Public Policy in a Time of Crisis: A Framework for Evaluating Canada's COVID-19 Income Support Programs.

Income support programs introduced for workers during the first wave of coronavirus disease 2019 (COVID-19) lockdowns faced criticism for their negative labour supply effects. We propose that these concerns about work disincentives are embedded in restrictive assumptions about work and led to suboptimal design of crisis support policies. We describe a framework for analyzing alternative crisis income support programs predicated on more realistic assumptions of labour markets and human motivation. Our framework proposes that balancing efficiency, equity, and voice objectives should be the goal of crisis labour market policies. We argue that adoption of a basic income targeted toward low-income workers, in combination with Canada's pre-existing Employment Insurance program, would have balanced efficiency, equity, and voice better than the combination of the Canada Emergency Response Benefit and Canada Emergency Wage Subsidy. A targeted basic income would also have been more effective at achieving stated public health objectives. © Canadian Public Policy / Analyse de politiques.

Introduction

The World Health Organization declared the novel coronavirus disease 2019 (COVID-19) a global pandemic in March 2020. That same month, the Canadian federal government closed its borders to all non-essential travel, and every provincial and territorial government declared a public health emergency that shuttered in-person operations for most non-essential businesses, government offices, and schools and some daycares. The public health restrictions and economic lockdowns resulted in one of the largest labour market shocks in Canadian history (Statistics Canada 2020) and had a negative impact on the ability of many workers and families to pay bills and meet basic needs.

In the earliest stages of the crisis, it appeared that the federal government was going to rely on slight modifications to Canada’s Employment Insurance (EI) program to provide income support for workers who lost their jobs as a result of pandemic lockdown orders. However, there were major limitations to this approach. Most notably, EI was unavailable to many of the most severely affected workers, and the program reportedly struggled to adjudicate the massive influx of new applications. To address these issues, the federal government subsequently introduced a new emergency income support program—the Canada Emergency Response Benefit (CERB)—as well as a wage subsidy program—the Canada Emergency Wage Subsidy (CEWS). In this article, we analyze and evaluate these two programs to inform the design of future crisis policies.

We have three broad goals. First, we present a framework for policy evaluation that highlights the importance of balancing efficiency, equity, and voice (EEV) in the optimal design of labour and social policy during a crisis. Our second goal is to explain why Canada’s flagship crisis support programs (i.e., CERB and CEWS) did not balance efficiency considerations with equity and voice objectives for workers. We argue that the design of these crisis policies was primarily driven by a broadly accepted, rarely questioned assumption that work is a utility-reducing activity. In mainstream economic models, work is viewed as a “bad,” not a “good,” and most people, in the absence of a cost for leisure (i.e., a wage for work), will prefer leisure to work. This basic assumption is premised on restrictive ideas about the operation of labour markets and human motivation, which in normal times results in a disproportionate focus on the mitigation of perceived work disincentives in both labour and social policy design. This preoccupation, we argue, carried over into the design of crisis income supports by tying them to participation in the labour market and attachment to employers. The third and final goal of this article is to use the efficiency–equity–voice framework to evaluate and compare the government’s main crisis support programs, specifically the CERB and CEWS, against alternative crisis policies, such as a universal basic income (UBI) or targeted basic income (TBI).

Our evaluation and comparison of crisis programs using the EEV framework leads us to conclude that an unconditional basic income targeted to low-income workers (those making less than $22,000 per year), as well as to all self-employed workers, introduced alongside pre-existing EI and provincial social assistance programs, would have produced an optimal balance among the three EEV objectives. A TBI introduced in combination with EI would have (a) minimized the direct and administrative costs of new crisis programs (efficiency); (b) resolved the issues of EI accessibility for low-income workers and served as a natural income “top-up” for low-income essential workers making less than $22,000 (equity); and (c) given low-income workers the freedom to choose their preferred level of exposure to COVID-19, the ability to raise concerns about safety with their employers, or both (voice). Decoupling new income supports from participation in the labour market during the pandemic under a TBI model would also have improved labour market efficiency during the crisis and into the recovery by avoiding the sharp cutoffs and earnings penalties embedded in the CERB and the stalling of inevitable economic structural adjustments present in the CEWS. A TBI would also have been administered more efficiently and with less cost by substantially reducing or eliminating the need for post-crisis audits of both workers and businesses.

Most important, a focus on balancing efficiency concerns with equity and voice for workers in the design of new COVID-19 income supports would have more effectively (and fairly) achieved stated public health goals. These goals, which included flattening the epidemic curve, reducing hospitalizations, and protecting vulnerable populations, required (a) having as many people as possible remain at home except to undertake essential activities and (b) providing all workers with an opportunity to choose not to work outside the home, on the basis of an assessment of their own COVID-19 illness or mortality risk and the safety precautions taken by their employers.

Analytical Framework for Evaluating COVID-19 Crisis Policies: Balancing Efficiency, Equity, and Voice

In this section, we present an analytical framework that can be used to guide the evaluation and comparison of COVID-19–related crisis policies. Our framework is premised on the long-standing normative concern in industrial relations with advancing efficiency and equity in labour markets (Barbash 1987; Budd, Gomez, and Meltz 2004).[1] However, because equity can theoretically be achieved without giving any voice to the individuals affected by administrative and distributive policy decisions, contemporary IR scholarship argues for the vital addition of voice to the equity–efficiency framework (Budd 2004, 23).[2] In the IR field, there is a unique preoccupation with balancing EEV objectives in the employment relationship and in the design of labour policies. What balance means is often in the eye of policy-makers. However, there is a consensus among IR scholars that too much emphasis on any one of these objectives—whether it is efficiency, equity, or voice—will lead to suboptimal outcomes for workers, businesses, and, ultimately, society as a whole (Budd et al. 2004).

The EEV framework provides a rich analytical foundation that academics and policy-makers can use to compare income support programs during the COVID-19 crisis and beyond. To this end, we discuss the key assumptions that underpin the EEV framework and then show how the framework can be applied to the analysis and evaluation of crisis support programs.

Key Assumptions of the Efficiency–Equity–Voice Framework

Two assumptions fortify the use of EEV as both an analytical and an evaluative framework, respectively: (a) labour markets are imperfect, leading to an inequality of bargaining power, particularly for the lowest-income workers, and (b) labour is embodied in human beings and should not be treated as a commodity. We discuss each in turn.

In the neoclassical theory of competitive labour markets, neither individual firms nor workers have the ability to significantly influence the market wage rate. The model assumes that there are no barriers to entering the labour market and that each employer is a wage taker striving to maximize profits in a competitive product market. Through the forces of labour supply (i.e., workers providing their labour to firms) and labour demand (i.e., firms hiring labour to produce goods or provide services), an equilibrium wage rate emerges, where the supply and demand for labour are equalized. At this equilibrium wage rate, when every individual who wants to work at that rate is able to, the labour market has “cleared.” The invisible hand of the market achieves an efficient outcome (i.e., Pareto optimality, discussed later), allowing firms to maximize profits and the value of consumption goods and workers to maximize earnings and leisure. Because both labour and capital are paid their productive value, outcomes are viewed as equitable or fair (i.e., marginal productivity justice).

