Article body

Introduction

In the 2018 federal budget, the Government of Canada announced major changes to a tax benefit for low wage workers, including changes that would increase benefits, increase eligibility and make enrolment automatic.[1] These, and similar changes to other public benefits, will have important effects on take-up rates of national social programs. Those take-up rates were also highlighted when the Prime Minister made increasing them a priority in his mandate letter to the Minister of National Revenue. More precisely, the Prime Minister instructed the Minister of National Revenue to “[e]nsure that CRA is a client-focused agency that will [...] proactively contact Canadians who are entitled to, but are not receiving, tax benefits.”[2]

Social programs aim to deliver cash transfers or services to those who are eligible for the offered benefits. To the extent that those eligible do not receive benefits, however, that goal is not realized. We demonstrate here that, while the agencies administering social programs make efforts to serve all those who are eligible, the intensity and effectiveness of their efforts vary considerably.

Low participation rates may make it more difficult for the social program to achieve its goals.[3] But for at least two reasons, governments might not want 100% participation in some kinds of programs. First, some of those eligible may not want to participate. For example, some might feel that getting benefits from the government compromises their independence. Others might feel that the available benefits do not justify the costs of applying for them. Second, 100% participation could greatly increase the current cost of some programs.

This second consideration affects the behaviour of government program administrators. Canadian government departments and agencies must make annual requests to Parliament for funding to administer social programs, even for those programs that are statutory and not at the discretion of the Crown.[4] This means that governments must make certain assumptions and projections about participation and the likely cost of cash benefits. In Canada’s expenditure management system, requesting more money than will be spent is treated as wasteful, just as requiring additional funding to cover unexpected costs is treated as a failure to plan.[5] Both overspending and underspending may attract unwanted attention from the opposition or the media. At an individual level, the performance pay of public service executives may be tied to managing an annual program budget to within a small range above or below funding allocated by Parliament, discouraging them from taking actions that might substantially increase the cost of the programs under their management.[6] Both at a micro level and at a whole of government level, there seem to be real disincentives to increasing participation in social programs if it results in program spending above historic trends.

Even if governments aim for 100% participation by those who are eligible, several kinds of barriers stand in the way. One kind of barrier is that many, if not most, programs require those eligible to apply for the benefits. If applications are complicated, if awareness of the program is low, or if the applicant must take intermediate steps as part of the application, some of those eligible will likely not apply.

In this paper we present case studies of three social programs in which the barriers to 100% participation take different forms. Each of the programs is a cash transfer of some kind, aimed at lower or modest income Canadians, and each is targeted to a different age group. In all three cases, as we describe in the later Parts of this paper, the federal government has no positive legal obligation to ensure access to eligible Canadians but has made some efforts to increase participation. Taken together, these programs offer an illustrative, though not exhaustive, set of examples of social programs in which voluntary non-participation is a challenge for policymakers. Brief introductions to these programs are as follows:

  • The Working Income Tax Benefit (WITB) is administered by the CRA and requires that potential beneficiaries file an income tax return. As part of that return, they must fill out a form specific to the WITB. That intermediate step—filing a proper tax return—seems to prevent a non-trivial number of potential participants from receiving the benefit.

  • The Canada Learning Bond (CLB) is essentially “free money” available to the parents or guardians of young children. But participation is low, partly because the program is not well-known and partly because those eligible must set up a Registered Education Savings Plan (RESP) account in order to receive the money.

  • Old Age Security (OAS) is Canada’s universal old age pension. Associated with OAS is a supplement for those with low income known as the Guaranteed Income Supplement (GIS). In these areas the government has partially implemented one of the most often mentioned solutions to high rates of eligible non-participation—automatic enrolment.

We define eligible non-participation as the share of the eligible target population that is not enrolled in a social program that would otherwise provide a financial benefit or social service. Non-participation, also referred to as under-subscription or as a take-up gap, may be voluntary (in which those eligible actively decide not to apply for the program) or involuntary (in which those eligible have not actively made such a decision and may not even be aware of the program’s existence).

We present descriptions of these three programs and outline what federal governments have done to increase participation by eligible Canadians. Canada’s taxation department, the CRA, plays a role in all of these programs even though it administers only one of them. We therefore highlight its role. We also take account of the legislation underlying these programs in order to see if there is a legal obligation to serve all those eligible. Finally, we discuss several ways to increase participation, including automatic enrolment and supporting third-party organizations’ efforts to enrol those eligible.

I. The Theory of Eligible Non-Participation

Significant numbers of people who are eligible for social benefits do not take them up. The academic literature seeking to explain eligible non-participation has generally focused on the United States (US) context and began in earnest with studies of non-participation in the federal welfare program once known as Aid to Families with Dependent Children (AFDC). In the early 1980s, Professor Robert Moffitt proposed an economic model of take-up in which potential recipients weigh the costs and benefits of participation and decide to participate only when the benefits exceed the costs.[7] Moffitt then emphasized “stigma” as the primary cost of participating in AFDC, arguing that for some people the cash benefits are not worth bearing the stigma attached to receiving them.[8]

In the early 2000s, Professor Janet Currie reviewed the then-extant literature on take-up.[9] She pointed out that the literature suggests “other costs associated with the take-up of social programs are more important than stigma.”[10] These “other costs” encompass two common explanations for non-participation—the lack of awareness of the benefit and the presence of administrative barriers to participation. That is, the time and energy needed to learn about programs and to overcome the associated administrative barriers are thought of as costs that are weighed against the benefit of program participation. Collecting all such costs under the name “transactions costs,” Currie asserted that the literature proposed three possible reasons for low take-up (high eligible non-participation): stigma, transactions costs and lack of awareness.[11] Of these, Currie argued that transactions costs were the major factor, citing a number of studies that failed to demonstrate that stigma was a decisive factor.[12]

The challenge created by labelling “transactions costs” as the major determinant of eligible non-participation is that almost any feature of a program can be interpreted as a cost. More recent work has not moved much beyond the analysis by Currie. For example, the extensive literature reviews by Finn and Goodship (2014)[13] for the United Kingdom and Daigneault, Jacob and Tereraho (2012)[14] for Canada list “awareness” as a critical factor, along with the complexity and difficulty of application as well as the anticipated benefits.

