Reflection

Expanding the Canada Workers Benefit to Design a Guaranteed Basic Income

Kourtney Koebel, Dionne Pohler

Canadian Public Policy, Volume 45, Number 3, September / septembre 2019, pp. 283-309 (Article)

Published by University of Toronto Press

For additional information about this article

[ Access provided at 10 May 2020 18:22 GMT from University of Winnipeg Library ]

https://muse.jhu.edu/article/734293

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Expanding the Canada Workers Benefi t to Design a Guaranteed

Basic Income

KOURTNEY KOEBEL AND DIONNE POHLER Centre for Industrial Relations and Human Resources, University of Toronto, Toronto, Ontario

Les auteures conçoivent pour les Canadiens en âge de travailler un programme hybride de revenu de base garanti et de subvention salariale qui résout les préoccupations relatives au fédéralisme et aux facteurs de désincitation au travail associés au revenu de base traditionnel grâce à l’extension de l’allocation cana- dienne pour les travailleurs. Établissant le coût de leur programme, elles proposent un modèle de fi nance- ment sans incidence sur les recettes, faisant appel à la consolidation des programmes d’aide sociale et à la suppression de plusieurs crédits d’impôt. Elles simulent les effets de redistribution de leur programme et de leur modèle de fi nancement sur le revenu disponible des ménages, selon les déciles et les types de famille, et en analysent l’incidence sur les taux effectifs marginaux d’imposition et les interactions avec les programmes pour personnes handicapées. Le programme des auteures réduit sensiblement les taux de pauvreté chez les familles biparentales et chez les célibataires ou les couples sans enfants, en âge de travailler.

Mots clés : allocation canadienne pour les travailleurs, incitatifs au travail, pauvreté, revenu de base ga- ranti, subvention salariale

We design a hybrid guaranteed basic income and earnings subsidy for working-age Canadians that ad- dresses federalism and work disincentive concerns associated with a conventional basic income by expand- ing the Canada Workers Benefi t. We cost our program and propose a revenue-neutral fi nancing model by consolidating provincial SA programs and eliminating several federal and provincial tax credits. We simu- late the distributional effects of our program and fi nancing on household disposable income across deciles and family types and discuss its impact on marginal effective tax rates and interaction with disability programs. Our program substantially reduces poverty rates among two-parent families and working-age singles and couples without children.

Keywords: guaranteed basic income, work incentives, earnings subsidy, Canada Workers Benefi t, poverty

Introduction A guaranteed basic income (GBI) is receiving renewed interest as a policy option among Canadian academics, policy-makers, and the broader public. 1 Although explora- tion of the adoption of a GBI in Canada is not new—for instance, a GBI was proposed by a Special Senate Commit- tee on Poverty as far back as 1971 ( Canada 1971 )—detailed design, costing, and implementation proposals are less common. Even more rare are GBI proposals that address both economic and political challenges associated with program design and implementation. We propose a program that addresses these challenges by expanding on the framework and design of the Canada Workers

Benefi t (CWB), a federal refundable tax credit that allows for provincial variation.

Specifi cally, we propose an individual GBI set at the current social assistance (SA) level for a single individual in each province, coupled with a 30 percent earnings subsidy, to be phased out at a rate of 50 percent once an individual reaches a disposable income threshold set at the market basket measure of poverty. We retain the Canada Child Benefi t (CCB) and Old Age Security/Guaranteed Income Supplement (OAS/GIS) as separate GBI programs for children and seniors, respectively. The key implication of our design is that every Canadian aged 18–64 years working approximately 9–16 hours per week will be

284 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

above the poverty threshold—the hours vary depending on the relevant market basket measure of poverty and the current minimum wage in each province. We cost our program and propose a revenue-neutral fi nancing model by consolidating SA and income-transfer programs for working-age individuals, as well as eliminating several tax credits. We discuss the implications of our program design and fi nancing on marginal effective tax rates (METRs) and people with disabilities, and we simulate its effects on household disposable income across deciles and family types. Our program substantially reduces poverty rates among two-parent families and working-age singles and couples without children.

Guaranteed Basic Income Proposals Proposed designs for a basic income vary widely. Al- though the term universal basic income (UBI) is often used interchangeably with guaranteed basic income, UBI generally refers to an unconditional demogrant. Unless the basic amount provided is very low, UBI proposals are cost pro- hibitive ( Hoynes and Rothstein 2019 ). 2 The most feasible policy proposals for a GBI in Canada are variations of a negative income tax (NIT). A NIT is structured so that the transfer is subject to a tax-back rate; individuals (or house- holds) who fall below a specifi ed income threshold receive a direct cash transfer from the government. As income increases, the amount of the transfer is gradually reduced to zero. Our proposed program is a variation of a NIT.

To date, most GBI policy proposals in Canada have focused on the fi scal impact on governments and taxpay- ers. Several thoughtful articles show different options for designing a revenue-neutral GBI, highlighting the distributional effects across household types and income deciles (e.g., Boadway, Cuff, and Koebel 2018b ; Stevens and Simpson 2017 ). The details of these proposals vary, but all identify the need to consolidate programs and re- structure the personal income tax and transfer system to be able to fund an adequate guarantee—one that provides for basic needs and lifts as many people as possible out of poverty. 3

We follow the general approach of Boadway et al. (2018b ), who propose a federal–provincial GBI by retain- ing the personal income tax system’s current progressive statutory tax rates and replacing select refundable and non-refundable tax credits with one income-tested refund- able tax credit delivered through the same tax system. Like prior proposals, we outline the overall cost, sources of funding, and tax-related implications of our design. However, our proposal extends previous work by includ- ing political considerations.

We posit that two specifi c political challenges are the key barriers to the realization of a GBI in Canada: (1) federalism—in particular, the need for federal– provincial–First Nations cooperation in implementation, administration, and fi nancing of a GBI—and (2) widely

held ideas about a GBI’s negative effect on work. The for- mer challenge is diffi cult to address given that federalism has been cited as “by far the strongest barrier to compre- hensive and integrated reform in social policy” ( Banting 1994 , 134). However, the latter challenge is even more diffi cult to address for two reasons. First, most feasible proposals from a cost perspective create net contributors and net benefi ciaries in moving from the current system to a new one. Second, moral expectations about reciprocity ( Rossiter 2018 ), deeply held beliefs about what constitutes a productive societal contribution, and assumptions about why people work ( Budd 2011 ) all contribute to the perceived unfairness of a GBI, particularly among net contributors. We seek to address both these challenges in our proposed design by expanding the CWB program.

Income Security in Canada Canada’s income security system is a complex mix of federal and provincial programs that provide assistance through direct cash transfers and tax subsidies to different categories of individuals and households. These include, but are not limited to, seniors’ OAS/GIS, the CCB, Goods and Services Tax (GST)/Harmonized Sales Tax rebates, SA, CWB, disability supplements, the Canada/Quebec Pension Plan (C/QPP), Employment Insurance (EI), worker’s compensation, and myriad non-refundable tax credits—for example, the personal, spousal, and depend- ent amounts.

These programs have varying requirements that de- termine individual and household eligibility and ongoing access. Eligibility for certain programs may also result in access to other supports, such as subsidized housing and child care and job-related training programs. The pro- grams differ in the extent to which they are (1) universal, universally accessible, or targeted; (2) means or income tested; and (3) considered a tax, social, labour, or insurance program. Although each program has been designed or has evolved to address particular problems—poverty, job loss, disability, horizontal or vertical tax equity, or facilita- tion of labour force participation—all are similar in that they provide a direct cash transfer, tax subsidy, or in-kind benefi t from the government on the basis of individual or household income, demographics, or both. Each program has different overall costs for different levels of govern- ment and results in specifi c (dis)incentive effects and (re) distributive benefi ts for different groups of individuals and households.

These programs vary in their effectiveness at achiev- ing stated policy goals, but a few have been particularly successful. For instance, the introduction of OAS/GIS and C/QPP markedly reduced the poverty rate among seniors ( Myles 2000 ). Although the parameters and administration of family allowances and child benefi ts have changed over time, these programs have reduced the number of children living in households below the poverty line ( ESDC 2018 ).

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 285

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

However, many low-income working-age individuals, particularly those without children, fall through the cracks of Canada’s patchwork system. Current levels of SA pro- vide inadequate support for raising individuals above the poverty line while at the same time creating a “welfare wall,” given that METRs for those on SA in some provinces can exceed 100 percent ( Torjman and Battle 1993 ; Wolfson 2018 ). A few refundable tax credits are available to low- income working-age individuals, such as the CWB and GST credit, but they are small compared with senior and child-related income transfers. EI is temporary and only available to those with recent access to a job who have worked enough to qualify for benefi ts. EI is arguably an insurance program intended to hedge against short-term job loss and not a social program; however, it is structured to redistribute from some types of workers to others ( Busby and Gray 2011 ). 4 Non-refundable tax credits are only available to those who have suffi cient taxes owing to benefi t from their application in the fi rst place, and they are regressive over the income range for which an individual’s tax liability is less than the total value of the individual’s claimable non-refundable tax credits. 5

Canada’s patchwork income security system is partly the result of earnest attempts to fi nd targeted policy solu- tions to social problems at different points throughout the country’s history. At the same time, however, the system consists of programs that were developed, at their worst, to cater to specifi c groups of infl uential vot- ers ( Wolfson 2018 ) and, at their best, to strike “a political compromise . . . among the relevant stakeholders both inside and outside government” ( Banting 1994 , 132). The system has been roundly criticized as an “incoherent hodgepodge that arose from a series of incremental policy changes . . . different [federal and provincial] priorities . . . and government [inability] to do the necessary horizontal policy analyses” ( Wolfson 2018 ).

A GBI is often proposed as a way to fi x the effi ciency- and equity-related challenges in the income security system. It is also promoted as a superior alternative to SA, which is inadequate, stigmatizes recipients through means testing, and creates work disincentives because of high clawback rates. Despite the effi ciency- and equity-related benefi ts that could be achieved through restructuring government programs and expenditures toward a GBI, addressing these issues fi rst requires consideration of two broad sets of political challenges—federalism and work disincentives. We outline each in turn.

Challenges of Federalism Canada’s treaties, constitution, jurisprudence, and cur- rently accepted policy give the federal government, provincial governments, and First Nations sole jurisdic- tion over certain policy areas and shared jurisdiction over others. For instance, provinces are primarily responsible for most labour and social policies, yet have entered into

various agreements with the federal government to ad- minister and fund particular income security programs (e.g., CPP, CWB). The federal government’s relationship with First Nations differs across communities, but First Nations, tribal councils, or both are responsible for ad- ministration of many on-reserve social programs ( Broad and Nadeau 2012 ).

The federal government plays a facilitative role in bringing the provinces and First Nations together and a redistributive role in reducing regional income inequal- ities across the country. Provinces and First Nations can benefi t from economies of scale in federal coordina- tion and implementation of programs. However, the federal government’s role and perceived effectiveness is politically contested across provinces and First Na- tions communities—for instance, there are often legal disputes and jurisdictional tensions between different governments over tax and transfer policies ( Béland et al. 2017 ). Because of the multi-level roles of governments in the tax and transfer system, we propose that a politically acceptable federal GBI is not possible without the full participation of provinces and First Nations in its design and implementation.

Perhaps more important, no adequate GBI is affordable without including both federal and provincial contribu- tions to its fi nancing. For instance, a full federal roll-out of a GBI based on the basic income pilot design in Ontario has been estimated at $76.0 billion for 2018–2019, yet the federal government spends less than half of that on direct transfers to groups that would benefi t from this particular GBI design ( Canada PBO 2018 ). Creating a revenue- neutral program requires consolidating and restructuring most current federal and provincial income security pro- grams, as well as some tax reform ( Boadway et al. 2018b ). The jurisdictional responsibilities for, and interaction ef- fects between, existing programs and federal–provincial tax collection agreements must be considered ( Tedds 2017 ), as well as the distributional implications across households and provinces. The challenges of fi scal federal- ism are different but no less pronounced for the federal government’s current funding and program arrange- ments with First Nations communities. Assessments of any newly proposed system impacts on First Nations outcomes should be undertaken with signifi cant engage- ment with First Nations communities, their leaders, and their members to avoid potential unintended negative consequences of adopting impact and evaluation criteria set by policy-makers ( Broad and Nadjiwon-Smith 2017 ).

There is also moral hazard if the federal government implements a federal GBI without full cooperation of provinces and First Nations. If provinces do not hold the amounts they spend on social programs constant (e.g., Boadway et al. 2018b ), or if federal and provincial income security programs lack integration, individuals may seize arbitrage opportunities across jurisdictions.

