employees with a focus on the benefit in question, it should be self-evident that there is a distinction between the former, who do not get full pension benefits, and the latter, who do. The more difficult issue is whether that distinction is based on an enumerated or analogous ground, which is the second part of the first step.

The Federal Court was correct that the distinction in receiving full pension benefits is “because of” part-time employment status, and that status has not been recognized as a protected ground under section 15(1) of the Charter. But whether a distinction is “because of” – words emphasized by Justice Kane (FC 2017 at para 107) – goes to direct discrimination and is only part of the story. As noted by the Federal Court of Appeal, adverse effects discrimination requires consideration of adverse treatment that is not apparent on the face of the provisions in question. However, the Court of Appeal also appears to have been looking for a direct connection to the claimed grounds when they said that the claimants were required to demonstrate that their adverse treatment as compared to others “results from the particular characteristics that the protected group possesses” (FCA 2018 at para 43, emphasis added; see also para 47 where they stated that the adverse treatment must be due to the protected grounds).

Let’s pause for a moment and think about the actual issue that the claimants have raised. Simply put, it is that more women than men work part-time as RCMP officers in order to be able to care for their children, resulting in women with parenting responsibilities receiving reduced pension benefits more often than men. Can we say that the disproportionate receipt of reduced pension benefits by mothers working as RCMP officers is because of or a result of or due to their gender or family / parental status? In some ways this seems self-evident, but there are at least two other problems embedded in the Court of Appeal’s analysis that may have prevented them from making this finding.

First, Charter claims for adverse effects discrimination are often subject to higher evidentiary and causation requirements. In the case at hand, there was expert evidence that working women in Canada disproportionately bear child-rearing responsibilities and that women in policing may experience these issues particularly acutely, especially if they work in areas with limited access to child-care. There was also evidence that in May 2010 and May 2014, all the RCMP members who were working part time in job-sharing arrangements were women, and a significant majority of these women job-shared as a result of child care needs. This should have been sufficient to prove that the adverse treatment experienced by the claimants was connected to their gender and their family / parental status. And being connected to their gender and their family or parental status should be sufficient without needing to prove that the adverse treatment was because of or due to these factors.

We can again find assistance in the human rights case law. In order to make out a prima facie case of discrimination, the claimant need only show that the protected ground(s) they are relying on were a factor– or in other words, connected to – the adverse treatment they received (see Moore at para 33). The Supreme Court has deliberately eschewed a causal connection approach in this context (see Quebec (Commission des droits de la personne et des droits de la jeunesse) v Bombardier Inc. (Bombardier Aerospace Training Center), [2015] 2 SCR 789, 2015 SCC 39 (CanLII) at paras 44-51). While it is true that it is sometimes more difficult to prove discrimination based on family status in the human rights context – with the Supreme Court unfortunately having denied leave to appeal on this very issue on August 8 – that line of cases deals with accommodation of child care time rather than receipt of reduced benefits that are connected to being a parent (and mother). Again, we maintain that based on the evidence available in Fraser, the Federal Courts should have found a sufficient connection between the adverse treatment and gender / family status. This is so even though the Court of Appeal stressed that not all the female RCMP officers who were job-sharing did so for reasons related to child-care. It is accepted under the Charter that not all members of a group need to be affected in the same way in order for an adverse distinction based on a protected ground to be made out. (see e.g. Vriend, supra)

A second problem with finding a link between the adverse treatment and the claimed grounds is the courts’ focus on choice. The Court of Appeal noted the lack of evidence that leave without pay was unavailable to female RCMP members with young children and that more men than women or more childless individuals had opted to take leaves without pay. In addition to focusing on the wrong comparator group, this passage suggests that it was simply a matter of individual choice to work part-time or to take a leave without pay. To repeat a passage quoted above, the Court of Appeal stated that the claimants “were not denied buy-back rights based on their personal characteristics of being female RCMP members with young children, but rather because they elected to job-share as opposed to taking care and nurturing leave” (at para 53). In other words, the claimants could have simply avoided the adverse treatment by making a different choice about how to manage their child-care responsibilities while remaining connected to the workforce.