One new possibility has arisen from the field of behavioral economics. Traditional economic analyses assume that decision makers are fully informed, taking into account all available information. Recent work in behavioral economics has instead focused on inattention, raising the possibility that human decision makers cannot pay attention to all aspects of all decisions that they must make. Fully informed decision makers would not need to be reminded to take beneficial actions but inattentive decision makers might not take those actions and reminders might alter their behaviour. To illustrate, we use the work of Guyton et al., who studied the effect of reminders on tax filing behavior among people who would be eligible for the US Earned Income Tax Credit (EITC) if they filed.

Guyton et al. conducted a randomized control trial using administrative tax data from the US Internal Revenue Service (IRS).[15] The first step was to create a data set consisting of individuals who had not filed taxes in 2011 or 2012 but who were potentially eligible for the EITC based on information in tax forms submitted by other parties, such as employers. Roughly 360,000 of those individuals became the treatment and control groups defined by Guyton et al. In 2014, the treatment group received reminders—a post card or a brochure sent by mail—about filing a tax return both in 2014 and for past years. The control group received nothing. The result was clear: “the outreach increased tax filing for both prior-year tax returns and current-year tax returns” by 0.5% to 1%.[16]

Given this review of the academic literature, it is clear why automatic enrolment appears to be a promising way to address the problem of eligible non-participation. It addresses informational issues (whatever their cause), greatly reduces the costs associated with application, and reduces (though does not eliminate) stigma.

II. The CRA and Tax Filing in Canada

A. Overview of the CRA’s Role

Our focus in this paper is on three federal benefits programs available to low-income Canadians. In Canada, several significant social welfare benefits—including the WITB, the Canada Child Benefit, and the GIS—are targeted at low-income Canadians and are administered through agencies and agreements as set out in the Income Tax Act and other pieces of intersecting legislation.[17] The CRA directly administers only the WITB but plays an important role in OAS/GIS because tax returns are used to determine eligibility for automatic enrolment. It also plays a role in the CLB because it certifies RESPs and does the income test described in Part III.

The central function of the CRA in these programs is to be a major source of information for what amounts to a centralized government database, albeit a fragmented and incomplete database. That information comes from individual tax returns and can be shared across departments as long as CRA’s enabling legislation allows it. Canadians who do not file a tax return may not receive benefits for which they are eligible. Qualitative research conducted for the CRA finds that low-income Canadians are generally aware of the desirability of filing an annual tax return, but may not be aware that many social programs and benefits are conditional on doing so.[18]

The CRA’s mandate limits its ability to share information collected from tax returns with other federal or provincial organizations for the purposes of administering benefit programs. The CRA is restricted to sharing information only, when pertinent, and to a specific list of agencies.[19] Entities on this list are strictly associated with the provincial and federal governments, thus preventing federally collected information from being transferred to private entities no matter the possible increase in administrative efficiency.[20] This translates into consumers being better protected from unscrupulous debt collectors, criminals, and questionable for-profit organizations.[21] The CRA’s involvement in collecting information reflects its historical role as a tax administrator; however, policy objectives, with an emphasis on government efficiency and income testing in government programs, have required CRA to assume a role in administering certain benefits. In theory, blurring the lines between CRA’s role as tax administrator and its role as an information provider encourages reduced constraints on information sharing within government.

There may be a shift underway, increasing pressure on the CRA to rethink its role in reaching eligible beneficiaries of social programs. As already mentioned, increasing the benefit take-up rate was explicitly listed as a priority of newly-elected Prime Minister Justin Trudeau in his 2015 mandate letter to the Minister of National Revenue.[22] Since the mandate letter, public pressure and investigative efforts by the media appear to have played a role in holding the government accountable for its commitment. For example, recent media coverage[23] of the role of the CRA in reaching women in shelters to inform them of the Canada Child Benefit[24] (CCB) led to an investigation by the Taxpayer’s Ombudsman, Sherra Profit.[25]

The shift in the CRA’s role also widens the scope to consider concerns surrounding a positive legal obligation to reach out to eligible beneficiaries. Under the Income Tax Act, benefits are not “automatically” paid by the CRA, even in programs it directly administers. If filing a tax return is a necessary step to access certain benefits and claimants fail to do this, the CRA has no legal obligation to pay. Similarly, if a tax return is filed, but the relevant application for a program is not completed, then CRA has not, historically, been required to assess eligibility and pay benefits. However, specific legislation pertinent to benefits may override this. Currently, only the Old Age Security pension[26] is a case of a CRA-related program where enrolment is automatic for some Canadians.

In the Parts that follow, three key social benefit programs: the WITB, now the CWB,[27] the CLB,[28] and the OAS/GIS pensions,[29] will be considered. The respective legal frameworks under which the programs operate and the legal frameworks for collecting information relevant to eligibility will be outlined. The WITB is directly administered by the CRA. The CRA must provide final approval of the WITB despite several provinces and territories—Alberta, Quebec, British Columbia, and Nunavut—having their own reconfiguration of the benefit.[30] The CLB is administered through the Canada Education Savings Program (CESP) at Employment and Social Development Canada. Only families receiving the Canada Child Benefit, a program administered by CRA, can receive the CLB. Employment and Social Development Canada (ESDC)[31] relies on CRA to verify income and eligibility for the CLB. Because the pensions provided by OAS and GIS are a function of the income of potential recipients, tax filings can be used to determine eligibility. Before delving into the structure of our three programs, however, we highlight the importance of filing taxes and provide an estimate of the proportion of Canadians who do not file.

B. Tax Filing in Canada: What Do We Know?

Canada’s Income Tax Act does not require all individuals in Canada to file an annual return unless they have income tax owing, have certain kinds of income to report or reconcile, or are requested to do so by the Minister.[32] Since 1942, personal income taxes have been withheld at source for individuals in standard employment or receiving taxable benefits. If workers in standard employment do not have additional tax owing, do not meet the other legislated tests, and are not asked by government to file a return, they are not legally required to complete a return.[33] However, the CRA tells Canadians they “should” file annual returns if they believe they may be owed a tax refund or want to claim certain refundable tax credits.[34] This message is also largely consistent with a view that it is up to individual tax filers to initiate an application for a benefit. In practice, even when enrolment into a benefit is automated, the Agency still relies on voluntary tax-returns to assess eligibility. Annual voluntary tax compliance is fundamental to the administration of many income-tested benefits, whether enrolment is discretionary or not.

