Employment Insurance (EI)

Decision

[1] The appeal is allowed. I find that section 153.17 of the Employment Insurance Act (Act) violates section 15 of the Canadian Charter of Rights and Freedoms (Charter) and that it is not saved by section 1 of the Charter.

Preliminary matters

[2] On the first day of the hearing, I asked the Appellant whether she had sent the required notice to the Attorney General of Canada and to those of the provinces.^{Footnote 1} I asked the question because a copy of the filing did not appear in the file. She said that she had sent it and that she would file a copy before the end of the day.

[3] But when I looked at the document that was filed, I realized that the ten days’ notice had not been respected. Even though the Canada Employment Insurance Commission (Commission) did not raise this issue at the hearing, I found I had no other choice but to make an interlocutory decision abridging the delays.^{Footnote 2} That decision was then served on the Attorneys General as a precaution.

[4] No attorney general asked to intervene, and the hearing continued as planned.

Overview

[5] The Appellant was working as a server in a restaurant in March 2020 when the COVID-19 pandemic began. Like millions of other Canadians, she was laid off. The economy was at a standstill because of the public health closures. She received the Canada Emergency Response Benefit (CERB) that was rapidly put in place by the Canadian government. This measure was set up because the Employment Insurance (EI) system was unable to process such a large volume of claims.^{Footnote 3}

[6] During the nearly three years of the pandemic, the Canadian government introduced measures to ensure that most Canadians who were negatively impacted could access some type of EI benefits. That meant, among other things, lowering the threshold requirements to access EI benefits, setting a uniform unemployment rate for all regions, and not requiring a medical certificate for EI sickness benefits.^{Footnote 4}

[7] Another measure was the introduction of a one-time credit of hours. That credit allowed people to qualify for benefits as long as they had at least 120 hours in their qualifying period. This measure, found in section 153.17 of the Act, was in place from September 2020 to September 2021, with a similar measure applicable retroactively to March 2020.

[8] This credit applied to all claims, even if claimants already had enough hours to qualify. That was the case for the Appellant in the fall of 2020. Once applied, it could not be used for a later claim. No mechanism was put in place to allow claimants to “bank” this credit of hours or choose to use it later.

[9] The Appellant became pregnant in the fall of 2020 and gave birth in July 2021. She then made a new claim for maternity and parental benefits, but the Commission decided that she did not have enough hours to qualify. She had accumulated only 131 hours of insurable employment since her last (and first) claim for regular benefits. She testified that she had been told twice by Service Canada agents that she could benefit again from the credit of hours. But the information she received was incorrect. Had she been able to benefit from the credit, she would have qualified for EI maternity and parental benefits. Now, she was left with nothing.

[10] The Appellant claims that the design of section 153.17 of the Act has a discriminatory impact on women, especially women who claimed both regular and special benefits during the pandemic.

[11] The Respondent argues that the disproportionate impact has not been demonstrated because the real reason for the distinction the Appellant suffered is not sex, but her work sector. Furthermore, it argues that the difficulties suffered were not caused by section 153.17 of the Act, but were the result of the Appellant’s personal circumstances. Finally, it argues that the perpetuation of a historic disadvantage is not enough to justify a finding of discrimination and that is, at most, the effect the contested section of the Act had on her.

[12] I must decide whether section 153.17 of the Act had a discriminatory impact on the Appellant and, if so, whether it can be justified by the application of section 1 of the Charter.

Issues

[13] This constitutional challenge raises the following issues:

1. a) Does section 153.17 of the Act violate section 15 of the Charter?
2. b) If so, can section 153.17 of the Act constitute a reasonable limit under section 1 of the Charter?
3. c) If not, what is the appropriate remedy?

Analysis

Introduction

[14] Charter challenges are complex cases. Those dealing with section 15 of the Charter, the equality rights provision, are particularly challenging.^{Footnote 5} That is because of the inherent difficulty in putting forward an analytical framework that is workable, fair, and sensitive to the fact that discrimination can be an unintended consequence of what first appears to be a neutral measure. The law has evolved greatly during the past 30 years.

[15] In keeping with this trend, at least six Supreme Court of Canada decisions have dealt with the equality rights provision in the past five years. All have been decisions where some of the judges wrote vigorous dissents.^{Footnote 6} At play here is the difficult balance to reach between exercising judicial reviewing powers under the Charter and deference to Parliament’s choices. Deference, in this context, is showing the respect that is owed to the decisions of Parliament as an elected body. This explains why the law on this question, particularly on the type and level of evidence needed at each step of the test, continues to be so difficult to grasp.

[16] Furthermore, although everyone agrees that section 15 should protect substantive equality, as opposed to formal equality, the understanding of what exactly is substantive equality varies significantly.^{Footnote 7} Many questions remain. What does it mean exactly for a law to treat everyone in a way that is substantively equal? How do you prove that the law has a substantively discriminatory impact on a person or on members of her group? What is the place of comparison in a discrimination analysis? And more importantly—but surely not more simply!—what is the place of deference to Parliament’s choices in designing benefits in this discrimination analysis? And at what step of the legal test should this deference be taken into account? Those are all difficult questions, and the answers will probably continue to evolve in the years to come.^{Footnote 8}

Legislative framework

[17] To better understand the parties’ positions and the analysis of the constitutional question that will follow, it is important to start by putting the specific measure set out in section 153.17 of the Act in its broader context. What is at issue in this decision is the second iteration of the credit of hours set out in section 153.17. As I will explain later, the parties referred to this as the “phase-two credit.”

[18] In explaining the framework, I will rely, in part, on the reports and testimony of experts Benoit Cadieux, who testified for the Respondent, and Professor Jennifer Robson and Dr Ruth Rose, who testified for the Appellant. The subjects for which each of them is qualified as expert for the purpose of this decision are as follows:

1. i. Dr. Ruth Rose
1. a) Data and trends in labour participation in Canada
2. b) Gendered employment
3. c) Pay equity
4. d) Design and access to benefits
2. ii. Professor Jennifer Robson
1. a) Public policy, policy design, political management, and delivery of security programs
2. b) Policy to support women in the workplace and support programs
3. c) Gender Based Analysis Plus (GBA+) and gendered impact of government benefits
4. d) Policy design and delivery of relief program by the federal government
3. iii. Benoit Cadieux
1. a) Design, development, and application of employment benefits policy

[19] The EI program has been described by the Federal Court of Appeal as “a contributory scheme which provides social insurance for Canadians who suffer a loss of income as a result of a loss of their employment or who are unable to work by reason of illness, pregnancy and childbirth or parental responsibilities for a newborn or newly-adopted child.”^{Footnote 9}

[20] The program is generally understood to be an insurance-based program because the benefits a claimant will receive are linked to their previous work history, instead of being set according to their needs only.^{Footnote 10} While this is true, some parts of the program, like maternity and parental benefits, are closer to a social support program because they do not presuppose a quick return to work or require an active job search.^{Footnote 11}

[21] When the pandemic hit in March 2020, the Canadian government, like many others around the world, responded by putting in place an economic safety net for its citizens and businesses.^{Footnote 12} This safety net evolved during the nearly three years of the pandemic to accommodate changing needs and an uncertain recovery.