In contrast, the EEV framework’s default assumption is that labour markets rarely adjust without frictions and are thus far from perfectly competitive (Kaufman 2010). Labour market frictions stem from myriad forces particular to labour (Rees 1993), such as mobility costs, positive job search costs, asymmetric information, segmented markets, lack of household savings, and excess labour supply. A similar argument also exists in labour economics, and the conclusions are clear: In other words, imperfect markets result in unequal bargaining power between employees and employers that usually favours the latter (Budd et al. 2004; Manning 2003).

The existence of [labour market] frictions means that there are generally rents to jobs—i.e., if an employer and worker are forcibly separated one or, more commonly, both of the parties would be made worse-off. This gives employers some market power over their workers as a small wage cut will no longer induce them to leave the firm. The assumption that employers set wages then tells us that employers exercise this market power. But, with these two assumptions, it is monopsony not perfect competition that is the best simple model to describe the decision problem facing an individual employer. (Manning 2003, 1)

IR scholars have argued that inequality of bargaining power leads to less equitable outcomes for workers (Barbash 1987; Budd 2004; Webb and Webb 1897). However, the extent of inequality in bargaining power is heterogeneous across workers. Workers who are low skilled, less educated, or in the bottom of the earnings distribution will generally have significantly less bargaining power than their higher-income or more-educated counterparts (especially in times of high unemployment such as during the pandemic lockdowns). Webb and Webb (1897) argued that for “workers at the bottom of the labor market … their reservation wage is grounded on family sustainability costs” rather than the marginal opportunity cost of leisure, and thus “absent a welfare state or substantial nonlabor family income, the choice for ordinary workers is less labor vs. leisure and more labor vs. homelessness and starvation” (cited in Kaufman 2013, 771). Low socio-economic status results in a lower reservation wage and thus a lower equilibrium wage rate for historically disadvantaged workers (Gomez and Lamb 2016; Kaufman 2013) than it would in a truly competitive market in which bargaining power approached equality among the parties. The first assumption of the EEV framework is thus that labour markets tend to diverge from competitive market structures and toward models of monopsonistic competition, with negative impacts on worker equity and voice, particularly for the lowest-income workers.[3]

The second assumption of the EEV framework is that labour is embodied in human beings who have “aspirations, feelings, emotions, needs and rights” (Budd et al. 2004, 197). Work therefore provides more than extrinsic, monetary rewards by “fulfilling important psychological and social needs,” and “a realistic analysis of consumer welfare requires explicit consideration of the nature of work and workers’ lives” (Budd et al. 2004, 197–98). The embodiment of labour in humans also means labour should not be treated as merely another factor of production, as it is in neoclassical economic models. Both analytical and normative questions about the treatment of workers, both in the employment relationship and in policy decisions, are unavoidable when considering the human functioning of the labour market (Budd et al. 2004).

These limitations of neoclassical models are not alien to economics. Indeed, as Thaler (1980, 39) described in a seminal article 40 years ago, The labour market is one of those well-defined situations in which the neoclassical model offers, at best, an incomplete description of how people actually behave.

the economic theory of [individual behaviour] is a combination of positive and normative theories. Since it is based on a rational maximizing model it describes how [individuals] should choose, but it is alleged to also describe how they do choose. … [However, it can be shown] that in certain well-defined situations many [individuals] act in a manner that is inconsistent with economic theory.

It is through the recognition that imperfect labour markets lead to unequal bargaining power and that workers are human beings that we evaluate crisis support policies for workers during COVID-19. In the next section, we describe the EEV framework in more detail and define efficiency, equity, and voice in relation to the current context. We also problematize key assumptions about work that fortify an over-emphasis on the efficiency objective in economics. Then, in the “Evaluating and Comparing COVID-19 Crisis Support Programs” section, we evaluate alternative crisis support programs against the standard that these programs balance labour market efficiency with equal consideration of equity and voice for workers.

Applying the Efficiency, Equity, and Voice Framework to Crisis Support Programs for Workers

Efficiency is the primary normative objective in economics. It is concerned with the effective use of scarce resources, and it captures goals related to encouraging competition and maximizing productivity, thereby generating economic prosperity. From the perspective of the social planner, efficiency can be formally conceptualized as Pareto optimality—that is, there is no way of reallocating resources that could make one party better off without making another party worse off. The simplest labour supply and demand models note that the invisible hand of competitive markets will, through the operation of freely adjusting prices (wages), result in Pareto optimality. However, monopsony-like conditions in labour markets, as well as market failures and inefficiencies in the private provision of public goods, mean that government intervention in the form of social insurance and income support can sometimes result in more efficient outcomes (Diamond 1977; Lipsey and Lancaster 1956).

However, programs themselves must also be efficiently run. An efficient social insurance or income support program would be one that is administratively simple, conserves and effectively deploys scarce financial resources, and rapidly provides support to those targeted.[4] Administrative inefficiency can arise from various features of a program such as complex eligibility criteria; the need for many government workers to design, administer, and audit a program; or the involvement of numerous organizations or intermediaries in delivery of program resources. Administrative complexity can both lead to higher financial costs and result in relatively high non-financial costs. For instance, targeted beneficiaries may suffer psychological and health costs related to delays and uncertainty in receipt of benefits. Complicated eligibility criteria may also lead to low take-up rates and erroneous receipt of benefits. Requirements to pay back benefits impose additional costs and also financial stress on those required to pay them back.

Another important aspect of efficiency is the extent to which a policy or program encourages (or at minimum does not discourage) productive societal activities. Although having people stay home (and undertake only essential activities outside the home) during the pandemic was a desired behaviour to achieve public health objectives, it also directly conflicted with ideas about what constitutes a productive contribution to society. Because market-based work is viewed as the primary form of productive participation, governments favour programs that make receipt of income support conditional on labour market participation (e.g., EI is a contributory program that provides income support in the case of job loss, and workers must continue to look for work while receiving benefits). This view of work, we argue, is the primary reason why macroeconomic support for unemployed persons is viewed through the lens of labour market stabilization and efficiency rather than income support and why concerns about work disincentives are central in debates among microeconomists about the design of both employment insurance and social assistance programs.

This, we propose, is also why the CEWS was introduced to facilitate labour attachment to employers and why the flagship crisis worker support program, the CERB, was introduced as a temporary program with time-limited benefits linked to a recent threshold level of participation in the labour market. It was intended to compensate for loss of a job or income resulting from involuntary unemployment, quarantine, illness, or caregiving responsibilities as a result of COVID-19. It is also why considerable concern has been raised about potential fraud and gaming in applications for the CERB, as well as the work-disincentive effects of CERB payments as the economy reopened over summer 2020 (e.g., Cross 2020).[5] For example, a report produced by the C.D. Howe Institute’s Crisis Working Group opines that the CERB is not tied closely enough to participation in the labour market, noting that

encouraging a gradual winding down of the CERB will require tackling disincentives to work. Unlike the Employment Insurance program, CERB recipients are not required to be able and available to work, or to seek suitable employment. There is already anecdotal evidence that some people on the CERB refuse to go back to work when employment opportunities arise. In some cases, this may be due to legitimate pandemic fears, but in many other cases employees may prefer receiving the CERB rather than working difficult shifts even if they would make more than $500 per week (Crisis Working Group 2020, 3).