The CRA reports annually on the number of returns it has received in a given year and, with approximately a two-year delay, reports aggregate tax statistics for individual tax years. These two data points do not necessarily align because some number of returns received in a given year will be for previous tax years. In 2017, the CRA reported that it had received and processed over 28 million tax returns.[35] The agency does not, however, report an official rate of tax filing.

One way to try to determine the overall population of potential tax filers is to use national census data. Later in this paper, we make use of census administrative documents that provide estimates of the number of adults, aged 20 to 64 years, who have not filed an income tax return in the census year. Using this approach, a national filing rate of between 85% and 96%, annually, can be estimated. Such estimates are based on the share of working-age adults recorded in the census for whom a tax record could not be identified by Statistics Canada. This does not mean that these individuals were required by law or would have benefited from filing an annual income tax return. It is, however, the best available approach to estimate the number of Canadians who are not filing a return and whose eligibility for a wide range of income-tested benefits, including those discussed in this paper, cannot be determined using income tax records.

III. Eligible Non-Participation in Three Social Welfare Programs

A. Working Income Tax Benefit[36]

The WITB is a refundable tax credit available for eligible working-age adults with low incomes. The credit is available to eligible taxpayers based on their employment income reported on the previous year’s tax return.[37] In the 2018 federal Budget, the Government of Canada announced significant reforms to the WITB including a change of name, changes to benefit levels and changes to the administration of the tax credit.[38] In this Part, we first discuss the tax credit as it has been operating, and then how the new Canada Workers Benefit (CWB) is expected to be structured. We also review available evidence on take-up rates in the WTIB as it operated prior to the new reforms.

To be eligible for the WITB, an individual must be at least nineteen years old, live in Canada throughout the year, have working income[39] within the eligible range, and may not be a full-time student (unless they are supporting a dependent, such as a child).[40] To qualify, an applicant must also have a family income within certain minimum and maximum amounts. The amounts described below are the general rules that apply in most provinces and territories.[41]

The basic WITB can be claimed if the individual’s working income is more than $3,000. The amount that an eligible person may receive depends on the following criteria: working income, family income, whether they have an eligible spouse, whether they have an eligible dependent, and their province of residence.[42] That basic design will not change under the new CWB. Under the WITB, the benefit is phased in at a rate of $0.25 per dollar of working income above $3,000, reaching a maximum value of $1,043 for single individuals when working income is just over $7,000, and reaching a maximum value of $1,894 for a claimant with a spouse or child when working income is just over $10,300. These maximum values had been increased in September 2016 to offset higher contributions to the Canada Pension Plan (CPP) and, in Budget 2018, have been further increased for the new CWB.[43] Under the new CWB, the maximum benefit will rise to $1,355 for singles and to $2,335 for a claimant with a spouse or child. Under the WITB, once benefits reached the maximum, they were reduced at a rate of $0.15 per dollar; under the CWB, that phase-out rate has been lowered to $0.12 per dollar. Benefits in the CWB will be fully phased out at $24,111 for singles (previously $20,734 under the WITB) and at $36,483 for couples or single parents (previously $32,339 under the WITB). In general, this means that a full-time, full-year worker making more than the minimum wage in Ontario[44] will be above the CWB income threshold and will therefore not receive CWB.

The WITB credit is claimed by completing Schedule 6 – Working Income Tax Benefit and filing it as part of a personal income tax return.[45] Filling out Schedule 6 requires that eligible workers file a tax return, be aware that they could receive the WITB, and remember to complete Schedule 6 as part of their return.[46] The paper version of Schedule 6 requires that the applicant complete no fewer than 42 different steps in the calculation of their working income, family net income, and their WITB benefit amount.[47] This complexity almost certainly reduces the number of low-wage workers who include Schedule 6 in a paper return. When taxpayers use software to file a return, the calculations on Schedule 6 are greatly simplified but taxpayers still need to proactively instruct the software to complete the application for the benefit.

Generally, WITB is paid out in a single lump sum amount with any other refund owing, once the applicant’s full tax return has been assessed. However, applicants can ask CRA to pay up to half of their total WITB benefit in quarterly payments over the government’s fiscal year, April to March. An application for advance payments has to be made using a second form and must be submitted to the CRA every year between January and September.

Under the new CWB, tax filers with low earned incomes will have their tax returns reviewed by the CRA to assess eligibility for the credit, even if they do not complete Schedule 6 with their return.[48] That review will then identify CWB-eligible tax filers and automatically enrol them into the benefit. The Government of Canada estimates that another 300,000 Canadians will receive the CWB, relative to the WITB, as a result of automatic enrolment and benefit changes.[49]

1. What Do We Know About WITB/CWB Take-up?

Because eligibility for the WITB is based on several criteria, arriving at a precise estimate of the take-up rate of the WITB using aggregate public data is not possible. The CRA has published annual statistics on the number of WITB claimants and the total dollar value of benefits paid for each year from 2013 to 2016. In 2016, the Department of Finance released additional data and an analysis of the WITB as part of its annual Report on Federal Tax Expenditures.[50]

Between 2009 and 2015, the number of beneficiaries receiving the WITB has been fairly constant, hovering between 1.4 million and 1.5 million Canadians.[51] Some number of these will be recurrent recipients who apply regularly, but there will be some who enter and exit the program. According to a Department of Finance evaluation, in the period 2009 to 2011, the entry and exit rate in the program was roughly 50%, suggesting a substantial amount of turnover in the population of beneficiaries.[52] This means that, in any given year, roughly half of WITB recipients may be new to the process of applying for and receiving the benefit.

The same departmental analysis estimated, using 2012 income tax data, that the WITB had an 85% take-up rate among eligible tax filers.[53] Finance arrived at that estimate of take-up by examining data in individual tax returns and determined that some 15% of tax filers had not completed Schedule 6 to apply for the WITB, even though their income (including family income where returns can be linked between spouses) suggested “they may have been eligible.”[54] The department’s analysis also suggested that eligible non-participation was related to the method of tax filing. Those using paper forms represented 13% of all WITB-eligible filers, with the remaining 87% from WITB-eligible tax filers using either a third-party tax preparer or tax-filing software. Among eligible tax filers using paper forms, take-up of the WITB was just 49% of all those eligible. Among tax filers using tax software or a third-party agent for their return, take-up was significantly higher at 86% and 94%, respectively.