[22] Although the nature of the EI program was not fundamentally changed by the modifications put in place during the pandemic, some of its key components were altered, and their proportional importance was greatly reduced. The design of those changes was influenced by many factors, including the vast number of Canadians affected by the pandemic, the ongoing economic uncertainty, the limitations of the existing system to process claims, and the need for impacted workers to be provided quickly and efficiently with income support.^{Footnote 13}

CERB—First period: March 15, 2020, to October 3, 2020

[23] During the first period, from March 15, 2020, to October 3, 2020, the federal government created the CERB. It was composed of two distinct programs.^{Footnote 14} The first one, usually referred to as the CERB, was intended for workers who would normally not have access to EI, like self-employed workers. The other one, called the EI Emergency Response Benefit (EI-ERB), replaced EI benefits for those who claimed regular or sickness benefits during that period. The benefit provided was $500 per week for all claimants. This amount was not correlated to the number of insurable hours worked or the amount earned in the previous year. Some of the admissibility criteria were also changed to make it easier to qualify.^{Footnote 15}

[24] For this first wave of measures, the regular rules and requirements to access special benefits (maternity, parental, compassionate care, etc.) were not changed.^{Footnote 16} In August 2020, after becoming aware that some mothers could not qualify for those benefits, the government put in place a first credit of hours, available once per claimant.^{Footnote 17} Parties referred to it, and so will I, as the “phase-one credit”. It had a retroactive effect to March 15, 2020, but was only available to those who needed it to qualify for an initial claim for special benefits.^{Footnote 18} Claimants had to apply to receive it—it was not automatically applied on each claim.

[25] The statistics provided show that more women than men received the CERB, in part because they were more likely to be unemployed. For the same reason, the accommodation and food services industry had the highest percentage of CERB recipients.^{Footnote 19} Also unsurprisingly, those with the lowest income in 2019 were the most likely to receive benefits, and to receive them for a longer period.^{Footnote 20}

CERB—Second period: September 26, 2020, to September 24, 2021

[26] Then, a second wave of measures was introduced at the end of August 2020 and implemented in early fall 2020. They were to last for one year. In his report, Benoit Cadieux (the Respondent’s expert) mentions as follows:^{Footnote 21}

The Government recognized that many workers at this point in time may not have had enough hours to qualify for EI benefits due to the pandemic. This included workers who did not cease working but experienced reduced hours due to COVID-19 and workers who returned to work shortly before COVID-19 related closures and restrictions occurred and did not have time to accumulate enough hours to qualify for EI benefits.

[27] Among those measures was the second iteration of the credit of hours found in section 153.17 of the Act. This section reads as follows:

Benefits under Part I

153.17 (1) A claimant who makes an initial claim for benefits under Part I on or after September 27, 2020 or in relation to an interruption of earnings that occurs on or after that date is deemed to have in their qualifying period

1. (a) if the initial claim is in respect of benefits referred to in any of sections 21 to 23.3, (https://www.canlii.org/en/ca/laws/stat/sc-1996-c-23/latest/sc-1996-c-23.html#sec21_smooth) an additional 480 hours of insurable employment; and
2. (b) in any other case, an additional 300 hours of insurable employment.

Limitation

(2) Subsection (1) does not apply to a claimant who has already had the number of insurable hours in their qualifying period increased under that subsection or under this section as it read on September 26, 2020, if a benefit period was established in relation to that qualifying period.

[28] The parties referred to it, and so will I, as the “phase-two credit.”^{Footnote 22} It was introduced as part of a series of measures aimed at facilitating access to benefits.^{Footnote 23} Unlike the phase-one credit, it was applied on all first claims for benefits, even if a claimant did not need those hours to qualify. Special benefits (sickness, pregnancy, parental, compassionate care, and critically ill benefits) were now treated the same as other benefits - only the number of hours credited differed.^{Footnote 24} Once applied, it could not be used on a new claim during the applicable period. There was no way to refuse this credit or to bank it for later use. The normal “stacking rules,” which deal with the combination of more than one type of benefits during a given benefit period, were not modified.

[29] When these measures were adopted, everyone hoped that the pandemic would end soon and that the economy would recover.^{Footnote 25} The Respondent’s expert witness testified that the government wanted to make sure the measures would be designed in such a way as to not discourage people from returning to work. The Government estimated that 400,000 more people would be able to qualify because of those measures.^{Footnote 26}

[30] What is clear from Benoit Cadieux’s testimony and the available government statements on the matter, is that both phases of the credit were aimed at making it easier for Canadians who needed support to access EI benefits. The government recognized that “some people still needed support and help to establish a first claim.”^{Footnote 27}

[31] It is important to note that claimants who had received the EI-ERB from the beginning of the pandemic were automatically transferred to EI in September 2020. This meant that, without any action on their part, they were then deemed to have made an initial claim for benefits. The evidence shows that in late September 2020 approximately 2.8 million Canadians transitioned to EI benefits.^{Footnote 28}

[32] At that point, the pandemic had been going on for six months, and women were severely hit by the closures. But it still meant that, at that time, they would have accumulated six months of insurable hours in a “normal” job market.

[33] When questioned in cross-examination, Benoit Cadieux answered that the emphasis was put on the first claim for benefits at the time the measure was implemented. The government could always have put new measures in place later if it proved necessary.

Third period

[34] This is exactly what was done in the fall of 2021, when a third set of one-year measures was put in place. These measures were meant “to ensure the system remained responsive to the needs of Canadian workers as the economy continued to reopen.”^{Footnote 29}

Return to regular rules

[35] It was only in the fall of 2022 that the regular rules and requirements for EI benefits were reactivated.

[36] The facts relating to the Appellant’s employment history and family situation are, essentially, not in dispute.

[37] At the beginning of the pandemic, the Appellant had two teenage children. I understand from her testimony that she was—and still is—the main caregiver and provider for those children.

[38] Since at least 2019, the Appellant has worked as a server in restaurants or pubs. She was working an irregular schedule but, on average, she worked between 30 to 35 hours per week. In February 2020, just before the pandemic, she started working full time for a new employer. On March 16, 2020, she was laid off when a country-wide lockdown was ordered.^{Footnote 30}

[39] She applied for EI benefits but was told she would be getting the CERB instead. She received $2,000 per month. It took a couple of weeks before she received the first payments.

[40] At the beginning of the lockdown, she moved in with her partner, who lived nearly an hour away. The move had been planned in the previous months but was pushed ahead by the pandemic. In July 2020, when restaurants and bars were allowed to reopen, she got called back by her former employer. She immediately agreed to return to work there, even if it meant she had to commute one hour each way. She negotiated with her employer to work longer hours, up to 12 hours per day, to reduce the number of days she would have to commute. This made it easier for her to be present for her children on the days she was not working. Because she was the employee with the least seniority and because there were still restrictions on indoor dinning, she was only able to work part-time.

[41] At the end of September 2020, the CERB ended, and she was therefore automatically transferred to the EI-ERB program.^{Footnote 31} She was only told that she needed to fill out weekly report cards. She was laid off again in November when new health orders came into effect and restaurants could no longer serve food indoors.^{Footnote 32} From July 2020 to November 2020, she accumulated 131 hours of insurable employment.

