L’instauration en 2009 des comptes d’épargne libres d’impôt (CELI) a transformé les habitudes d’épargne des Canadiens. Leur création visait notamment à procurer aux citoyens à faible revenu un instrument qui leur permettrait d’améliorer leur revenu de retraite. Dix ans plus tard, beaucoup de ces petits épargnants n’utilisent toujours pas les CELI de manière à profiter pleinement des programmes de transferts gouvernementaux conçus pour leur retraite, par exemple le Supplément de revenu garanti. Résultat : les avantages escomptés des CELI restent largement inexploités. La sensibilisation du public et une meilleure littératie financière aideraient sans doute à résoudre ce problème, mais il serait plus efficace de privilégier des incitatifs intégrés et des ajustements fiscaux.
The introduction of Tax-Free Savings Accounts (TFSAs) in 2009 provoked a sea change in Canadians’ savings behaviour. TFSAs have become almost as popular as, and ultimately might surpass, Registered Retirement Savings Plans (RRSPs), which have existed for over 50 years. Although TFSAs are a welcome policy innovation for Canadian savers, TFSA benefits are tilted toward the wealthy, and many low-income Canadians fail to use TFSAs effectively to optimize their retirement income.
Equity concerns first came to light during the 2015 federal election debates, when other parties criticized the Conservative government’s doubling of TFSA contribution limits in the 2015 budget. This policy would be reversed by the subsequent Liberal government on the basis that inequities would grow considerably under the higher contribution limits. Nevertheless, remaining equity shortcomings of TFSAs need to be addressed.
TFSAs were designed, in good part, to assist lower-income seniors to make better retirement saving decisions, but progress in this regard has been limited. Since 2008, around 36 percent of workers without an employer-sponsored pension plan have opened a TFSA. Many future low-income retirees, however, are still not using TFSAs in ways that would allow them to benefit fully from government transfer programs intended for them — such as the Guaranteed Income Supplement (GIS) — and increase their retirement income.
In 2016, about one-third of Canadian seniors qualified for the GIS, and around 14 percent earned less than Statistics Canada’s low-income measurement. This proportion ranged from a low of about 4 percent of seniors in Alberta to a high of nearly 27 percent in Newfoundland and Labrador. As Canada’s population ages, growing numbers of seniors will need to rely on programs like the GIS, in addition to personal retirement savings, to boost their income and support a reasonable standard of living.
This paper proposes a combination of increased public education, nudges and tax revisions to help low-income Canadians take full advantage of TFSAs to increase their access to government transfers. Without action to improve TFSA use among low-
income Canadians, intended TFSA benefits will go untapped.
The retirement income system for seniors is based on three pillars: 1) Old Age Security (OAS) and the GIS; 2) the Canada/Quebec Pension Plan (CPP/QPP); and 3) private savings, including employer pension plans. Employer-sponsored pension plans, RRSPs and TFSAs each provide tax-assisted mechanisms to support retirement saving. CPP/QPP benefits are paid to almost all retired Canadians based on their prior earnings and years of contributions. Full OAS benefits, of around $7,200 per year in 2019, go to the majority of seniors. Some seniors will access only partial OAS benefits if they have limited years of residence in Canada or have a very high income: individuals with annual income over $74,800 must pay back OAS benefits at the rate of 15 percent on income above this amount, with the full OAS paid back once income reaches around $121,300.
The GIS is narrowly targeted to lower-income seniors, and benefits are reduced quickly, depending on income other than OAS benefits and a modest earnings exemption, with different maximum benefit amounts and reduction rates for singles and married or common-law couples. GIS benefit reduction rates are roughly $50 per $100 of income for most singles and couples. However, seniors with very low incomes — for instance, singles with annual income below $8,400 — can access a GIS “top-up” that is reduced at the rate of just under $75 per $100 of income. Table 1 (column 4) presents average GIS benefit reduction rates over the full qualifying income range.
A TFSA is a tax-prepaid savings account, meaning that contributions to the account are not tax deductible but investment returns from the account are not subject to income tax. And because TFSA withdrawals are not included in net income for tax purposes, they are not considered in determining eligibility for income-tested benefits, such as OAS, the GIS, the goods and services tax (GST) credit and the age credit.
In contrast, an RRSP is a tax-deferred savings plan, which means that contributions to the plan are deducted from current taxable income, and taxes owing on amounts contributed and earned on these investments are deferred to when savings are withdrawn. RRSP withdrawals are included in net income for tax purposes, and therefore taken into account in determining eligibility for income-tested benefits.