This reliance on individual choice fails to see the systemic challenges that (mostly) mothers have in balancing work and child-care – including continued pay inequality when compared with men, lack of a national child-care program, and so on. Adverse impact claims require that systemic discrimination be seen. Furthermore, the Supreme Court has recognized that it is inappropriate to consider personal “choices” such as marital status in analyzing discrimination claims under the Charter. In Quebec v A, supra, Justice Abella’s majority decision on section 15(1) noted that marital status may not be the subject of actual choice in fact (at paras 316-17). She also relied on a human rights decision, Brooks v Canada Safeway Ltd., 1989 CanLII 96 (SCC), [1989] 1 SCR 1219, for the point that “this Court has repeatedly rejected arguments that choice protects a distinction from a finding of discrimination” (at para 336). In Brooks, not only was pregnancy rejected as a matter of “true choice”, choice was also “irrelevant to the question of discrimination” (Quebec v A at para 336). Similarly, the courts in Fraser should have seen any “elections” the claimants or other female RCMP officers made about working part-time as irrelevant to whether their reduction in pension benefits was related to their parental status (and gender).

Step Two Problems

As noted earlier, Justice Kane considered step two in the alternative, and there are also some problems – typical problems, for the most part – with her analysis, which relied almost exclusively on principles set out in Withler (2017 FC at paras 156-166).

In first examining whether the differential impact perpetuated disadvantage, Justice Kane’s primary concern was with the lack of very specific types of evidence. She acknowledged the more general evidence of the historic disadvantage and diminished financial security of women in the workforce due to their primary responsibility for child care (at paras 168-169). However, what she wanted was evidence “that the RCMPSA was or is a disincentive to recruitment of women to the RCMP,” and no such evidence of deterrence had been provided (at para 170). Just why an ability to buy-back pension benefits in a job-sharing position must be a deterrent to women thinking about joining the RCMP in order to find a perpetuation of disadvantage was not articulated. How would such evidence be gathered when, if it exists, it exists in the minds of women, or women who are parents, who decided not to become RCMP members because the pension plan benefits would not be good enough if they became parents and if they job shared as a result of assuming the primary responsibility for child care? Justice Kane assumed that women or women who are parents contemplating joining the RCMP would recognize the overall pension plan as a “stable and reliable pension plan” and a reason to join (at paras 170-171).

After setting out very high expectations for a particular type of evidence, Justice Kane then engaged in some very formalistic reasoning – another typical problem for adverse impact claims. She explicitly noted that the “RCMPSA treats all members the same,” using a formula “without regard to sex or any enumerated or analogous ground” as though that was a reason for finding no disadvantage instead of the very definition of adverse effects discrimination (at para 172). This same point is reiterated. According to Justice Kane, the claimants could not make pension contributions at the full-time rate simply “because they did not meet the requirements of the plan” (at para 179). Again the plan is presented as neutral, without consideration of the fact that the claimants could not make full contributions because they worked part-time, and they worked part-time for systemic reasons related to their gender and family/parental status.

The next problem we see is the use of section 1 considerations in the section 15 analysis, caused by the use of the contextual factors from Law v Canada (Minister of Employment and Immigration), [1999] 1 SCR 497, to assess a broadly-based social benefit. Justice Kane seemed to want evidence that the claimants had been “excluded or left out of the RCMPSA” (at para 175). Being excluded from the group of members who could buy back full-time benefits and thus having reduced pension benefits upon retirement was not a disadvantage – or not enough of a disadvantage because “perfect correspondence between a benefit program and the actual needs and circumstances of the claimant group is not required” (at paras 160, 174, relying on Withler at para 67 and Miceli-Riggins at paras 76-79). After all, “nothing is perfect” (2017 FC at para 186). It was enough that the “overall goal of providing retirement income is met”; the “unique needs” of retirees such as the claimants – women who were parents – were characterized as just one type of many possible retiree needs “that are not perfectly met” (at para 177).

Here Justice Kane relied upon Withler’s comments specific to social benefit programs, in which the Supreme Court stated that the ameliorative effects on others and the multiplicity of interests that governments had to balance, as well as their policy goals for the plan and its intended beneficiaries, had to be taken into account when analyzing the rights violation – and not as part of the government’s justification under section 1, as we might expect (at paras 173-174). Nevertheless, Justice Kane found that the pension plan had “an overall ameliorative effect on all members who contribute and later receive a pension, who are the intended beneficiaries” (at para 177). Perhaps the government led evidence on the plan’s overall ameliorative effect, but there is nothing in the Federal Court judgment to indicate the existence of such evidence.