The department updated its estimate of the take-up rate of the WITB in a 2017 internal memorandum prepared for the Deputy Minister and released under the Access to Information Act.[55] The department again used confidential tax data to estimate take-up and arrived at an estimate of approximately 85% for the 2014 tax year, replicating the earlier evaluation that had used 2012 data. They further estimated that approximately 240,000 tax filers were likely eligible for the WITB but had not applied for it in 2014. This would suggest that, overall, 1.6 million people or 6% of all tax filers are eligible for the WITB.[56]

But what about those low-wage workers who are not filing taxes but would otherwise likely be WTIB (or CWB) eligible? The analysis of take-up from the Department of Finance does not consider non-filers. To arrive at an estimate of the number of Canadians who might be eligible but are not receiving the WITB/CWB, we need to know the number of non-filing workers. The technical documents from the most recent national Census provide some estimates of the overall rate of non-filing in Canada.[57] For the working-age population (aged 20 to 64), Statistics Canada reports that they are able to find a tax-return or other record with the CRA for between 85% and 96% of Canadian residents, suggesting a non-filing rate of between 4% and 15% of working-age adults. Not all non-filers, however, would be eligible for the WITB/CWB. If we assume that WITB/CWB-eligibility is the same among those who file a tax return and those who do not, then the WTIB/CWB eligible population could rise by the same 4% to 15%. Such increases would imply that there are between 46,000 and 172,000 working-age adults who are eligible for the WITB/CWB but do not file a return.[58] If the incidence of WITB/CWB eligibility is higher or lower than among those who file tax returns, then this range would likewise under- or over-estimate the number of persons who are eligible for the WITB/CWB but not filing tax returns. Regardless, it is almost certain that tens of thousands of low-wage workers in Canada are eligible but not receiving the WITB; they will not benefit from the proposed automatic enrolment system of the CWB because they are not filing tax returns.

2. How Does the WITB Take-up Compare to Similar International Examples?

Like the WITB, both the EITC (in the US) and the Working Tax Credit (in the UK) require that applicants actively apply for the program. Like the WITB, the EITC requires applicants to file an annual income tax return and complete a schedule to apply for the credit. In the UK, annual personal income tax returns are less common as many tax measures are dealt with, at least for workers in standard employment, through employer payroll systems. In fact, just 11 million,[59] or roughly one third, of the more than 31 million taxpayers in the UK file an annual tax return. The Working Tax Credit, however, does require a positive application, whether or not a return is required.

Both the US and UK governments regularly publish official statistics on the take-up rates for these WITB/CWB-comparable benefits; in contrast, the Canadian government does not publish take-up rates.[60] It is also important to note that the denominator used by both US and UK governments appears to include tax filers as well as non-filers, again in contrast to the Canadian government’s approach of counting only tax filers.

More specifically, the US government estimates, using administrative records and representative surveys of the US population, that 80% of eligible persons in the US are receiving the EITC. There is substantial variation across US states with the lowest observed take-up in the District of Columbia at 71.9%, and the highest in Mississippi at 84.1%, for the most recent tax year.[61] In the UK, take-up of the Working Tax Credit is much lower at just 65% for the most recent tax year.[62] By contrast, the WITB has had a relatively high take-up rate, estimated by the government at 85% among tax filers.[63] Under the automatic enrolment provision of the CWB, take-up among tax filers is expected to be effectively 100%. But when otherwise eligible non-filers are included in the denominator, take-up among low-wage workers is likely to be considerably lower, if rates of tax filing do not also rise. The automatic enrolment feature planned for the CWB is clearly a powerful tool in reducing eligible non-participation, even if automatic enrolment cannot eliminate it completely.

3. What Has the Government of Canada Done to Address Eligible Non-participation in the WITB?

It seems that requiring separate applications, whether as part of, or in addition to, a personal income tax return, reduces program take-up. As already mentioned, the Department of Finance reports that tax filers who appear to be eligible for the WITB are significantly less likely to apply for the benefit if they are using a paper return. In fact, the paper version of the WITB application requires an individual to complete a complex form with many steps needed to determine their potential eligibility and benefit amount.[64] By contrast, those tax filers who use software to complete an electronic return, or who use a third-party tax-filer (who, since January 1, 2013, have been required to electronically file tax returns if they accept payment for their services as a tax agent)[65] are much more likely to apply for the WITB.[66]

The transition from the WITB to the CWB with automatic enrolment will significantly improve take-up rates, but only among eligible tax filers. The problem of reaching otherwise eligible non-filers will persist. In fact, some important number of eligible WITB participants are new entrants or re-entrants to the workforce who are unaware of the availability of the tax credit and of their potential eligibility for it. They may be unaware of the need to file a return at all.[67]

In early 2018, the Minister of National Revenue announced changes intended to simplify the tax-filing process for some low-income Canadians.[68] In 2018, the agency invited 950,000 persons with low and stable income to complete a simplified return by responding to a short series of questions. If the target population overlaps with low-income workers who are eligible for the WITB, then this may lead to some increase in take-up of the benefit. The agency also reports that it has sent letters to 260,000 individuals for whom it had not received a tax return but, presumably, had some other administrative record on file.[69] The agency reports an 8% response rate to the letters and states that, as a result, $27.2 million in benefits and credits were paid once the returns were filed.[70] An unknown portion of those returns and benefits might relate to the WITB.

Previous authors have asserted that the relatively low value of the WITB/CWB to the average claimant, in comparison to the transaction and information costs involved in claiming the benefit, is also likely to reduce participation.[71] If so, then the increase in the maximum and average value of benefits paid under the new CWB, as compared with the WITB, may also increase take-up. Maximum benefits under the CWB are projected to be approximately $1,355 for singles and $2,353 for a recipient with a spouse or dependent child. But, this is still low by comparison with the maximum EITC value, which is upwards of USD $5,600 for recipients with a family.[72] Experimental studies of EITC participation suggest that one of the most powerful ways to improve take-up is to promote visibility of the dollar value of the expected benefit.[73] And yet, despite the modest dollar value of the benefit, take-up rates of the WITB have been comparable to the EITC and are expected to rise significantly under the CWB. We believe that, in contrast to increases in the dollar value of the benefit, automatic enrolment is the more effective avenue to reduce eligible non-participation.