[42] Also in November 2020, one week before she was again laid off, she learned she was pregnant. This added difficulty because pregnancy is one of the factors that increases the risk of developing more severe symptoms of COVID-19. She did not work from November 2020 to July 2021 and continued to receive the EI-ERB.

[43] She testified that she called Service Canada twice to find out whether she would be entitled to maternity and parental benefits. She said she was told that, because the hours credit would be applied to her claim, she would have the required number of insurable hours to qualify for maternity and parental benefits.^{Footnote 33}

[44] She gave birth in July 2021 and continued to receive the EI-ERB until her benefit period ended in September 2021. She made a new claim to receive maternity and parental benefits on September 13, 2021, but it was refused because she did not have enough hours to qualify.^{Footnote 34} She had only 131 hours.

[45] She testified that it then became difficult for her because her finances and those of her partner were separate. She was giving him money for rent and food. She could not go back to work at that time because she had no one to look after her newborn and there was a one-year waiting list for daycare. Her partner went back to work in October 2021, so he could not stay home either. She testified that she was under enormous financial and psychological pressure and that the whole situation made difficulties in her relationship with her partner worse.

[46] Even though she was no longer receiving benefits, she still had to provide the essentials to her two older children. She also had phone and insurance payments. She would have liked to be able to give more opportunities to her children as the economy reopened, but she could not afford to.

[47] On cross-examination, she agreed that she did not look for other jobs while receiving benefits. She was waiting to be called back by her former employer. She said that she did not believe anything would be available because of the lockdown and the strict rules that applied in the food and drink industry. This is why she did not search for other jobs.

[48] She also testified that the information she received from Service Canada, to the effect that she could qualify for maternity benefits with only 120 hours, did not influence her willingness to look for work. It was the reality of the labour market in her field of work that discouraged her from looking for other jobs. The Commission never asked her to prove her availability for work or to provide evidence of her job search.

[49] She finally testified that she had found other women online who were in situations like hers. She also saw news stories on the same subject. Those women came from across Canada. They all had different circumstances but ended up in situations similar to hers.

[50] The Respondent filed an affidavit from Joycelyn De Souza, Business Expertise Consultant with the Benefits Delivery Services Branch of Service Canada. She was cross-examined at the hearing.

[51] Her testimony was essentially on what type and how many weeks of benefits the Appellant received. For example, she testified that the Appellant received, in total, 74 weeks of benefits in addition to the 4 weeks of advance payments, for a total of $40,111.^{Footnote 35} Those numbers are not contested.

[52] On cross-examination, she confirmed that, even if the Appellant had converted her claim to maternity benefits when she gave birth, she would only have been able to receive 9 weeks and 1 day of maternity benefits because of the “stacking rule.” To be able to receive the full 15 weeks of maternity and 35 weeks of regular parental benefits, she would have had to qualify for a new claim.

Discrimination analysis—legal principles

[53] As I mentioned earlier, the discrimination analysis has evolved greatly over time. It is important to keep in mind that some of the main principles of those decisions are no longer accepted law. That means that relying on past decisions of the Supreme Court of Canada on section 15 of the Charter must be done only with extreme caution and great discernment.

[54] The unique context of the pandemic, and the exceptionally broad measures that were put in place to respond to the needs of millions of Canadians, means that we must take caution when applying earlier cases that dealt with the Act or the Canada Pension Plan.^{Footnote 36} Contrary to what was at issue in some past decisions, the measures put in place during the pandemic were meant to cover a large number of Canadians with temporary, but highly simplified, measures that did not draw as many distinctions as the scheme would normally do. Considering the far-reaching scope of the special measures put in place and the amounts the government was willing to spend on these measures, the question of “allocation of scarce resources” that is usually important in these types of cases is much less of a consideration in this matter. So, less deference will be made to that question.^{Footnote 37}

[55] The legal test has not changed in the last 10 years. It remains a two-part test, where each part should be analyzed, as much as possible, distinctly.^{Footnote 38} The focus must be on the situation of the claimant’s group. The law can affect the group in two ways: It can either create obstacles for members of the protected group or fail to accommodate members of the protected group.^{Footnote 39}

[56] In R v Sharma, the most recent Supreme Court of Canada decision on section 15, the Court frames the test in this way:^{Footnote 40}

[The test] requires the claimant to demonstrate that the impugned law or state action: a) creates a distinction based on enumerated or analogous grounds, on its face or in its impact; and b) imposes a burden or denies a benefit in a manner that has the effect of reinforcing, perpetuating, or exacerbating disadvantage.

[57] Two types of discrimination exist. It can be direct (“on its face”), when the law itself makes a distinction based on a protected ground. It can also be indirect (“adverse impact”) when the law has a disproportionate adverse impact on a claimant or on a claimant’s group. Adverse impact discrimination generally requires more work from the claimant to prove the discrimination.^{Footnote 41} Since not all distinctions are discriminatory, many factors, including the overall context of the law, must be considered to establish discrimination.^{Footnote 42}

[58] I will now explain what the Supreme Court teaches us for each part of the test.

Step 1— “creates a distinction based on enumerated or analogous grounds, on its face or in its impact”

1. a. The distinction must be based on an enumerated or analogous ground. More than one ground can be invoked, as can intersecting grounds of discrimination.^{Footnote 43}
2. b. It is not appropriate to search for a “mirror comparator” group since it too often leads to a search for sameness, which is not the goal of substantive equality.^{Footnote 44}
3. c. The law must have a disproportionate impact on members of a protected group. Not all members of the protected group must be impacted, or impacted in the same way, for the first step to be met.^{Footnote 45}
4. d. Some sort of link, or nexus, must be established between the law and the adverse impact experienced by the protected group. This means that the contested norm or law must have, in some way, contributed to the disproportionate impact. It does not have to be the main, or the only, cause of the impact.^{Footnote 46}
5. e. The disproportionate impact can result from a law (or a norm) that has the effect of putting members of a protected group at a disadvantage. Or, it can result from a law that fails to provide the accommodation needed by members of that group to access the full benefits of the law.^{Footnote 47}
6. f. To prove the disproportionate impact in adverse impact cases, two types of evidence should be presented: “The first is evidence about the situation of the claimant group. The second is evidence about the results of the law.”^{Footnote 48}
7. g. Statistics, expert evidence, judicial notice and inferences can, among other things, be used to show a disproportionate impact, but no specific form of evidence is required. It all depends on the context and on what is available.^{Footnote 49}
8. h. This first step of the test should not be an onerous hurdle for claimants to meet.^{Footnote 50}

Step 2— “imposes a burden or denies a benefit in a manner that has the effect of reinforcing, perpetuating, or exacerbating disadvantage”

1. a. A claimant must show that the historic or systemic disadvantage suffered by their protected group is reinforced or exacerbated: “Leaving the situation of a claimant group unaffected is insufficient to meet the step two requirements.”^{Footnote 51}
2. b. Arbitrariness, prejudice, and stereotyping may assist, like other factors, in showing the negative impact of the law on the protected group.^{Footnote 52} No specific factor is required; it all depends on the specific circumstances of the case.
3. c. Intention to discriminate by the legislator and choices made by the member of the protect group are not relevant considerations at this stage of the discrimination analysis.^{Footnote 53}
4. d. The broader context, including the legislative context, must be considered to assess the nature and extent of the disadvantage.^{Footnote 54}
5. e. Courts (and tribunals) may take judicial notice of “notorious and undisputed facts.”^{Footnote 55}
6. f. When Parliament chooses to act, it must do so in a way that does not have an adverse impact on a protected group.^{Footnote 56} The state will always have the possibility of justifying, under section 1 of the Charter, the choices it made, and the disadvantage thus created or exacerbated.^{Footnote 57}

Application of the legal principles to the facts of this case

Step 1—Distinction based on an enumerated ground

Protected ground

[59] The parties do not dispute the ground of discrimination that applies in this case. It is the ground of sex. Maternity has been linked to the fact of being a woman, and therefore to the ground of sex.

