Radical Incrementalism and Trust in the Citizen: Income Security in Canada in the Time of COVID-19

Abstract

This article documents Canada’s main public policy responses to promote income security among working-age adults during the coronavirus disease 2019 (COVID-19) crisis between March and early June 2020. This period of rapid policy change unfolded broadly in three phases, starting with minor adjustments to existing policy instruments, followed by larger amendments to a wider range of programs, and finally ending with the creation of new and quite generous benefits. The pathway of policy change is best described as incremental, but it resulted in a more radical shift to “trust but verify” to administer benefits rather than the pre–COVID-19 practice of verifying eligibility before paying benefits. The reasons and precedents for this decision are discussed. I conclude with some observations on the applicability and limitations of trust but verify for income security policy in the post–COVID-19 period.

Introduction

This article has two primary goals. The first is to document the main government policy responses in Canada intended to promote income security among working-age adults during the economic shutdowns ordered because of the coronavirus disease 2019 (COVID-19) pandemic. This review is not exhaustive; instead, it aims to capture the essential features and evolution of a period of very fast-moving policy change between March and early June 2020. This period covers the introduction of mandatory shutdowns of non-essential businesses, schools, child care providers, and government and community services. It also covers the start of the gradual lifting of these restrictions and a turn in the public discourse toward getting people back to work rather than keeping people at home.

Income security is an area of shared jurisdiction in Canadian federalism, but most of the policy measures were announced by the national government rather than provincial governments. My aim here is to ask what governments in Canada did to ensure that individual Canadians and families could make ends meet when the usual opportunities to earn income through paid work were suddenly interrupted across the country. The necessary public health restrictions to contain the spread of the virus also led to a sudden and unprecedented spike in layoffs and a decline in paid hours worked among those who remained employed (Statistics Canada 2020). Although the focus here is primarily on working-age adults, the instruments used have extended beyond working age, both up and down the life cycle.

The second goal of this article is to review the trend in COVID-19 policy-making to understand how and why governments suddenly embraced a “trust-but-verify” approach to income support. Income support programs in Canada, such as provincial social assistance, Employment Insurance (EI), the Canada Pension Plan (CPP), and child benefits, generally require a detailed application with documentation that the applicant meets the eligibility criteria. This usual “verify-then-trust” approach places a burden on applicants to collect the proof necessary to support their claim and may lead to eligible people not participating as a result of stigma, gaps in information about or understanding of the program, and the effort required to apply relative to the expected income benefits that could be made available (Bhargava and Manoli 2015; Currie 2004; Moynihan, Herd, and Harvey 2015). The verify-then-trust approach, however, also comes with a heavy burden on program administrators, who must verify each claim and determine the appropriate amount, if any, to pay out. By retracing the brief but reasonably complicated trajectory of Canada’s income security policy during COVID-19, one can see that the switch from verify then trust to trust but verify came incrementally and as a matter of necessity. It is a policy shift meant to be temporary, but one that might have lessons for post–COVID-19 policy-making.

Evolution of Income Support Policy during the Pandemic

Canada’s income security response to COVID-19 has not, despite the rapid pace of change, been a story of a sudden and sweeping overhaul to its social safety net. Instead, policy responses on income security evolved in a more stepwise and incremental fashion, building on past practice and responding iteratively to emerging challenges and concerns from external stakeholders. Studies of path dependency and incrementalism in Canadian social policy usually describe shifts over years or decades (Banting 2005; Banting and Myles 2013; Battle 2001; Béland et al. 2014), not days or weeks as has been the case during the COVID-19 crisis. This policy shift occurred, broadly, in three overlapping phases: a first phase of using existing policy instruments, a second phase of amendments to existing policy instruments, and a third phase of developing new and national tools to cope with the shortcomings of existing policy. It is worth noting that this third phase was iterative and included several waves of adjustments and amendments as the shortcomings of newly created policy instruments gained attention. These three phases, as I describe them, are thematic but still largely chronological.

Phase 1: Modest Changes to Existing Instruments

On 11 March 2020, in the first major announcement of federal policy response to COVID-19, the prime minister announced new spending for public health and testing, as well as the first income support measures (Office of the Prime Minister 2020b). This announcement made some modest changes to both the EI sickness benefits and the EI work-sharing program.

In normal times, EI sickness benefits are paid as special benefits to qualifying workers. For workers with at least 600 hours of insurable employment, sickness benefits can be paid for up to 15 weeks at the normal EI rate of 55 percent of maximum insurable income (currently $54,200) as long as applicants can prove they are unable to work for medical reasons, as certified by a medical practitioner (Employment and Social Development Canada 2020a).¹ According to the most recent EI annual report, $1.7 billion in sickness benefits are claimed annually by roughly 412,000 claimants and represent 9 percent of all EI benefits paid per year (Employment and Social Development Canada 2019).² The policy change announced on 11 March waived both the usual one-week waiting period (equivalent to a co-pay in insurance terms) and the requirement to provide a medical certificate to prove illness. These two policy changes were expected to cost $5 million (Office of the Prime Minister 2020b). These changes would, in principle, accelerate the speed and ease with which EI-eligible workers could receive benefits and reduce pressure on health care systems to provide medical certificates. Amid a global pandemic, when policy-makers were concerned about excess demand on the health care sector, there would likely have been limited marginal value in asking sickness benefits claimants to obtain a medical certificate to prove their illness. This change, although modest, is notable because it converted EI sickness benefits into a benefit available by attestation, rather than verification, at least on a temporary basis. Research on comparable insurance systems suggests that medical certificates can be of limited value as a tool for policing benefits eligibility (Söderberg and Alexanderson 2005)

The other incremental change made to the EI system on 11 March was to extend work-sharing agreements. Under agreements between the federal government and employers, EI benefits can be paid to workers who collectively agree to reduce their paid hours of work and share jobs in a workplace to avoid layoffs (Employment and Social Development Canada 2020b). On 11 March, the government announced that the maximum duration of such agreements would be doubled from 38 to 76 weeks and that firms who had recently used the program would no longer have to observe a cooling-off period before reapplying (Office of the Prime Minister 2020b). Claims for EI benefits under a work-sharing agreement are generally very small, at just 3,700 of a total of 1.82 million EI claims in fiscal year 2017–18 and paying $13.1 million in total benefits per year (Employment and Social Development Canada 2019). The government estimated that the changes announced on 11 March would cost $12 million, essentially doubling the annual spending on the program.