286 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

More specifi cally, low-income individuals may move to provinces that provide a GBI and, to the extent that a GBI is funded through tax increases, high-income individuals may move to provinces that do not (similar effects may occur on and off reserve), increasing income inequality within the country. The potential for arbitrage among high-income individuals is a well-known problem in a political federation (e.g., Musgrave 1971 ; Oates 1968 ), and the concern about jurisdiction hopping among low-income individuals has been raised elsewhere in discussions of a global basic income guarantee ( Van Parijs and Vander- borght 2017 ).

The aforementioned issues require consideration of whether universal cooperation between the federal and provincial governments and First Nations communities is necessary and, if the answer is yes—which we believe it is—the extent to which it is realistic. Given jurisdic- tional realities, the development of any framework will need to provide the provinces and First Nations control over key aspects of the program, with a few caveats. First, to ensure some parity across the country and ad- dress the moral hazard problem, in return for the federal government committing to fund a set percentage of the relevant income threshold measure in the province or on reserve, provinces and First Nations must agree to fund the remaining percentage of the relevant income threshold measure. Second, provinces and First Nations should only be able to deem individuals or households ineligible for the program as long as these individuals or households already have access to an equivalent program that would be maintained.

Economies of scale could be achieved by the federal government’s administration of the program through the Canada Revenue Agency and personal tax system. Participation by First Nations leaders and organizations would determine the specifi c details of the funding, ad- ministration, and delivery arrangements that would need to be adjusted on reserve to refl ect differences between how the provinces and First Nations operate vis-à-vis the federal government. 6

Given the opportunities and challenges that we have outlined with federalism, it is critical to know where to begin in developing a framework for a federal–provincial– First Nations GBI. What type of framework would encourage cooperation? What type of framework could be developed and implemented in a short period of time to avoid changes in governments and shifts in policy priorities?

We propose that the best approach would be to expand upon a well-established program that already requires cooperation between different levels of government and is a variant of a NIT. A program that fi ts this description is the CWB (formerly the Working Income Tax Benefi t [WITB]), which operates as an earnings subsidy for those who earn between a minimum and maximum amount. 7

The benefi t of using the CWB as the foundation for a GBI is that it already addresses many of the challenges of federalism that we raise in this section. For instance, although the CWB is a national program, provinces can tailor it ( Mah 2017 ), subject to certain general constraints such as preserving actions to improve work incentives for low-income individuals and families, provision of a minimum benefi t, cost neutrality to the federal govern- ment, and harmonization with existing federal programs ( Kesselman 2019 ). Currently, British Columbia, Alberta, Quebec, and Nunavut have all made changes to the initial federal design. For example, individuals with children in Quebec receive a much smaller CWB payment than those residing in other provinces. This was permitted by the federal government because other Quebec programs already provide generous transfers to families with chil- dren ( Finance Canada 2016 ). The CWB has advantages for addressing the challenges of federalism, but it also addresses both economic and political concerns about work disincentives, which we turn to next.

Challenges of Work Disincentives Addressing the challenges associated with work disincen- tives of a GBI are arguably even more problematic than addressing federalism challenges. Also, to complicate matters further, there are both real and perceived work disincentives that affect political support for the introduc- tion of a GBI.

The extent to which individuals would rely on a GBI in lieu of market-based work depends largely on its generosity. Prior proposals by public economists for a targeted basic income of $20,000 (adjusted for household size) have modeled a small labour supply response at the intensive margin (e.g., Boadway et al. 2018b ), suggesting they are not overly concerned about large negative effects. 8 However, if enough people stopped working in the labour market (or reduced their labour supply) and subsisted on the GBI, it would both raise the costs of the program and reduce government tax revenues. Benefi ts that result from improved population health outcomes (e.g., Forget 2011 ) and reduced crime rates (e.g., Marinescu 2017 ) are less likely to be realized immediately, and any negative impacts on labour supply would occur in the shorter term, reducing the likelihood that a government sensitive to the election cycle and concerned about work disincentives would implement a GBI.

Although Canada has implemented various basic income and welfare-to-work pilot experiments over the years (e.g., the basic income experiment in Dauphin, Manitoba, in the 1970s; the self-suffi ciency projects in British Columbia and New Brunswick in the 1990s; and, most recently, the basic income pilot in Ontario, which was cancelled in 2018), no universally accessible, income- tested but otherwise unconditional, cash transfer program that gets individuals above a poverty threshold has ever

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 287

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

been introduced at scale. We cannot know for certain what the long-run equilibrium effects might be, and un- fortunately, conducting short-term pilots cannot help us much in this regard ( Hoynes and Rothstein 2019 ; Pohler and Koebel 2018 ).

Some scholars suggest societies could follow a more incremental approach, by introducing a smaller GBI that may not get everyone above the poverty line in a fi rst-stage introduction but that allows any adverse effects to be ob- served before implementing a more generous guarantee ( Van Parijs and Vanderborght 2017 ). Many economists also suggest implementing a low tax-back rate for social benefi ts to mitigate work disincentives (e.g., Torjman and Battle 1993 ).

Governments have also introduced refundable earnings tax credits to incentivize work. Although SA negatively affects labour force participation, reviews of the empirical evidence on the effects of working tax credits on labour supply show any negative effects to be minimal and to mostly affect mothers ( Brewer et al. 2006 ; Eissa and Hoynes 2004 ; Eissa and Liebman 1996 ). In fact, most of the effects of earnings tax credits on labour supply are positive at the extensive margin of labour force partici- pation, with small negative or non-existent effects on the intensive margin of hours worked ( Meyer 2010 ). There has been very little research in Canada on the labour supply effects of refundable working tax credits, but the research on child benefi ts similarly shows positive effects on the labour supply of single, divorced mothers and small nega- tive effects on married mothers ( Koebel and Schirle 2016 ).

Two groups with particularly high work disincentives in the current income security system are working-age Canadians on SA, who face a “welfare trap” associated with high METRs on entering the labour market (or work- ing more than a nominal amount), and those in the lowest income tax bracket—the working poor—who often fail to qualify for SA and related benefi ts yet also receive little benefi t from non-refundable tax credits and contribution- based programs tied to work such as EI and C/QPP. The higher relative costs associated with working for individ- uals in these groups, and the often low-paid, part-time, temporary, and contingent nature of their employment, requires that any proposed GBI program concerned about work disincentives fi rst consider its effect on them.

It is also important to consider not only the potential for real work disincentives of a GBI on individuals, but also the perception of work disincentives among the pub- lic. Although these perceptions are deeply embedded in somewhat problematic assumptions about the disutility of work (e.g., Kaplan and Schulhofer-Wohl 2018), the perceived impact of a GBI on work can lead to perceptions of unfairness among those who are net contributors (or even those who are not) if the public believes that those benefi ting from a GBI are lazy or undeserving ( Young and Mulvale 2009 ). Beliefs about work disincentives reduce

support for a GBI and enhance support for means-tested approaches to delivering income security programs.

Whether one agrees or disagrees with these beliefs, they should not be discounted because they are rooted in societal norms and values about reciprocity—that is, what individuals should be required to contribute (or not) to their communities in exchange for access to public benefi ts ( Rossiter 2018 ). Societies should not dismiss expectations that citizens will give back; all citizens benefi t from the production of public goods. Beliefs about the import- ance of work and expectations of reciprocity are not the problem; the problem is that on the basis of these beliefs, public benefi ts provided to low-income individuals are more heavily monitored than those provided to higher income individuals (e.g., Wolfson 2018 ).

Changing attitudes about these issues is diffi cult, and an alternative approach is required if a GBI for working- age individuals will ever be introduced in this lifetime. We propose that including a work incentive in the design of a GBI program would help alleviate these concerns. Supporting this, the Earned Income Tax Credit (EITC) is among the largest cash transfer programs in the United States ( Jones and Michelmore 2018 ), and although other benefi ts and entitlements are often derided by American politicians, the EITC has faced relatively little criticism. The program receives wide public support, and there have been bipartisan proposals to expand it (e.g., Cass 2018 ; Hoynes 2014 ; Marr, Horton, and Duke 2017 ). The program is popular because the EITC is universally ac- cessible and designed to encourage and support work, an activity that Americans strongly value (e.g., Porter 2010 ). Again, whether one agrees with these values or not, what constitutes a contribution that is seen by a community as deserving of social benefi ts is ignored at the peril of improvements in social benefi t provision.

We propose that, like the EITC in the United States, expanding out from the CWB would mitigate both real and perceived concerns about work disincentives due to its design, which incorporates an earned income subsidy. All previously proposed NIT variants of a GBI include a tax-back rate (also referred to as a phase-out or claw- back rate ). However, by incorporating elements from the CWB, our hybrid GBI program also includes an earnings subsidy (i.e., a phase-in component), which builds in some conditionality to achieve the maximum amount of the transfer.

We mitigate work disincentives in our proposed GBI design by (1) providing a modest basic guarantee; (2) in- corporating a work incentive focused on labour supply at the extensive margin (i.e., the decision to enter the workforce); and (3) minimizing the tax-back rate, subject to other constraints. Full optimization of these parameters is primarily constrained by the revenues we have available from our proposed fi nancing model. In the next section, we outline our program in more detail.

288 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Program Design Our proposed design is a hybrid program consisting of both an employment earnings subsidy along the lines of the current CWB in Canada and the EITC in the United States and a basic income guarantee. We apply our pro- gram to individuals aged 18–64 years and retain OAS/ GIS and the CCB outside of our framework as the GBI for seniors and children, respectively. We base our design on the individual unit, both for the income used to measure benefi t entitlement and for the provision of the transfer (i.e., the benefi t is unadjusted for family or household size). 9 Individualizing the design rather than adjusting the benefi t for family economies of scale increases the cost, but we chose this design for several reasons.

First, although norms are changing, women still per- form most of the unpaid care, domestic, and volunteering work in Canada ( Moyser and Burlock 2018 ) and may not have equal access to family income if intrahousehold bargaining power is dependent on women earning their own independent income (e.g., Anderson and Eswaran 2009 ). As such, there is a gender-based equity rationale for recognizing this societally valuable work through a GBI. Second, basing the benefi t’s phase-in and phase-out on individual income avoids economic distortions to intrahousehold bargaining and avoids distorting work incentives for married individuals, because their transfer is not affected by their spouse’s income (particularly for married women: Schirle 2017 ). Third, we did not want to disincentivize family formation or create incentives to misreport family or household status. Fourth, targeting the GBI at the individual level softens the net household tax and disposable income implications of the proposed fund- ing of our program that occurs as a result of the removal of certain federal and provincial non-refundable tax credits (e.g., for individual, spousal, and dependent amounts), reducing opposition from some families. This occurs because some spouses or dependents may now receive our refundable benefi t in place of the non-refundable tax credits that the primary income earner in the household was able to claim before the introduction of our program. Fifth, there are many liberal arguments that provide support for an individual GBI, because it preserves the dignity, autonomy, and freedom of individuals to have the opportunity and real material freedom to make their own decisions (see Van Parijs and Vanderborght 2017 ; Young and Mulvale 2009 ). Finally, as a pragmatic matter, using individual income is most consistent with the design of Canada’s personal income tax system, which is based on the individual rather than the household or family. 10

Program Parameters In this section, we describe the following parameters of our program: income threshold for receipt of maximum benefi t transfer, basic guarantee, employment earnings subsidy rate, and phase-out or claw-back rate.

Income Threshold for Receipt of Maximum Benefi t The income threshold for receipt of the maximum benefi t transfer is a function of other parameters we chose to set in our design (which we describe later), as well as the combined current federal and provincial tax rates. The formula for calculating this income threshold is as follows (see Appendix A for the derivation):

I I TM e B

w

0

1( ) ,

where I eM, an individual’s net employment earnings before tax at the entry of the maximum transfer (i.e., beginning of the plateau range), is a function of I B , the market basket measure of poverty for that individual on the basis of their geographic location; T o , the maximum benefi t transfer received when income is zero dollars (i.e., the basic guar- antee); and φw , the employment earnings subsidy rate.

There are different income thresholds for receipt of the maximum benefi t transfer across individuals. Our target disposable income was the market basket measure of poverty (MBM; ESDC 2018 ); we code the relevant MBM for each individual in our simulation (i.e., on the basis of their location). Individuals reach a maximum benefi t transfer when the combination of the transfer (i.e., the basic amount plus the earnings subsidy) and their own disposable labour market earnings (net employment earnings minus income tax) results in an income that places them to their set poverty threshold (MBM). Once an individual’s disposable income reaches the MBM, there is a very small benefi t plateau range for $2,000 additional earnings in all provinces, which was arbitrarily chosen.

Basic Guarantee The basic guarantee represents the amount received by individuals who have no employment earnings. It is set at the maximum income that a single, able-bodied indi- vidual (without children) currently receives on SA in each province. For example, in 2017, individuals on SA received $9,461 per year in Ontario and $8,124 per year in British Columbia; this amount would automatically go to every individual who earns zero dollars of income (we adopt different measures of earnings for the phase-in and phase- out rates, which we discuss later). The basic guarantee was selected so that the lowest income individuals targeted in our program would not be worse off.