We attribute these concerns about work disincentives, particularly during a deadly pandemic, as arising from deeply embedded neoclassical assumptions about work as a disutility—that is, work as a “necessary curse” that assumes people facing zero wages will always prefer leisure to work (Budd 2011; see also Kaplan and Schulhofer-Wohl 2018). In the simplest neoclassical models, leisure includes all non-market work, such as child care and domestic responsibilities. More complex economic models recognize differences between market work, non-market work, and sitting on a beach or skiing down a mountain slope (Kahneman, Knetsch, and Thaler 1991; Thaler 1980), but the underlying assumption of market-based work as disutility remains. The standard economic model assumes that workers make trade-offs between consumption and leisure, such that government benefits will unambiguously reduce labour supply because workers will use them to purchase more leisure (i.e., the “income effect”). Targeted government benefits can also induce a substitution effect through high clawback rates.

There are several important reasons why concerns about the income and substitution effects of government benefits (and work disincentives more generally) were problematic during the COVID-19 crisis. One reason is that public health guidance was focused on encouraging everyone who could stay at home to do so. However, many low-income essential workers (such as personal support workers in health care and food production and retail workers, among others) were unable to perform their jobs from home. Low-income essential workers therefore faced a greater risk of exposure to COVID-19 than higher-income workers because of the nature of their employment in public-facing roles or large workplaces where physical distancing was difficult.[6] Moreover, encouraging non-essential unemployed workers to return to work prematurely (an unfortunate feature of the Canada Recovery Benefit [CRB]) could have contributed to increased transmission of the virus during the fall months. Other less obvious reasons are linked to the need for a reimagined view of work itself that does not rely on a restrictive neoclassical model of human behaviour (Budd 2011; Gibbons 1998; Rees 1993; Thaler 1980)[7] and instead takes its inspiration from the field of behavioural economics (Akerlof and Shiller 2010; Ariely 2009; Kahneman et al. 1991) and the long-standing multi-faceted view of why people work drawn from the field of IR (Budd 2011). We return to a discussion of this reimagined view of work in a later section of the article.

In the context of income support programs, equity is a standard of fairness in both distributive outcomes and process, and it also encompasses public economic notions of vertical and horizontal equity. Horizontal and vertical equity considerations imply, respectively, that those in similar circumstances (i.e., job loss) should receive similar supports and that those in dissimilar circumstances (i.e., with greater needs) should receive differing treatment (i.e., more supports). Notions of horizontal equity might require that income support provide relief when a job is lost, whereas notions of vertical equity might require that income support be provided when someone is unable to work.

Payments in an equitable income support system are also made without favouritism toward one person or one group over another, and equitable systems treat individual recipients with respect and sensitivity and protect their privacy. Equity also includes the existence of safeguards—such as the ability to appeal a denial of benefits to a neutral party—and transparency to prevent arbitrary or capricious decision making and enhance accountability. A more equitable income support system would also deliver benefits to targeted recipients independent of resources or expertise (e.g., asking participants to apply to programs with complex eligibility criteria or using an Internet-based application may lead to unequal access) and is equally accessible to targeted beneficiaries irrespective of gender or race.

Voice is the third objective, and it is the one that is most often ignored in the design and evaluation of public policies. Although equity and voice are closely related, they are treated as distinct concepts in contemporary IR. Because equity in distribution can be achieved without providing beneficiaries any real power or input into decisions that affect them, Budd (2004, 23) adds voice to the equity–efficiency framework’s evaluation criteria as a separate “standard of participation or involvement [and] the ability [for individuals] to have meaningful input into decisions” that affect their lives. Viewed through a lens of the voice objective, public policies can have a dampening effect on the ability of individuals to have meaningful input, such as when programs create inequalities in bargaining power among different labour market actors. For instance, providing wage subsidies or top-ups for firms’ essential workers may place more power in the hands of employers, whereas direct income support paid to individuals may place more power in the hands of workers.

There is a massive empirical literature in both management and IR documenting how workplace voice can lead to higher productivity via greater employee engagement and suggestions for efficiency-enhancing improvements (see Freeman and Medoff 1984; Morrison 2011; Pohler and Luchak 2014). Notwithstanding its impact on efficiency, opportunities for workers to use their voice meaningfully are intrinsically important, regardless of whether they are efficient (Budd and Colvin 2008). Voice as an objective is tied to an understanding that labour should not be treated as a commodity because it is embodied in a human being (Budd 2004) and that meaningful input requires relatively equal bargaining power between employers and workers (Webb and Webb 1897).

Whether the voice objective is balanced alongside efficiency and equity objectives in an income support program is determined by the extent to which it provides workers with a realistic fall-back option to losing a job, because this enables them to express concerns, for instance, about personal protective equipment in their workplace during the pandemic. Research has shown that legislated employment rights (e.g., to refuse unsafe work) are unfortunately not always enough to ensure protection for workers, particularly in non-unionized workplaces (Lewchuk 2013; Pohler and Riddell 2019; Reilly, Paci, and Holl 1995; Walters 1996; Weil 1999). Knowing that you may not be eligible for income replacement or support benefits if you are terminated from your job might also prevent you from airing such grievances to your employer or to a Ministry of Labour official. Even though workers across Canada have access to joint health and safety committees and an internal responsibility system that empowers them to refuse unsafe work, workers who lack union or professional licensing protection, and who are otherwise fearful or insecure about their employment, are less likely to voice their concerns to managers or third parties (Harcourt and Harcourt 2000; Kissinger 2017; Walters and Denton 1990; Walters and Haines 1988).

What options do workers have if their employer does not address their concerns, if they are at higher risk of COVID-19 mortality, or if they are unable to perform their job from home? In non-union settings, the answer is “not many” (see, e.g., Pohler and Riddell 2019; Weil 1999). Thus, crisis income support programs should also be evaluated on how they affect the ability of workers (particularly low-income workers) to express (voice) concerns about workplace safety and their own risk preferences.

In summary, to achieve optimal social and economic outcomes and, in particular, stated public health objectives during a crisis such as COVID-19, it is important to carefully design labour and social policies in such a way that they balance efficiency with equity and voice. Table 1 provides a description of the three EEV objectives with some of the suggested measures adapted to the crisis income support programs outlined in this section. In the next section, we apply these ideas to qualitatively evaluate the crisis programs introduced by the federal government and then compare these programs with alternative crisis proposals.