As well, a TFSA provides more flexibility than an RRSP because, when the owner reaches age 71, the RRSP must be converted to a Registered Retirement Income Fund or an annuity subject to minimum annual withdrawals. A TFSA, in contrast, is not subject to rules on when funds must be withdrawn. Further, a TFSA allows the individual to recontribute past withdrawals, whereas an RRSP does not.
The introduction of TFSAs was originally proposed as part of an effort to shift the tax system away from taxes on income toward taxes on consumption and to increase net savings. TFSAs were also meant to expand the room for tax-assisted retirement savings for Canadians, and provide a more appropriate savings vehicle for low-income savers. TFSAs were considered especially advantageous for low-income seniors who would receive the GIS. In fact, the value of TFSAs for low-income seniors was trumpeted in a number of newspaper op-eds, and the 2008 budget speech further emphasized this point:
A TFSA will provide greater savings incentives for low- and modest-income individuals because neither the income earned in a TFSA nor withdrawals from it will affect eligibility for federal income-tested benefits and credits, such as the Canada Child Tax Benefit, the GST credit, the Age Credit, and Old Age Security and Guaranteed Income Supplement benefits…In the first five years, it is estimated that over three-quarters of the benefits of saving in a TFSA will go to individuals in the two lowest tax brackets.
What, then, were the intended effects of TFSAs on lower-income savers and retirees? Prior to the introduction of TFSAs, Canadians who were likely to have low incomes and to qualify for the GIS in retirement had no effective tax-assisted method to save for retirement because GIS benefits are reduced by at least 50 cents for each dollar of income in retirement. If they have an RRSP, low-income seniors pay income tax on their withdrawals from the plan and also have their GIS benefits clawed back — and possibly other income-tested benefits as well. In other words, RRSP savings serve little effective purpose for low-income retirees because nearly all their registered savings are clawed back directly through taxes and indirectly through benefit reductions.
A large proportion, roughly one-third, of individuals ages 65 and older qualify for the GIS. The ability of these individuals to navigate access to this program through TFSAs can have a major influence on their income. Avoiding the GIS clawback was seen as a central feature of the TFSA when it was introduced, but if it is to work effectively for lower-income Canadians, TFSAs must substitute for RRSPs so that seniors’ savings are not subject to high effective tax rates.
Consider the example of a single individual who has both RRSP and TFSA savings and $16,000 per year in retirement income. For this individual, a $2,000 RRSP withdrawal would reduce their GIS benefits by over $1,000 per year and also be subject to income tax such that, in the end, they likely would be left with less than $500 — or less than one-quarter of the amount of the original withdrawal. In contrast, the same amount withdrawn from a TFSA would have no effect on GIS benefits and no tax implications, which means the individual’s income would be boosted by $2,000, the full amount of the withdrawal.
Unlike for most middle- to high-income seniors, financial planning for retirement for low-income seniors should emphasize how to get the most out of entitlement programs such as the GIS. For single individuals earning roughly $18,200 and couples earning less than roughly $24,100 per year (near the maximum income levels at which one would qualify for the GIS; see table 1), optimizing their retirement income means making savings and withdrawal decisions — prior to and during retirement — that take account of GIS qualification rules. For instance, ideally, many lower-income Canadians would be well advised to cash out any RRSPs around age 65 and shift those funds to a TFSA.
Moreover, low-income seniors with modest RRSP savings in retirement would be better off rapidly withdrawing funds than making gradual withdrawals. Cashing out RRSPs in the first year of retirement would mean that an individual likely would not have access to the GIS in the following year, but in subsequent years would qualify for the GIS for the rest of their life.
For example, suppose a single individual’s annual retirement income consists of OAS/GIS and $5,000 of CPP benefits, and the individual also has $50,000 in an RRSP. As illustrated in figure 1, if the individual draws down RRSP savings fully in the first year of retirement (the rapid RRSP withdrawal scenario), they would be ineligible for GIS benefits for one year, but subsequently would access much higher annual GIS benefits — around $1,850 higher — than if they take $2,500 per year out of the RRSP (the slow RRSP withdrawal scenario) over the full length of retirement.