Phase 2: Amending Existing Tools

A second phase of policy responses for income support during COVID-19 was announced just a week later on 18 March. By this time, provincial and local officials had begun implementing a series of emergency orders to significantly restrict or close places of work, as well as closing child care and schools (Robson 2020). The sectors directly affected earliest by these measures, including arts, entertainment, accommodation, food services, air transportation, and the retail trade, accounted for approximately two million jobs alone, or roughly 10 percent of the Canadian labour force.³ During this time, applications for EI benefits started to surge and created a significant backlog. As the minister responsible for EI later testified in Parliament, at the normal pace of processing claims, and applying the usual process of verifying application details and determining the right level of benefits to be paid, estimates were that it could take as long as 18 months to respond to the millions of claims received (Canada, Parliament 2020a). In addition to the backlog of likely EI-eligible claimants, many millions of workers were now facing economic impacts from COVID-19 but would not be eligible for benefits under the EI system. More than one-third of unemployed workers in Canada will not be covered by any EI benefits (Statistics Canada 2019). According to the same Statistics Canada report, coverage in EI during unemployment is also lower among younger workers and women, subpopulations who would see proportionally larger losses of paid employment during the COVID-19 shutdowns.

The announcement on 18 March was regarding a mix of measures intended to improve temporary income security for Canadians by reducing ongoing or immediate financial obligations, providing one-time top-ups to existing benefits, and creating a Temporary Wage Subsidy (TWS) for employees. Each of these are discussed in more detail later. Before discussing this policy mix, it is important to note that the 18 March announcement proposed the creation of two temporary emergency benefits, outside of EI, to provide assistance to people unable to work as a result of COVID-19. These benefits were not, in fact, ever implemented but were instead merged into the Canada Emergency Response Benefit (CERB) in the third phase of the policy response on income security. As such, they are discussed later, as part of the background of the CERB. One other policy change increased the flexibility in Registered Retirement Income Funds (RRIFs), by reducing the required RRIF withdrawals by older Canadians at a time at which financial markets had plummeted.

To reduce ongoing or immediate financial obligations, the federal government announced that payments on Canada student loans would be suspended for six months, without interest or penalties (Office of the Prime Minister 2020a). Provincial loan programs followed suit, either tied to the federal announcement and federal administration of their student loans or through their own independent announcement. The federal government expected this would cost $190 million in foregone interest payments on outstanding loan balances. Student loan debt is reported by 27.2 percent of households with a major income earner aged younger than 35 years, and mean student loan debts for these households is reported at $13,000.⁴ Typically, students who want to request a reduction in or other relief to their federal loan repayments must apply to the Repayment Assistant Program (RAP) and must renew their application every six months. Research by Lochner, Stinebrickner, and Suleymanoglu (2018) finds significant non-participation by eligible borrowers in the program. That is, many borrowers who are otherwise eligible do not apply for and so do not receive repayment relief, even if it would significantly reduce their debt and increase their monthly cash flow. In the case of the COVID-19 policy response, federal (National Student Loans Service Centre 2020) and provincial (cf. Quebec 2020b) policy-makers instead made the relief automatic and proactive so that no application would be required. This approach was no doubt possible because the loan administration systems had complete and up-to-date information about borrowers and could unilaterally suspend payments to all borrowers without requiring any action on the part of individual Canadians. Another contrast with the RAP program is that the repayment holiday was applied universally, whereas RAP repayment relief is normally needs tested and requires users to demonstrate that they are continuously eligible to be able to receive assistance.

Two other policy measures in the 18 March announcement also aimed at increasing immediate household income by deferring some financial obligations. The government announcement included an agreement for six of Canada’s largest financial institutions to provide deferrals on financial obligations such as private loans and mortgages for up to six months (Office of the Prime Minister 2020a). Using the Public Use Microdata File of the 2016 Survey of Financial Security, I estimate that mean mortgage debt among working-age⁵ Canadian homeowners who have a mortgage is $211,800 and that mean monthly mortgage payments are $1,371. Unlike the Canada Student Loans Program relief, deferrals on mortgage and other loans would continue to incur interest. This agreement with the banks must be viewed, however, in the context of another measure from the same date: the government’s announcement that the Canada Mortgage and Housing Corporation (CMHC) would purchase up to $50 billion in CHMC-insured mortgages from the banks. This would provide greater liquidity to lenders, and the government expressed willingness to do even more through the CMHC to increase liquidity among mortgage lenders. Consumers who wanted to access this new flexibility were asked to negotiate with their own financial institution.

At the time of writing, there is almost no public information on Canadians who did or did not successfully negotiate a credit or mortgage deferral with their bank. In testimony to the Standing Committee on Finance of the House of Commons, CMHC reported that 13 percent of all CMHC-insured mortgages had received some form of deferral (Canada, Parliament, 2020b). The estimate provided by the industry association of banks at the same time, and accounting for both insured and uninsured mortgages, was only slightly higher at 15 percent of mortgages (Canadian Bankers’ Association 2020).⁶ In late May, CMHC also stated that it expects that as many as one in five homeowners with an insured mortgage may take a deferral on payments by September. In normal times, Canadians are rarely delinquent in their mortgage payments. Using the same 2016 Survey of Financial Security, I find that just 4.2 percent of working-age households report delaying or missing a mortgage payment in the past 12 months. Official estimates of delinquency from CMHC range between 0.17 percent and 1.93 percent (CMHC 2019). This rate may increase in 2020 and 2021 if homeowners who sought a deferral are not able to resume regular payments (plus the additional interest) when deferral arrangements expire. Nevertheless, the additional liquidity for homeowners will have offered some relief, but this assistance will have gone to households who were more likely to have liquid savings to make ends meet when their incomes were interrupted. Previous research on the rate of asset poverty⁷ in Canada suggests that homeowners are significantly less likely to be asset poor than are renters (Rothwell and Robson 2017). Relief for renters came instead in the form of provincial actions to temporarily suspend eviction proceedings, with only two provinces (British Columbia and Prince Edward Island) offering new temporary income-tested financial assistance to residential renters (Robson 2020).