B. Canada Learning Bond[74]

Since 2005, the Government of Canada has offered the CLB to low-income families who open an eligible RESP for their eligible children. For those eligible, the government deposits up to $2,000 per child into the RESP to be used by the children for post-secondary education. RESPs are tax-benefited accounts, available regardless of income, intended to allow those who hold the accounts—known as “subscribers”—to save for a beneficiary’s education without incurring tax on the investment income.[75] Funds deposited by subscribers into the RESP divide into two buckets: contributions and Education Assistance Payments (EAPs). Contributions by the subscriber are not tax-deductible given that their deposits are net of tax.[76] Contributions by the subscriber may also receive a federal matching grant to the RESP, at a rate between 20% and 40%, depending on the beneficiary’s family income. The CLB is payable regardless of other contributions into the RESP. The EAPs are revenues generated from the contributions and are subject to a marginal tax rate in the hands of the beneficiaries who, as students, are presumed to have low or no taxable income.[77] Money held in RESPs can be withdrawn only for the education expenses of the beneficiaries. The CLB program is aimed at children in low income families; the federal government deposits the CLB in an RESP established for the children and can therefore only be spent when they enter post-secondary education. Otherwise, the CLB and any other government funds are returned to the government.

To be eligible for the CLB, a child born on or after January 1, 2004 must have an eligible RESP in their name and their family must meet an annual income test, based on net adjusted income.[78] The income test is administered by the CRA, based on annual personal income tax data and applications, made through CRA, for the Canada Child Benefit. Normally, the CLB can only be paid to a RESP opened and controlled by the parent or guardian who receives federal child benefits for the child in question. Those who establish RESPs and want to also receive the CLB must be careful: not all RESPs available to consumers participate in the CLB. In fact, at least seven current RESP providers do not currently offer plans that participate in the CLB.[79]

Although the CLB is not directly administered by the CRA and is not applied for through a tax return, eligibility depends on tax return data for the parents or guardians of the child and data-sharing between the CRA and ESDC, the federal department responsible for the CLB program. Before the government will pay the CLB into a RESP, a subscriber (usually, though not exclusively, a parent or guardian)[80] must first obtain a Social Insurance Number for an eligible child, apply for and receive federal child benefits to verify their guardianship for the child, then select from among approximately 90 financial institutions providing RESPs (making sure to choose one that participates in the CLB) and enter into a legal contract with the RESP promoter.[81] This contract will require the subscriber to also have a Social Insurance Number, to file a tax return (so that CRA can assess eligibility for the CLB and related grants) and some may also require the subscriber to agree to make minimum monthly deposits into the RESP to qualify for additional grants.[82] The CLB itself does not require subscriber contributions to a RESP. Once a RESP is opened and a child is verified by CRA as eligible for the CLB, ESDC and the RESP promoter administer the program through a complex system of electronic funds transfers and monthly financial reporting. In short, the administrative obstacles to participating in the CLB are numerous for eligible families and for RESP promoters as well. Administrative barriers have been shown to significantly limit access, even where the programs were designed to assist people in meeting their basic needs, such as access to health services.[83] In the case of the CLB, these administrative barriers are in addition to the likely structural barriers to participation, such as differences in parental education or expectations.[84] It is those very same structural barriers to higher education that policymakers hoped to address in creating the CLB.

The CLB consists of an initial payment of $500 and up to 15 subsequent payments of $100, to a maximum of $2,000 per child.[85] Children who have been eligible for the CLB remain eligible for retroactive payments of the benefit until the age 21.[86] In total, 2.5 million children have been eligible for the CLB since the beginning of the program.[87] By best estimates, a little more than 126,000 children in Canada become newly eligible for the $500 CLB in any given year, another 1.3 million children are eligible for a subsequent $100 CLB payment, and 580,000 who were previously eligible will stop being eligible because of an increase in family income.[88] In short, there is substantial annual churn in the program as children move in and out of eligibility based on year-to-year variability of family incomes.

1. What Do We Know About CLB Take-up?

Take-up of the CLB differs from the WITB in at least two ways. First, compared to the WITB, CLB take-up rates are very low. Second, cumulative take-up rates of the CLB are voluntarily published on a regular basis, while take-up rates of the WITB have not been.

Each year, ESDC publishes an estimate of cumulative take-up based on the number of children who have ever been eligible for the CLB (that is, eligible in any year since the program started) based on CRA records of their family income.[89] ESDC reports on the proportion of those children who have ever received a CLB payment, including retroactive claims back to 2004.[90] According to the most recent statistical report, this proportion is estimated to be 38.3%.[91] The same statistical report also provides a within year estimate of the number of beneficiaries eligible in a given year and the number of CLB payments made in that year. Recognizing that payments made in a year may not correspond to eligibility for that year (given the option to claim retroactive payments), the trend in the ratio of beneficiaries to eligible children within a year nevertheless gives us an indication of the within-year take-up rates for the CLB. Cumulative take-up has inched slowly upward while within-year take-up has been relatively flat at an average of just 7.3% since 2006.

Again, using the same ESDC published data,[92] we estimate that a total of nearly 1.7 million children in Canada have been eligible for the CLB in a previous year and are still eligible in the most recent year. Another half a million children have been eligible but are no longer so in the most recent year.

2. What Has the Government Done Recently to Improve CLB Take-up?

In 2015, the Government of Canada committed to working with provincial and territorial governments to improve take-up of the CLB by simplifying enrolment and improving promotion of the program.[93] Since that time, the government has commissioned both external[94] and internal research to better understand obstacles to participation and identify avenues to increase take-up.[95]

Beginning in Spring 2018, new parents in Ontario will be able to start the process of applying for the CLB by opening an RESP online when they register a child’s birth.[96] Because the initiative is so new, there is not yet any data on the effectiveness of this new system. The interface to encourage RESP participation is available to all parents of newborn children in Ontario, and not specifically those who have lower or modest income and likely eligible for the CLB. In other words, the impact of the new system on CLB participation may be modest. Furthermore, the system will only be available to newborn children and will not address the backlog of eligible children who have not yet received the CLB.