Under the slow RRSP withdrawal scenario, cumulative GIS benefits and total after-tax income would be much less than the additional income from cashing out the RRSP quickly. It would amount to a difference of approximately $19,000 over a 20-year retirement period. The financial gain to be derived from a rapid RRSP withdrawal is even greater once other income-tested provincial and federal government transfers — which are not included in the figure 1 scenarios — are taken into account.
By withdrawing their RRSPs rapidly, prudent low-income savers could improve not only their cumulative retirement income but also their annual income. For instance, under this scenario if an individual places the $40,000 left over after tax ($50,000 RRSP withdrawal minus $10,000 in taxes) into a TFSA, they could top up their annual income by $2,000 per year over a 20-year period. This would give them an annual income of $21,830, compared with $20,480 in the slow RRSP withdrawal scenario.
Given the design of low-income benefits for Canadian seniors, we should be seeing, a decade after TFSAs were created, a significant movement away from savings in RRSPs and toward TFSAs by individuals who are likely to qualify for the GIS, but this has not occurred.
Are TFSAs Meeting Policy Objectives for Low-Income Workers?
Despite being available for only 10 years, TFSAs are being used by many Canadians and are approaching the popularity of RRSPs, although assets in TFSAs are still much lower. According to the 2016 Survey of Financial Security, about 42 percent of families have a TFSA, while 58 percent have an RRSP (table 2). By comparison, Registered Pension Plans (RPPs) are held by roughly half of families, and the average and median dollar amounts of assets in these plans are much higher than in either RRSPs or TFSAs.
To what extent are TFSAs being used by seniors who are most likely to have low income in retirement — that is, those without an employer-sponsored pension plan? As table 3 indicates, about 36 percent of seniors without an employer pension plan have a TFSA and about 42 percent have an RRSP; the average value of assets held in these accounts is $12,000 and $84,000, respectively. On average, these seniors have close to 90 percent of their total savings in RRSPs, with the rest in TFSAs. Further, the savings levels of seniors without an employer pension plan are very low, with average retirement savings of $97,000 and a median value of $5,000. The majority of those without an employer pension plan or adequate retirement savings are GIS bound: one-third of seniors are GIS recipients, but GIS recipients make up more than one-half of seniors without an employer pension plan.
Since the introduction of TFSAs, some lower-income Canadians have changed their saving behaviour by opening a TFSA and making deposits into it. This is a positive development. Yet TFSAs were expected to play a special role for lower-income Canadians: because it is advantageous for them to draw down their RRSPs prior to retirement or to draw them down rapidly once retired, TFSA savings should have begun to replace RRSPs. Although about 23 percent of GIS recipients have a TFSA, similar proportions also have an RRSP. Available data make it clear that, although a substantial minority of low-income savers have taken up TFSAs, the majority still have most of their funds in RRSPs.
How should one interpret the persistence of RRSP savings among low-income earners soon to qualify for the GIS? It might be that the transition to TFSAs will take several decades because seniors are reluctant to withdraw savings from previously existing RRSPs. It is more likely, however, that financial institutions are not tailoring appropriate financial advice for low-income savers. For instance, John Stapleton suggests that the advice GIS-bound Canadians obtain from banks is likely wrong. He mentions having accompanied, on a number of occasions, a GIS-bound senior to a bank, where they were told they would be foolish to cash in their RRSP before age 65 — advice that is demonstrably wrong for most people. Such anecdotal evidence might explain the very modest changes in TFSA and RRSP savings patterns since 2008.
Another major concern is that, in the long run, the benefits of TFSAs are likely to accrue mainly to high-income savers. Indeed, a study by the Office of the Parliamentary Budget Officer has found that, over time, TFSA benefits — in terms of lower taxes or higher government transfers received — shift more toward high-wealth households. The same study projects that, by 2060, TFSA benefits accruing to high-wealth households will be twice those going to median-wealth households and 10 times those going to low-wealth households (see figure 2). These higher benefits will come mainly at the expense of forgone government revenues, which means TFSAs will have a major effect on public finances in the decades to come. (See box 1 for a discussion on the long-term government revenue costs of TFSAs.)
Indeed, not only do TFSAs allow those with available after-tax cash to accumulate significant wealth free of income tax, the additional risk is that, in retirement, they could also take advantage of targeted benefit programs such as the GIS and GST credits, which were explicitly designed to exclude those with income above a certain level. Some analysts have cast doubt that seniors with significant TFSA withdrawals could get their income low enough to receive the GIS — under $20,000 for a single senior. However, even a few well-publicized cases of GIS benefits going to “well-off” seniors risk undermining support for benefits such as the GIS that are targeted to low-income seniors.