Table 1:

An Efficiency–Equity–Voice Framework for Analyzing and Evaluating Federal Income Support Programs

Dimension and Definition	Income Support Program Metric
Efficiency:Effective use of scarce resourcesPromotion of productive activity	Low cost Rapid distribution of benefits Administratively easy to run (not complex) Minimizes errors Facilitates productive societal activities Minimal auditing requirements
Equity:Standards of fairness in distribution, process, and interaction	Horizontal equity (treating those in similar circumstances, such as job loss, the same) Vertical equity (treating those in different circumstances, such as low- and high-income workers, differently) Ease of accessibility for targeted beneficiaries Protection of privacy Transparent and accountable Clear eligibility criteria and ability to appeal denial of benefits
Voice:Standard of involvement or participationMeaningful input into decisions (ability to affect change at the workplace)	Balances power in the labour market by providing workers an outside option Does not supress the ability to voice concerns (i.e., about safety) Encourages meaningful participation in society without fear of losing the material means by which to live

Source: Authors’ compilation.

Evaluating and Comparing COVID-19 Crisis Support Programs

Key Features of Canada’s COVID-19 Crisis Support Programs

The federal government’s initial income support response to the COVID-19 economic lockdowns included using minor modifications and increases to existing programs such as EI, the Goods and Services Tax credit (GST-C), and the Canada Child Benefit (CCB). In March 2020, the federal government announced that the EI program would waive the one-week waiting period to cover individuals who were involuntarily unemployed as a result of mandatory self-isolation requirements and removed the requirement for a medical certificate.

It soon became apparent to anyone watching long lines of unemployed workers waiting to apply for EI benefits that the program (as designed) did not have the administrative capacity to handle the massive influx of new applications. The bigger problem, however, was that relying primarily on EI would have effectively excluded many low-income workers, who were disproportionately affected by jobs and hours losses during the shutdown (e.g., Koebel and Pohler 2020; Lemieux et al. 2020). EI is a contributory insurance scheme, with eligibility based on hours worked over a certain time period. During the pandemic, many low-income workers would not have qualified for EI benefits or would have received inadequate benefits (Macdonald 2020a). Moreover, the planned increases and one-time payments for the GST-C, CCB, OAS–GIS, and disability tax credit programs were quite small and were not well targeted to low-income workers (Campion-Smith 2020). For these reasons, critics of the early supports argued that additional cash payments for low-income workers would be necessary complements to EI (Pohler et al. 2020).

To its credit, the federal government introduced a new set of income support programs that addressed some of these concerns. The CERB provided workers who had lost their jobs (or income)[8] with $2,000 every four weeks, starting on 15 March 2020 for a maximum of 28 weeks (Canada 2020). However, workers who made less than $5,000 in 2019 (or during the prior 12 months) were ineligible. The government also (later) introduced the Canada Emergency Student Benefit (CESB), an income support program targeted toward students, and a wage supplement for essential workers, negotiated with the provinces. The CEWS was also introduced, which provided businesses with up to 75 percent of their employees’ wages (to a maximum amount of $847 per week)—a major increase over the initial announcement of a 10 percent wage subsidy. Canada’s crisis income support programs for workers, individuals, and households are summarized in Table 2.

Table 2:

Summary of Major Federal Supports Before October 2020

Support Measure	Eligibility Criteria	Amount ($)
EI	Canadian tax residents who are involuntarily unemployed (layoff, sickness, isolation, or caregiving responsibilities) and actively looking for work. Worked 420–700 insurable unemployment hours within the past 52 wk (number of hours depends on several factors); 1-wk waiting period and medical certificate needed to access sickness EI waived.	Based on income and geographic location
CERB	Canadian tax residents who stopped working due to COVID-19 (involuntary unemployment, sickness, caregiving) or an income of less than $1,000 in 14 or more consecutive days within the 4-wk benefit claim period. Income of at least $5,000 in 2019 or past 12 mo.	$2,000/mo for up to 7 mo
CESB	Students (Canadian citizens, permanent residents, or protected persons) who are involuntarily unemployed or earn less than $1,000 within the 4-wk benefit claim period and plan to enroll in a post-secondary institution in September 2019 or have graduated after December 2019 from a post-secondary institution.	$1,250/mo (May–Aug); additional $750/mo disability/dependent supplement
TWSP	Workers (full-time, part-time, or casual workers) in in-person roles considered essential (care facilities). Eligibility determined by each province.	Additional monthly supplement to essential workers
CEWS	Private employers (for-profit, non-profit, and charitable) who experienced an eligible reduction in revenue.	Up to 75% wage subsidy, up to a maximum of $847/wk per eligible employee
Increase in the CCB	Canadian tax residents who are low- or middle-income individuals and households with children.	Additional $300/child (one-time increase in May)
Increase in the GST-C	Canadian tax residents who are low- or middle-income individuals or households.	Double maximum annual GSTC amount (one-time increase in May)
Increase in OAS–GIS	Low or middle-income seniors who resided in Canada for at least 10 y.	$300 one-time increase in OAS; $200 one-time increase in GIS (June)
One-time payment to persons with disabilities	Persons with disabilities who applied for the disability tax credit by 31 December 2020.	One-time non-taxable payment up to $600

Notes: EI = Employment Insurance; CERB = Canada Emergency Response Benefit; COVID-19 = coronavirus disease 2019; CESB = Canada Emergency Student Benefit; TWSP = Temporary Wage Supplement Program; CEWS = Canada Emergency Wage Subsidy; CCB = Canada Child Benefit; GST-C = Goods and Services Tax credit; OAS–GIS = Old Age Security–Guaranteed Income Supplement.

Source: Authors’ compilation from various government websites, such as Canada’s (2020) COVID-19 Economic Response Plan.

With the exception of the small top-ups to the GST-C and the CCB (and also to OAS–GIS and disability payments), the majority of other individual and household crisis support programs introduced during the pandemic were directly tied to labour market participation (i.e., being employed or having lost a job). Although the CERB had been touted by some as moving toward a basic income (Coyne 2020; UBI Works n.d.), in reality it was a more generous form of EI, because it removed the hours and contribution requirements and the scaling of benefits but still targeted workers whose jobs or income had been affected and provided them with time-limited benefits. In terms of a crisis support program, CERB was an improvement over EI in that it was more generous in terms of the availability and adequacy of benefits, particularly for low-income workers (Robson 2020a), and it relied on a less onerous “trust, but verify” application process (Robson 2020b). Unlike EI, for which eligibility and benefit amounts are conditional on prior contributions and hours worked, the CERB provided the same amount to everyone who lost their job or income as a result of COVID-19 and did not require workers to look for work. However, neither EI nor CERB were available to workers who earned less than $5,000 in 2019 or over the past year.[9]

Moreover, neither EI nor the CERB allowed workers to collect benefits if they had quit their job for any number of relevant COVID-19–related reasons. Indeed, there is a grey area in what counts as voluntary separation. For instance, quitting one’s job due to loss of child care would be eligible under both the EI and CERB programs. However, there is understandably confusion over the interpretation of these kinds of eligibility criteria on the part of employers and workers. Accordingly, it is not at all clear that workers would be eligible for benefits if they quit their jobs for another COVID-19–related reason—for instance, because of the lack of personal protective equipment in their workplace or fear of contracting the virus more generally (e.g., if a worker is reliant on public transportation to get to work). Moreover, the CERB website clearly specified that workers could not voluntarily quit their jobs to access CERB. We can (and should) assume that most workers would take this stated eligibility criteria at face value.