On 18 March, the federal government also announced that it was postponing the due date for filing an personal income tax return from 30 April to 1 June 2020 and providing a penalty-free deferral on any tax amount owing to the end of August (Office of the Prime Minister 2020a). This reduced immediate pressures to file a return, given that in-person fee-for-service and volunteer tax filing services could not operate under the emergency orders. Because the federal government serves as the tax collection agent for nine provinces, this decision also effectively deferred collection of provincial income taxes, with Quebec following suit.⁸ Tax filing has two distinct but important financial impacts on households. First is the possibility that filing will reveal a balance owing to government. The option to defer tax payments until the end of August may have increased immediate cashflow for those taxpayers required to make a payment. A second impact of filing a return, particularly for low- and modest-income Canadians, is on accessing or retaining several income-tested benefits, including the Canada Child Benefit (CCB), the Goods and Services Tax (GST) Credit, the Canada Workers’ Benefit, Old Age Security, and the Guaranteed Income Supplement. Lower- and modest-income Canadians may see an important increase in their net income when they file a tax return, and, like other Canadians, a majority (55 percent) rely on third-party tax preparers to complete their return (CRA 2017a).⁹ Delaying the tax-filing deadline appears to have enabled CRA and organizations that hold volunteer tax clinics to develop new models for distance or virtual clinics to reduce person-to-person contact (CRA 2020f). However, the later filing date has also affected the assessment of income benefits. In normal years, tax returns filed by 30 April are assessed for benefits, and amounts payable to eligible filers are adjusted starting with payments in July. With a delayed tax deadline, CRA was also eventually required to assure Canadians receiving certain benefits that payments would be grandfathered, based on the previous year’s tax return, until at least the end of September (CRA 2020b).

However, tax filers who did not file by 1 June will have key benefits halted in October and until some time after they do file a tax return. As Saul Schwartz and I discuss in a forthcoming article in this journal (Robson and Schwartz forthcoming), an important share (approximately 12 percent) of working-age adults do not file a tax return. It is possible and even likely that, in the context of a pandemic, with more limited to access to third-party help to file a return, the share of non-filing Canadians will increase in 2020 for the 2019 tax year. As many have argued, this problem of non-filing and disruption in important income benefits could be alleviated if the CRA were to take more responsibility for completing tax returns, or even deeming returns to have been filed, for low-income Canadians when adequate third-party information (from information slips submitted to the agency)¹⁰ is readily available to the agency (Cameron et al. 2020; Sanger 2020; Schwartz and Robson 2020; Stapleton 2013).

Another key policy change made during this second phase was to create one-time top-ups to income benefits and to extend access to job-protected leave for COVID-19–related reasons. The 18 March announcement included temporary top-ups to the GST Credit and to the CCB. The GST Credit is an income-tested refundable benefit paid quarterly to approximately 10.5 million low- and modest-income tax filers at an annual cost of just less than $5 billion (Department of Finance Canada 2020c). The CCB is also income tested but reaches well into the upper-income distribution as a monthly cash transfer to 3.5 million families with children (Department of Finance Canada 2020b). As discussed earlier, both credits are determined in relation to family income for the previous tax year. At the time of the prime minister’s announcement, the government estimated that the payments could be issued roughly 60 days later, in May (Fife and Curry 2020). No applications were required, and CRA simply used existing tax filer data to determine eligibility. Because individual recipients might not yet have filed their 2019 tax return, eligibility (by income as well as family composition) and contact information (including mailing address or banking information for direct deposit payments) would largely have been from 2018 tax returns filed nearly a year earlier.¹¹ On the one hand, this assumes that the 2018 circumstances of Canadians was indicative of their Spring 2020 circumstances in the midst of a global pandemic. If family income had suddenly decreased in the wake of the COVID-19–related shutdowns, that change would not make them eligible for more support than they had received throughout 2019. On the other hand, this approach meant that CRA could issue emergency payments reasonably quickly and without requiring action on the part of Canadians. In fact, payments for the GST credit were made on 9 April (CRA 2020h), a month sooner than the prime minister had announced on 18 March. The agency’s capacity to issue payments quickly, conditional on adequate taxpayer information being on record, would become an important early indication of administrative capacity to deliver new emergency cash benefits later on.

The 18 March announcement also included a new federal TWS payable to employers to reduce payroll costs. The subsidy would allow eligible employers to claim up to 10 percent of the remuneration paid to employees during the period of 18 March–19 June, to a maximum of $1,375 per employee and $25,000 per employer (Office of the Prime Minister 2020a). To claim the subsidy, employers would be permitted to reduce their monthly remittances to the CRA for income, but not payroll, taxes withheld at source (CRA 2020g). The goal of the subsidy was to encourage employers to continue paying as many employees as possible, even if the business had to temporarily shutter to comply with local emergency orders. In a sense, this would function as a form of temporary and cost-shared paid leave outside of the EI system that would try to avoid layoffs (even if temporary) and preserve more employer–employee relationships. The literature in labour economics is clear that the costs of employee turnover are non-negligible for both the firm and the individual worker (Bertola 1992; Kugler 2004; Faia, Lechthaler, and Merkl 2014).

Although it may be argued that work-sharing agreements that use EI to offset wage losses are a form of wage subsidy, direct subsidies payable to employers have not been widely used in Canada in recent memory. Wage subsidies in Canada have instead been piloted more often as a means to encourage employment and welfare exit among very-low-income working-age adults, and with some success (Connolly and Gottschalk 2003; Lacroix and Brouillette 2011). The logic of the TWS, however, was not necessarily about increasing the performance of paid work. Rather, in the context of emergency orders, it was intended to preserve worker–employer relationships in the short term (ending after just 90 days) in the hopes that unsubsidized paid work could more easily resume when restrictions are lifted, as some observers had called for (Yalnizyan and Robson 2020). The TWS, however, appears to have imposed a ceiling on employers that was too low to be useful. Targeting relief to smaller firms would have been supported by pre–COVID-19 evidence that access to paid leave for sickness or other reasons (among other indicators of job quality) is significantly lower among workers in small firms (Chen and Mehdi 2018). Subsequent downward revisions to the projected fiscal cost of the TWS, however, suggest that take-up was not as large as originally planned. The government’s initial projection of the cost of the program was $3.8 billion (Department of Finance Canada 2020b), but it was soon revised downward to just $975 million (Department of Finance Canada 2020a). Assuming as an approximation that each firm receives the maximum subsidy of $25,000 per firm, this suggests a downward projection in take-up from 152,000 firms to just 39,000. More recent analysis from the Parliamentary Budget Officer projects that the final cost of the program will be even lower at $844 million (Ammar and Ahmed 2020), suggesting that just 33,760 firms will receive the subsidy.

Why would firms refuse an opportunity to retain staff and avoid costly turnover when restrictions could be lifted? First, the share covered by government was likely too low to be attractive to employers who could otherwise reduce their wage costs to $0 by temporarily laying off employees without pay. Second, the subsidy was received passively by firms not remitting employee personal income tax deducted at source¹² instead of being paid as a cash transfer. It was a small subsidy that could be delivered without an application and nearly zero administrative burden on CRA. Conversely, however, it could only be delivered this way because it was small in value. Although a somewhat larger benefit could have been delivered passively by not requiring remittance of EI and CPP/Quebec Pension Plan premiums as well, a more generous benefit that exceeded the value of employer remittances would have required a positive application to CRA so that firms could self-identify as eligible and request a positive cash transfer from the agency. This is, in fact, what happened in the third wave.