The government also states that it has allocated money to fund pilot projects with private and voluntary sector stakeholders to bolster program participation.[97] As of the time of writing, no announcements have yet been made regarding those agreements. The government has also signaled its intent to amend the Canada Education Savings Act to make it easier for families to apply for the CLB.[98] It is also noteworthy that the government’s message, both in the 2015 mandate letter for the responsible minister, and again in public reporting on mandate commitments, is that parents will continue to have to proactively apply for the bond. This is in contrast to the direction taken with the WITB/CWB and OAS/GIS.

C. Old Age Security[99]

Most Canadians over the age of 65 are eligible to receive the OAS pension. Receipt is conditional on certain residency requirements but does not require any history of paid employment. OAS and the associated GIS are commonly referred to as the “first pillar” of Canada’s retirement income system.[100] The other two “pillars” are: (1) employment-based pension plans; and (2) personal retirement savings. OAS and GIS are the only pensions available to Canadians who have not worked in the paid labour market. The aim of these programs is to provide an amount sufficient to keep older Canadians out of poverty even if they have no other sources of income. The OAS and GIS programs are administered by ESDC.

To qualify for OAS, applicants living in Canada must: 1) be older than 65 years of age; 2) be Canadian citizens or legal residents at the time of application; and 3) have resided in Canada for at least ten years from the age of 18.[101] If applicants are living outside of Canada, they must meet the same requirements except that they must have resided in Canada for at least 20 years from the age of 18.[102]

The current maximum OAS payment is $891.18 per month.[103] Applicants can opt to delay their OAS pension by 60 months, with their monthly pension increasing by 0.6% for each month of deferral.[104] The OAS payment is income-tested with benefits phased out as a function of other income. Currently, seniors with annual net income above $123,302 have their OAS pension reduced to zero.[105]

Canadians may be eligible for a full or partial OAS pension. Applicants can qualify for a full pension in one of two ways. The first way is to have resided in Canada for at least 40 years after turning 18.[106] The second way to receive a full pension is to be 65 as of July 1 in a particular year and have evidence of living in Canada before being 25 years of age.[107] In addition to these requirements, the full pension requires the applicant to have resided in Canada continuously for ten years immediately before the approval of the OAS pension.[108]

Alternatively, the applicant may receive a partial pension. The difference between receiving a full and a partial pension rests on how long the applicant has lived in Canada after the age of 18. A partial pension is calculated at a rate of 1/40 of the full OAS pension for each complete year of residence in Canada after age 18. The minimum period needed to qualify for a partial OAS pension is ten years.[109]

In addition to the basic OAS pension, three additional benefits exist within the overall OAS program: the GIS, the Allowance, and Allowance for the Survivor. All three provide additional financial assistance to low-income recipients. The non-taxable GIS is available to applicants who live in Canada and have low incomes.[110] To qualify for GIS, the applicant must: 1) receive an OAS pension, and 2) possess an annual income (or in the case of a couple, a combined annual income) lower than a maximum threshold.[111]

The Allowance benefit addresses the financial needs of the spouses (or common-law partners) of GIS recipients. To be eligible for this benefit, the applicant must meet all of the following conditions: 1) be between 60 and 64 years of age (and thus not yet eligible for their own OAS/GIS benefits); 2) have a spouse or common-law partner who receives the OAS pension and is eligible for GIS; 3) is a Canadian citizen or legal resident; 4) if a legal resident, have resided in Canada for at least ten years since age of 18; and 5) have a combined annual income that is less than the maximum allowable threshold.[112] The applicant should apply for the Allowance, in writing, six to eleven months before their 60th birthday.[113]

The Allowance for the Survivor benefit[114] is available to those who are aged 60 to 64, who have a low income, who are living in Canada, and whose spouse or common-law partner has died. Applicants must meet all of the following criteria: 1) be between 60 and 64 years of age; 2) be a Canadian citizen or legal resident; 3) if a legal resident, have resided in Canada for at least ten years since the age of 18; 4) have a spouse or common-law partner who has died but the applicant has not remarried or entered into a common-law relationship; and 5) have an income less than the maximum annual threshold.[115] The amount received for these benefits is based on the previous year’s income.[116] As with the Allowance, applicants should apply in writing six to eleven months before their sixtieth birthday.[117]

1. Automatic Enrolment

As programs, OAS and GIS are particularly well-suited to automatic enrolment. Eligibility for most potential recipients is a function only of age, taxable income, and marital status, information that is readily available for all those filing tax returns. Moreover, the availability of OAS is widely known and there seems to be little stigma attached to its receipt.

Historically, in order to receive an OAS pension, it was required that seniors manually complete an application and mail it in.[118] Automatic enrolment into OAS (but not GIS) began in 2012 but the foundation for its introduction was established in 2001 by a survey of seniors using a Toronto food bank, an analysis of eligible non-participation in GIS by the CRA and a December 2001 parliamentary hearing.

The survey was conducted by Toronto’s Daily Bread Food Bank with the help of Ottawa social policy analyst Richard Shillington.[119] As reported on the front page of the August 23, 2001 edition of the Toronto Star, one quarter of the 800 seniors surveyed at the food bank were eligible for the GIS but were not receiving it.[120] The same article reported that HRDC had estimated in 1998 that about 1.4 million seniors were receiving the GIS but another 380,000 were eligible for the GIS but were not receiving the benefit.

The resulting public attention led to a hearing of the Standing Committee on Human Resources Development in December of 2001 and its subsequent report, entitled “The Guaranteed Income Supplement: The Duty to Reach All”.[121] The committee heard that HRDC had been aware of significant “under-subscription” of the GIS since 1994 and was taking “active measures” to reduce the number of eligible non-recipients.