Likewise, statutory provisions in provincial occupational health and safety (OHS) programs that provide workers with the right to refuse unsafe work were established decades ago and have been shown in other contexts to not be widely known about by workers (Walters and Haines 1988), especially traditionally disadvantaged workers who “are least likely to be aware of their rights” (Walters and Denton 1990, 531). Moreover, where OHS rights are understood, they are rarely exercised by workers (particularly in non-union settings) who are fearful of losing their job for raising OHS concerns (Kissinger 2017). This perception on the part of workers that their concerns may take a back seat to management rights appears to be based on an accurate depiction of how things actually play out at the workplace. One study using Canadian court decision data found that

employees had to satisfy many rigid conditions to qualify for protection from discipline [for raising OHS concerns]. These conditions appear to have been based on the notions that health and safety are properly management’s prerogative and that obedience to management authority is essential to efficient production. (Harcourt and Harcourt 2000, 684)

The government’s preferred approach to crisis income supports appeared to place more weight on policy objectives such as supporting labour market efficiency over worker voice regarding workplace safety. Other objectives, such as providing workers the ability to stay home on the basis of their own risk assessment, were given even less weight, or ignored altogether. Moreover, many low-income workers still do not have any access to paid sick days. It is entirely plausible that providing workers with these options could have helped reduce the transmission of COVID-19 in workplaces.

Assumptions Embedded in the Canada Emergency Response Benefit and Canada Emergency Wage Subsidy: Work as Disutility

The CERB and the CEWS are the largest and most prominent of Canada’s federal crisis support programs. As with EI, these programs place primary emphasis on the efficiency objective of labour market stabilization. In other words, the combination of CERB and CEWS provides short-term support to unemployed persons while at the same time encouraging worker attachment to employers (because these programs are unavailable to workers who voluntarily quit their jobs).

Embedded in this labour market stabilization and efficiency objective are assumptions about work as a “bad” and workers’ related preferences for leisure over market-based work. This assumption leads to a prediction that the introduction of an unconditional government transfer would reduce labour supply for beneficiaries—that is, the income effect in basic labour supply models. Yet, the empirical evidence for the income effect of government cash transfers (such as the Mincome experiment in Dauphin, Manitoba, in the 1970s) is sparse (Hum and Simpson 1993). Any adverse labour-supply effects of more permanent cash transfers in Canada have been shown to mostly affect specific sub-samples of the population, such as married women with children (Koebel and Schirle 2016). Although there is plenty of evidence for the substitution effects of government cash transfers in the labour supply literature (i.e., work disincentives created via benefit clawback rates), these are usually the result of very sharp cutoffs or extremely high clawback rates that make people financially worse off when they choose to work and do not provide any direct evidence that work itself provides disutility.

Moreover, “work as disutility” is only one conception of work that relies on a relatively narrow view (i.e., the neoclassical economic model) of human behaviour and of work itself (Budd 2011). Behavioural economic research (Heyman and Ariely 2004) and systematic empirical evidence from industrial and social psychology (Schwartz 2015; Wrzesniewski et al. 1997) suggests that people find dignity and purpose in almost any job. In a particularly illuminating study of hospital janitors, Schwartz and Sharpe (2010) found that the ways in which workers described their duties (e.g., helping patients, families, doctors and nurses) differed greatly from the job descriptions listed in employment contracts and collective agreements. In other words, a multitude of intrinsic rewards could explain the lack of turnover in these so-called low-status jobs and differentially compensate for the relatively low pay. Indeed, as a result of the COVID-19 crisis, we have all, it is hoped, recognized that the value of a job to society is not synonymous with what workers are paid. It is also clear that even jobs that require lower levels of education are performed by workers (often young, female, racialized, or all of these) who are engaged in vital tasks such as feeding and caring for people during a lockdown (e.g., food production workers, delivery drivers, grocery clerks, and personal support workers).

This same conclusion can be inferred from studies examining the effects of COVID-19 on working hours. What these data show is striking: namely, the impact on labour supply has been quite heterogeneous, even among low-income workers. A recent study by Koebel and Pohler (2020) highlights that a significant group of Canadian workers in the bottom quintile of the earnings distribution—those most likely to be in low-paid jobs—were working significantly more hours than usual and relative to workers at the top of the earnings distribution after COVID-19 restrictions were introduced.[10] Preliminary analyses specific to the CERB also suggest a lack of evidence of any short-term work-disincentive effect (Schirle and Skuterud 2020).

People work for a multitude of reasons—out of a sense of pride, obligation, and loyalty to customers, co-workers, and employers or because they know they are making a contribution (however small) to help society in a time of crisis. Just as the biblical adage notes that people “do not live by bread alone,” so it is that work encompasses more than pecuniary rewards. Work brings intrinsic rewards—the friendships, the mastery of a craft, the feeling of making a contribution, and so forth (Budd 2011).

There is also well-established evidence that extrinsic rewards (i.e., money) are not always effective motivators. Beyond a satisficing minimum, extrinsic rewards give way to intrinsic rewards, which are more effective in increasing output and improving quality. In an field-based study of more than 200 urban schools across three cities where financial incentives were given to students to boost achievement (e.g., in Dallas, students were paid to read books; in New York, students were rewarded for performance on interim assessments; and in Chicago, students were paid for classroom grades), the estimated impact of financial incentives on student achievement was statistically zero (Fryer 2011).

Similar results for teacher performance were prefigured in a psychological study of school outcomes that found that both students and teachers performed best when they were intrinsically motivated by their work (Sylvia and Hutchison 1985). These findings stretch out beyond education to areas as diverse as predicting who engages in civic behaviour—for example, people are less likely to show up for jury duty when offered money (Bloeser, McCurley, and Mondak 2012)—and who is more apt to act in an environmentally friendly way—for example, when hotel guests made a brief but specific commitment at check-in and received a lapel pin to symbolize their commitment, they were more than 25 percent more likely to hang at least one towel for reuse than those who received a monetary incentive, increasing the total number of towels hung by more than 40 percent (Baca-Motes et al. 2013).

People also know of the power of intrinsic rewards associated with work from casual observation, because so many people are willing to work for free. Hundreds of millions of hours are donated by people all around the world every day, whether it is through formal volunteering efforts or spending extra hours on a job without the expectation of remuneration (Gomez and Gunderson 2003). Indeed, during the crisis, thousands volunteered as contact tracers (Miller 2020).

Work serves purposes beyond the conventional income–leisure trade-off of the neoclassical labour supply model. To borrow the language of consumer economics, work encompasses a bundle of attributes, some of which are negative and others of which provide positive intrinsic benefits. Work, in other words, should not be treated by default as a “bad,” as it is under neoclassical labour supply models, but rather as a multi-attribute “good” in the consumer behaviour sense (see Bruno and Wildt 1975; Kleine, Kleine, and Brunswick 2009), with features that provide negative, but also positive, utility.