Planning for retirement is complicated at the best of times, requiring some understanding of compound interest, investment risk and longevity risk. It also raises questions about what mix of savings instruments is most efficient at what stages of life and the choice between paying down a mortgage and contributing to RRSPs or TFSAs. For most middle- and higher-income Canadians, the advice is relatively straightforward: save early in an RRSP or pension plan because your marginal tax rate will fall at retirement along with a decline in taxable income.
The effect of retirement on earnings and taxes, however, is quite different for low-income seniors. While choosing incorrectly between an RRSP and a TFSA will have some marginal effect on income at retirement for seniors who are better off, for those who might qualify for the GIS the clawback makes saving in an RRSP, compared with a TFSA, extremely wasteful. For many lower-income workers who are approaching retirement and are GIS bound, the best advice is to use a TFSA instead and cash out their RRSP at around age 65. For new generations of low-income workers, not opening an RRSP in the first place is best. But who is going to give them this advice?
The complex retirement income system increases the need for financial advice and better financial literacy for Canadians. Given the cost of expert advice, however, lower-income seniors will often rely on bank employees and/or family and friends for advice. In this context, there are two general approaches: active measures such as improving financial education and literacy for seniors and advisers; and passive measures that simplify the system and “nudge” savers into desirable savings vehicles based on their financial circumstances.
One way to increase the effectiveness of TFSAs is to educate lower-income Canadians on how to navigate Canada’s complex tax-and-transfer system to their best advantage. Recent history, however, especially regarding GIS enrolment, is not encouraging. In 2001, 300,000 seniors were not getting the GIS, even though they qualified, because they did not apply. The federal government’s first response was to write to these seniors encouraging them to apply, but this proved ineffective. Finally, over a decade later, Ottawa abandoned the attempt to increase financial literacy and outreach in favour of automatic enrolment in the GIS and OAS.
The Canada Revenue Agency is able to identify individuals at or near retirement, including those who potentially could face steep reductions in GIS benefits. The agency could show these individuals directly the different outcomes of withdrawals from either a TFSA or an RRSP. Encouraging the use of TFSAs by raising awareness and knowledge alone, however, has its limits.
A more effective approach than trying to educate low-income seniors about optimal RRSP use might be to implement measures to simplify and mitigate how RRSP withdrawals interact with other government programs — for instance, by easing annual income exemptions from GIS qualification calculations. Currently, the first $3,500 of wage income is ignored in determining eligibility for the GIS. Budget 2019 announced that this exemption will rise to $5,000 annually for wage and self-employment income. It also introduces a partial exemption of 50 percent on up to $10,000 of annual wage and self-employment income that is above the new $5,000 threshold for GIS recipients. Although this new measure is an attempt to improve GIS recipients’ incentives to work, it does not help address the RRSP withdrawal dilemma faced by many retirees who qualify for the GIS.
One could imagine a scenario whereby the first $3,500, say, of income, regardless of source — whether wages, investment income, pension benefits, RRSPs or CPP/QPP benefits — could be ignored for GIS purposes. As a rough estimate, the added program cost would be about $2.8 billion per year, based on the current number of GIS recipients. Although this is arguably a crude policy solution for the GIS-reduction problem that low-income savers with RRSP savings face, it would reduce their RRSP/TFSA dilemma.
Another possibility is to explore how behavioural nudges and cleverly designed policies could shift savings behaviour toward a more beneficial savings vehicle. Consider a concrete example such as group RRSPs, which are usually sponsored by employers and held with a designated financial institution. Even though most individuals who earn less than $50,000 per year almost always will be better off putting their savings into a TFSA, many are steered toward group RRSPs via employer-sponsored plans. On top of this, many Canadian employers will match a certain share of individuals’ contributions to group RRSPs, potentially further exacerbating the problem.
A “nudge” solution could be to offer employer-sponsored group TFSAs — currently provided by a few financial institutions — to employees who earn less than $50,000 per year, making that savings vehicle the default option for these employees, although, of course, they could opt out of the group TFSA in favour of the group RRSP. Similarly, workers earning over $50,000 a year could be placed in the group RRSP but given the option to opt out in favour of the group TFSA. As with group RRSPs, employers who sponsor group TFSAs would have the option to match employee contributions, but they would not be obligated to do so.