[60] What the parties do not agree on is who belongs to the group and the role the protected ground played in the way the law applied to the Appellant. The Appellant argues that the protected group should be described as women who suffered employment loss and also gave birth during the pandemic.^{Footnote 58} The fact that only a subset of women is claimed to be adversely disadvantaged by the measure is not something to consider at this stage.^{Footnote 59}

[61] The Respondent argues that the distinction in this case is not based on sex, but on the Appellant’s personal circumstances. In other words, it argues that she could have received all of her maternity and parental benefits if she had given birth earlier in the pandemic.^{Footnote 60} Therefore, it is not the measure that caused the disadvantage, but the timing of her pregnancy and childbirth.

[62] In my opinion, the Respondent’s argument is similar to the one put forward in Fraser, where it was argued that it was not the pension plan provision that discriminated against women. Rather, it was argued that it was their choice to job-share that caused the disadvantage. ^{Footnote 61} This argument, related to the choices made by those claiming adverse impact, was squarely rejected by the majority in Fraser, as it had been in previous Supreme Court decisions.^{Footnote 62} It must also be rejected in this case.

[63] I find that, at this first part of the test, it is enough to find that the Appellant’s pregnancy, and therefore her sex, played a role in the fact that she was treated differently. There is therefore a distinction based on the ground of sex. The fact that other factors played a role, such as her personal circumstances or the timing of her pregnancy, does not negate this finding.

Disproportionate impact

[64] In this case, the disadvantage is economic. It was impossible for some new mothers to benefit from the credit of hours they needed to qualify for a second claim for EI benefits, in this case maternity and parental benefits. This meant they were prevented from receiving all or part of those benefits. Those mothers could not benefit from the credit of hours because it had been applied on their previous claim for regular benefits, even if they did not need those hours to qualify then.

[65] To demonstrate the impact of a legislative measure, two types of evidence can be used: evidence that explains the full context of the group’s situation and evidence that demonstrates the outcome the law has produced in practice for that group.^{Footnote 63} In adverse effect discrimination cases, it can be useful to have both types of evidence. The emphasis must be on the protected group or subgroup’s historic and present situation and on the effects that the contested measure had on them.

[66] The parties have produced evidence that falls in both categories mentioned above. For the Appellant, Dr. Ruth Rose and Dr. Jennifer Robson were heard as expert witnesses. For the Respondent, Benoit Cadieux was heard as expert witness.^{Footnote 64} I will summarize their evidence in the following paragraphs.

Full context of a claimant’s group situation

General situation of women in the workplace

[67] It is generally recognized that women face specific disadvantages in the workplace. The most recent Supreme Court of Canada decision on the subject is Fraser, where the majority of the Court said: “Recognizing the reality of gender divisions in domestic labour and their impact on women’s working lives is neither new nor disputable.”^{Footnote 65}

[68] The evidence presented by Dr. Rose is to the same effect.^{Footnote 66} Her report shows that the participation rate of men in the labour force has consistently been higher than that of women in Canada. The same is true for the employment rate.^{Footnote 67} Statistics also show that women continue to do a larger share of the unpaid household chores, even if it tends to become less pronounced in the younger generation.^{Footnote 68}

[69] The evidence shows the following:

1. i. Women still work part-time two and a half times more often than men, and this has not been fundamentally altered by the pandemic.^{Footnote 69}
2. ii. More women than men work part-time for involuntary reasons. And when they do choose to work part-time, it is more often to care for children or other family members.^{Footnote 70}
3. iii. Women work, on average, less hours per week than men.
4. iv. Women mostly work part-time to be able to accomplish domestic work and to take care of children or family members.^{Footnote 71}

[70] There is still a significant difference in the average annual employment income, weekly wages, and hourly rate in favour of men.^{Footnote 72} Single mothers with children under three years old face the most difficult situation, in terms of both unemployment and earnings.^{Footnote 73} The impact of having young children on women’s employment is often referred to as “the motherhood penalty,” and it is greatest for less-educated and single mothers.^{Footnote 74}

[71] The workforce is still highly segregated based on occupation. Women still work mostly in “jobs [that] are stereotypically identified with women.”^{Footnote 75} This is the case with the “Sales and service occupations,” which include the “food and services occupations” subcategory.^{Footnote 76} But the workforce is also segregated within broad occupational categories. For example, Dr. Rose testified that “Chefs and cooks” is a predominantly male category where workers are better paid than the other job classifications in the same category.^{Footnote 77} For example, in the “Food and beverage servers” occupational category, 78.8% of workers were women in 2015 and the average yearly salary was significantly lower than that of chefs, which is a male-dominated category.^{Footnote 78}

[72] Women are historically less likely than men to receive unemployment insurance benefits, for a variety of factors, including the fact that they are more likely to work part-time and less hours per week, on average, than men. As a result, the fact that admissibility is calculated in hours instead of weeks has had a detrimental effect on women’s access to EI benefits.^{Footnote 79}

[73] Finally, but not surprisingly, women receive the vast majority of maternity, parental, and family caregiving benefits. For example, women represent 76.5% of recipients of parental benefits in Canada, excluding Quebec. They also represent the vast majority of claimants who receive a family supplement for low-income families.^{Footnote 80} Finally, women are three times more likely than men to combine more than one type of benefit during one benefit period—in this case, regular and maternity/parental benefits.^{Footnote 81}

Situation of women during the pandemic

[74] The evidence shows that, throughout the pandemic, women were hardest hit. And those who got pregnant or had young children were the most affected of all. They lost their jobs more often, worked less hours per week, and needed support longer than men. This was so evident that the government recognized it and took steps to respond.^{Footnote 82} Their historic disadvantage in the workplace was therefore increased by the pandemic closures and work restrictions.

[75] Employment (and unemployment) rates were greatly affected by the pandemic. Unemployment rates rose to a peak in May 2020 and decreased gradually over the following months.^{Footnote 83} It was only in February 2022 that unemployment rates were back to what they were in January 2020.^{Footnote 84}

[76] Evidence shows that here, like elsewhere in the world, women were more affected by the pandemic, especially during the early phases.^{Footnote 85} Women experienced more job loss than men, took longer to return to employment, and were more likely to lose their job again.^{Footnote 86} The increase in EI claims for regular benefits was far greater for women than it was for men between October 2020 and October 2021.^{Footnote 87}

[77] While men recovered their pre-pandemic number of hours paid per week in September 2020, women only did so eight months later, in May 2021.^{Footnote 88} Finally, women were more likely than men to be on and off work because of unpaid care.^{Footnote 89}

[78] Also of relevance is the fact that single-parent families, often led by women, where the hardest hit in all sectors. This was, in part, because women took on greater childcare responsibilities because of school or daycare closures.^{Footnote 90} For example, mothers with school-age children were four percentage points less likely than fathers in the same situation to be at work when schools were closed because of health restrictions.^{Footnote 91} This is because it was mostly mothers who stayed home to care for children when the schools were closed.