Finally, we note that although the assumptions have yet to make their way into the design of labour market policies, the contest between intrinsic and extrinsic motivators is not new, nor is it even new to the standard economic modelling of work (Akerlof and Shiller 2010; Rees 1993). On the dangers embedded in an unwavering commitment to the neoclassical view of agents influenced only by extrinsic (monetary) rewards, Gibbons (1998, 132) makes the following observation:

One simple possibility is that economic models that ignore social psychology are incomplete (but perhaps still useful) descriptions of incentives in organizations. A more troubling possibility is that management practices based on economic models may dampen (or even destroy) non-economic realities such as intrinsic motivation and social relations.

Comparing Alternative COVID-19 Basic Income Proposals with the Canada Emergency Response Benefit and Canada Emergency Wage Subsidy Using the Efficiency–Equity–Voice Framework

To assume that work is, on average (or even on the margin), disutility is not well supported by the quantitative and qualitative evidence during the COVID-19 pandemic or otherwise. The assumption that work is disutility leads to the belief that most people (and, in particular, low-income workers), if not engaged in market-based work, are shirking. At the best of times this is a flawed assumption; however, during COVID-19 this assumption is particularly problematic because individuals may refuse to work simply because they are following public health orders to shelter in place or are staying at home out of fear of contracting COVID-19 or exposing loved ones to the virus.

This assumption of work as a “bad” leads to the design of income support programs that may be simultaneously inequitable and inefficient, as policy-makers link receipt of crisis income supports to unemployment and labour market participation (as can clearly be seen in both the CERB and the CRB program that replaced it), attachment to employers (as can be clearly seen in the CEWS), or both. Inequities arise when workers are unable to access COVID-19 income support, for instance, if they quit their job or made less than the threshold amount of $5,000 in 2019 (or the previous 12 months). Explicitly linking income supports to job or income loss leads to the corollary that those who did not lose their jobs or income should not receive supports. This, however, is difficult to justify when many low-income essential workers (e.g., grocery store clerks, food production workers, and personal support workers) bear the greatest risk of COVID-19 exposure on the job because they are less likely to be able to perform their jobs from home and are more likely to work in public-facing roles (Lu 2020).

Although an economy needs workers to function, provincial and territorial governments believed that the threat of COVID-19 was so serious that it warranted widespread lockdowns and business closures on multiple occasions. Throughout the various waves of the pandemic, working from home became the new norm for many professionals who were able to do so. Indeed, many high earners (such as university professors) were encouraged to work from home or collectively organized against the perception of being forced back into the workplace (Samba 2020). Inefficiencies also arise because crisis programs too closely linked to labour market participation may be more administratively complex and costly to implement quickly or to audit post-crisis, and they can stall healthy structural economic adjustments that may be inevitable post-pandemic.

In addition to the inequities and inefficiencies that arise from linking crisis supports to participation in the labour market or unemployment, there are also implications for worker voice. Recall that the CERB was officially not available to workers who quit their jobs (e.g., because of fear of the virus) or to those who earned less than $5,000 in the previous year. The effect of these eligibility restrictions is that they do not allow these workers to access support, nor do they give them the freedom to choose whether to continue working during a time when there is a high risk of exposing oneself (and one’s family) to a potentially fatal virus.

The CEWS, a program that encourages attachment to employers, may be particularly problematic for worker voice. Low-income essential workers, who are less likely to have a job that can be performed from home and whose employers are likely receiving a wage subsidy through CEWS or a wage supplement for their essential workers, may face even greater inequality in bargaining power during the crisis vis-à-vis their employers. Without an alternative income source or access to savings, these workers may be less able to say no to unsafe work even if they or their family members are at a greater risk from COVID-19 mortality. They may also be less willing to voice concerns about workplace safety.

One of the more hotly debated alternative policy ideas for addressing the crisis involved delivering some form of a basic income through the tax system. A basic income is an unconditional income transfer from the government to individuals or families. During the COVID-19 crisis, politicians (McCullough 2020), academics (Pohler et al. 2020), pundits (Boessenkool 2020), non-profit organizations (UBI Works n.d.), and individual citizens (Monsebraaten 2020) voiced support for the introduction of some form of a basic income program. Two types of basic income programs emerged from the flurry of proposals: one in the form of a UBI and the other a TBI. The common characteristic of both types of basic income proposals was that income support was not conditional on (un)employment, lost hours, or ongoing participation in the labour market. However, they also have important differences.

The crisis UBI proposal involved distributing $2,000 to all Canadians who filed taxes in 2018 (Boessenkool 2020). Boessenkool (2020) argues that distributing a crisis UBI through the tax system would have been relatively quick and efficient, although it would have come at a high cost: $57 billion per month. Because the duration of the CERB was seven four-week periods, the total UBI program cost over this period would have been almost $400 billion. Boessenkool notes that, to reduce the net cost of the policy, the UBI could be clawed back during the following year’s tax season (as we write this during the 2021 tax season, we are still in the midst of pandemic-related lockdowns in many provinces).

Our TBI proposal similarly relies on distributing emergency cash transfers through the tax system, but rather than complete universality, direct cash payments of $1,000 per month (non-taxable) would be targeted toward low-income workers (aged 15 y or older) who earned between $0.01 and $22,000 in 2019[11] and who filed taxes. This monthly amount (i.e., $1,000) was chosen because it is close to the average monthly social assistance provided to single working-age people across the country (Pohler et al. 2020), and the $22,000 threshold was chosen because it is roughly the amount of income at which monthly EI benefits would also have been $1,000, assuming EI eligibility.[12] Workers who earned more than $22,000 in 2019 would be ineligible for the TBI. Workers who made more than $22,000 and found themselves unemployed could still apply for EI benefits under the normal eligibility rules. Because self-employed workers voluntarily opt in to EI benefits, and many may not do so, in our modelling we also provide the $1,000 transfer to all self-employed workers regardless of their earnings.

The estimated cost of the TBI alone is $7.7 billion per month (over seven four-week periods, this would have a total cost of $53.9 billion). The cost of EI under this proposal is difficult to estimate, but we use the number of CERB applications by federal income tax brackets to calculate what the EI cost might have been under our proposal. Our highest program cost estimate assumes maximum monthly EI benefits for the number of individuals within each tax bracket that applied for the CERB for a cost of $6.1 billion per month.[13] For CERB applicants in the first tax bracket, we exclude those who earned less than $22,000 in our EI cost estimates because these individuals would receive our TBI. Assuming each CERB applicant received full EI benefits for all seven months, the total cost of the EI component of our proposal is $42.7 billion. The gross cost of our combined TBI–EI program is thus $96.3 billion (note that we include students in all computations). By comparison, over the seven-month period, the GST-C, CCB, and OAS–GIS top-ups are estimated to cost approximately $10 billion in total; the CESB is estimated at $2.9 billion and the CERB at $76.7 billion. Although it is more difficult to estimate the cost of the CEWS from March to October, the 2020/21 fiscal year cost is estimated to be $74 billion. The estimated total cost of all these programs is therefore about $164 billion (Office of the Parliamentary Budget Officer 2020). Our proposed program thus costs substantially less at $96.3 billion than either the UBI ($400 billion) or the government’s mix of crisis supports ($164 billion).