No doubt, the decline in projected take-up of the TWS is also because, on 11 April, the federal government revealed the details of a much more generous wage subsidy, the Canada Emergency Wage Subsidy (CEWS; Department of Finance Canada 2020a). This was less than two weeks after the TWS became available. The description of the new subsidy (CEWS) made it clear that the amount claimed under TWS would reduce the amount people could receive for CEWS for the same period. The CEWS is among the new programs that were launched in what I describe, next, as the third phase of policy response to the effects of the pandemic on household finances.

Phase 3: Creating New National Programs

In the third phase of policy response, provincial governments began to take action on the household financial effects of the public health orders. In normal times, provincial governments play a key role in income support through residual welfare programs and child benefits or, in the case of Quebec, through a separate and parallel system for some EI benefits. In the COVID-19 response, however, provincial responses on income support were decidedly subordinate to the federal policy package, a dynamic that might have been unthinkable even a few months before.

Between the 18 March announcement and the 6 April implementation of the federal government’s largest emergency income support program, four provincial governments (Alberta, Saskatchewan, Quebec, and Nova Scotia) introduced temporary, one-time cash benefits to tide their residents over (Robson 2020). These benefits were generally in the range of $1,000 and had very short-lived application periods, even if they had very inclusive eligibility based largely on self-identification of economic hardship as a result of COVID-19. It is not clear that provincial governments had the administrative or technological capacity to deliver the cash benefits quickly to the large number of applicants in need. According to media reports in Alberta, many applicants found that the online application portal was closed at least twice because it could not handle the volume of demand (Bowen 2020). In Quebec, applications for the provincial temporary assistance were instead outsourced to the Red Cross (Canadian Red Cross 2020). In normal times, applications for the largest provincial benefit programs are often administered by the CRA (in the case of provincial child benefits) or through high-touch and detailed assessment processes in which people generally meet in person with a government employee, even if they complete an online application, before benefits can be paid. Applicants to provincial welfare are also required to provide extensive documentation to support their claim before benefits can be paid.¹³ It is noteworthy that provinces did not choose to use their existing delivery networks for social assistance to provide the temporary cash benefits for COVID-19. It is possible that they recognized the potential stigma associated with presenting for aid through residual welfare as a political risk. It is also possible, given the previously mentioned technical difficulties and outsourcing, that social assistance administration systems were simply not well suited to issuing a large volume of payments with only a light touch on verification.

Although the federal government had moved early to negotiate relief for homeowners with mortgages, regulation of rent falls within provincial jurisdiction. Provincial governments did universally act to temporarily suspend rental evictions, but only two (British Columbia and Prince Edward Island) introduced new, if modest, rental assistance programs during COVID-19 (Robson 2020). At the time of writing, the federal and provincial governments had not yet moved ahead to implement the cost-shared Canada Housing Benefit proposed in the 2017 National Housing Strategy. This portable and income-tested benefit might otherwise have acted as an automatic stabilizer for an important share of lower- and modest-income renters who are also less likely to have pools of financial assets to cover rent or other essential costs if their income is suddenly interrupted (Rothwell and Robson 2017).

Another area of provincial policy action that will likely have an important impact on household incomes during COVID-19 is the introduction of new job-protected leave related to the pandemic. Labour laws covering public servants as well as 900,000 employees in the federally regulated workforce (Statistics Canada 2016) were also amended to offer a new pandemic-related unpaid leave entitlement. In Table A.1 of the Appendix, I summarize the new unpaid leave entitlements.

Whether employers and employees will adhere to the intent and letter of the new emergency rule is as yet unclear. Among the new leave rules created for COVID-19, the federal policy change (at just 16 wk of unpaid leave) is actually much less generous than that of most provinces. The federal COVID-19–related leave provision does seem to have been harmonized with the new emergency income support program, the CERB. The CERB and the CEWS, along with a variant for post-secondary students and recent graduates (the Canadian Emergency Student Benefit) became the workhorses of the policy response to the effects of the pandemic on the household finances of Canadians. These are the focus of the remainder of this section of the article.

Canadian Emergency Response Benefit

As discussed earlier, the federal government announced its intent, on 18 March, to create two new emergency benefits for those workers without paid sick leave or access to EI and who need to take time off from paid work to self-isolate or care for a dependent. A week later, the government instead introduced legislation to create the CERB, both as a special benefit within EI and for those not covered by the EI system (Canada 2020b). In legislation, the CERB created an entitlement for those aged 15 years and older to a taxable payment of $2,000 per four-week period between 15 March and 3 October, to a maximum of 16 weeks. It did not differentiate between persons who could not work because of illness, layoff, or caregiving responsibility. Unlike EI, access was not dependent on past social insurance contributions or number of insurable hours worked but instead on a uniform threshold of just $5,000 in work income in 2019 or over the 12 months previous to the CERB application. Work income included traditional employment, self-employment, non-eligible dividends (small business dividends), and EI benefits paid to parents on maternity or parental leave. The legislation denied CERB eligibility to workers who quit their employment to take CERB and made it clear that double dipping between CERB and EI would not be permitted (Canada 2020b, Part 2, s6[1]b.ii and s6[2]). Initial claims for CERB allowed, per the legislation, workers to work in up to two of the four weeks for the claim period. However, the same legislation, passed unanimously by all parties in the House and Senate, also required that an eligible recipient cease working (Canada 2020b, Part 2, s6[1][a]). In practice, this initially meant that claimants could not have any employment or self-employment income if they intended to apply for a second or subsequent payment. There was considerable public criticism of this feature,¹⁴ and on 15 April the federal government announced that CERB would allow claimants to continue to earn up to $1,000 in employment or self-employment income. In the context of a policy intended to allow workers to stay home and avoid infection and illness, the potential work disincentive of a flat earnings disregard does not seem to have been a key consideration in the policy decision. Although a worker making $1,001 might technically be disqualified from receiving CERB, the goal of the policy was not to encourage work but to allow workers to reduce their supply to protect public health. As Canada enters a new phase in the emergency, it has become clear that the flat earnings disregard is not as well suited to the task of encouraging aggregate increases in paid and safe work (Sterling 2020).