Employment Earnings Subsidy Rate Our program incentivizes individuals to enter the labour market by implementing a 30 percent employ- ment earnings subsidy. 11 That is, individuals receive an additional $0.30 for every dollar earned in wages and self- employment. The pre-tax employment earnings subsidy applied in each province varies between 50 percent and 60 percent to account for our removal of the basic personal

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 289

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

amount in our proposed funding plan (described later) and the resulting application of the federal and provin- cial tax rates on the fi rst dollar earned. In other words, to ensure a 30 percent after-tax earnings subsidy in every province, the phase-in rate is set so that after federal and provincial tax rates are applied to the fi rst dollar earned, the individual receives a net subsidy of 30 percent (i.e., the METR is −30 percent). For example, the phase-in rate of our design in Ontario is set at 50.05 percent; on the fi rst dollar earned, 15 percent of an individual’s income is col- lected by the federal government, 5.05 percent is collected by the Ontario government, and 30 percent is distributed to the individual as an earnings subsidy.

Phase-Out or Claw-Back Rate The phase-out rate was set as a function of the chosen plateau range and the dual goals of minimizing METRs and approximating revenue neutrality based on our pro- posed fi nancing (described later). Unlike the phase-in rate, which is applied to employment earnings, the phase-out rate is applied to individuals on the basis of their total in- dividual net income. 12 Thus, if an individual’s net income is greater than I eM , the phase-out will apply regardless of the value of I eM . Although this reduces the strength of the employment earnings incentive for these individuals, it also reduces the program cost, improves its targeting to low-income workers, and avoids providing a benefi t to higher income individuals with little or no employment earnings. 13 Once an individual’s net income places the

individual at the end of the benefi t plateau range, the transfer begins to be clawed back at a rate of 50 percent. 14

This program design is graphically illustrated in Figure 1 below using Ontario as an example.

Cost and Financing We use Statistics Canada’s Social Policy Simulation Data- base and Model (SPSD/M; Version 27.0) to estimate the cost of our proposed program and to develop the tables showing the average tax implications and distributional impacts of implementing our program across household income deciles and family types. 15

The estimated gross cost is $90.51 billion. 16 We follow the general approach of Boadway et al. (2018b ), who propose a revenue-neutral fi nancing option through the consolidation and replacement of provincial SA programs, most current federal and provincial refundable tax credits, and certain federal and provincial NRTCs. Refundable tax credits and SA already target low-income individuals/households and Boadway (2015) provides a case for the inequity of non-refundable tax credits. 17 Our choice of NRTCs differs slightly from Boadway et al.’s (2018b), because we remove fewer NRTCs to avoid incurring large costs for middle- income households, who benefit from many NRTCs. Similar to Boadway et al., we do not remove NRTCs associ- ated with incentivizing behaviours such as charitable and political donations and retirement and education savings. See Appendix B for a complete list of federal and provincial non-refundable tax credits that are removed.

Figure 1 : Proposed Design (Illustrative Example of Ontario)

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

$0

$2,000

$4,000

$6,000

$8,000

$10,000

$12,000

$14,000

$16,000

$18,000

$0 $5,000 $10,000 $15,000 $20,000 $25,000 $30,000 $35,000 $40,000

G B

I/C W

B A

m ou

nt ($

)

Earnings (Phase-In)—Net Income (Phase-Out)

Phase-Out: 50% Phase-In: 50.05%

290 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Table 1 : Impact on Household Disposable Income by Household Net Income Decile, Ages 18–64 Years

Decile ($2018)

Average GBI/

CWB, $

Average Household Disposable Income, $

Average Change in Disposable Income

Pre-GBI Post-GBI $ %

0–4,147 9,805 9,067 12,920 3,853 42.49 4,148–12,504 11,969 12,862 21,207 8,345 64.88 12,505–20,737 10,392 19,410 25,949 6,539 33.69 20,738–30,893 7,166 27,203 30,229 3,026 11.12 30,894–42,303 3,656 35,856 35,236 −620 −1.73 42,304–55,154 3,430 44,213 43,542 −671 −1.52 55,155–72,434 3,780 55,522 54,815 −707 −1.27 72,435–94,689 2,933 70,397 68,480 −1,917 −2.72 94,690–131,915 2,118 92,468 89,318 −3,150 −3.41 131,916+ 1,319 170,315 166,128 −4,187 −2.46 Aggregate 5,502 56,513 57,340 827 1.46

Note: Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

In addition to the basic personal deduction, most of the NRTCs we eliminate are based on the presence of a dependent in the household: married, married equivalent, infi rm dependent, and caregiver. Our program, which provides an average of $8,790 at zero dollars of income across the provinces, could be partially conceptualized as restructuring current tax expenditures on NRTCs within the household from taxpayers benefi ting from these NRTCs to the spouses or dependents themselves (who are aged between 18 and 64 years). We also remove all education-related NRTCs; students who benefi t from the tuition and education amounts would be eligible to receive our transfer instead. In this way, we reduce the tax increases associated with removing NRTCs in the household unit.

Like Boadway et al. (2018b), we leave equalization payments and the Canada Health and Social Transfers to the provinces in place, and we allow the federal ex- penditures on income security programs and NRTCs to be redistributed across the provinces to fund the federal portion of our program. Given federal revenues and the cost of the program, the federal government fi nances 55 percent of the cost within each province. However, unlike Boadway et al., we constrain the provincial savings from the consolidation of SA and removal of the provincial ex- penditures on NRTCs to remain in the provinces to fund their own programs, rather than allocating these amounts across provinces. This fi nancing approach should reduce provincial opposition to the program.

See Appendix C for the total revenue available from the federal and provincial funding sources (i.e., refundable tax credits, NRTCs, and direct cash transfers for provincial SA). The federal revenues from these funding sources are $53.70 billion and provincial revenues are $39.30 billion. These consolidations, tax reforms, and funding choices lead to small provincial defi cits in Newfoundland and Labrador, Prince Edward Island, Nova Scotia, Ontario, and British Columbia. However, overall, our program is revenue neutral; we model a federal surplus ($3.92 billion) that is higher than the combined cost of the provincial defi cits ($1.43 billion).

Distributional Impacts In this section we present tables of the distributional impacts across income deciles for Canada. 18 We used individuals aged 18 years and older to determine the household net income deciles, except when presenting the results by family type, where we use income deciles for the relevant subgroup of households.

Table 1 shows the average distributional impacts on our target population of Canadians aged between 18 and 64 years across household income deciles. There are a few key impacts of our program. Households in the lowest four income deciles are net benefi ciaries, whereas house- holds in the highest six income deciles are net contributors.

Although we do not raise statutory tax rates to fund our program, there are de facto tax increases associated with our program as a result of the removal of the tax credits. A limitation for the political feasibility of our design is the average negative effect on the disposable incomes of middle-income households, although the tax increases in our program are much smaller than those in other GBI proposals (e.g., see Boadway et al. 2018b ), on average in- creasing the taxes for households in income deciles fi ve to seven by approximately $650, or 1.5 percent of disposable income; the largest absolute tax increase occurs for house- holds in the top income decile at $4,187 (approximately 2.5 percent of disposable income).

The negative effect on middle-income households could be mitigated by fi nding additional sources of rev- enue outside the personal tax and transfer system that would eliminate the need to remove as many NRTCs to fi nance the program. Alternatively, our preference would be to use any additional sources of revenue to fund a reduction in the claw-back rate, which would have the dual effects of reducing the marginal effective tax rate and pushing access to the benefi t higher up in the income distribution, thereby mitigating work disincentives and offsetting the tax increase in the middle deciles that arises from elimination of the NRTCs.

Tables 2 – 7 show the distributional impacts across income deciles by family type. Although seniors are not eligible for our program, we show their distributional impacts, because seniors are affected by the loss of the tax credits. The distributional impact tables by family type show that on average there are negative effects on the dis- posable incomes of senior and single-parent households

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 291

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Table 2 : Impact on Household Disposable Income by Family Type, Senior Single Households

Decile ($2018)

Average GBI/

CWB, $

Average House- hold Disposable

Income, $

Average Change in Disposable

Income

Pre-GBI Post-GBI $ %

0–7,850 0 17,817 16,639 −1,178 −6.61 7,851–12,080 0 21,559 20,422 −1,137 −5.27 12,081–14,840 0 22,462 20,847 −1,615 −7.19 14,841–18,548 0 23,658 21,611 −2,047 −8.65 18,549–23,084 0 25,696 23,078 −2,618 −10.19 23,085–29,579 0 28,343 25,488 −2,855 −10.07 29,580–37,846 0 34,005 31,080 −2,925 −8.60 37,847–48,149 0 40,374 37,593 −2,781 −6.89 48,150–67,584 0 50,019 47,381 −2,638 −5.27 67,585+ 0 98,819 96,142 −2,677 −2.71 Aggregate 0 36,288 34,041 −2,247 −6.19

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0).

Table 3 : Impact on Household Disposable Income by Family Type, Senior Couple Households

Decile ($2018)

Average GBI/

CWB, $

Average House- hold Disposable

Income, $

Average Change in Disposable

Income

Pre-GBI Post-GBI $ %

0–21,914 3,269 26,874 27,211 337 1.25 21,915–31,948 1,819 36,926 35,504 −1,422 −3.85 31,949–42,906 1,558 42,292 39,535 −2,757 −6.52 42,907–51,172 1,188 48,197 44,435 −3,762 −7.81 51,173–62,410 1,040 55,799 51,555 −4,244 −7.61 62,411–73,473 646 63,657 59,069 −4,588 −7.21 73,474–88,058 680 74,130 69,549 −4,581 −6.18 88,059–106,868 633 87,521 82,803 −4,718 −5.39 106,869–140,679 354 107,637 102,746 −4,891 −4.54 140,680+ 213 246,381 241,196 −5,185 −2.10 Aggregate 1,140 78,940 75,359 −3,581 −4.54

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

Table 4 : Impact on Household Disposable Income by Family Type, Single Parent Households

Decile ($2018)

Average GBI/

CWB, $

Average House- hold Disposable

Income, $

Average Change in Disposable Income

Pre-GBI Post-GBI $ %

0 9,080 23,672 22,349 −1,323 −5.59 1–3,312 9,431 23,243 26,101 2,858 12.30 3,313–12,297 9,795 31,122 32,058 936 3.01 12,298–21,272 8,901 35,108 38,653 3,545 10.10 21,273–29,437 5,422 38,814 38,310 −504 −1.30 29,438–39,258 1,440 45,392 40,872 −4,520 −9.96 39,259–47,563 135 52,193 46,203 −5,990 −11.48 47,564–56,840 28 54,667 49,374 −5,293 −9.68 56,841–78,967 0 65,479 60,250 −5,229 −7.99 78,968+ 0 102,814 97,658 −5,156 −5.01 Aggregate 4,271 48,002 45,598 −2,404 −5.01

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0).

Table 5 : Impact on Household Disposable Income by Family Type, Two-Parent Households

Decile ($2018)

Average GBI/

CWB, $

Average House- hold Disposable

Income, $

Average Change in Disposable

Income

Pre-GBI Post-GBI $ %

0–33,161 17,714 39,241 49,595 10,354 26.39 33,162–51,570 11,614 53,704 58,701 4,997 9.30 51,571–66,378 8,384 64,112 66,645 2,533 3.95 66,379–79,559 6,442 72,144 72,644 500 0.69 79,560–93,026 4,747 82,280 80,974 −1,306 −1.59 93,027–107,897 3,821 91,612 89,505 −2,107 −2.30 107,898–124,891 2,653 102,736 99,523 −3,213 −3.13 124,892–147,587 2,055 117,612 113,810 −3,802 −3.23 147,588–187,196 1,503 139,111 134,799 −4,312 −3.10 187,197+ 1,568 243,827 239,517 −4,310 −1.77 Aggregate 5,988 101,469 101,344 −125 −0.12

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

( Tables 2 – 4 ). Our program primarily benefi ts two-parent families and working-age singles and couples ( Tables 5 – 7 ).