In its report, the committee wrote “[t]hat not all eligible seniors receive the GIS is disturbing and unnecessary” and recommended that HRDC, working in conjunction with the CRA, develop an automatic enrolment system.[122] Noting that HRDC officials had argued that privacy concerns prevented them from using CRA tax return information, the report cited the opinion of the then-Privacy Commissioner, George Radwanski, that the Income Tax Act explicitly authorized CRA to provide information necessary to administer the Old Age Security Act, so that cooperation between CRA and HRDC would not raise privacy concerns.[123]

Subsequently, a comprehensive study of the take-up of OAS benefits (including GIS), again involving Richard Shillington, was completed in 2009.[124] The main finding was that there remained a significant number of Canadians who were eligible for OAS or GIS (or both) but were not receiving it.[125] The study noted, however, that the extent of the eligible non-participation, however, had declined significantly since 1999.[126]

To study eligible non-participation in OAS and GIS, the study used the Longitudinal Administrative Database (LAD). The LAD is an anonymized 20% sample of the tax returns of Canadians where the same tax-filer is recorded year after year. The study estimated that 116,000 Canadians, or about 3% of those eligible for OAS, were not receiving it. This number had remained about the same over the time period covered by the study (1999 to 2006). A much greater proportion were eligible for GIS but not receiving it. In 2006, between 135,000–150,000 were eligible but not receiving benefits.[127] Those numbers had fallen significantly since 1999, when about 25% were eligible non-participants.

Most reliable estimates of eligible non-participation are derived by analyzing tax returns; for example, only Canadians who filed tax returns are included in the LAD database. As noted above, however, some Canadians do not file tax returns and the proportion of non-filers who might be eligible for OAS and GIS but not receiving them is therefore difficult to estimate.[128]

In the 2012 budget, the federal government introduced automatic enrolment provisions for OAS.[129] The Minister of Employment and Social Development Canada was given discretion to waive the application requirements for the OAS if the Minister was satisfied that the person would meet the eligibility criteria for pension upon reaching age 65.[130]

Seniors who are eligible for automatic enrolment in OAS receive a notification letter the month after they turn 64.[131] Those who are not eligible for automatic enrolment must apply in writing for the OAS pension; that is, they must fill out the Application for the Old Age Security Pension.[132] The OAS automatic enrolment procedure now enrolls about 60% of those eligible.[133]

As of December 2017, automatic enrolment was expanded to include some of those eligible for the GIS.[134] Individuals not selected for automatic enrolment, however, must still apply in writing.[135] Further, for the purpose of the GIS, both the applicant and his or her spouse or common-law partner must certify their incomes.[136]

The transition to automatic enrolment has served to significantly increase the number of seniors qualified for a full pension and holds merit to be applied widely.[137] For OAS itself, the participation rate among those eligible is likely close to the maximum possible, currently standing at about 95%.[138] The participation rate among those eligible for GIS and the Allowances is lower but has risen in the wake of automatic enrolment.

While the legislation governing the OAS gives ESDC the option to waive certain requirements to ensure that commencement of the payment is automatic, the legislation does not prescribe a positive obligation to do so. Instead, the introduction of automatic enrolment seems to have been an acknowledgement that non-participation was creating hardship for seniors and that the hardship was publicly visible.

IV. Potential Solutions and Conclusion

Social programs provide assistance to individuals and families in need of support. When those eligible fail to participate, for whatever reason, the social programs have not fully achieved their policy goals. The loss resulting from a lack of take-up might not just be individual if the alternatives to receiving an income-tested transfer are socially costly—lower employment, lower participation in education, and greater demand for public services such as health or emergency care.

Governments may now recognize, when targeting vulnerable populations, that it is not enough simply to put a program in place and to hope that those eligible will sign up. To be sure, some believe that it is up to the individual to take the necessary steps to obtain access to the benefits without intense government assistance. Nonetheless, in all three of the programs discussed above, the federal government has made some efforts to increase participation despite the governing legislation containing no positive obligation to do so. Outreach efforts have been made, applications simplified, and automatic enrolment undertaken.

In this Part, we discuss avenues by which government could seek greater participation in the programs discussed here. Automatic enrolment is already underway in two of the three programs (OAS/GIS and WITB/CWB). Third party intermediaries are already involved in outreach efforts. Informational campaigns have been extensively used in the past.

A. Automatic Enrolment

Perhaps the most radical method of increasing participation is automatic enrolment. The key element required is information about those potentially eligible. In the two programs, OAS/GIS and the WITB/CWB, where automatic enrolment is being introduced, governments still rely on potential beneficiaries to file an annual tax return to provide essential information. But automatic enrolment might also be designed to use other sources of information so that an annual return is no longer needed. This broader approach to automatic enrolment, however, raises concerns about privacy as well as respect for the autonomy of individuals.

For example, suppose the government maintained a database that brought together public records of all residents in the country. Such a database could include birth dates, residence records, family structure information, education histories, income, and employment records, to name a few. For program administrators, it would then be a relatively simple matter to determine eligibility for OAS/GIS (only age, family income, family status, and residence history are needed), WITB/CWB (only personal earnings, education and employment status, family income, and family status are needed), and CLB (only the age of children and family income are needed if deposits did not require enrolment in a RESP).

While the database might simplify and largely automate the determination of eligibility for social programs, the Canadian public may not trust its government to construct, maintain, and properly use such a centralized government database. The creation and use of a centralized government database would be an additional threat to privacy. In addition, the feasibility of constructing and maintaining such a database is uncertain. Government is currently constrained from creating such a database by a patchwork of legislation preventing the sharing of information across departments, depending on the department and the nature of the information.[139] Previous efforts to assemble a centralized database on individual Canadians have been abandoned in the face of public opposition and criticism from the federal Privacy Commissioner.[140] Furthermore, high profile breaches or losses of personal data have occurred in several federal departments and agencies, including the Canada Revenue Agency.[141] In addition to concerns about the security of personal information, a central database that effectively automates applications for social programs removes an element of choice and control from individuals. Canadians have rights that generally (though not always) ensure the requirement for informed consent before their private data can be retained and used. That principle of autonomy also applies to participation in public benefits. The government may be embracing a positive obligation to deliver benefits to eligible Canadians, but eligible Canadians can still refuse to participate—either by refusing payments issued to them or by refusing to file a tax return and thereby denying the government access to personal information.