Regulatory requirements could make it incumbent on an employer, or perhaps even the financial institution the employer uses for its group retirement savings plan, to make the default choice of enrolling an employee in either a TFSA or an RRSP, depending on the employee’s annual pay. Further, this administrative criterion could be built into the Canadian Association of Pension Supervisory Authorities’ guidelines for pension plan governance to oversee and ensure compliance.
Given how central behavioural economics has been in the design and evaluation of pension programs and retirement savings decisions, a broader evaluation of the existing nudges in the RRSP-versus-TFSA decision-making process is worthy of federal policy-makers’ efforts. For example, policy-makers could consider a “savings credit” that would match a certain amount of TFSA contributions for lower-income savers. Using Registered Education Savings Plans as a model, the federal government could contribute a top-up amount to TFSAs that would vary according to family income. The matching credit could be up to 20 percent for very low-income families, and be reduced to zero as family income approaches the median annual income — roughly $50,000. A clear rule to lock in previously saved funds and matching contributions, perhaps until retirement, would also need to be put in place. A strong “nudge” such as this would cost roughly $1.5 billion, and would induce banks to encourage TFSA contributions rather than RRSPs for lower-income savers. This initiative no doubt would cost the federal government money, but the amounts required would pale in comparison with the long-run costs of TFSAs in terms of forgone tax revenue.
Both the take-up and maxing out of contributions to TFSAs increase with income, and as a result they disproportionately benefit higher-income individuals — and increasingly so over time. The government has two options to prevent retired Canadians whose total (taxable and nontaxable) income exceeds transfer program eligibility limits from using TFSAs to access benefits meant for low-income households. It can either modify the clawback rules on these income-tested programs, or change the tax code so that a portion of (nontaxable) TFSA withdrawals is taken into account in determining benefit eligibility.
Options to that effect include applying an asset threshold on an individual’s TFSA balance, whereby any withdrawals from the TFSA above the asset threshold would have to be reported as income in determining eligibility for GIS benefits. This asset limit would need to be high enough — say, a threshold of $100,000 for the GIS program — so as to provide maximum flexibility for low-income savers.
Further, asset limits could be considered in the context of a discussion on how to limit the growing government costs associated with TFSAs over time, including imposing lifetime contribution limits (see box 2). For instance, one analyst proposes a lifetime contribution limit of $150,000 and a limit of $300,000 on the TFSA assets that are sheltered, which would mean that assets over that amount would no longer grow tax-free. In targeting both contribution limits and assets held in TFSAs, the objective would be to limit the future effect on government finances as well as prevent too much inequity in the distribution of TFSA benefits.
Ten years after their introduction, Tax-Free Savings Accounts are having a positive effect. Usage is increasingly common and savings are up. TFSAs have yielded limited improvements, however, as a tax-deferred vehicle to assist low-income individuals save for retirement, even though this was one of the main reasons for their introduction. Ensuring new generations of low-income savers choose TFSAs over RRSPs to save for retirement and helping existing ones better manage asset allocation between their prior RRSP savings and new TFSA contributions is critical to achieve the intended benefits for this group of savers.
The evidence suggests that growing numbers of low-income Canadians are using TFSAs, but RRSPs continue to dominate. Too many future GIS recipients are not getting the advice they need to shed their RRSPs and some are still, wastefully, saving in them. Policy-makers’ efforts should focus on improving TFSA use among low-income savers, where a combination of improved education, nudging, perhaps a “saver’s credit” and other tax changes could boost low-income retirees’ incomes. Legitimate equity concerns will continue as TFSAs evolve, making it critical to follow through on a central motivation for creating TFSAs: to ensure that more low-income Canadians save effectively and optimize their retirement income.
 J. Kesselman and F. Poschmann, “A New Option for Retirement Savings: Tax-Prepaid Savings Plans,” C.D. Howe Institute Commentary 149 (Toronto: C.D. Howe Institute, February 2001).
 St. Christopher House, “A Proposal for a Tax Prepaid Savings Plan Exempt from Welfare Restrictions on Assets and Income,” Pre-budget Submission to the House of Commons Committee on Finance (Toronto: St. Christopher House, November 2003).