Two caveats are in order. The first caveat is that the UBI was proposed as taxable income, which would reduce the program cost (the CERB was also taxable income, and the cost estimates provided by the Parliamentary Budget Officer for the government programs that we use here represent the net cost to the government). It would take a substantially higher tax rate on UBI benefits to place the net cost below the TBI–EI program, and the CERB program costs also do not take into account the post-crisis auditing costs. The second caveat is that if the CEWS had not existed, more people may have had to rely on EI, increasing the cost of our TBI–EI program; however, the CEWS will require more substantial post-crisis auditing. To facilitate comparison, Table 3 places the CERB and CEWS alongside the UBI and TBI–EI proposals in terms of eligibility criteria, payments to recipients, and total estimated program costs.

Table 3:

Comparison of CERB and CEWS with a Universal and Targeted Basic Income

Support Measure	Eligibility Criteria and Payments	Estimated Total Program Cost
CEWS	Eligibility: Private employers (for-profit, non-profits, and charitable) who experienced an eligible reduction in revenuePayment: Pays up to 75% wage subsidy, up to a maximum of $847/wk per employee	$74 billion
CERB	Eligibility: Canadian tax residents who stopped working as a result of COVID-19 (involuntary unemployment, sickness, caregiving) and an income of less than $1,000 in 14 or more consecutive days within the 4-wk benefit period of the claim. Income of at least $5,000 in 2019 or past 12 mo.Payment: $2,000/mo up to 7 mo (taxable)	In combination with other income support programs (see Table 2), $90 billion.CEWS + CERB ~ $164 billion
UBI	Eligibility: All persons (regardless of labour market status) who filed taxes in 2018Payment: $2,000 to all Canadians (taxable)	$400 billionUBI would replace all other supports but still require an additional $236 billion, minus the tax collected back.
TBI	Eligibility: All persons who filed taxes (regardless of labour market status) earning between $0.01 and $22,000 in 2019Payment: Direct cash payments of $1,000 per month (non-taxable)	In conjunction with EI, $96.3 billionTBI would replace all other supports so cost savings would be $68 billion.

Notes: CERB = Canada Emergency Response Benefit; CEWS = Canada Emergency Wage Subsidy; COVID-19 = coronavirus disease 2019; UBI = Universal Basic Income; TBI = Targeted Basic Income; EI = Employment Insurance.

Source: Estimated program costs for CERB and CEWS are from Office of the Parliamentary Budget Officer (2020). Estimated program costs for the TBI–EI are based on the authors’ own calculations using Statistics Canada’s Social Policy Simulation Database and Model (2019).

Critiques of basic income proposals have been made on equity- and efficiency-based grounds. For example, Cameron and Petit (2020, 1) note that the implementation of a basic income would have been too slow and that This argument lends supports to our earlier point that programs that are slow to implement are inefficient, particularly during a pandemic when individuals require immediate income support. However, the flagship CERB–CEWS programs implemented by the federal government also did not exist before the crisis, and there is no reason to believe that a basic income would have required any more up-front investment of time. Indeed, it may even have been more efficient in the long run: it would have more directly addressed the limitations of EI (particularly its inaccessibility for many low-income and self-employed workers) by introducing a much smaller number of programs than were ultimately implemented (see Table 2). Moreover, it would not have required extensive auditing of CERB and CEWS claims post-crisis, nor the creation of the CRB to eliminate the highly criticized work disincentives within the CERB design. Although the CERB application process was a major improvement over the EI system, there are further administrative efficiencies that could be made using approaches that relied on information the government already had, requiring less administrative oversight and delivery of multiple programs and post-crisis audits as a result of less room for individual errors in interpretation of eligibility across numerous programs.[14]

policymakers know that designing and delivering new programs takes time; as we have seen … even applying top-ups to existing programs cannot be done overnight. … The best option for governments is to avoid creating new programs, and to instead use systems and infrastructure already in place.

Another critique of the basic income proposal is that some individuals would be missed using tax files as a result of changes in banking and address information and because approximately 10 percent of Canadians do not file taxes (Robson and Schwartz 2020). However, the 90 percent of Canadians who do file taxes could theoretically update their bank and mailing address information with the Canada Revenue Agency (CRA), and those who had not previously filed their taxes could do so if they wanted to be eligible for the benefit. The CERB also required all eligible individuals to apply for benefits through either the CRA or Service Canada, and CERB is a taxable benefit, so recipients who have not previously filed taxes would still be required to do so in the subsequent year.

Cameron and Petit (2020) argue that basic income proposals often lack a targeting feature and thus do not account for heterogeneity of need (i.e., they fail to achieve vertical equity). Although we agree that the UBI is not targeted, our TBI proposal explicitly prioritizes vertical equity during the crisis, because not everyone who lost their job was in need of emergency income support. It was more likely that low-income workers who lost their jobs would have been in need of emergency income support because they would have been less likely to qualify for EI or to have savings to help them weather the crisis in the short term. Because the CERB provides $2,000 a month to recipients regardless of their actual need, that program is also relatively blunt and so Cameron and Petit’s vertical equity critique applies to the CERB as well. Although the CERB is taxable income, there are no plans to claw it all back from the highest income workers.

Putting It All Together

In Figure 1, we adapt a triangle graphic from Budd’s (2004) application of EEV to labour market institutions and public policies to visually summarize the preceding sections in which we evaluate and compare actual and proposed COVID-19 income support systems on the basis of how they meet and balance the three EEV objectives. The equilateral triangle represents the equal weight of each of the three EEV objectives in our evaluation framework. A program primarily concerned with efficiency would reside closest to the bottom left of the triangle. Similarly, a program would be located at the other two extreme corners if only equity or voice was the primary objective. A more balanced program would be located in the middle because it is equally concerned with achieving all three objectives—efficiency, equity, and voice.

Figure 1:

Geometry of Income Support Programs

The relative locations of specific COVID-19 income support programs should be apparent on the basis of our analysis and comparison of the different programs outlined in the previous sections. Disagreements over these locations are encouraged because one of the major contributions of this framework is to provide a basis for such debates and to adopt an evaluation lens for labour policies that relies on explicit criteria for equity and voice as well as efficiency.

A UBI would have provided all workers greater voice than the CERB or CEWS, and a TBI in combination with EI would have provided low-income workers (those making less than $22,000) greater voice than the CERB or CEWS. This is because the distribution of a basic income unconditional on labour market participation would have made employment-related decisions during the pandemic more voluntary. By giving workers an alternative income source to meet basic needs, a basic income would have equalized bargaining power in the labour market and thus provided workers with the ability to voice concerns about unsafe workplace practices, to say no to unsafe work, or both without fear of losing an income source. Because the essential economy relied in particular on many low-income workers to function, a TBI would have allowed these workers to continue working if they were able to do so while not feeling coerced to put themselves or their families in harm’s way. The basic income approaches would also have more efficiently provided a de facto top-up for low-income essential workers who continued working, eliminating the need for the government to directly intervene in the labour market with yet another separate crisis program—a temporary wage supplement—that required collaboration with the provinces.[15] Finally, a basic income crisis policy would have avoided the work disincentives built into the CERB via the sharp income cutoffs by decoupling crisis benefits from employment and income during the crisis.