One possible way to address the threats posed by a centralized government database is to give individuals control over information about themselves. The information could be stored in an electronic “ledger”, essentially a history of personal events including, for example, educational achievement, marriages and divorces, and income sources. The events would be agreed upon and verified by the person and by relevant third parties (e.g., schools, courts and employers). Once the events had been entered into the ledger, they could not easily be altered.[142] Furthermore, the ledger could be “permissioned”, meaning that only the individual person could grant access to all or part of the information in it. For example, the individual could grant the government temporary access to the information for the purposes of determining eligibility for social programs.

The technology to create such an individual-level database, tamper-proof and under the control of the individual, already exists in principle under the name of the “blockchain system” and governments are already looking into its possible uses.[143] If blockchain systems containing encrypted individual data under the control of each individual become available, the information relevant for automatic enrolment could be released by individuals to the relevant departments. Of course, it would be up to the individual to maintain the completeness and accuracy of the personal databases and to voluntarily grant access to government departments and agencies that administer social programs. The same administrative barriers and structural issues that currently lead to eligible non-participation—including lack of awareness and a lack of the necessary skills or access to resources (namely the equipment and network services needed to securely access a “blockchain” database)—might also carry over to the maintenance of the individual databases and to providing access for its use in administering benefits.

B. Third-party Intermediaries

Non-governmental organizations are currently being asked, and sometimes subsidized, to encourage low-income Canadians to sign up for social programs.[144] The idea is that such organizations are closer to the groups likely to be eligible, whether they are immigrant integration agencies, medical professionals, community organizations, or anti-poverty groups, and thus will likely have better success in promoting and increasing access to these programs than a representative of the government. Since 2005, the Government of Canada has periodically offered grants to community organizations to promote education savings incentives, including but not limited to the CLB.[145] Groups such as Prosper Canada and SEED Winnipeg have designed programs in which community workers help low income Canadians fill out application forms or book appointments with their preferred RESP provider. The Omega Foundation, using a mix of federal and other funding, has instead developed an online platform that simplifies the RESP and CLB applications and seamlessly shares information with participating RESP providers. In the 2017 federal budget, the government allocated another $12.6 million over six years to support similar community outreach efforts. On the one hand, spending on these programs does not seem to have had a clear or measurable effect on take-up of the CLB. On the other hand, the amount allocated represents less than 0.3% of the total program costs and is considerably less than the Government of Canada has historically spent on campaigns to raise awareness of other government benefits.[146]

C. Informational Campaigns

One clear barrier to participation is the lack of awareness of the existence of the relevant programs or a misunderstanding of the eligibility requirements. When awareness is high, as it is for the OAS pension, participation rates are high. When awareness is low, as it is for the CLB, participation rates are low. A well-funded and well-designed advertising campaign could, at least in principle, raise awareness and raise participation rates. As noted above, some believe that the dollar value of benefits should be widely known and advertised, as participation seems to rise with the level of available benefits. The main challenge with trying to increase participation rates using either third-party intermediaries or informational campaigns is that they have been tried in the past with limited success.[147]

D. Two Practical Innovations

Neither increasing the extent of interdepartmental data sharing nor developing a blockchain system will come easily or quickly. In the short run, two smaller innovations may help to increase participation.

First, as mentioned above, both the US and the UK publish take-up rates for various social programs. Canada generally does not.[148] As we saw with the furor created when low GIS participation rates became publicly known in the early 2000s, public attention can lead to significant improvements. Publishing Canadian take-up rates would increase transparency and perhaps lead to improvements in those rates.

Second, more funding and public access to low and free electronic tax filing intermediaries should be introduced. We know from the WITB/CWB experience that the participation rate among those who are eligible and who file their taxes using paper forms is far lower than the rate among those using third-party intermediaries or who use electronic software. Increasing access to electronic filing will not only increase participation rates among tax filers but may also increase the rate of tax filing, and thus reduce the number of eligible non-participants.

E. Positive Obligations

One overriding measure that requires regulatory action and clarification is a positive obligation on the part of the government to ensure that payment of social benefits is made to those Canadians who need it. No explicit positive obligation provisions exist in any of the legislation we have explored, but automatic enrolment provisions have been instituted because eligible participants do not collect all the benefits to which they are entitled.[149] That is to say, automatic enrolment is the first step to satisfying a positive obligation, with the following steps being support, consolidation, and clarification through legislation in order to make sure that the positive obligation is met. Without such steps in place, the government of the day would have room to interpret these obligations through a political lens.

In sum, the OAS/GIS and the CWB are two of the three programs discussed here which have tried to increase participation through a form of automatic enrolment; however, the need to file a tax return with the CRA effectively prevents a number of potential beneficiaries from receiving benefits from these programs. For benefits like the CLB, there is still work to be done to ensure these programs are structured to be accessible to intended participants.

Automatic enrolment, potentially using a blockchain system, could collect information from eligible participants while simultaneously alleviating concerns surrounding privacy and access to information. Other countries have demonstrated an interest in this approach to sharing and utilizing information, and if instituted in a Canadian context, it could address the issue of benefit take-up rates. However, a fully automatic enrolment system appears quite distant. Our review of the current state of the enabling legislation and comparative take-up rates of each of the programs reviewed suggests there is a need for additional legislation to clarify the way that the government and CRA currently operate with respect to take-up rates. Advertising can further be used to accelerate this change and raise awareness about the need for automatic enrolment. Publicizing the dollar value of benefits to which claimants are entitled might motivate potential participants to apply. In the short-term, however, successfully utilizing third-party intermediaries to increase access to benefits for eligible claimants seems like the best alternative.

One of the key challenges that remain for developing solutions to the problem of eligible non-participation is that we do not know if people are making a mistake by not applying, or if they have rationally decided it is not in their own interests to apply. Nor, in fact, do we have official and reliable estimates of the number of eligible non-participants. It seems clear that tax filing presents an important obstacle, and one that not even recent efforts to automate enrolment have been able to overcome. It also seems clear that efforts to address benefit take-up vary considerably across government. We believe this is a public policy challenge that will outlast the political commitments of any one elected government. Further empirical work is needed to document the nature and extent of the problem of gaps in take-up of public programs, to identify both individual and institutional obstacles to improving take-up, and to consider the feasibility of potential solutions, including those described in this paper. This paper is, we hope, the beginning a longer conversation.