[79] Dr. Rose showed that 21% of jobs lost in the peak months of the pandemic were in the “Accommodation and food services” sector, even if it counts for only 6% of total employment in normal circumstances.^{Footnote 92} Jobs in this sector are occupied predominantly by women. Also, the “Food and beverage service” was one of the occupational categories that was the most severely hit throughout the length of the pandemic. For example, figure 4.2^{Footnote 93} of Dr. Rose’s report shows that restrictions for in-person dining in restaurants was consistently higher between August 2020 and May 2021 than the overall restriction index.^{Footnote 94} The impact of the pandemic restrictions on indoor dining, among other things, was felt long after the first wave. As late as January 2022, the pre-pandemic level of employment in this sector had not been regained.^{Footnote 95}

[80] Soon after the start of the pandemic, women began expressing concern about their capacity to accumulate the number of hours needed to qualify for maternity and parental benefits. Benoit Cadieux testified that the government was aware of this situation and that it was one of the reasons why the credit of hours was put in place.^{Footnote 96}

Impact of the law in practice

[81] The second type of evidence we need to evaluate is the impact of the law, or the contested legislative measure, on the members of the group. This evidence can come from many sources: witness testimony, material facts, statistics, expert evidence, and, in appropriate cases, inferences.

[82] The evidence has shown the following:

1. i. Women were more likely than men to experience work interruptions caused by several factors.
2. ii. Women more often work part-time and for fewer hours per week than men.
3. iii. Women claim more special EI benefits, such as maternity and parental benefits.
4. iv. From March to September 2020, at least 400 women qualified for maternity benefits only because they benefited from the credit of hours.

[83] The Canadian government put in place measures to make the EI program accessible to more workers, including the CERB and other temporary measures like the credit of hours.^{Footnote 97} Those measures helped a large number of Canadians access EI benefits from March 2020 to September 2022.

[84] Evidence provided by Dr. Rose shows that the lower threshold requirements put in place for EI benefits between September 2020 and September 2021 generally gave more women access to benefits than men.^{Footnote 98} Women were also more likely to gain from the $500 uniform benefits under the new EI rules because they historically have lower weekly earnings than men. So, they would not necessarily have received this amount but for the government measures. It is important to note that Dr. Rose mentioned in cross-examination that these statistics do not specifically take into account the credit of hours because, to her knowledge, no statistics are available to show their specific impact.

[85] The only statistics that were introduced on the specific impact of the credit of hours were provided by Benoit Cadieux in his report. His report shows that the special measures put in place in September 2020 allowed 13.2% of all claims for regular benefits to be established. In other words, 13.2% of claims for regular benefits during this period would not have succeeded if the Canadian government had not adopted special measures to facilitate access to benefits. Out of those claims, approximately 70% were established due to the credit of hours. And the other 30% were established due to the unemployment rate being fixed at 13.1%.^{Footnote 99}

[86] The numbers provided in the report also show that 7.6% of all claims for special benefits were established because of the credit of hours. 98.8% of those were established on or after September 27, 2020. 4.8% of all claims for special benefits that were established were for maternity benefits. The other claims were established because of the “phase-one credit,” representing approximately 400 claims.^{Footnote 100}

[87] Benoit Cadieux testified in cross-examination that those statistics only cover first claims for benefits. Therefore, they do not tell us the number of people who, like the Appellant, would have needed the credit of hours to qualify for a second claim for maternity or parental benefits. They also do not tell us how many claims were rejected or not made, for example, because women received incorrect information (as was the case for the Appellant), or for other factors.

[88] The Appellant also testified about how hard her situation became because she could not qualify for maternity benefits without the application of the credit of hours. Obviously, taking care of a newborn and two teenagers with no income is not an easy situation. Other Social Security Tribunal decisions about situations similar to the Appellant’s also document the hardship created by the contested measure.^{Footnote 101}

[89] Considering what the evidence has shown, it is possible to conclude that at least as many women, and probably many more, were not able—or would not have been able if they had made a claim—to qualify for maternity and parental benefits later on in the pandemic.

Conclusion on the evidence

[90] The question I must address at this stage is this: Does the evidence on the record establish that section 153.17 of the Act has a disproportionate impact on women?

[91] I find that there is enough evidence on the record to support the Appellant’s arguments at this first step of the section 15 test. This conclusion is supported by evidence from both the Appellant’s and the Respondent’s expert witnesses and by the Appellant’s testimony.

[92] I find there is a distinction (or disproportionate impact) on the ground of sex for the following reasons:

1. i. Women were more affected by the pandemic, and the government knew this.
2. ii. Women, historically, receive fewer regular benefits and more special benefits (maternity and parental) than men.
3. iii. Women were in a more precarious situation in the job market during the pandemic than men were.

[93] Women were hit harder by the pandemic. The evidence has shown that women are still disadvantaged in the workplace. They are more likely than men to work part-time, work less hours, and earn less per week. They are more likely than men to have been unable to work during the pandemic either because of school closures or because they had to care for family members. Throughout the pandemic, women were more impacted than men by the shutdowns and pandemic-related work restrictions. This was in part because they are more likely to work in sectors and jobs categories that were hit harder by the pandemic.

[94] The evidence has shown that women are historically less likely than men to receive unemployment insurance benefits. Women make up the vast majority of maternity and parental benefits recipients. Birth rates stayed relatively stable during the pandemic (excluding a sharp decline in 2020), but maternity and parental benefits claims were lower than before the pandemic.^{Footnote 102}

[95] Measures put in place did help both women and men make a first claim for benefits. The evidence shows that claims for both types of benefits (regular and special) were facilitated. It also shows that at least 4.8% of all claims for special benefits that were established because of the credit of hours were for maternity benefits.

[96] It is clear from this evidence that women are, as a group, in a more precarious situation than men and that this situation was amplified during the pandemic. Therefore, it is possible to conclude that some women would also be more likely than men to have been unable to qualify for special benefits later on in the pandemic. But because the credit of hours was available for only the first claim, it would not have been available for that later claim. The fact that only a subgroup of women is disadvantaged in this way does not mean there is no discrimination.^{Footnote 103} So, the distinction based on an enumerated ground (sex) is clear.

Nexus

[97] This brings me to the question of the nexus or link between the contested measure and the adverse impact suffered, namely that some women were unable to qualify for maternity and parental benefits because of the design of section 153.17 of the Act.