Applications for CERB opened on 6 April. Canadians were asked to apply through either the CRA or Service Canada, depending on their likelihood of being eligible for EI regular or sickness benefits.¹⁵ Applicants who were directed to the Service Canada site completed an application as though they were applying for EI sickness or regular benefits but without the requirement to submit a Record of Employment to support their claim.¹⁶ The EI application asks for information on the reason for job separation, education, ethnicity, union membership, family composition, and recent work history. Canadians receiving CERB through Service Canada actually receive their payment biweekly and must complete biweekly reports, as is the norm for EI, to retain their benefits. Applicants directed to the CRA instead completed a much shorter application that required only a valid social insurance number, contact and banking information, and information from a past tax return.¹⁷ Applicants to the CRA are paid a lump sum for the four-week benefit period. Applicants to both Service Canada and the CRA are required to complete an attestation of their eligibility, to consent to information collection to verify their claim, and to apply again in each benefit period in which they want to receive payment. The initial estimate of the fiscal cost of CERB was $24 billion (Department of Finance Canada 2020a). The Parliamentary Budget Officer (Perrault and Worswick 2020) estimated, at the end of April, that CERB would cost $35 billion. At the time of writing in early June, the current cost of benefits paid is already more than $43 billion.¹⁸ The government has also provided an updated estimate of the projected program costs that, at $60 billion, is more than double its original projection (House of Commons 2020).

Relative to legislation and regulation governing EI and provincial social assistance, the rules for CERB were remarkably light and inclusive. There were no rules on permissible reasons for application, just a requirement that paid work had stopped or fallen below $1,001 in a four-week period. By comparison, EI divides applicants and pays different amounts depending on whether they have been laid off, cannot work as a result of illness, or need time away for caregiving responsibilities. There was a very loose and flat minimum participation rule that recognized the widest possible range of paid work. The closest model in pre–COVID-19 policy might be Quebec’s parental insurance program that demands only $2,000 minimum work earnings before benefits can be paid but still requires past contributions to the social insurance program (Quebec 2020a). Perhaps the most radical change in CERB, relative to pre–COVID-19 income support policy, was a move to pay benefits on the basis of attestations by applicants rather than holding benefits until documentary evidence could be submitted, verified, and approved to support the claim. The federal and provincial governments had taken modest steps in this direction in earlier COVID-19 policy responses through temporary changes to EI sickness benefits, as well as the design of the TWS and temporary benefits introduced by provincial governments.

Kevin Milligan (2020) has argued that the trust-but-verify feature of CERB shortens the time between application and payment delivery. In the context of a crisis, this is a non-negligible issue. As discussed earlier, the government estimated that processing EI claims using the regular system would take up to 18 months, leaving millions of Canadians without income assistance when they needed it. In normal times, the EI system processes a little more than 1.8 million claims per year, of which 1.7 million are claims for regular or sickness payments (Employment and Social Development Canada 2019). Ten days after the launch of CERB, the Government of Canada began publishing rolling summaries of the number of CERB claims made and processed. In Table 1, I summarize these data for the period from 16 April (the earliest available) through 2 June, the most recent data available as of the time of writing. What these results show is that CERB applications in the first ten days were more than three times larger than EI applications, for all types of benefits, in a full year. This suggests that the minister’s public statement on the projected delay to process claims using usual methods was very plausible.

Table 1:

CERB Application Data

Date	Applications in Week to Date (1000s)	Unique Applicants, Cumulative (millions)	Total Applications Received, Cumulative (millions)	Total Applications Processed, Cumulative (millions)	Total Benefits Paid, Cumulative ($billions)	Mean Benefits Paid per Unique Applicant ($1000s)	Notes
15 Mar 2020	NA	NA	NA	NA	NA	NA	Start of 1st CERB-eligible payment period
6 Apr 2020	NA	NA	NA	NA	NA	NA	CERB applications open
11 Apr 2020	NA	NA	NA	NA	NA	NA	End of 1st CERB- & CEWS-eligible payment period
12 Apr 2020	NA	NA	NA	NA	NA	NA	Start of 2nd CERB- & CEWS-eligible payment period
16 Apr 2020	NA	6.39	7.9	7.55	17.35	2.715
19 Apr 2020	NA	6.73	8.68	8.38	19.8	2.942
21 Apr 2020	528	6.82	9.2	8.9	21.3	3.123
23 Apr 2020	971	7.12	9.65	9.51	22.4	3.146
28 Apr 2020	316	7.26	10.31	10.15	25.63	3.530	CEWS applications open
5 May 2020	218	7.59	11.02	10.84	28.57	3.764
7 May 2020	401	7.76	11.21	11.12	29.05	3.744
9 May 2020	NA	NA	NA	NA	NA	NA	End of 2nd CERB- and CEWS-eligible payment period
10 May 2020	569	7.83	11.38	11.29	30.48	3.893	Start of 3rd CERB- and CEWS-eligible payment period
14 May 2020	2400	7.98	13.77	13.72	35.88	4.496
19 May 2020	128	8.11	14.35	14.31	38.41	4.736
21 May 2020	340	8.16	14.57	14.54	38.98	4.777
24 May 2020	516	8.21	14.74	14.71	40.33	4.912
26 May 2020	175	8.25	14.92	14.9	41.03	4.973
28 May 2020	318	8.29	15.06	15.05	41.44	4.999
1 Jun 2020	403	8.33	15.18	15.17	42.59	5.113
2 Jun 2020	142	8.37	15.32	15.32	43.18	5.159

Notes: CERB = Canada Emergency Response Benefit; CEWS = Canada Emergency Wage Subsidy; NA = not available.

Source: Canada (2020); Internet Archive Wayback Machine.

The data in Table 1 also suggest that, in the first ten days of the program, nearly 1.6 million Canadians filed two CERB claims. Without access to as-yet unpublished internal data, it is difficult to explain this result. First, by 16 April, the first CERB payment period (15 March–11 April) had passed and the second one had started. The program rules allow for claims to be made both retrospectively and prospectively. Some share of the 1.6 million double claims may have been claims for the second CERB period (12 April–9 May). As described earlier, CERB was implemented as a parallel program at both CRA and Service Canada, depending on expected EI eligibility. It is possible, and in fact likely, that some share of applicants were confused during the online application process and applied to both CRA and Service Canada in error. In normal times, the EI system experiences errors in about 4 percent of application files in a year, and the single largest source of these errors are claimants themselves (Employment and Social Development Canada 2019). The federal government has acknowledged that double payments have occurred and has assured Canadians such payments will be reconciled, including applications to both CRA and Service Canada (Service Canada 2020).