Single parents and seniors appear worse off under our program for several reasons. The primary reason is that our model is static; if the lowest income single parents

are currently not working, then this result is expected. Single parents on SA receive a higher SA benefi t amount in most provinces, and we set the basic amount at the current level received by single employable individuals without children. In a dynamic model, for single parents

292 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Table 7 : Impact on Household Disposable Income by Family Type, Non-Senior Couple Households

Decile ($2018)

Average GBI/CWB,

$

Average House- hold Disposable

Income, $

Average Change in Dis- posable Income

Pre-GBI Post-GBI $ %

0–28,599 18,680 22,518 33,529 11,011 48.90 28,600–48,451 12,502 40,046 46,600 6,554 16.37 48,452–62,805 8,573 51,618 54,282 2,664 5.16 62,806–76,859 5,921 61,743 61,978 235 0.38 76,860–88,644 3,743 72,316 70,270 −2,046 −2.83 88,645–101,895 3,134 81,092 78,411 −2,681 −3.31 101,896–119,219 2,193 93,703 90,109 −3,594 −3.84 119,220–142,913 1,675 109,285 105,093 −4,192 −3.84 142,914–180,034 1,121 131,162 126,511 −4,651 −3.55 180,035+ 1,378 239,634 235,062 −4,572 −1.91 Aggregate 5,893 90,311 90,185 −126 −0.14

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

Table 6 : Impact on Household Disposable Income by Family Type, Non-Senior Single Households

Decile ($2018)

Average GBI/

CWB, $

Average House- hold Disposable

Income, $

Average Change in Disposable

Income

Pre-GBI Post-GBI $ %

0 8,958 7,116 9,738 2,622 36.85 1–4,828 9,677 5,556 12,649 7,093 127.66 4,829–10,094 11,279 10,005 18,323 8,318 83.14 10,095–15,915 11,134 14,719 22,300 7,581 51.50 15,916–22,256 8,179 19,304 23,998 4,694 24.32 22,257–30,966 4,308 24,950 25,755 805 3.23 30,967–40,476 579 31,724 28,948 −2,776 −8.75 40,477–51,862 0 38,805 35,782 −3,023 −7.79 51,863–72,690 0 50,553 47,620 −2,933 −5.80 72,691+ 0 91,051 88,062 −2,989 −3.28 Aggregate 5,400 29,405 31,266 1,861 6.33

Note: CWB = Canada Workers Benefi t; GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

who are not working, our program benefi ts would provide an incentive to work (rather than a disincentive because of the welfare wall) and help overcome fi xed costs as- sociated with working (consistent with prior empirical research on cash transfers for single parents, e.g., Koebel and Schirle 2016 ), mitigating this negative effect on single- parent households over time. Indeed, Table 4 shows that low-income single parents in income deciles two, three, and four, who are more likely to be working, gain from our program. To the extent that our benefi t transfer does not provide enough to help single parents overcome fi xed costs (e.g., daycare) of working at least a few hours per week, these individuals would, on average, be worse off under our system.

Seniors are worse off under our program because of the inaccessibility of our benefi t scheme for this group and the removal of tax credits. To mitigate this, our program could be expanded to include working seniors, or current GIS benefi ts for low-income seniors could be increased a small amount to offset reduction in disposable income resulting from the elimination of NRTCs. Because we do not remove tax credits associated with the age or pen- sion amounts, another option could be to increase the age amount, keeping it phased out with higher incomes, although this would not help the lowest income seniors.

Notwithstanding the average negative effects on sen- iors and single parents, there is still a small reduction in the poverty rate (i.e., the MBM) for single parents and senior couples in our program when we account for both the benefi t transfer and the removal of tax credits. Table

8 shows the impact of our program on the poverty rate by household type. The small increase in the poverty rate for senior singles could be addressed by making the previously suggested adjustments to GIS. Overall, our program reduces poverty by 40 percent, and the reduction in poverty rates for two-parent families and working-age singles and couples without children is substantial. 19 Table 9 shows the percentage of households that gain from our program across income deciles.

Implications of Program Design We consider the implications of our program design on labour supply and demand in this section, as well as its im- pact on people with disabilities. We also highlight several other administrative and implementation considerations of our program in Appendix D.

Work Incentives Under this design, individuals reach the MBM poverty threshold working between 9.3 and 16.5 hours per week at the current provincial minimum wage in each province (or earn between $5,656 and $9,613 annually depending on the province). 20 In addition to the benefi ts that this design provides in addressing the political challenges associated with real and perceived work disincentives, we justify incentivizing a certain amount of labour force participation because it has well-documented economic and psychological benefi ts for individuals while at the same time allowing people more fl exibility and freedom to choose hours spent in market-based work versus pursuing

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 293

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Table 8 : Impact on Poverty Rate (Market Basket Measure of Poverty), Age 18 Years and Older

2018 Poverty Rate, % Impact

Pre-GBI Post-GBI % Point %

Single Parent 15.2 14.8 −0.40 −2.63 Two Parent 3.4 1.1 −2.30 −67.65 Senior Single 2.2 3.9 1.70 77.27 Senior Couple 1.5 0.8 −0.70 −46.67 Non-Senior Single 17.0 10.4 −6.60 −38.82 Non-Senior Couple 5.3 2.2 −3.10 −58.49 Aggregate 7.7 4.6 −3.10 −40.26

Note: GBI = guaranteed basic income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

Table 9 : Percent Gaining by Net Income Decile, 18+

Decile Households, %

1 60.2 2 68.0 3 62.3 4 47.8 5 8.1 6 5.2 7 2.5 8 0.4 9 0.7

10 0.3 Aggregate 25.5

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)

other societally valuable, non-market-based activities. Just as important, however, is that as a proportion of their income, lower income individuals face higher fi xed costs of working relative to both higher income individuals and those who do not work, which may require additional income to offset these costs and overcome barriers to participation.

One consequence of using different program par- ameters across the provinces is that there is provincial variation in terms of the employment earnings necessary to obtain the maximum benefi t. For example, in Quebec, individuals must on average earn roughly $6,188 (or work an average of 9.9 hours at the provincial minimum wage) to qualify for our maximum benefi t amount; in contrast, individuals in Nova Scotia need to earn $9,212 (i.e., 16.1 hours a week at the provincial minimum wage). Thus, residents of Nova Scotia must work more than those in Quebec to obtain an income at the poverty line. This outcome is a function of the current provincial minimum wages as well as setting different basic guarantee amounts across provinces (i.e., at the current provincial level of SA) for individuals with no income. There is a trade-off between designing a system whereby every individual, regardless of province of residence, has to work exactly the same hours to get to the poverty line and setting the guaranteed amount such that current SA recipients are not worse off after the implementation of our program. Our program design places more weight on the latter priority.

One other important point on the work incentives in our design is that using two different individual income measures for the earnings subsidy (i.e., employment in- come) and phase-out (i.e., net income) rates means that if an individual’s net income is approaching the threshold for the maximum benefi t, even if they have no or few employment earnings, they would receive part of the benefi t if their net income is below the exit threshold and no transfer if their net income exceeds the exit threshold.

This reduces the work incentive for those who have higher net incomes but no employment earnings, but it also reduces the cost and improves the program’s targeting toward low-income working individuals.

Marginal Effective Tax Rates One of the unavoidable challenges of a NIT design is that at the point at which the claw-back rate kicks in, there will be an increase in the METRs for individuals who are subject to it and now face a possible work disincentive at the intensive margin ( Kesselman 2018 ). However, there are a few improvements of our design over the status quo of current income security programs for the METRs faced by individuals.

First, one of the ways to mitigate high METRs is through adopting a low tax-back rate, which softens the reinforcing substitution effect that theoretically reduces labour supply; we have set our phase-out at 50 percent in every province. This is lower than the current claw-back rates for all provincial SA programs apart from Ontario (which is exactly 50 percent). Almost all SA recipients will face lower METRs in our program compared with the status quo, 21 and they can receive a higher transfer under our program than in the current system.

Second, the complete phase-out of our program’s benefi ts occurs much higher up in the income distribution relative to current SA benefi ts. These individuals are less likely to be subject to a work disincentive arising from the complete phase-out of benefi ts for three reasons: (1) they have already overcome the fi xed costs associated with working (participation decision); (2) at the point at which the benefi t fully phases out, they are more likely to be working in full-time jobs and thus are less able to reduce hours worked; and (3) reduction in their labour supply would reduce their overall disposable income—the net effects of working on disposable income are more salient for most low-income individuals (and even accountants;

294 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

see Chetty and Saez 2013 , 2–3) than the income and sub- stitution effects that primarily concern tax and labour economists.

Our proposed design also addresses critiques of prior proposals that do not consider the interaction between federal and provincial tax rates and the resulting impact of eliminating tax credits on the METRs for targeted groups (see Kesselman 2018 ). We do not fund our program through removal of administrative costs, employment assistance and training programs, or in-kind provision of social services (e.g., housing, medical and dental benefi ts), so these do not factor into our METR calculations.

Demand-Side Considerations and People with Disabilities Our proposal does not directly address labour demand, and it assumes that individuals would be able to fi nd at least 9–16 hours of work per week at the current minimum wage in each province. This is a fairly low income–work threshold required to receive the maximum benefi t in our program; however, it could still pose challenges for those who live in low-development areas with few employers or who may not have the necessary skills for the jobs available.

Perhaps most problematic is that our program design does not provide a basic income above the MBM poverty threshold for those who are unable to work. According to the Canadian Survey on Disability, of the almost 3.8 mil- lion Canadians who reported some type of a disability, just fewer than half reported a severe or very severe disability (approximately 5 percent of the entire Canadian popula- tion; Statistics Canada 2015 ). The number is much lower if we exclude seniors (for whom the prevalence of disability is substantially higher than for working-age Canadians) and children, who both receive disability supplements through other programs that we do not change. However, a small number of working-age individuals with severe disabilities may end up slightly worse off under our pro- posed system restructuring.

Medical certifi cates are already required to receive cer- tain types of disability supplements and services, so one option to address the inadequacy of our program for those who cannot work would be to create an additional dis- ability supplement. Another option would be to increase the basic guarantee for everyone to the amount currently provided to people with disabilities to reduce any stig- matization associated with applying for and obtaining a disability supplement. However, this option substantially increases the cost, as well as the real and perceived work disincentives for employable individuals, and it thus reduces the political feasibility of the GBI.

Ensuring that those unable to work are not made worse off under our program would not be that costly. If only public assistance and not disability supplements col- lected under pension and social insurance programs such

as C/QPPD and worker’s compensation are considered, the current income transfer and public assistance pro- grams supporting working-age people with disabilities currently provide very low supplementary benefi ts ( De Jong 2018 , 235).

However, we propose that responsibility for additional support for individuals with severe disabilities should be shifted from the tax and transfer system to the health care system, with associated funding increases to that system. The tax and transfer system is an inappropriate place to deal with the specifi c needs of the relatively small propor- tion of Canadians who are completely unable to work at least nine hours per week at the minimum wage as a result of severe physical and mental disabilities. People with se- vere disabilities may face additional costs compared with other Canadians that could still be inadequately covered by adopting a design with either a higher basic guarantee or an additional disability supplement. People with dis- abilities have different needs, and some may require much larger supplements than others to help pay for specialized supports. These supports cannot be adequately provided within the design of a less intrusive and less stigmatizing income security program administered by ensuring hori- zontal equity in the tax and transfer system; indeed, “to achieve a broader social policy objective, direct support measures undertaken outside the tax system are likely to be more effective” ( Smart and Stabile 2006 , 420). Thus, we maintain all additional administrative costs and in-kind services for people with disabilities in our program. 22

Interaction with Employment Insurance More detailed consideration will need to be given to the interaction of our program with EI. Appendix F outlines some adaptations that could be made to our program if we eliminate EI benefi ts and employee EI contributions but retain the employer portion of the EI contributions to fund our program. This option allows us to reduce the phase-out rate in our program from 50 percent to 40 per- cent, reduces the tax burden of our program on the middle class, and improves the fi scal position of the provinces.

Conclusion In this article, we propose a hybrid program that integrates an employment earnings subsidy with a basic income guarantee. We sought to address both economic and political challenges associated with GBI implementation in Canada by expanding on the framework and design of the CWB. Our primary goal was to design a program that considered economic issues addressed in other propos- als, such as adequacy of support, cost, fi nancing, and the impact on average and marginal effective tax rates, while at the same time addressing major critiques that limit the political feasibility of prior proposals. More specifi cally, we consider the political economy implications related to federalism and work disincentives. What we have

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 295

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

ultimately proposed is comprehensive tax, social, and labour policy reform for working-age Canadians, and it remains to be seen whether provincial governments, First Nations, and the federal government can agree on the de- velopment and implementation of a universally accessible income security program that addresses the challenges in Canada’s current patchwork system.

The major critiques of our program from public econo- mists and basic income advocates, respectively, are that (1) a more targeted household design or scaling the benefi t for family size would reduce cost and improve redistribu- tive effi ciency and (2) the basic guarantee provided in our design may be inadequate and the maximum benefi t is partially conditional on working, thus we are not propos- ing a true basic income. We do not dispute these critiques. However, we believe that certain trade-offs are required to enhance the political feasibility of a universally accessible GBI in Canada, and the choices we made in our design are a result of our continued struggle to evaluate these trade- offs against each other. Others would inevitably choose different trade-offs. Any serious policy proposal for a basic income, however, should consider both economic and political implications of design implementation, as the devil is in these details. We hope our proposal will cre- ate debates about our chosen trade-offs and inspire other proposals that move us toward the actual implementation of a GBI in Canada.