The problem of choice is also evident in the Federal Court’s step two analysis of the perpetuation of disadvantage. Justice Kane noted that the claimants were fully informed of the pension consequences of job-sharing before or shortly after they made the choice to job-share. The pension plan “did not interfere with the choices made by the Applicants at all”; options such as LWOP or savings or RRSPs were all open to them (at paras 179-180). The evidence about the context for these “choices” and the Quebec v A cautions about relying on the notion of freedom of choice in a section 15 analysis were forgotten.

As for the perpetuation of stereotypes, Justice Kane noted there was no evidence of stereotyping such that “women who seek to combine the roles of wife and mother are less worthy of respect” (at para 181) or are “singled out” for adverse treatment (at para 185). The idea that the law must single out members of a group is a measure of direct discrimination, not adverse effects discrimination. More generally, and as we argued in our Adverse Impact article (at 214-216), it is difficult to prove stereotyping in respect of facially neutral measures unless the failure to accommodate difference is intentional.

Neutrality also makes an appearance in the stereotyping analysis. Justice Kane characterized the reason that the claimants were denied benefits as one of “technical requirements” not being met (at para 184). She concluded her step two analysis of discriminatory impact by twice noting how well the claimants had done as female RCMP members, finding ways return to full-time employment and have long careers with the RCMP (at paras 185-186). This may be both true and praise-worthy, but the claimants were not challenging the job-sharing pension provisions only for themselves. This individualization of persevering in the face of challenge is another signal of the lack of recognition of systemic discrimination.

Conclusion

In addition to the problems we have pointed out above with respect to the courts’ application of the two steps of the test for section 15(1), we note that the Federal Court of Appeal framed the claim as one for “a constitutional right to increased pension benefits” (FCA 2018 at para 61). It is inappropriate to see access to pension buy-back rights as “increased pension benefits.” The claimants are simply seeking equal access to the pension benefits that full-time workers receive, which – it bears noting – the claimants would have to pay for.

We also note that the Federal Court of Appeal referenced Griggs v Duke Power Co.(1971), 401 US 424 as a “quintessential case of adverse effect discrimination” (at para 45). Griggswas relied on by the Supreme Court in its last adverse effects discrimination case as well (see Taypotat, at para 23). It is important to point out that Griggs– a case involving a facially neutral rule that required job applicants to have a high school diploma or pass intelligence tests and that was found to adversely impact African Americans – was decided in a very different legal context. Griggs did set the precedent for “disparate impact” racial discrimination cases, but it was brought under Title VII of the Civil Rights Act of 1964, which applied to private employers with 15 or more employees. It was not a constitutional law case implicating government action such as the pension rules at issue in Fraser. Disparate impact claims can be defended by showing that they are “job related for the position in question and consistent with business necessity” (section 2000e-2(k)(1)(A)(i)), a consideration that is not apt in Charter discrimination cases. The claimants in Griggs used statistics to prove disparate impact on a prima facie basis, which shifted the burden to the employer to prove the business necessity of the job requirements. In Fraser, the Federal Court explicitly held that the fact that more women job-share is not enough to find discrimination based on sex, saying a claim “cannot succeed based only on the numbers … [and] a more qualitative assessment is required” (2017 FC at para 122).

The courts’ reliance on Griggs highlights the sad fact that Canadian courts do not have many home grown adverse effects discrimination cases to rely on, at least under section 15(1) of the Charter. Even Eldridge and Vriend could be seen as direct discrimination cases, at least in part. There are also no Charter cases where a majority of the Supreme Court has found adverse effects discrimination based on gender. We hope that the Supreme Court will pay close attention to the problems that have traditionally plagued adverse effects cases and “buy back” into this concept in Fraser.

This post may be cited as: Jennifer Koshan and Jonnette Watson Hamilton, “Time for Buy-Back: Supreme Court Set to Hear Important Adverse Effects Discrimination Case” (September 5, 2019), online: ABlawg, http://ablawg.ca/wp-content/uploads/2019/09/Blog_JK_JWH_Fraser.pdf

To subscribe to ABlawg by email or RSS feed, please go to http://ablawg.ca