Another possibility, raised by some, has been that an important share of claims for CERB have been ineligible or fraudulent claims (Jones 2020; Poilievre 2020). Similarly, news media report that as many as 200,000 claims for CERB had been red flagged as likely ineligible but paid anyway (Blackwell 2020) and that payments have even been issued to prison inmates (Cullen 2020). In testimony to the Commons Finance Committee, officials stated that the 200,000 figure was for applications made, perhaps in error, to both the CRA and Service Canada for the same benefit period (Canada, Parliament 2020c). Generally, studies of cheating in social welfare and taxation find that deliberate non-compliance is quite rare (Andreoni, Erard, and Feinstein 1998; CRA 2017b; Kleven et al. 2011; Mosher and Hermer 2005; Saad 2014). Deliberate cheating in social welfare programs can also be hard to establish, given changing and complex rules and difficulty proving intent to contravene those rules (Mosher and Hermer 2005).

Using more robust methods to compare CERB claims with contemporaneous data on job losses, labour force exits, and lost hours of paid work, Skuterud (2020) found that an important share (just more than one million) of CERB claims could not be explained by labour market outcomes using best available data. Skuterud allowed, however, that CERB-eligible self-employed persons would not be captured by Canada’s official labour market data.¹⁹ In late May, the CRA amended its standing program for reporting suspected fraud and non-compliance to encourage Canadians to report cases of suspected misuse of CERB and other COVID-19–related benefits (Forani 2020).

Although CERB is still available at the time of writing, are concerns of widespread fraud justified? There are at least two significant difficulties in assessing allegations of widespread fraud in CERB claims. The first is, as highlighted by Skuterud (2020), that there are significant gaps in the arm’s-length data sources that might otherwise allow governments and outside stakeholders to estimate the extent of fraud. The second is that Parliament created a statutory benefit for which the eligibility criteria may not be easily proven or disproven by either administrators or claimants. For example, the requirement to have $5,000 in paid work income to qualify for CERB might be verified against past income tax records, but the CERB claimant may not have filed a tax return for the period in question (particularly because that period may be in the first part of 2020). Likewise, the legislative requirement to have not quit paid employment to collect CERB will likely prove difficult to verify without a Record of Employment and even then will be subject to considerable subjective judgment.²⁰ On 10 June, the Government of Canada introduced a bill to amend the legislation governing CERB (Canada, Parliament 2020a). The proposed legislation created severe new penalties, including fines of up to triple the amount of CERB payments received and six months imprisonment. The draft legislation also proposed withdrawing eligibility for CERB if the government determined that a beneficiary had refused to return to work, including self-employment, when it was “reasonable” to do so. In the concluding section of this article, I return to the question of fraud and the respective trade-offs among speed of payment, ease of application, and adjustments to benefit amounts.

As CERB was launched, it became apparent that many students and recent graduates would be unlikely to qualify for the benefit and would be more likely to face weaker job prospects even after a broader economic recovery (Neill and Foley 2020). After significant pressure from student organizations (Canadian Alliance of Student Associations 2020; Canadian Federation of Students 2020), the federal government created a separate emergency benefit for students. This was on top of millions in funding for youth employment programs, job placement, enhancements to student financial assistance, and extension of research grant funding supporting students (Department of Finance 2020d). As a sort of scaled-back version of the CERB, the Canada Emergency Student Benefit (CESB) pays $1,250²¹ per four-week period from May to August to post-secondary students (including secondary students moving to post-secondary in Fall 2020) and recent graduates who do not otherwise qualify for the CERB and have either lost work or been unable to start work as a result of COVID-19 (CRA 2020c).

Unlike CERB, CESB applicants are required to search for work while receiving the benefit, either by registering with a federal job bank or through other means. As with CERB, applications for CESB are made through the CRA My Account system that verifies basic information such as name, social insurance number, and address against a recent income tax return. Many students will not, however, ever have filed a tax return (Robson and Schwartz forthcoming). In these cases, students must first call the CRA to register their social insurance number. This small detail of CESB is important because it highlights the degree to which the tax agency is blind when it comes to non-filing Canadians. Moreover, the CRA does not have independent access to the registry of social insurance numbers issued to Canadians by Service Canada (the Social Insurance Registry) and instead relies on tax filers and third parties to issue returns and tax records with a valid social insurance number (Employment and Social Development Canada 2015). In fact, the CRA’s internal data also appear to be partitioned into separate purpose-built databases (called data marts), depending on the intended use (CRA 2016). Relative to the CERB, the CESB is a small program with an estimated cost of just $5.2 billion. At the time of writing, claims for CESB have amounted to $713 million (CRA 2020a).

Canadian Emergency Wage Subsidy

The other main income security workhorse that was created during COVID-19 is the CEWS. In the previous section, I described the first wage subsidy and note that it was perhaps too small to have much impact. The CEWS program does not suffer the same problems of the TWS, but neither does it have its virtue of skipping any application process. This larger wage subsidy is available to most employers with payroll expenses,²² regardless of size (with public institutions excluded, including local government, hospitals, schools, and publicly funded post-secondary institutions), if the employer can demonstrate a decline of 30 percent²³ in revenues relative to a comparison period. As with CERB and CESB, the CEWS is also paid in renewable four-week periods to a maximum of 24 weeks, retroactive to March 15 (CRA 2020d).²⁴ The subsidy is worth 75 percent of employee wages or salary to a maximum of $847 per week in subsidy per employee and is not conditional on the employee working, only that the employee is being paid. This means that it can be applied to employees on paid leave and employees recalled from a temporary layoff and paid retroactively. In addition to the wage or salary costs, the federal government also refunds 100 percent of the employer share of payroll taxes for any employees on paid leave. Unlike the initial 10 percent subsidy, there is no upper limit on the total amount of subsidy that any one employer can claim during the program. The federal government projected that the cost of CEWS would be more than three times that of CERB at $73 billion (Department of Finance 2020a). Analysis by the Parliamentary Budget Officer suggested a similar cost at $76 billion but noted that the behavioral response of employers was highly uncertain (Ammar, Segel-Brown, and Ahmed 2020). The most recent public data suggest that claims for CEWS have been relatively soft, with 172,123 claims to date for the first four-week period and 150,782 claims for the second four-week period, for a total of $9.4 billion in CEWS payments made, or 13 percent of the projected total cost paid during the first third of the program’s total duration (CRA 2020e). It is difficult to assess, from public data, how employers are responding to the CEWS and whether policy-makers overestimated demand or whether there is widespread non-participation by eligible persons.

Like CERB and CESB, CEWS relies on an application process to the CRA. Also like CERB and CESB, the program largely operates on a trust-but-verify model by asking employers to provide relatively detailed declarations, as well as their own calculation of the subsidy they qualify for, but it does not require that documentation be submitted with the application (CRA 2020d). Instead, employers must keep records and make these records available on request to the CRA to verify eligibility for claims as well as reconcile any overlap between CERB and CEWS at the employee level. However, the burden of verifying eligibility does not fall on only the employer or CRA.