Acknowledgements The authors thank Robin Boadway, J. Rhys Kesselman, Harvey Stevens, David Green, Yiannis Kipouros, Arthur Sweetman, Michael Veall, Colin Busby, Elliot Rossiter, Katherine Cuff, Morley Gunderson, Jen Budney, Kelly Foley, Kevin McLeod, and Ken Coates for helpful com- ments on drafts of this article. They are also grateful to Statistics Canada for the availability of the Social Policy and Simulation Database and Model. All simulations were undertaken by the authors and any errors are their own.

Notes 1 For instance, see a recent exchange on the topic by Kessel-

man (2018 ), Boadway, Cuff, and Koebel (2018a ), and Ste- vens and Simpson (2018) in Canadian Public Policy .

2 The cost estimate for an individual UBI of $22,000 is ap- proximately $650 billion in Canada.

3 The factors that contribute to poverty cannot all be ad- dressed by income transfers; consideration of poverty re- duction strategies across regions and groups of people have been reviewed elsewhere (e.g., see Banerjee, Benabou, and Mookherjee 2006 ; Danziger and Haveman 2002 ; ESDC 2018; Native Women’s Association of Canada 2017 ).

4 For instance, EI provides benefi ts during parental leave (see Canada 2018 ).

5 For those income ranges with tax liabilities greater than the value of the non-refundable tax credit, the tax benefi t is uniform; thus, over this income range, the benefi t would

be greater for lower income individuals or households than higher income individuals or households as a percentage of income. See Boadway (2015 ) for a more detailed analysis of the inequity of non-refundable tax credits and Macdonald (2016 ) for a broader discussion of the unequal distribution of federal tax expenditures.

6 We are not experts on the complex details of federal frame- works with First Nations communities, and there is unlike- ly to be a one-size-fi ts-all approach that could be adopted. However, it is important that negotiations occur with their participation and that an acceptable framework be devel- oped so as to achieve widespread support for this program among as many First Nations as possible. The same oppor- tunity to access social benefi ts should be provided to First Nations people as to other Canadians.

7 WITB (now CWB) has been more diffi cult for First Nations people to access than other Canadians because it is adminis- tered through the personal income tax system, which could be a limitation of our proposed design; see AFOA Canada and Prosper Canada (2018 ) for a recent proposal with rec- ommendations for helping improve benefi t take-up among Indigenous people.

8 Modeling dynamic labour supply effects relies on a num- ber of controversial assumptions. Benefi ts of a GBI are also rarely modeled when determining net costs or down- stream savings (e.g., improved health, reduction in crime, increased education).

9 This creates some ineffi ciency in redistribution for poverty reduction. If family income was used to determine eligibil- ity for the benefi t instead of individual income, for instance, by allocating half of the family income to each spouse, as has been proposed elsewhere (e.g., Boadway, Cuff, and Koebel 2018c ), this could reduce the cost of our program, improve its targeting, or allow us to increase the basic amount. Other options include adjusting the basic amount provided for household income using the square root scale (e.g., OECD n.d. ; Stevens and Simpson 2017 ). However, our preference is to use individual income for the reasons we outline.

10 Some transfers or non-refundable tax credits (NRTCs) are based on family, for cost reasons or to achieve certain types of tax fairness.

11 The inclusion of self-employment earnings may require great- er consideration by policy-makers, because self-employment earnings can be more easily manipulated to maximize ben- efi t receipt than wages in these types of designs (e.g., Chetty, Friedman, and Saez 2013).

12 The phase-in rate is applied to the individual’s wages and self-employment earnings and the phase-out rates are ap- plied to the individual’s net income using the variables “immemp” and “iminet” in the SPSD/M, respectively. The net income variable corresponds to CRA’s defi nition of net income—total income minus deductions.

13 Using these two different income measures for the phase- in and phase-out rates has the possible effect of distribut- ing high benefi ts to individuals who have low employment earnings but a net income that places them beyond the ben- efi t plateau (but below the exit threshold). Another option could be to base the entire program on employment earn- ings only; however, this would have a greater inequitable

296 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

effect of distributing high benefi ts to individuals who have low employment earnings but higher net income. To ad- dress both of these problems, the benefi t transfer could be adjusted if it is being calculated on the basis of net income and the individual has no or low employment earnings. These modeling choices affect a very small number of in- dividuals, however, because net income and employment earnings tend to be equal or very similar for most low- income individuals.

14 We adopt an income threshold rather than a wealth thresh- old for access to our program to remove the requirement for means testing and enable our program to be more easily administered through the personal tax system; however, to the extent that wealth information is or could easily be col- lected through personal tax returns, an additional wealth threshold could be adopted as a cut-off.

15 All calculations were performed by the authors. 16 Note that one of the limitations of the SPSD/M is that we

cannot provide models of the cost for First Nations people living on reserves, or for individuals living in the territories. A back-of-the-envelope calculation suggests that if every sin- gle person living in the territories (2016 Census population numbers) and on-reserve (2016 Census numbers for First Nations people with registered or treaty Indian status living on reserve) received a maximum benefi t of $12,968, the addi- tional gross cost would be approximately $5.7 billion. This is a very small fraction of the cost of our program, and the most conservative estimate because it ignores overlap between those who live both on-reserve and in the territories and the fact that many people in the territories/on-reserve would be above the income threshold and so would not receive any benefi t. Moreover, similar to our proposed design with prov- inces, the current SA transfers to individuals/households on- reserve and in the territories should be taken into account in determining the net cost of the program.

17 See also Macdonald (2016) for a discussion of the unequal distribution of federal tax expenditures.

18 Provincial distributional impact tables are available on re- quest.

19 Approximately 60% of the reduction in the poverty rate among non-senior singles occurs from providing a benefi t transfer to those between ages 18 and 25 years. If only those aged between 25 and 64 years were eligible for the benefi t, the gross cost of our program would decrease by $27.42 billion. To enhance political feasibility of this proposal, all individuals aged younger than 25 years could be exclud- ed from receiving the benefi t, and these savings could be used to reduce the impact on seniors and the middle class by raising the GIS amount and reducing the tax-back rate, respectively.

20 We calculate the hours using the current provincial mini- mum wages and the annual net employment earnings indi- viduals would need to earn to obtain the maximum transfer in each province. Our program design and these calcula- tions assume that provinces would not lower minimum wages after program introduction.

21 There is a very small range of income in both Manitoba and Ontario over which our METRs are higher for those on SA. See Appendix E for calculations of the METRs under our program compared with the current METRs for those cur-

rently receiving SA and those not receiving SA. We calcu- late these separately for each province and take into account provincial and federal tax rates as well as current claw-back rates on SA, GST credits, and WITB.

22 We also do not recommend eliminating the non-refundable disability tax credit or the refundable and non-refundable medical expense credits. These tax credits serve the impor- tant social policy objective of horizontal equity ( Smart and Stabile 2006 ), although like other non-refundable tax credits, they are less benefi cial for low-income people with disabili- ties, which could be addressed by making them refundable.

References AFOA Canada and Prosper Canada. 2018. “Increasing

Indigenous Benefi t Take-Up in Canada: 2018 Federal Budget Submission.” Toronto: Prosper Canada. At http:// prospercanada.org/getattachment/f4add5df-0edb-4883- b804-60661f500c56/Increasing-Indigenous-benefi t-take-up- in-Canada.aspx .

Anderson, S., and M. Eswaran. 2009. “What Determines Female Autonomy? Evidence from Bangladesh.” Jour- nal of Development Economics 90(2):179–91. https://doi. org/10.1016/j.jdeveco.2008.10.004.

Banerjee, A.V., R. Benabou, and D. Mookherjee, eds. 2006. Understanding Poverty . Oxford: Oxford University Press.

Banting, K. 1994. “The Way Beavers Build Dams: Social Policy Change in Canada.” In A New Social Vision for Canada? Perspectives on the Federal Discussion Paper on Social Secu- rity Reform, ed. K. Banting and K. Battle, 131–38. Kingston and Ontario: School of Public Policy Studies and Caledon Institute of Social Policy.

Béland, D., A. Lecours, G. Marchildon, H. Mou, and M.R. Olf- ert. 2017. Fiscal Federalism and Equalization Policy in Canada: Political and Economic Dimensions. Toronto: University of Toronto Press.

Boadway, R. 2015. “Tax Policy for a Rent-Rich Economy.” Canadian Public Policy/Analyse de politiques 41(4):253–64. https://doi.org/10.3138/cpp.2015-044.

Boadway, R., K. Cuff, and K. Koebel. 2018a. “Can Self- Financing Redeem the Basic Income Guarantee? Disin- centives, Effi ciency Costs, Tax Burdens, and Attitudes: A Rejoinder.” Canadian Public Policy 44(4):447–57. https:// doi.org/10.3138/cpp.2018-044

Boadway, R., K. Cuff, and K. Koebel. 2018b. “Designing a Basic Income Guarantee for Canada.” In Federalism and the Welfare State in a Multicultural World , eds. E. Goodyear- Grant, R. Johnston, W. Kymlicka, and J. Myles, 101–29. Kingston: Queen’s School of Policy Studies and McGill- Queen’s University Press.

Boadway, R., K. Cuff, and K. Koebel. 2018c. “Implementing a Basic Income Guarantee in Canada: Prospects and Prob- lems.” Working Paper, Queen’s University, Toronto.

Brewer, M., A. Duncan, A. Shephard, and M.J. Suarez. 2006. “Did Working Families’ Tax Credit Work? The Impact of In-Work Support on Labour Supply in Great Britain.” Labour Economics 13(6):699–720. https://doi.org/10.1016/j. labeco.2005.11.002.

Broad, G., and M. Nadeau. 2012. “Renewing Funding Rela- tionships: Certifying First Nations Social Service Adminis- trators.” Canadian Review of Social Policy 67:74–79.

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 297

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Broad, G., and J. Nadjiwon-Smith. 2017. “B.I.G. and First Nations: Cautions for Implementation.” Northern Policy Institute Research Paper No. 19. Thunder Bay, ON: North- ern Policy Institute. At https://www.northernpolicy.ca/ upload/documents/publications/reports-new/broad- nadjiwon-smith_big-and-fn-en.pdf .

Budd, J.W. 2011. The Thought of Work . Ithaca, NY: Cornell University Press.

Busby, C., and D. Gray. 2011. “Mending Canada’s Employment Insurance Quilt: The Case for Restoring Equity.” Com- mentary No. 144, C.D. Howe Institute, Toronto. At https:// www.cdhowe.org/sites/default/fi les/attachments/ research_papers/mixed//Backgrounder_144.pdf .

Canada. 1971. “Poverty in Canada: Report of the Spe- cial Senate Committee on Poverty.” At http://www. albertasenator.ca/fl ashblocks/data/BT%20Poverty/ Croll%20Report%201971.pdf .

Canada. 2018. “EI Maternity and Parental Benefi ts: What These Benefi ts Offer.” At https://www.canada.ca/en/ services/benefi ts/ei/ei-maternity-parental.html .

Canada, Parliamentary Budget Offi ce. 2018. “Costing a Na- tional Guaranteed Basic Income Using the Ontario Basic Income Model.” Ottawa: Offi ce of the Parliamentary Bud- get Offi cer, Government of Canada. At https://www.pbo- dpb.gc.ca/web/default/fi les/Documents/Reports/2018/ Basic%20Income/Basic_Income_Costing_EN.pdf .

Cass, O. 2018. The Once and Future Worker: A Vision for the Renewal of Work in America . New York: Encounter Books.

Chetty, R., J. Friedman, and E. Saez. 2013. “Using Differences in Knowledge across Neighborhoods to Uncover the Im- pacts of the EITC on Earnings.” American Economic Review 103(7):2683–721. https://doi.org/10.1257/aer.103.7.2683.

Chetty, R., and E. Saez. 2013. “Teaching the Tax Code: Earn- ings Responses to an Experiment With EITC Recipients.” American Economic Journal: Applied Economics 5(1):1–31. https://doi.org/10.1257/app.5.1.1.

Danziger, S., and R.H. Haveman, eds. 2002. Understanding Poverty . Cambridge, MA: Harvard University Press.

De Jong, C.H. 2018. “Privatisation in Practice: A Study from an Insurance Perspective into the Effects of Privatization of Public Sickness and Disability Programs in the Neth- erlands, Germany and Canada.” PhD thesis, University of Amsterdam, Faculteit Economie en Bedrijfskunde, Amsterdam.

Eissa, N., and H.W. Hoynes. 2004. “Taxes and the Labor Mar- ket Participation of Married Couples: The Earned Income Tax Credit.” Journal of Public Economics 88(9–10):1931–58. https://doi.org/10.1016/j.jpubeco.2003.09.005.