[98] This is a contentious point in the Supreme Court of Canada’s jurisprudence on section 15 of the Charter. While it has been a long-standing requirement in human rights discrimination cases, the necessity to show a nexus in Charter discrimination cases was not adopted as a requirement by the majority of the Supreme Court before its decision in Sharma.^{Footnote 104}

[99] The majority of the Supreme Court in Sharma says that this link must be more than relational—it must be causal. This means that it must be shown that the contested measure contributed, at least in a small way, to the adverse impact suffered by the group. Again, many different types of evidence can be used to show this nexus, and the link will sometimes be so evident that it requires no evidence at all. The necessity to show a nexus does not transform the test. Therefore, the first step of the discrimination analysis should not place a heavy burden on claimants. This first part of the test is there to make sure that only relevant claims, namely those where there is a distinction based on a protected ground, are addressed.

[100] In this case, the Respondent argues that the nexus was not established. It says that it has not been shown that women suffered an adverse impact because of the design of the credit of hours. It also argues that the evidence presented by the Appellant is too general to show a real link between the impact on women and the contested measure. It says that the disadvantage suffered by women is essentially due to the historic disadvantage they already face and not the contested legislative measure.^{Footnote 105}

[101] I disagree. The question is this: Did the design of section 153.17 contribute to a greater disadvantage for women, compared to men?

[102] The evidence has shown the following:

1. i. Women were not able to qualify for another type of benefits later on in the pandemic because the credit of hours applied only to the first claim, even if it was not needed to qualify.
2. ii. Women were more likely than men to need this credit. This was because they were more affected by the pandemic in terms of possibilities to earn enough qualifying hours. They were also more likely to be absent from work for family and caregiving obligations.
3. iii. Women were more likely than men to take maternity and parental leave and, therefore, to need benefits during those leaves.

[103] The link between the design of the credit and the disadvantage is clear: It contributed, along with many other factors, to increase the historic disadvantage women already suffered.

Step 2—imposes a burden or denies a benefit in a manner that has the effect of reinforcing, perpetuating, or exacerbating disadvantage

[104] The second part of the section 15(1) test is concerned with the discriminatory nature of the distinction. The focus is on “members of the group and the negative impact of the law on them. The analysis is contextual, not formalistic, grounded in the actual situation of the group and the potential of the impugned law to worsen their situation.”^{Footnote 106} Their historic position and disadvantage are therefore relevant to the analysis. So is their situation within the legislative scheme as a whole.^{Footnote 107}

[105] Even if proof of prejudice, stereotyping, or arbitrariness is not necessary, it can be useful in demonstrating the perpetuation of a historic disadvantage. An arbitrary disadvantage is one that “fails to respond to the actual capacities and needs of the members of the group.”^{Footnote 108} The impact on the dignity of the person or group is, since Kapp, no longer relevant to the section 15 analysis.^{Footnote 109} This is why any reference to the Court’s analysis in Gosselin^{Footnote 110} or Granovsky^{Footnote 111} must be made with extreme caution.^{Footnote 112}

[106] Finally, the Court in Sharma mentioned the importance of considering the purpose of the measure within the broader legislative context when going through the second step of the section 15 analysis.^{Footnote 113}

Purpose of the benefit within the legislative scheme

[107] I find that the purpose or objective of section 153.17 is to facilitate access to benefits.

[108] The specific objective of the phase-two credit is contested. The Appellant argues that the objective of this measure was to ensure that those who could not accumulate enough hours because of the pandemic could still qualify for EI benefits when they needed them. The Respondent, on the other hand, argues that, even if it was part of a series of measures aimed at facilitating access to EI, the credit of hours, in itself, was strictly aimed at facilitating the transition from the EI-ERB to first claims for EI.

[109] The purpose of the contested measure needs to be examined within the legislative context explained above, which departs greatly from the normal workings of the EI scheme. Viewed in the broader context presented in the previous paragraphs, it becomes clear that the purpose of section 153.17 was to facilitate access to EI benefits for all those who required help to qualify.^{Footnote 114}

[110] This is a departure from the Supreme Court’s jurisprudence on this question because the focus of the section 15 analysis should be on the claimant’s group situation, not on the policy concern of the government. Traditionally, the finer analysis of the purpose of a contested legislative measure is done at the first step of the section 1 analysis, where the burden of proof rests with the Respondent. This being said, I now turn to the context in this case.

[111] The Supreme Court of Canada has established consistent guidelines on how to identify the purpose of a legislative measure.^{Footnote 115} These can be summarized as follows:

1. i. It is the purpose of the contested measure that must be identified, not the purpose of the legislative scheme that it is a part of.^{Footnote 116}
2. ii. The purpose of the legislative scheme can be a tool in identifying the purpose of the contested measure.^{Footnote 117}
3. iii. The purpose must stay the same throughout the analysis.
4. iv. The purpose must be defined with an adequate balance between too much precision and too much generality.^{Footnote 118}
5. v. The purpose of a measure must be distinguished from the means used to attain it or from the effect it creates.^{Footnote 119}

[112] The Respondent has proposed a different formulation of the purpose of section 153.17 of the Act throughout its factum, as noticed by the Appellant.^{Footnote 120} No direct evidence, such as a legislative statement of purpose, is available in this case.

[113] The witness and Respondent offered many different iterations of the purpose of the measure in their factum and arguments. Benoit Cadieux was not able to point to any external evidence to support those claims. But, since he was recognized as an expert on this subject and helped develop this legislative measure, I find his testimony to be persuasive. According to him, the purpose of the measure was as follows:

1. a) “to recognize the fact that some people still needed support and help to establish a first claim as labour market conditions continued to be uncertain”^{Footnote 121}
2. b) to allow those who would need to transition from regular benefits to special benefits to do so, even if they were not able to accumulate the required number of hours^{Footnote 122}
3. c) to provide support to claimants while not hindering the return to work if the market conditions were to improve significantly^{Footnote 123}
4. d) to ensure a quick and efficient transition to the EI regime while considering the volume of claims and the capacity of the system to process them more complex propositions^{Footnote 124}

[114] On the other hand, the Appellant defines the purpose of section 153.17 of the Act in a more general way. She claims the purpose of the measure “was to assist claimants who could not work the necessary hours to qualify for benefits in light of the circumstances of the pandemic.”^{Footnote 125}

[115] The Appellant argues that this is the proper definition of the purpose of the benefits because the narrower definition proposed by the Respondent confuses means with objectives. She also argues that defining the purpose of section 153.17 as facilitating a first claim for benefits is simply a description of the measure and not its objective.^{Footnote 126}

[116] How, then, should the purpose of section 153.17 of the Act be defined? First, it needs to be viewed in the context of the purpose of the Act as a whole. As explained in paragraph 16 above, the purpose of the Act is essentially to provide temporary income replacement to workers who have lost their employment through no fault of their own, or who need temporary income replacement because they are temporarily prevented from working. Defining the purpose of the measure as facilitating access to benefits is in line with the objective of the Act.

[117] Second, we need to distinguish between the objective, the means used to attain it, and the policy considerations that influenced those choices. In one part of his report, Benoit Cadieux described the objective of the phase-one credit in this way:^{Footnote 127}

To address this issue of workers, through no fault of their own, being unable to qualify for benefits due to insufficient hours, a one-time credit of hours was temporarily created to facilitate access to benefits while ensuring claimants still had some labour force attachment as part of the first set of EI temporary measures.