The government has been clear that individual employees cannot claim CERB and receive payment (even up to the $1,000 earnings disregard) from an employer through the CEWS. The current policy direction is to reconcile this overlap by recovering CERB payments from individual workers (Service Canada 2020). It is not immediately clear how the CRA intends to do this reconciliation. The CEWS application does not collect individually identifiable employee information (such as a social insurance number). Although tax slips issued by an employer will allow the agency to identify workers who received any compensation from a firm, this information source does not include within-year information that would allow the tax agency to differentiate between months of compensation paid using CEWS and other compensation paid without the subsidy. Without the monthly information, linked to individual social insurance numbers, it is not clear how the agency will identify periods of overlap between CERB and CEWS. It is possible that all or some sample of Canadian taxpayers who claimed CERB and whose employer collected CEWS will be subject to a file review by the agency and required to submit documentation or repay some or all CERB amounts they received. This would be an extraordinary administrative burden to impose on individual Canadians.

It is also not clear, given past reviews of the agency’s conduct by the taxpayers’ ombudsman, that all individuals who are subject to a review and receive a collection notice will be treated equally, depending on which part of CRA they interact with (Office of the Taxpayers’ Ombudsman 2019). The ombudsman’s review found that the agency is not always consistent or clear in its communications and that agents do not always provide full information to taxpayers on options available to them for responding to a request for payment. An earlier study, also conducted by the taxpayers’ ombudsman, concluded that taxpayers are often confused about the kinds of documentation needed to verify eligibility for benefits (in this case, child benefits) when required by the CRA as part of a file review (Office of the Taxpayers’ Ombudsman 2010). In practice, unevenness in CRA’s file review and collections processes may prove to be the most important weakness of the trust-and-verify approach that has been so widely used for income security during the COVID-19 emergency.

Conclusion

In my review of emergency income support policies for working-age Canadians during the initial wave of the COVID-19 pandemic in Spring 2020, I see at least two patterns: incrementalism in policy adaptation and adoption of a trust-but-verify model of delivery. The former describes the process or pathway of policy change, and the latter describes a significant outcome of that policy change.

The policy responses on income security shifted incrementally, and often iteratively, as governments responded to external stakeholders or changed course in the face of emerging challenges. The two largest workhorses of the income security policy response that emerged in the third phase of the policy responses, CERB and CEWS, both had predecessors at earlier phases and seem to have been designed more as evolutions of past programs that themselves had some origins in pre–COVID-19 social policy. The CERB, for example, replaced an overwhelmed EI system. However, one part of the program, payments, relies on the EI administrative architecture for benefit applications and has even maintained EI’s mandatory regular reports from recipients. What has been very unusual is the speed and frequency with which policy has changed and been revised in such a short period of time. Still, the chronology and details of Canadian income support policy in Spring 2020 show a familiar pattern of adjustments and evolution over time rather than emerging fully formed in one single sweep. This is not in any way to suggest that the incremental pathway of change has not resulted in some important or even radical departures from pre–COVID-19 practice. Incremental processes in policy-making can still yield dramatic outcomes in policy change, and it is worth distinguishing between process and results. The end result of the incremental change has been some important departures from the pre–COVID-19 norms. Governments in Canada did not immediately launch new support programs, and policy change took place in a rapid but still stepwise fashion. The net result of this accelerated incrementalism, however, is that both CERB and CEWS, as the two largest programs, have been far more inclusive than narrowly targeted in eligibility and more flexible in their use by Canadians because they have not required applications to different programs for different uses.²⁵ Perhaps the most striking change has been a shift, made incrementally from the earliest COVID-19 policy changes, away from traditional application processes intended to verify eligibility and toward a more widespread use of trust-but-verify systems to administer emergency benefits. Seeing this incremental process helps in understanding how and why trust but verify was adopted. I conclude this article with a discussion of the lessons that might be learned from this temporary experiment for post–COVID-19 policy.

Consistent with my finding of an incremental evolution of income support policy during the COVID-19 crisis, trust but verify is not without precedent in Canadian policy-making and has been used extensively in other policy areas, namely in tax policy. Trust but verify is the underlying logic of many long-standing non-refundable tax credits and deductions, such as the child care expenses deduction, the moving expenses deduction, and the medical expenses tax credit. For these and similar tax expenditures, tax filers are not asked to submit detailed receipts but simply to retain them in case they are requested by the CRA as part of a review. Credit for the expenses is claimed as an attestation of eligibility in which the assumption is that fraud and non-compliance will be deterred by the risk of being caught by the CRA. In fact, Canada’s personal income tax system largely relies on self-assessment and voluntary compliance (Brooks and Doob 1990). Models of non-compliance in tax policy generally assume that rational actors will only avoid taxes if the net gain of minimizing their taxes is large enough to offset the financial and social risks of being caught and penalized by the tax collection agency (Andreoni et al. 1998; Kleven et al. 2011). Using internal administrative records, the CRA has issued one national report on non-compliance in federal income tax filing (CRA 2017b). That report concluded that the total dollar value of uncollected personal income taxes owing was quite modest at just $2.2 billion for the 2014 tax year. In recent years, the CRA has significantly increased its access to third-party information on income from a wider range of sources. This has enabled the agency to, on the one hand, provide new services to tax filers such as automatically completing important portions of electronic tax returns by using tax information slips already on file and, on the other hand, to improve enforcement of tax rules.

Canada’s verify-but-trust administrative systems were simply not equipped to handle the volume of demand during the crisis. Earlier, I noted the statement by the minister responsible for EI about the projected delays in issuing benefits if the regular EI rules and application process had been left in place (Canada, Parliament 2020a), despite the fact that some three-quarters of EI applicants are processed using a partially or fully automated system (Employment and Social Development Canada 2019). I also noted that the minister’s statement is consistent with the significant difference between the volume of CERB claims in a matter of months and the normal volume of EI claims in a full year and that many Canadians experience unemployment or loss of income because of illness or caregiving but do not qualify for EI benefits under pre–COVID-19 rules. Add to this the finding that provincial governments faced administrative problems or needed to outsource applications to their single-payment and temporary COVID-19 income benefits programs, and the general picture is one in which Canada’s existing income support programs were not able to cope with the level of demand seen during COVID-19. An international comparison of COVID-19–related income support programs, including Canada’s, likewise concluded that “poor accessibility already before the crisis may signal significant bottle-necks during the claiming process. In these cases, existing safety-nets may not be able to cope with significant additional demand” (OECD 2020, 8). From this perspective, moving to trust but verify may be a matter of necessity if the alternative is delaying payments by weeks or even months when government help is no longer timely and relevant.