Eissa, N., and J.B. Liebman. 1996. “Labor Supply Response to the Earned Income Tax Credit.” Quarterly Journal of Eco- nomics 111(2):605–37. https://doi.org/10.2307/2946689.

Employment and Social Development Canada (ESDC). 2018. “Canada’s First Poverty Reduction Strategy.” At https:// www.canada.ca/en/employment-social-development/ programs/poverty-reduction/reports/strategy.html .

Finance Canada. 2016. “Report on Federal Tax Expenditures: Concepts, Estimates and Evaluations.” At https://www. fi n.gc.ca/taxexp-depfi sc/2016/taxexp-depfi sc16-eng.pdf.

Forget, E. 2011. “The Town with No Poverty: The Health Effects of a Canadian Guaranteed Annual Income Field

Experiment.” Canadian Public Policy/Analyse de politiques 37(3):283–305. https://doi.org/10.1353/cpp.2011.0036.

Hoynes, H. 2014. “Building on the Success of the Earned Income Tax Credit.” At https://www.brookings.edu/ research/building-on-the-success-of-the-earned-income- tax-credit/ .

Hoynes, H., and J. Rothstein. 2019. Universal Basic Income in the US and Advanced Countries . NBER Working Paper No. w25538. Cambridge, MA: National Bureau of Economic Research.

Jones, L.E., and K. Michelmore. 2018. “The Impact of the Earned Income Tax Credit on Household Finances.” Jour- nal of Policy Analysis and Management 37(3):521–45. https:// doi.org/10.1002/pam.22062

Kaplan , G., and S. Schulhofer-Wohl. 2018. “The Changing (Dis-)Utility of Work.” Journal of Economic Perspectives 32(3):239–58. https://doi.org/10.1257/jep.32.3.239.

Kesselman, J.R. 2018. “Can ‘Self-Financing’ Redeem the Basic Income Guarantee? Disincentives, Effi ciency Cost, Tax Bur- dens, and Attitudes.” Canadian Public Policy/Analyse de poli- tiques 44(4):423–37. https://doi.org/10.3138/cpp.2017-064.

Kesselman, J.R. 2019. “Policy Options for Retargeting the Canada Child Benefi t.” Canadian Public Policy/Analyse de politiques 45(3):308–28. https://doi.org/cpp.2019-003.

Koebel, K., and T. Schirle. 2016. “The Differential Impact of Universal Child Benefi ts on the Labour Supply of Married and Single Mothers.” Canadian Public Policy/Analyse de poli- tiques 42(1):49–64. https://doi.org/10.3138/cpp.2015-049.

Macdonald, D. 2016. “Out of the Shadows: Shining a Light on Canada’s Unequal Distribution of Federal Tax Expendi- tures.” Ottawa: Canadian Centre for Policy Alternatives. At https://www.policyalternatives.ca/sites/default/ fi les/uploads/publications/National%20Offi ce/2016/11/ Out_of_the_Shadows.pdf.

Mah, B. 2017. “Working Income Tax Benefi t.” Toronto: Maytree. At https://maytree.com/wp-content/uploads/ Working-Income-Tax-Benefi t.pdf .

Marinescu, I. 2017. “No Strings Attached: The Behavioural Ef- fects of U.S. Unconditional Cash Transfer Programs.” New York: Roosevelt Institute. At http://www.marinescu.eu/ Marinescu_UBI_review_2017.pdf.

Marr, C., E. Horton, and B. Duke. 2017. “Brown-Khanna Proposal to Expand EITC Would Raise Incomes of 47 Mil- lion Working Households.” At https://www.cbpp.org/ research/federal-tax/brown-khanna-proposal-to-expand- eitc-would-raise-incomes-of-47-million-working .

Meyer, B. 2010. “The Effects of the Earned Income Tax Credit and Recent Reforms.” In Tax Policy and the Economy , ed. J. Brown, Vol. 24, 153–80. Chicago: University of Chicago Press.

Moyser, M., and A. Burlock. 2018. “Time Use: Total Work Burden, Unpaid Work, and Leisure.” Cat. No. 89-503-X. Ottawa: Statistics Canada.

Musgrave, R. 1971. “Economics of Fiscal Federalism.” Ne- braska Journal of Economics and Business 10(4):3–13.

Myles, J. 2000. “The Maturation of Canada’s Retirement Income System: Income Levels, Income Inequality and Low Income among Older Persons.” Canadian Jour- nal on Aging 19(3):287–316. https://doi.org/10.1017/ s0714980800015014.

298 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Native Women’s Association of Canada. 2017. “Poverty Reduction Strategy: The Native Women’s Association of Canada Engagement Results.” Akwesasne, ON: Native Womenès Association of Canada. At https://www.nwac. ca/wp-content/uploads/2018/08/Poverty-Reduction- Strategy-Revised-Aug23.pdf .

Oates, W. 1968. “The Theory of Public Finance in a Federal System.” Canadian Journal of Economics/Revue canadienne d’économique 1(1):37–54. https://doi.org/10.2307/133460.

Organisation for Economic Co-operation and Development. n.d. “What Are Equivalence Scales?” At http://www.oecd. org/eco/growth/OECD-Note-EquivalenceScales.pdf .

Pohler, D., and K. Koebel. 2018. “A Critique of Basic Income Pilots Following the Ontario Government’s Announced Cancellation.” At https://www.dionnepohler.com/ blog/a-critique-of-basic-income-pilots-following-the- ontario-government-s-announced-cancellation .

Porter, G. 2010. “Work Ethic and Ethical Work: Distortions in the American Dream.” Journal of Business Ethics 96(4):535– 50. https://doi.org/10.1007/s10551-010-0481-6.

Rossiter, E. 2018. “Reciprocity, Capabilities, and Basic In- come.” Working Paper, Douglas College, Vancouver.

Schirle, T. 2017. “Economic Security of Women in Canada.” Presentation to the House of Commons Standing Com- mittee on the Status of Women. Ottawa, 9 February. At http://www.tammyschirle.org/researchdocs/FEWO- 2017-Economic-Security-of-Women-in-Canada.pdf .

Smart, M., and M. Stabile. 2006. “Tax Support for the Disabled in Canada: Economic Principles and Options for Reform.” Canadian Tax Journal 54(2):407–25.

Statistics Canada. 2015. “Table 3: Severity of Disability, by Sex, Aged 15 Years or Older with Disabilities, Canada, 2012.” At https://www150.statcan.gc.ca/n1/pub/89- 654-x/2015001/tbl/tbl03-eng.htm .

Stevens, H., and W. Simpson. 2017. “Toward a National Uni- versal Guaranteed Basic Income.” Canadian Public Policy/ Analyse de politiques 43(2):120–39. https://doi.org/10.3138/ cpp.2016-042.

Stevens, H., and W. Simpson. 2018. “Is Canada Ready for Real Poverty Reduction through a Universal Guaranteed Basic Income? A Rejoinder to Kesselman’s ‘Can “Self-Financing” Redeem the Basic Income Guarantee? Disincentives, Ef- fi ciency Costs, Tax Burdens, and Attitudes’.” Canadian Public Policy/Analyse de politiques 44(4):438–46. https://doi. org/10.3138/cpp.2018-043.

Tedds, L. 2017. “Implementing a Basic Income Guaran- tee through the Personal Income Tax System: Benefi ts, Barriers, and Bothers.” Research Paper 25. Thunder Bay, ON: Northern Policy Institute. At https://www. northernpolicy.ca/upload/documents/publications/ reports-new/tedds_big-and-pit-en.pdf .

Torjman, S., and K. Battle. 1993. “Breaking Down the Wel- fare Wall.” Ottawa: Caledon Institute of Social Policy. At https://maytree.com/wp-content/uploads/488ENG.pdf .

Tweddle, A., and H. Aldridge. 2018. “Welfare in Canada, 2017.” Toronto: Maytree. At https://maytree.com/ welfare-in-canada/

Van Parijs, P., and Y. Vanderborght. 2017. Basic Income: A Radical Proposal for a Free Society and a Sane Economy . Cam- bridge, MA: Harvard University Press.

Wolfson, M. 2018. “How a Guaranteed Income Could Work.” Policy Options . At http://policyoptions.irpp.org/magazines/ july-2018/how-a-guaranteed-income-could-work/ .

Young, M., and J. Mulvale. 2009. “Possibilities and Prospects: The Debate over a Guaranteed Income.” Ottawa: Cana- dian Centre for Policy Alternatives. At https://mspace. lib.umanitoba.ca/bitstream/handle/1993/32344/CCPA_ Guaranteed_Income_Nov_2009.pdf?sequence=1 .

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 299

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

The income threshold at which individuals obtain the maximum benefit transfer (i.e., the beginning of the plateau range) is calculated by determining the point at which after-tax employment earnings, together with the maximum benefi t transfer, place the individual at the set poverty line (i.e., the relevant provincial market basket measure [MBM] of poverty).

In other words, the maximum benefi t transfer an in- dividual can receive, T M , is a function of the individual’s employment earnings at the entry of the maximum trans- fer (i.e., beginning of plateau range), I eM ; the federal and provincial taxes the individual pays on those employment earnings, I eM *τ; and the relevant MBM for the individual, I B . This relationship is expressed in Equation (A.1).

TM = IB − (1 − τ)I e M ( A.1 )

Equation (A.2) provides additional information required to calculate I eM in our program. Here, the phase-in rate, expressed by the sum of the federal and provincial tax rates, τ, and the employment earnings subsidy rate, φw , is a function of the maximum benefi t transfer, TM ; the max- imum benefi t transfer at which individual net income is $0, T O ; and the employment earnings that represent the beginning of the plateau range, I eM .

w M M e

T T I

0 . ( A.2 )

Substituting (A.1) into (A.2), and rearranging and solving for I eM , we obtain

(τ + φw )I e M = IB − (1 − τ)I

e M − T0

(τ + φw )I e M + (1 − τ)I

e M = IB − T0

(τ + φw + 1 − τ)I e M = IB − T0

(1 + φw )I e M = IB − T0

I I TM e B

w

0

1( ) ( A.3 )

That is, an individual’s employment earnings at the entry of the maximum benefi t is equivalent to the maximum transfer at which income is $0 subtracted from the MBM, divided by one plus the employment earnings subsidy rate.

Credit Federal NF PE NS NB QC ON MB SK AB BC

Basic personal amount X X X X X X a X X X X X Married amount X X X X X X X X X X Married equivalent amount X X X X X X X X X X Infi rm dependent X X X X X X X X Tuition tax credit X X X X X X X Education amount X X X X X X Interest on student loans credit X X X X X X X X X X Caregiver tax credit X X X X X X X X X X Home buyers’ amount X X Canada Employment Credit X GST/HST credit X Working Income Tax Benefi t X Working Income Tax Benefi t Supplement for Disabilities

X

Note: GST = Goods and Services Tax; HST = Harmonized Sales Tax.

a Because of the high Basic Personal Amount in Quebec, it is reduced by half rather than eliminated completely.

Source: Authors.

Appendix B: Federal and Provincial Tax Credits Removed (Refundable and Non-Refundable)

Appendix A: Derivation of Formula for Calculating Income Threshold for Receipt of Maximum Benefi t

300 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Appendix C: Federal and Provincial Revenues and Surpluses and Defi cits

Table C.1: Federal Revenues

Federal Programs $Billions

Basic personal amount 39.726 Married amount 2.215 Married equivalent amount 1.022 Canada Employment Credit 3.198 Home buyers’ amount 0.127 Caregiver Tax Credit 0.239 Interest on Student Loans Tax Credit 0.044 Tuition Tax Credit 1.074 Federal GST Credit 4.618 Working Income Tax Benefi t 1.407 Working Income Tax Benefi t Supplement for Disabilities 0.028 Total 53.698 Federal surplus 3.919

Note: GST = Goods and Services Tax.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0).

Table C.2: Provincial Revenues

Province NRTCs SA HST/PST Total Surplus (Defi cit)

NF 0.263 0.225 0.000 0.488 (0.116) PE 0.098 0.023 0.007 0.128 (0.033) NS 0.556 0.282 0.005 0.843 (0.113) NB 0.503 0.227 0.090 0.820 0.131 QC 6.113 3.465 0.003 9.581 1.195 ON 4.865 7.378 1.310 13.553 (3.827) MB 0.909 0.591 0.000 1.500 0.112 SK 1.246 0.542 0.000 1.788 0.739 AB 5.850 1.442 0.000 7.292 2.310 BC 1.599 1.703 0.000 3.302 (1.830) Provincial total 22.002 15.878 1.415 39.295 (1.434)

Note: All values are in $Billions. HST = Harmonized Sales Tax; NRTCs = non-refundable tax credit; PST = Provincial Sales tax; SA = social assistance.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0).