[118] From this description, we can see that the objective is to “facilitate access to benefits.” The means is “a one-time credit of hours,” which would be “temporary.” And the policy considerations were to ensure that “claimants still had some labour force attachment” when, “through no fault of their own,” they were “unable to qualify for benefits due to insufficient hours.”

[119] This same technique can be used, with the same result, for the phase-two credits. This is true simply because they shared the same objective: facilitating access to benefits. It was the means, the policy considerations, and the technical aspects that differed between the two phases of the credit.

Exacerbating the disadvantage

[120] I find that the distinction established by the automatic application of the credit of hours on the first claim for benefits is discriminatory because it fails to respond to the actual needs of the members of the group.^{Footnote 128} Instead, it denies them a benefit which has the effect of exacerbating their historic disadvantage in the job market and in accessing EI benefits, and specifically special benefits (maternity and parental benefits).

[121] Women’s socioeconomic context and historic disadvantage have been extensively covered in previous paragraphs and are not really at issue. It is common knowledge that women face historic disadvantage in the job market and that their caregiving responsibilities limit their presence in the workplace.^{Footnote 129} This is correlated to more difficulty accessing EI benefits, especially during the uncertain and difficult context of the pandemic.^{Footnote 130} It is also common knowledge, and the evidence in this case has shown it, that they are the primary recipient of maternity and parental benefits. During the pandemic, many women lived with the fear of not being able to qualify for those important benefits because of a situation beyond their control.

[122] Even with that knowledge, the evidence shows that no GBA+ was done for the phase two hours credit and that the GBA+ analysis that was done for all the measures adopted in the fall of 2020 has not been made public.

[123] Supreme Court jurisprudence tells us that the analysis at the second step must be adapted to the particular situation of the case. The criteria developed in Law are not relevant in all cases.^{Footnote 131} In this case, I find it is not necessary to spend much time on the question of the government’s allocation of resources within the legislative scheme or on the importance of being flexible when analyzing this question in the context of a benefits scheme.^{Footnote 132} It is well known that massive amounts of money were devoted to pandemic relief, and the question of cost or resource allocation was not part of the Respondent’s arguments in this case.

Conclusion on step 2 of the section 15 analysis

[124] I find that the design of section 153.17 of the Act prevents women from accessing maternity and parental benefits in a way that perpetuates and exacerbates their historic disadvantage, both in the workplace and in accessing EI benefits. It does this in a way that does not correspond to, or even consider, their actual needs and circumstances. As a result, I find that the Appellant has satisfied both steps of the section 15 test and established discrimination based on sex.

Section 1 analysis

[125] Section 1 of the Charter allows the state to show that “the infringement is demonstrably justified in a free and democratic society.”^{Footnote 133} The legal test was first established in Oakes and has not really changed since.^{Footnote 134} At this stage of the analysis, the burden of proof rests with the state. The focus of the analysis must be on the infringing measure, not on the law as a whole.^{Footnote 135} The test, in two parts, is as follows:

1. i. Is the legislative objective of the measure pressing and substantial?
2. ii. Are the means chosen proportional to that objective? To answer this question, three other questions must be asked:
1. a) Are the means adopted rationally connected to that objective?
2. b) Are the means minimally impairing of the protected right?
3. c) Is there proportionality between the deleterious and salutary effects of the law?

[126] If the answer to any of those questions is “no,” the state will have failed to meet its burden of proof and the infringing measure will not be “saved” by the section 1 analysis.

Part I – Pressing and substantial objective

[127] As part of the second step in the section 15 analysis, I have established that the real purpose of section 153.17 of the Act is to facilitate access EI benefits.

[128] In this case, the Appellant does not explicitly contest the fact that the infringing measure had a pressing and substantial objective. Rather, she focuses her arguments on how the purpose of the contested section of the Act should be defined. Since this argument was dealt with earlier, I will not go into further details here.

[129] It seems clear to me that no one can seriously say that assisting Canadians in need of economic support during a stressful, uncertain, and disruptive time was not a pressing and substantial objective. I therefore find that this first part of the test has been met.

Part II – Proportionality analysis

[130] The second part of the test, often referred to as the proportionality analysis, is where a closer analysis of the means chosen and the alternatives available to the government must be done.

Rational connection

[131] The first question to ask is this: Is there a logical connection between the means chosen and the objective sought by the government? Again, this step of the test does not impose a heavy burden on the government. The government must only show that “it is reasonable to suppose that the limit may further the goal, not that it will do so.”^{Footnote 136}

[132] It is important to distinguish between the means chosen and the objective sought to be able to put the two into a meaningful relation. Here, the objective was to facilitate access to EI benefits. And the means chosen was to give a one-time hours credit, irrespective of needs, on all first claims for benefits.

[133] The Appellant argues that the government fails at this stage of the test because the design of the infringing measure is what caused the disproportionate impact on women. Therefore, it cannot be rationally connected to the objective of allowing greater access to EI benefits. The Respondent, on the other hand, argues that section 153.17 of the Act “was rationally connected to the objective of providing EI benefits to Canadians quickly and efficiently during the ongoing crisis.”^{Footnote 137}

[134] I find it was reasonable for the government to presume that giving a credit of hours to everyone would further the goal of facilitating access to benefits. The evidence has shown that it did indeed help thousands of people to qualify. That is enough at this stage.

Minimal impairment

[135] To answer the second question, I have to identify whether the means chosen are minimally impairing of the protected right. It is at this step of the section 1 analysis that I will have to consider what the other options available to the government were. I will assess how those other options were evaluated by the government and to what extent they would have been able to meet the objective of the measure. Again, the burden of proof rests here with the Respondent.^{Footnote 138}

[136] The Respondent did not introduce evidence on the availability and evaluation of other means of achieving the objective of facilitating access to benefits at the time the infringing measure was put in place. This might be due, in part, to the fact that the Respondent has maintained that the objective should be described in a narrower way, to include both its one-time and its automatic nature. When described in this way, it is indeed difficult to imagine other less impairing measures.

[137] The other reason that can explain this lack of evidence is the privileged nature of government deliberations that was cited by the Respondent’s expert witness when questioned in cross-examination about the other options that the government studied.

[138] On why the government chose to apply the credit automatically on all claims, regardless of whether it was needed to qualify, Benoit Cadieux wrote:

Allowing claimants to choose whether or not to apply the credit when claiming benefits would have been highly complex for clients to navigate and for Service Canada to administer and would have resulted in significant delays in the payment of benefits.^{Footnote 139}

[139] The problem is that the Respondent provided no evidence to demonstrate why it would have been highly complex for clients to navigate, what could have been done to make it simpler, or why it would have been this complex for Service Canada to administer. The Respondent did not explain why it would have been more complex than what was done for the phase-one credit, except to say that it applied at a time where a vast number of people were transitioning back to EI. It also did not explain why it would have been more complex than other measures tailored for specific groups put in place before or after the fall of 2020.

[140] This contradicts the testimony of Professor Robson, who found that at least three other less impairing measures could have been explored or put forward by the government.

[141] The first option would have been to allow both the one-time credit for regular benefits and the one for special benefits to apply during the one-year period when they were in effect.^{Footnote 140} It would have been easy to apply since no choice or banking would have been needed.