Looking ahead, could trust but verify be a useful principle for reforms to income support in the post–COVID-19 period? My answer is a cautious and conditional yes. One reason why trust but verify could be implemented during COVID-19 is that the benefit amounts and conditions for CERB have been relatively simple, with the acknowledged trade-off that the flat earnings disregard creates a potential work disincentive. As provincial reopening plans progress, there are growing calls to return to a benefit structure that encourages work through a graduated claw-back mechanism. This may be more challenging to administer using a trust-but-verify system but perhaps not impossible if, as with CEWS, individual claimants are supported to calculate the level of support they are personally entitled to and required to voluntarily declare verifiable information in support of their claim. This is, in normal times, much how the Working While on Claim program of EI operates, although it is important to note that the program seems to have very mixed effects on job search and returns to employment, particularly full-time employment (Lluis and McCall 2019). Although interesting as a proof of concept that trust but verify can work with gradual claw-backs and variable benefit rates, Working While on Claim may not have the optimal design for employment incentives.

The main source of my caution is instead on the verification side and the important principle in welfare policy of procedural fairness. As I noted in the discussion of CERB, Parliament created a statutory benefit in which the eligibility criteria may not be easy for administrators to verify or for claimants to demonstrate. Absent a Record of Employment showing that beneficiaries voluntarily quit their job, it is unclear how verification systems at either Service Canada or CRA will be able to identify ineligibility for having quit without cause. What is more, EI benefits do allow voluntary quitting when employees have reasonable concerns about the safety of their workplace. In the context of a global health pandemic, it is not impossible that some share of workers will need to exercise that right to leave a job but still receive benefits. Earning more than the $1,000 disregard is also likely to be challenging to verify. With existing sources of third-party information on income shared with the CRA, it seems unlikely that the CRA will be able to easily identify CERB claimants who earned more than the $1,000 disregard during the four-week periods in which they received benefits. If instead some share of CERB claimants are selected for review, then the findings of the taxpayer’s ombudsman ought to raise concerns about whether CRA will be transparent and clear in how reviews are conducted and what forms of evidence CERB recipients will be expected to produce to retain their benefits and avoid penalties at tax time next year. In short, even if Canada returns to some sense of normal economic activity by next spring, Canadian policy-making and administration will likely need to continue to be responsive to the unique circumstance created by the pandemic in Spring 2020.

In the context of an unprecedented economic crisis brought on by a global health pandemic, policy-makers faced trade-offs between the need to make timely payments and the administrative systems available to complete the task. They chose to square the circle by simplifying the benefit and adopting the trust-but-verify approach that had already been used in tax policy and, in small ways, at other phases of the COVID-19 policy response. Looking ahead, there is a clear imperative to introduce more targeting and to verify eligibility for the millions of Canadians who will continue to need income support until the labour market has truly recovered. The federal government, through Service Canada and the CRA, has a clear comparative advantage to be able to handle the scale of this ongoing demand. To efficiently and effectively complete the “verify” in trust but verify, however, the government is likely to find it needs new sources of information, such as within-year income information instead of annual amounts. This could be done, for example, by asking employers to share more detailed payroll information on monthly remittances, as is done in countries such as the United Kingdom (HM Revenue and Customs 2020), or, as Saul Schwartz and I propose in our forthcoming article for this journal, by accessing better data through Payments Canada. The burden of verification has to be reasonably shared between individual Canadians and program administrators. Access to verifiable third-party information would lift much of the burden off individual benefit claimants, as long as both government and individuals have access to the same information and individuals have a reasonable opportunity to dispute and correct errors in that information. The proposed new threat of hefty fines and even jail time for false CERB claims must be matched with procedural fairness and transparency in the government’s end of the verification systems for COVID-19–related benefits and, in fact, income support going forward (Canada 2020a).

Appendix

Table A.1 summarizes new leave provisions created for COVID-19 or, where no new leave was introduced, existing provincial labour laws allowing job-protected leave for caregiving. These provisions include caregiving for someone who is ill or for dependent children because of school and child care closures. Demand for job-protected leave for caregiving will have vastly exceeded demand for job-protected sick leave related to COVID-19. At the time of writing, there are just more than 93,000 presumed and confirmed cases of COVID-19 infection in Canada (Esri Canada 2020), but school closures will have affected nearly four million families across the country.²⁶

Table A.1:

Summary of Job-Protected Leave Introduced during COVID-19 Emergency

Jurisdiction	Duration of Job-Protected Unpaid Leave Related to COVID-19	Available for Caregiving?
Federal	16 wk unless more provided by regulation	Available for general reasons related to COVID-19
British Columbia	For as long as the circumstances described in legislation apply, retroactive to 27 January	Yes
Alberta	14 days unless more provided by ministerial order	No, for quarantine only
Saskatchewan	For the duration of an order by an employer, government, doctor, or chief medical health officer, retroactive to 6 March	Yes
Manitoba	For as long as the circumstances described in legislation apply, retroactive to 1 March	Yes
Ontario	Until the emergency order is terminated or disallowed, retroactive to 25 January	Yes
Quebec	No new leave created. Pre–COVID-19, 10 days available, of which 2 will be paid after 3 mo or more in service	Yes
New Brunswick	Sooner of duration of agreement between employer and employee or for as long as the circumstances described in legislation apply, retroactive to 12 March	Yes
Nova Scotia	14 days if ordered to self-isolate	No, for quarantine only
Prince Edward Island	No new leave created Pre–COVID-19: 3 unpaid days for family caregiving after 6 mo of service	Yes
Newfoundland and Labrador	For as long as the circumstances described in legislation apply, retroactive to 14 March	Yes

Note: COVID-19 = Coronavirus Disease 2019.

Source: Author’s summary from British Columbia, Employment, Business and Economic Development (2020); CNESST (2020); Alberta (2020); Manitoba (2020); Manitoba (2020); New Brunswick (2020); Nova Scotia (2020); Ontario (2020); Prince Edward Island (2020); Saskatchewan (2020); Newfoundland and Labrador (2020).

What the information in Table A.1 makes clear is that 11 of 13 jurisdictions acted to create new but unpaid leave entitlements in labour codes to ensure workers could retain a legal right to employment when their opportunity to work was interrupted because of COVID-19. Among the 11 jurisdictions, 9 ensured that caregiving would be included as a reason for unpaid absence, whereas 2 were more restrictive in their approach and covered only suspected or actual illness for individual workers. More striking is the variation in the duration and conditionality of the leave. Four jurisdictions appear to have created a fairly open-ended and inclusive leave entitlement, at least on paper.