Appendix D: Additional Administrative and Implementation Considerations

Administration In principle, the determination of eligibility and the bene- fi t transfer amount in our program could be administered by the Canada Revenue Agency on top of the current income tax system by changing the Canada Workers Benefi t (CWB) parameters. The proposed funding model for our program, which eliminates many federal and provincial non-refundable tax credits and social assist- ance, requires more substantial tax and social reform, particularly if not all provinces participated, and certain elements of the tax harmonization agreements would need to be revisited.

Applications and Claims Following recently announced changes in the future administration of the CWB, no application should be required for participation in the program (eligibility can be determined on the basis of the information collected on the personal income tax return). A major concern about consolidating the income security system into one program administered through the personal tax system is its lack of responsiveness to within-year fl uctuations in income ( Tedds 2017 ). The program should be structured to allow individuals to apply for advanced, regular pay- ments of the benefi t, as is currently allowed under the CWB. Alternatively, the current Employment Insurance (EI) framework that determines eligibility and payments for individuals experiencing temporary income instability could be used. If EI is retained, all individuals who experi- ence a short-term loss in employment income could apply for EI. If they are not eligible to receive EI, they would automatically be considered for advance payments from the Guaranteed Basic Income program.

Access Other concerns raised about social transfers and benefi ts administered through the personal income tax system include the observation that some individuals do not have a consistent address or may not regularly fi le taxes. Specifi c concerns have also been raised about on-reserve First Nations individuals’ lack of access to benefi ts that are administered through the tax system ( AFOA 2018 ). Given that our proposal does not eliminate any admin- istrative funds associated with current social programs, this funding could be reallocated away from screening people into (and out of) welfare and toward helping those who currently struggle to access their benefi ts to be able to better do so. For instance, government case workers could provide help with tax fi ling, a location to pick up benefi ts cheques, assistance with fi nding jobs, and so forth.

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 301

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Non-Compliance Another implementation challenge is how to avoid adjust- ments to reported self-employment earnings to maximize the benefi t, an issue with the Earned Income Tax Credit in the United States ( Chetty, Friedman, and Saez 2013 ). We note that there are also compliance issues in the current income security system, and these should be reduced and more easily monitored by the implementation of a con- solidated program with consistently applied parameters across all individuals. Also, with the elimination of the basic fi ler deduction in our proposal, individuals who attempt to maximize the total benefi t received by misre- porting their income would pay taxes on those reported earnings. Another benefi t of our design is that lower income individuals now have an incentive to report any previously unreported earnings.

General Equilibrium Effects If fi rms and landlords could capture the earnings subsidy by paying lower wages or increasing rent, respectively, the impact of our program on the targeted individuals may be attenuated. Our program may be more effective with complementary strategies around access to afford- able housing and in-kind services for the lowest income Canadians. We assume current provincial minimum wages would not be lowered; also, the earnings subsidy may generate a greater supply of labour willing to work at the current maximum wage that fi rms are prepared to offer, which is benefi cial for both fi rms and workers.

Appendix E: Marginal Effective Tax Rate Implications of Proposed Program In this appendix, we compare the marginal effective tax rates (METRs) before and after the introduction of our program. The METR calculations provide approximations for single, able-bodied, working-age individuals with no children. Calculations assume that an individual only re- ceives the federal and provincial basic personal amounts to reduce taxes owed; calculations do not include reduction rates on the following non-refundable tax credits: mar- ried amount, married equivalent amount, and caregiver amount. Calculations also assume that eligible individuals claim and receive the Working Income Tax Benefi t and Goods and Services Tax credits. Table E.1 summarizes the 2017 direct income assistance program information across the provinces, and the information on federal and prov- incial tax rates used in Tables E.2 and E.3 were compiled from 2017 tax schedules. Because our program design varies within provinces, Table E.3 refl ects only one market basket measure region for each province. We chose either the region with a population between 100,000 and 499,000 or an otherwise arbitrarily chosen city in each province. The choice of the region affects the income ranges at which changes in the METR schedule occur.

Table E.1 : Annual Income Assistance, Claw-Back Rates, and Exit Level by Province, 2017

Province SA Income ($) Reduction Rate, %

(Earnings Exemption; $) Exit Level ($) a

AB 8,027 75 (2,760) 13,462.67 BC 8,124 100 (2,400) 10,524 MB 9,494 70 (2,400) 15,962.86 NB 7,122 70 (1,800) 11,974.29 NL 11,379 80 (900) 15,123.75 NS 7,433 70 (1,800) 12,418.57 ON 9,461 50 (2,400) 21,322 PEI 7,900 90 (900) 9,677.78 QC 9,083 100 (2,400) 11,483 SK 8,820 100 8,820

Note: SA = social assistance.

a The following equation is used to calculate the exit level: (SA Income / Reduction Rate) + Earnings Exemption.

Source: Tweddle and Aldridge (2018).

302 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Table E.2 : Calculation of Pre-GBI METRs for Single, Able-Bodied, Working Age Individuals with No Children, 2017

Province and Income Range ($) SA GST Credit WITB Provincial Tax Federal Tax

METR Pre-GBI

SA No SA

AB 0 – 2,760 0 0 0 2,761–8,963 75 −19.7 55.3 −19.7 8,964–11,634 75 0 75 0 11,635–12,402 75 0 15 90 15 12,403–13,462.67 75 14.8 15 104.8 29.8 13,463.67–18,690 14.8 15 29.8 18,691–19,984 14.8 10 15 39.8 19,985–36,975 10 15 25 36,976–45,915 5 10 15 30 45,916–91,831 10 20.5 30.5 BC 0–2,400 0 0 0 2,401–4,750 100 100 0 4,751–10,208 100 −19.4 80.6 −19.4 10,209–10,524 100 −19.4 5.06 85.66 −14.34 10,525–10,935.57 −19.4 5.06 −14.34 10,936.57–11,634 0 5.06 5.06 11,635–12,965 0 5.06 15 20.06 12,966–20,609 15.7 5.06 15 35.76 20,610–36,975 5.06 15 20.06 36,976–38,898 5 5.06 15 25.06 38,899–45,915 5 7.7 15 27.7 45,916–77,797 7.7 20.5 28.2 MB 0–2,400 0 0 0 2,401–3,000 70 70 0 3,001–7,172 70 −25 45 −25 7,173–9,271 70 0 70 0 9,272–11,634 70 0 10.8 80.8 10.8 11,635–11,838 70 0 10.8 15 95.8 25.8 11,839–15,962.86 70 15 10.8 15 110.8 40.8 15,963.86–18,792 15 10.8 15 40.8 18,793–31,465 10.8 15 25.8 31,466–36,975 12.75 15 27.75 36,976–45,915 5 12.75 15 32.75 45,916–68,005 12.75 20.5 33.25 NB 0 – 1,800 0 0 0 1,801–3,000 70 70 0 3,001–7,172 70 −25 45 −25 7,173–9,894 70 0 70 0 9,895–11,634 70 0 9.68 79.68 9.68 11,635–11,838 70 0 9.68 15 94.68 24.68 11,839–11,974.29 70 15 9.68 15 109.68 39.68 11,975.29–18,792 15 9.68 15 39.68 18,793–36,975 . 9.68 15 24.68

(Continued)

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 303

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Province and Income Range ($) SA GST Credit WITB Provincial Tax Federal Tax

METR Pre-GBI

SA No SA

36,976–41,059 5 9.68 15 29.68 41,060–45,915 5 14.82 15 34.82 45,916–82,119 14.82 20.5 35.32 NL 0–900 0 0 0 901–3,000 80 80 0 3,001–7,172 80 −25 55 −25 7,173–8,977 80 0 80 0 8,978–11,634 80 0 8.7 88.7 8.7 11,635–11,838 80 0 8.7 15 103.7 23.7 11,839–15,123.75 80 15 8.7 15 118.7 38.7 15,124.75–18,792 15 8.7 15 38.7 18,793–35,851 8.7 15 23.7 35,852–36,975 14.5 15 29.5 36,976–45,915 5 14.5 15 34.5 45,916–71,701 14.5 20.5 35 NS 0–1,800 0 0 0 1,801–3,000 70 70 0 3,001–7,172 70 −25 45 −25 7,173–8,481 70 0 70 0 8,482–11,634 70 0 8.79 78.79 8.79 11,635–11,838 70 0 8.79 15 93.79 23.79 11,839–12,418.57 70 15 8.79 15 108.79 38.79 12,419.57–18,792 15 8.79 15 38.79 18,793–29,590 8.79 15 23.79 29,591–36,975 14.95 15 29.95 36,976–45,915 5 14.95 15 34.95 45,916–59,180 14.95 20.5 35.45 ON 0–2,400 0 0 0 2,401–3,000 50 50 0 3,001–7,172 50 −25 25 −25 7,173–10,170 50 0 50 0 10,171–11,634 50 0 5.05 55.05 5.05 11,635–11,838 50 0 5.05 15 70.05 20.05 11,839–18,792 50 15 5.05 15 85.05 35.05 18,793–21,322 50 5.05 15 70.05 20.05 21,323–36,975 5.05 15 20.05 36,976–42,201 5 5.05 15 25.05 42,202–45,915 5 9.15 15 29.15 45,916–84,404 9.15 20.5 29.65 PE 0–900 0 0 0 901–3,000 90 90 0 3,001–7,172 90 −25 65 −25 7,173–8,160 90 0 90 0

(Continued)

Table E.2 : Continued

304 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Province and Income Range ($) SA GST Credit WITB Provincial Tax Federal Tax

METR Pre-GBI

SA No SA

8,161–9,677.78 90 0 9.8 99.8 9.8 9,678.78–11,634 0 9.8 9.8 11,635–11,838 0 9.8 15 24.8 11,839–18,792 15 9.8 15 39.8 18,793–31,984 9.8 15 24.8 31,985–36,975 13.8 15 28.8 36,976–45,915 5 13.8 15 33.8 45,916–63,969 13.8 20.5 34.3 QC 0–2,400 0 0 0 2,401–10,506 100 −20.5 79.5 −20.5 10,507–11,483 100 0 100 0 11,484–11,634 0 0 11,635–11,816.44 0 15 15 11,817.44–14,890 20 15 35 14,891–20.125.09 20 15 15 50 20.126.09–36,975 15 15 30 36,976–42,705 5 15 15 35 42,706–45,915 5 20 15 40 45,916–85,405 20 20.5 40.5 SK 0–3,000 100 100 0 3,001–7,172 100 −25 75 −25 7,173–8,820 100 0 100 0 8,821–11,634 0 0 11,635–11,838 0 15 15 11,839–16,604 15 15 30 16,605–18,792 15 10.75 15 40.75 18,793–36,975 10.75 15 25.75 36,976–42,225 5 10.75 15 30.75 42,226–45,915 5 12.75 15 32.75 45,916–91,831 12.75 20.5 33.25

Note: Values in table body are percentages. Empty cells indicate the absence of a value. GBI = Guaranteed Basic Income; METR = marginal effective tax rates; SA = social assistance; WITB = Working Income Tax Benefi t.

Source: Calculations by authors.

Table E.2 : Continued

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 305

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Table E.3 : Comparison of METRs Pre-GBI and Post-GBI, 2017

Province and Income Range ($) GBI Provincial Tax Federal Tax

METR Pre-GBI (from Table E.2)

METR Post-GBI SA No SA

AB, population 100,000–499,000 0–2,760 −55 10 15 0 0 −30 2,761–8,963 −55 10 15 55.3 −19.7 −30 8,964–9,313 −55 10 15 75 0 −30 9,314–11,314 0 10 15 75 0 25 11,315–11,634 50 10 15 75 0 75 11,635–12,402 50 10 15 90 15 75 12,403–13,462.67 50 10 15 104.8 29.8 75 13,463.67–18,690 50 10 15 39.8 75 18,691–19,984 50 10 15 39.8 75 19,985–36,975 50 10 15 25 75 36,976–37,612 50 10 15 30 75 37,613–45,915 10 15 30% 25% 45,916–91,831 10 20.5 30.5% 30.5% BC, population 100,000–499,000 0–2,400 −50.06 5.06 15 0 0 −30 2,401–4,750 −50.06 5.06 15 100 0 −30 4,751–8,715 −50.06 5.06 15 80.6 −19.4 −30 8,716–10,208 0 5.06 15 80.6 −19.4 20.06 10,209–10,524 0 5.06 15 85.66 −14.34 20.06 10,525–10,716 0 5.06 15 −14.34 20.06 10,717–10,935.57 50 5.06 15 −14.34 70.06 10,936.57–11,634 50 5.06 15 5.06 70.06 11,635–12,965 50 5.06 15 20.06 70.06 12,966–20,609 50 5.06 15 35.76 70.06 20,610–35,690 50 5.06 15 20.06 70.06 35,691–36,975 5.06 15 20.06 20.06 36,976–38,898 5.06 15 25.06 20.06 38,899–45,915 7.7 15 29.15 22.7 45,916–77,797 7.7 20.5 29.65 28.2 MB, Winnipeg 0–2,400 −55.8 10.8 15 0 0 −30 2,401–3,000 −55.8 10.8 15 70 0 −30 3,001–6,751 −55.8 10.8 15 45 −25 −30 6,752–8,752 0 10.8 15 45 −25 25.8 8,752–9,271 50 10.8 15 70 0 75.8 9,272–11,634 50 10.8 15 80.8 10.8 75.8 11,635–11,838 50 10.8 15 95.8 25.8 75.8 11,839–15,962.86 50 10.8 15 110.8 40.8 75.8 15,963.86–18,792 50 10.8 15 40.8 75.8 18,793–31,465 50 10.8 15 25.8 75.8 31,466–35,275 50 12.75 15 27.75 77.75 35,276–36,975 12.75 15 27.75 27.75 36,976–45,915 12.75 15 32.75 27.75 45,916–68,005 12.75 20.5 33.25 33.25