[142] The second option would have been to apply the credit of hours only on claims without enough hours to qualify and to allow the other claimants to “bank” the credit for later use. Professor Robson’s argues that, since the system already allows for tracking of certain characteristics on an individual basis, it could probably have supported such an option.^{Footnote 141}

[143] Finally, the legislation could have simply transferred the responsibility to claimants to ask to bank the credit if they did not want it automatically applied on their claims.^{Footnote 142} It would certainly have lessened the number of files to treat individually, while allowing those most likely to need the credit to make an informed choice.

[144] When cross-examined on those proposed alternatives, Benoix Cadieux said that they would not have met the policy intent because it would not recognize “the continuing and uncertain impact of the pandemic.” In other words, giving “too much” access to benefits might have hindered some claimants’ desire to return to work when—and if—the economy recovered. He testified that “it was more prudent to design it like this and adjust as things evolve.”

[145] But he also recognized that, even if it was aware that many women could not qualify for maternity benefits because of the pandemic and that they were far more likely than men to claim both types of benefits, the government did not adjust its policy before September 2021. Even though, in September 2021, a uniform 420 hours threshold requirement was put in place to recognize that the pandemic still affected workers’ ability to accumulate the required number of hours to qualify for all types of EI benefits.

[146] The Respondent also did not explain why, at the same time as the phase-two credit of hours, the government was able to introduce a number of benefits specifically adapted to the realities of other groups, like fishers, or other measures that were not universally applicable, but that it was not possible for the phase two hours-credit.

[147] Furthermore, although a GBA+ was done for this measure, the analysis is confidential and was therefore not entered into evidence. The only statement we can rely on relates to Interim Order No. 1, which says that the EI ERB “[does] not target persons of any gender or identified group.”^{Footnote 143} Unfortunately, this shows a very truncated and formal view of equality and does not help to demonstrate how, or even if, other measures were seriously considered. It only means that the government found, after the measure was put in place,^{Footnote 144} that it did not target women of other protected groups.

[148] I conclude that section 153.17 fails to meet the minimal impairment requirement because the Respondent has failed to show that this measure impairs the right to equality as little as reasonably possible to achieve the legislative purpose.

[149] Because of this finding, the infringing measure cannot be saved by section 1.

Remedy

[150] I now have to decide what the appropriate remedy is in this case. The parties have argued extensively over my jurisdiction to grant a remedy under section 24(1) of the Charter. But I find that it is not necessary to decide this question in this case. I will instead focus on the application of section 52 of the Charter.^{Footnote 145}

The legal principles

[151] To begin with, the Supreme Court of Canada jurisprudence is clear. When tribunals have the power to decide general questions of law, they also have the power to decide whether the laws they have to apply respect the Charter. But their jurisdiction is limited: They only have the power to grant a declaration of invalidity that will apply to the case before them.^{Footnote 146} This means that I do not have the power to make a general declaration of invalidity that would apply to other cases before the Tribunal or to other women in the same situation.^{Footnote 147}

[152] Section 52(1) of the Charter reads as follows:

52 (1) The Constitution of Canada is the supreme law of Canada, and any law that is inconsistent with the provisions of the Constitution is, to the extent of the inconsistency, of no force or effect. (my emphasis)

[153] This means that I do not have to declare the whole law inapplicable because only one section, or part of a section, is found to violate the Charter. I can declare only that specific section inapplicable. As we will see, this is also true of a section of a law: I do not have to refuse to apply the whole section of the Act if only part of it violates a Charter-protected right.

[154] In Ontario v G, the Supreme Court of Canada set out the following “fundamental remedial principles” to guide decision-makers “in determining the appropriate remedy” in cases such as ours:^{Footnote 148}

1. A. Charter rights should be safeguarded through effective remedies.
2. B. The public has an interest in the constitutional compliance of legislation.
3. C. The public is entitled to the benefit of legislation.
4. D. Courts and legislatures play different institutional roles.

[155] This decision teaches us that the first step in crafting an appropriate remedy under section 52 is to clearly identify the extent of the inconsistency with the Charter. This is an important step because “the breadth of the remedy ultimately granted [should] reflect at least the extent of the breach.”^{Footnote 149}

[156] But the remedy must also be crafted in a sufficiently narrow way as to not prevent the parts of the law that do not violate the Charter from applying. This rule ensures that principles C and D cited above are respected.

[157] This is not always possible when striking down a section of legislation. Sometimes, even more precise mechanisms are needed. This is why “remedies of reading down, reading in, and severance, tailored to the breadth of the violation, should be employed when possible so that the constitutional aspects of legislation are preserved.”^{Footnote 150}

[158] Because they interfere with the wording of a law, a matter generally left to the legislator, those remedies must be used with caution. Two rules guide their application. The first is that they can be used only “where it can be fairly assumed that ‘the legislature would have passed the constitutionally sound part of the scheme without the unsound part’” and “where it is possible to precisely define the unconstitutional aspect of the law.”^{Footnote 151}

[159] These principles are generally applied by the courts. But I see no reason why an administrative tribunal with the power to apply section 52 of the Charter would not have the corresponding power to craft a remedy that fully respects the fundamental remedial principles listed in paragraph 156 above.

Application to this case

[160] As we have seen, the first step is to determine the extent of the law’s inconsistency with the Charter. Here, I have found that section 153.17 of the Act violates section 15 of the Charter in two ways:

1. i. by making it automatically apply on the first claim for benefits, even if it was not needed to qualify
2. ii. by negating the possibility of applying the credit that was not used when making a second claim for a different type of benefits during the same period

[161] This design perpetuated and reinforced the disadvantage suffered by women, and more precisely by pregnant mothers or those who recently gave birth. It did this by preventing them from receiving benefits specifically provided for them in the Act.

[161] The second step of the inquiry is to determine whether a declaration of invalidity applying to the whole law must be made or whether a tailored remedy would be more appropriate. Here, clearly a tailored remedy is preferable. No one has argued that the Act in its entirety should be struck down, and this is obviously not a desirable result.

[162] It would also not be a desirable result to strike down the entirety of section 153.17 of the Act. This would go against the purpose of facilitating access to benefits because then, no credit of hours would be available.^{Footnote 152} As stated before, the remedy must be sufficiently large to cover the whole Charter breach, while being sufficiently narrow to cover only that breach.

[160] In this case, the appropriate tool to remedy the constitutional deficiency is severance.^{Footnote 153} The Court describes this concept as follows:^{Footnote 154}

Severance is when a court declares certain words to be of no force or effect, thereby achieving the same effects as reading down or reading in, depending on whether the severed portion serves to limit or broaden the legislation’s reach. Severance is appropriate where the offending portion is set out explicitly in the words of the legislation.

[165] I find that the best way to remedy the impairment of the Appellant’s right to equality in the present case is to declare the second paragraph of section 153.17 of the Act to be of no force or effect.^{Footnote 155} This is the paragraph that limits the possibility of applying the credit on more than one claim for benefits and is the only part of the section that impairs the right to equality.

Conclusion

[163] The appeal is allowed. I find that section 153.17 of the Act violates section 15 of the Charter and is not saved by section 1.

[164] I also find that, to remedy this violation, the second paragraph of section 153.17 of the Act is of no force or effect.

[165] Finally, I find that this decision applies only to the Appellant’s claim for benefits made on September 13, 2021.