Notes: CCB = Canada Child Benefit; OAS/GIS = Old Age Security/Guaranteed Income Supplement; GST-C = Goods and Services Tax credit; CERB = Canada Emergency Response Benefit; TBI/EI = Targeted Basic Income/Employment Insurance; EI = Employment Insurance; UBI = Universal Basic Income; CEWS = Canada Emergency Wage Subsidy.

The extent to which basic income proposals outperform CERB and other programs tied to the labour market in balancing EEV, however, varies across the type of basic income. By treating all individuals identically, a UBI somewhat addresses horizontal equity concerns: low- and high-income individuals would have had equal access to the crisis basic income during the pandemic (although not all of them were in a similar situation in needing to be able to say no to unsafe work, because many people were able to work from home). In contrast, the combined TBI–EI proposal would have provided a transfer to workers with employment income less than $22,000, and workers with employment income greater than this would have relied on EI in the event of a job loss. This may arguably have created some inequities for workers making more than $22,000, similar to the critiques we outlined for CERB, because EI does not provide benefits to those who voluntarily quit their jobs. However, because higher-income individuals were more likely to (a) receive maximum EI benefits, (b) be less affected by job loss during COVID-19, (c) find jobs once the restrictions were lifted, and (d) work from home, we posit that this is less problematic than the lack of voice for workers who made less than $22,000, and the lack of vertical equity for workers who made less than $5,000, embedded in the CERB.

The enormous cost of the UBI also reduces its efficiency relative to both CERB–CEWS and TBI–EI. However, the greater efficiency of the CERB and CEWS over the UBI comes at the expense of equity and voice. The favoring of labour market efficiency is seen most strongly in the CEWS, although it did possibly impede healthy structural adjustments in the economy and “very little of the funds spent [went] to saving jobs at risk” (Smart and Eisen 2021). Although the CEWS created higher income support for workers (i.e., workers could receive higher payments through the CEWS than they would have through the CERB), if an employer received a wage subsidy for a worker who was receiving the CERB, the worker is the one who would be required to repay the benefit. However, the CEWS does provide the ability for workers to maintain attachment to their employers if they desire to do so.

The inclusion of expanded and more accessible income supports for low-income workers with the CERB led to some gains over solely relying on EI on the equity and voice dimensions, as well as improved efficiency as a result of simplified application processes. However, the combination of the CERB and CEWS (and all other income-related supports to individuals and households) were ultimately more costly than the combination of TBI and EI. Direct cash transfers such as the CCB, OAS–GIS, and the GST-C were administratively efficient and perhaps helped to address the limitations of the CERB for workers making less than $5,000, but are evaluated less favorably in terms of equity or voice because they are not well targeted to low-income workers—a group that was both more likely to lose their jobs and less likely to be able to work from home (Koebel and Pohler 2020). A basic income scheme targeted toward low-income workers, introduced alongside the pre-existing EI program, can therefore be viewed as optimizing the relative mix of all three objectives—efficiency, equity, and voice—hence its central position in the triangle in Figure 1.

Conclusion

A range of possible program ideas could have been used during the COVID-19 crisis. Each has important implications for labour market supports that are implemented in future crises and even for permanent changes to Canada’s pre-crisis income support system. Policy choices made during the pandemic generated strong debates, even as criteria for evaluating and comparing support programs for workers were largely incomplete (e.g., by adopting limited notions of efficiency and equity or ignoring voice entirely). There is a need for metrics that go beyond labour market efficiency and the related goal of reducing work disincentives that is so prevalent in the labour and social policy evaluation literature and that was present in the public discourse that surrounded the CERB and CEWS. We argue that balancing efficiency, equity, and voice provides a consistent framework and set of criteria for evaluating and comparing income support programs.

Our framework was adopted to analyze, evaluate, and compare COVID-19 crisis income support programs, but this framework can also be used to design future income support programs. Although pluralist IR thought emphasizes balancing (sometimes) competing objectives by giving efficiency, equity, and voice considerations equal weight (Budd et al. 2004), policy-makers could similarly use this analysis to design institutions and programs that achieve other desired mixes of efficiency, equity, and voice. In fact, the introduction of the CERB after the inadequacies of EI in the face of the COVID-19 government-mandated shutdowns could be viewed as a struggle among policy-makers to find their own desired balance among efficiency and equity in their COVID-19 responses. For most policy-makers, however, there is still a strong overriding concern with the efficiency objective, and with particular aspects of efficiency: namely, labour market attachment and the potential work disincentives created through income support programs. Equity is given somewhat less emphasis, and voice is almost entirely ignored.

One of the goals of this article was to highlight why concerns about work disincentives are largely misplaced, because they are built on mixed empirical support and a restrictive understanding of what motivates people to work. Concerns about work disincentives originate out of neoclassical ideas about worker motivation, and the structure of labour markets, and have been used as the default assumptions in the design of labour market policy. These assumptions may even contribute to a self-fulfilling prophecy; that is, during the COVID-19 pandemic these assumptions may have led to programs that inadvertently created work disincentives while at the same time generating other inefficiencies and undermining public health objectives. A lack of credible worker voice embedded in the crisis support programs made it difficult for workers to assess and act on their own COVID-19 risk, as well as to raise any concerns about the (lack of) safety precautions put in place by their employers.

The elimination of CERB in Fall 2020 and its replacement with the CRB also failed to balance EEV, especially in the context of stringent lockdowns re-imposed in three provinces (Ontario, Quebec, and Alberta) shortly after the CERB was mothballed. The federal government either assumed that people would be returning to work in the sectors most affected during the first wave of lockdowns in Spring 2020 or was concerned about CERB’s work disincentives. Either way, to help “persuade” workers to get back to work, CERB was eliminated in favour of the CRB, which is available on application for two weeks at a time (up to 26 weeks). The CRB covers workers too sick or immune compromised to work, but subsequent lockdowns introduced after Fall 2020 threw some people out of work again and made it very difficult for other workers to re-enter the labour market. The modifications to EI and the creation of the CRB, although more generous and easier to access than what existed before COVID-19, still exclude workers who have been unable to find work or have accumulated too few hours of work in the intermezzo between lockdowns.

Even if one disagrees with the analyses and conclusions herein—specifically, that a combination of a TBI alongside the pre-existing EI program would have been most effective at balancing efficiency, equity and voice objectives—such disagreements underscore the need for an analytical framework and set of normative evaluation criteria that are made explicit and are defined much broader than efficiency. Although our recommendations are specific to the COVID-19 crisis, our article illustrates the analytical integrity and applicability of balancing efficiency, equity, and voice as a framework for analyzing and comparing alternative income support programs and proposals—in the time of COVID-19 and beyond.