(Continued)

306 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Province and Income Range ($) GBI Provincial Tax Federal Tax

METR Pre-GBI (from Table E.2)

METR Post-GBI SA No SA

NB, St. John and Moncton 0–1,800 −54.68 9.68 15 0 0 −30 1,801–3,000 −54.68 9.68 15 70 0 −30 3,001–7,172 −54.68 9.68 15 45 −25 −30 7,173–8,578 −54.68 9.68 15 70 0 −30 8,579–9,894 0 9.68 15 70 0 24.68 9,895–10,579 0 9.68 15 79.68 9.68 24.68 10,580–11,634 50 9.68 15 79.68 9.68 74.68 11,635–11,838 50 9.68 15 94.68 24.68 74.68 11,839–11,974.29 50 9.68 15 109.68 39.68 74.68 11,975.29–18,792 50 9.68 15 39.68 74.68 18,793–34,204 50 9.68 15 24.68 74.68 34,205–36,975 9.68 15 24.68 24.68 36,976–41,059 9.68 15 29.68 24.68 41,060–45,915 14.82 15 34.82 29.82 45,916–82,119 14.82 20.5 35.32 35.32 NL, St. John's 0–900 −53.7 8.7 15 0 0 −30 901–3,000 −53.7 8.7 15 80 0 −30 3,001–5,872 −53.7 8.7 15 55 −25 −30 5,873–7,172 0 8.7 15 55 −25 23.7 7,173–7,873 0 8.7 15 80 0 23.7 7,874–8,977 50 8.7 15 80 0 73.7 8,978–11,634 50 8.7 15 88.7 8.7 73.7 11,635–11,838 50 8.7 15 103.7 23.7 73.7 11,839–15,123.75 50 8.7 15 118.7 38.7 73.7 15,124.75–18,792 50 8.7 15 38.7 73.7 18,793–35,851 50 8.7 15 23.7 73.7 35,852–36,938 50 14.5 15 29.5 79.5 36,939–36,975 14.5 15 29.5 29.5 36,976–38,005 14.5 15 34.5 29.5 38,006–45,915 14.5 15 34.5 29.5 45,916–71,701 14.5 20.5 35 35 NS, Halifax and Cape Breton 0–1,800 −53.79 8.79 15 0 0 −30 1,801–3,000 −53.79 8.79 15 70 0 −30 3,001–7,172 −53.79 8.79 15 45 −25 −30 7,173–8,481 −53.79 8.79 15 70 0 −30 8,482–8,811 0 8.79 15 70 0 23.79 8,812–10,812 0 8.79 15 78.79 8.79 23.79 10,813–11,634 50 8.79 15 78.79 8.79 73.79 11,635–11,838 50 8.79 15 93.79 23.79 73.79 11,839–12,418.57 50 8.79 15 108.79 38.79 73.79 12,419.57–18,792 50 8.79 15 38.79 73.79 18,793–29,590 50 8.79 15 23.79 73.79 29,591–35,158 50 14.95 15 29.95 79.95

Table E.3 : Continued

(Continued)

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 307

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Table E.3 : Continued

(Continued)

Province and Income Range ($) GBI Provincial Tax Federal Tax

METR Pre-GBI (from Table E.2)

METR Post-GBI SA No SA

35,159–35,989 14.95 15 29.95 29.95 35,990–36,975 14.95 15 29.95 29.95 36,976–45,915 14.95 15 34.95 29.95 45,916–59,180 14.95 20.5 35.45 35.45 ON, population 100,000–499,000 0–2,400 −50.05 5.05 15 0 0 −30 2,401–3,000 −50.05 5.05 15 50 0 −30 3,001–6,903 −50.05 5.05 15 25 −25 −30 6,904–7,172 0 5.05 15 25 −25 20.05 7,173–8,904 0 5.05 15 50 0 20.05 8,905–10,170 50 5.05 15 50 0 70.05 10,171–11,634 50 5.05 15 55.05 5.05 70.05 11,635–11,838 50 5.05 15 70.05 20.05 70.05 11,839–18,792 50 5.05 15 85.05 35.05 70.05 18,793–19,122 50 5.05 15 70.05 20.05 70.05 19,123–21,322 50 5.05 15 70.05 20.05 70.05 21,323–34,736 50 5.05 15 20.05 70.05 34,737–36,975 5.05 15 20.05 20.05 36,976–42,201 5.05 15 25.05 20.05 42,202–45,915 9.15 15 29.15 24.15 45,916–84,404 9.15 20.5 29.65 29.65 PEI 0–900 −54.8 9.8 15 0 0 −30 901–3,000 −54.8 9.8 15 90 0 −30 3,001–7,172 −54.8 9.8 15 65 −25 −30 7,173–8,160 −54.8 9.8 15 90 0 −30 8,161–9,082 −54.8 9.8 15 99.8 9.8 −30 9,083–9,677.78 0 9.8 15 99.8 9.8 24.8 9,678.78–11,083 0 9.8 15 9.8 24.8 11,084–11,634 50 9.8 15 9.8 74.8 11,635–11,838 50 9.8 15 24.8 74.8 11,839–18,792 50 9.8 15 39.8 74.8 18,793–31,984 50 9.8 15 24.8 74.8 31,985–36,837 50 13.8 15 28.8 78.8 36,838–36,975 13.8 15 28.8 28.8 36,976–45,915 13.8 15 33.8 28.8 45,916–63,969 13.8 20.5 34.3 34.3 QC, population 100,000–499,000 0–2,400 −45 15 0 0 −30 2,401–5,977 −45 15 79.5 −20.5 −30 5,978–6,187 0 15 79.5 −20.5 15 6,188–7,506 0 15 79.5 −20.5 15 7,507–7,978 0 15 15 79.5 −20.5 30 7,979–8,188 50 15 15 79.5 −20.5 80 8,189–10,506 50 15 15 79.5 −20.5 80 10,507–11,483 50 15 15 100 0 80

308 Koebel and Pohler

© Canadian Public Policy / Analyse de politiques, September / septembre 2019 doi:10.3138/cpp.2019-016

Province and Income Range ($) GBI Provincial Tax Federal Tax

METR Pre-GBI (from Table E.2)

METR Post-GBI SA No SA

11,484–11,634 50 15 15 0 80 11,635–11,816.44 50 15 15 15 80 11,817.44–14,890 50 15 15 35 80 14,891–20.125.09 50 15 15 50 80 20.126.09–31,523 50 15 15 30 80 31,524–36,975 15 15 30 30 36,976–42,705 15 15 35 30 42,706–45,915 20 15 40 35 45,916–85,405 20 20.5 40.5 40.5 SK, Saskatoon 0–3,000 −55.75 10.75 15 100 0 −30 3,001–7,172 −55.75 10.75 15 75 −25 −30 7,173–7,872 −55.75 10.75 15 100 0 −30 7,873–8,820 0 10.75 15 100 0 25.75 8,821–9,873 0 10.75 15 0 25.75 9,874–11,634 50 10.75 15 0 75.75 11,635–11,838 50 10.75 15 15 75.75 11,839–16,604 50 10.75 15 30 75.75 16,605–18,792 50 10.75 15 40.75 75.75 18,793–36,291 50 10.75 15 25.75 75.75 36,292–36,975 10.75 15 25.75 25.75 36,976–42,225 10.75 15 30.75 25.75 42,226–45,915 12.75 15 32.75 27.75 45,916–91,831 12.75 20.5 33.25 33.25

Note: Values in table body are percentages. Empty cells indicate the absence of a value. GBI = Guaranteed Basic Income; METR = marginal effective tax rates; SA = social assistance; WITB = Working Income Tax Benefi t.

Source: Calculations by authors.

Table E.3 : Continued

Expanding the Canada Workers Benefi t to Design a Guaranteed Basic Income 309

doi:10.3138/cpp.2019-016 © Canadian Public Policy / Analyse de politiques, September / septembre 2019

Appendix F: Option that Eliminates Employment Insurance, Reduces the Phase- Out Rate, and Improves the Fiscal Position of the Provinces In this appendix, we illustrate some alternative ideas by modeling a slightly more costly option that eliminates Employment Insurance (EI) benefi ts and individual EI con- tributions, reduces the phase-out rate in our program from 50 percent to 40 percent, and improves the fi scal position of the provinces. The motivation for considering the elimina- tion of EI benefi ts is that our program will provide income to the unemployed, albeit at lower levels for some individuals, but with three important advantages: (1) the benefi ts under our program are not terminal, (2) the benefi ts under our program are universally accessible to working-age people, and (3) qualifi cation for our program is more transparent and removes the complex formulas and criteria associated with determining EI benefi t (in)eligibility. EI is only avail- able to those who have had recent access to a job and have worked enough to be eligible for benefi ts. Although these slightly regressive aspects of EI are possibly justifi able because it is structured as an insurance program and not a redistribution program, it currently redistributes from some types of workers and toward others in a way more similar to a social program than a labour insurance program. For instance, EI not only provides insurance benefi ts for short- term job loss, but also provides income-replacement benefi ts during parental leaves. If all employers and workers are col- lectively subsidizing parental leaves, from a redistribution perspective, it is unclear why higher and middle-income parents (many of whom also benefi t from top-ups through their employer benefi ts) should receive more from a public insurance program than lower income parents.

In this option, we retain the same program parameters, with the exception of the phase-out rate, which is reduced

from 50 percent to 40 percent, further reducing marginal effective tax rates (METRs) under our program and in- creasing the number of households that receive some benefi t. These changes to the phase-out rate increase the gross cost in this option to $100.91 billion. We eliminate the provision of EI benefi ts to individuals, as well as all employee premium payments (another METR reduction) and the associated non-refundable tax credits for those premiums. We retain all the previously proposed funding sources and incorporate the additional funding from the employer EI contributions to fund our program, although the collection of this revenue could be removed from pay- roll taxes and instead be fi nanced from other taxes such as the corporation tax. The latter approach has the additional benefi t of removing hiring disincentives.

If we use the additional revenue from the current em- ployer portion of the EI contributions (~$8 billion) as well as removing the EI insurance premium tax credits feder- ally and provincially (~$2 billion), we can increase the federal portion of funding the GBI program cost from 55 percent to 62 percent, which improves the fi scal position of the provinces. In Ontario, the defi cit is reduced by almost $1.5 billion, and overall now there is a small surplus at the provincial level (as opposed to a small defi cit).

Table F.1 shows the distributional impacts of these changes. The household tax implications of the redis- tribution from those with higher incomes to those with lower incomes substantially improves when compared with the option proposed in the main article. In this op- tion that eliminates EI, average disposable income now increases for households in the bottom fi ve deciles (rather than the bottom four), and the average tax implications for households in deciles six and seven are quite small. The major tax burden falls to those households in the top three income deciles.

Table F.1 : Impact on Household Disposable Income by Household Net Income Decile, Age 18–64 Years

Decile ($2018) Average GBI/CWB ($)

Average Household Disposable Income ($) Average Change in Disposable Income

Pre-GBI Post-GBI $ %

0–4,147 9,805 9,067 12,900 3,833 42.27 4,148–12,504 12,020 12,862 21,114 8,252 64.16 12,505–20,737 11,039 19,410 26,102 6,692 34.48 20,738–30,893 8,637 27,203 30,876 3,673 13.50 30,894–42,303 5,254 35,856 36,200 344 0.96 42,304–55,154 3,996 44,213 43,745 –468 –1.06 55,155–72,434 4,592 55,522 55,281 –241 –0.43 72,435–94,689 3,680 70,397 69,013 –1,384 –1.97 94,690–131,915 2,619 92,468 89,733 –2,735 –2.96 131,916+ 1,563 170,315 166,767 –3,548 –2.08 Aggregate 6,135 56,513 57,729 1,216 2.15

Note: CWB = Canada Workers Benefi t; GBI = Guaranteed Basic Income.

Source: Calculations by authors using Statistics Canada’s Social Policy Simulation Database and Model (Version 27.0)