 Op-eds and letters to the editor suggested TFSAs would assist low-income seniors, as shown by the following quote from a letter to the editor: “The class division question, whether TFSAs favour the rich over the poor, lies at the heart of why we thought developing TFSAs was good policy: because the old system did not serve the poor well”; Financial Post, April 8, 2011.
 Finance Canada, Responsible Leadership: The Budget Plan 2008 (Ottawa, February 26, 2008), 81.
 The general reduction rate is 50 percent; it is much higher in some income ranges, and can be 100 percent or over when combined with various provincial income-tested benefits, principally GIS top-ups.
 See B. Baldwin and R. Shillington, Unfinished Business; Pension Reform in Canada, IRPP Study 64 (Montreal: Institute for Research on Public Policy, 2017); A. Laurin and F. Poschmann, “Saver’s Choice: Comparing the Marginal Effective Tax Burdens on RRSPs and TFSAs,” e-brief (Toronto: C.D. Howe Institute, January 27, 2010); J. Stapleton, “Down but Not Out: Reforming Social Assistance Rules That Punish the Poor for Saving,” e-brief (Toronto: C.D. Howe Institute, March 2, 2010).
 Further, the individual could move any RRSP withdrawals that exceed annual expenses to a TFSA.
 The first $13,000 of RRSP withdrawals (when combined with $5,000 of CPP/QPP income, it would bring total income to $18,000 — roughly the maximum income to qualify for the GIS) would translate into a full reduction of GIS benefits and would be taxed at a fairly low rate — say, 20 percent.
 For simplicity, I make no assumptions about RRSP investment growth during the retirement period.
 This figure is not discounted to net present-value terms, which would make the rapid TFSA withdrawal schedule more advantageous.
 R. Shillington, “An Analysis of the Economic Circumstances of Canadian Seniors” (Toronto: Broadbent Institute, February 2016).
 Finance Canada, “Tax Evaluations and Research Reports: Tax-Free Savings Accounts: A Profile of Account Holders” (Ottawa, 2012), accessed May 17, 2018, https://www.fin.gc.ca/taxexp-depfisc/2012/taxexp1202-eng.asp#toc346014054
 J. Stapleton, “Planning to Retire on a Low Income: What You Need to Know” (presentation at the Lillian Smith Branch, Toronto Public Library, November 2, 2016), accessed May 17, 2018, https://openpolicyontario.s3.amazonaws.com/uploads/2016/10/Retiring-on-a-Low-Income-libraries-september-2016new-oct30.pdf
 See Stapleton, “Planning to Retire on a Low Income.”
 Office of the Parliamentary Budget Officer, “The Tax-Free Savings Account” (Ottawa, February 24, 2015), https://www.pbo-dpb.gc.ca/web/default/files/files/files/TFSA_2015_EN.pdf
 Income-tested benefits such as the age credit, OAS and other targeted provincial seniors’ benefits could be affected.
 J. Robson, “TFSAs (As They Are Today) Aren’t Terrible,” Maclean’s, March 1, 2015, accessed May 17, 2018, https://www.macleans.ca/economy/economicanalysis/tfsas-as-they-are-today-arent-terrible/
 Moreover, the opportunities to use TFSA assets flexibly around retirement could also allow individuals to delay the receipt of CPP/QPP benefits and withdrawals from RRSPs until age 70, and therefore they would have virtually no income for the purposes of determining GIS eligibility. Some experts and advisers have explained how, under the current rules, this could be done; see, for example, F. Vettese, “Even the Rich Can Qualify for Guaranteed Income Supplement — Here’s How,” Financial Post, November 11, 2014, http://business.financialpost.com/personal-finance/tfsa/even-the-rich-can-qualify-for-guaranteed-incomesupplement-heres-how
 For more on Budget 2019’s changes to the GIS earnings exemption, see Government of Canada, A Secure and Dignified Retirement for Canadians, https://www.budget.gc.ca/2019/docs/themes/seniors-aines-en.html
 This would also help with a similar problem associated with enhanced CPP/QPP benefits in the case of low-income earners; see Baldwin and Shillington, Unfinished Business.
 J. Stapleton and R. Shillington, “No Strings Attached: How the Tax-Free Savings Account Can Help Lower-Income Canadians Get Ahead,” e-brief (Toronto: C.D. Howe Institute, September 30, 2008).
 For a recent example of this option, see Common Wealth and Maytree, The Canada Saver’s Credit: A Proposal to Build Financial Security for Lower- and Modest-Income Canadians (Toronto: Maytree, February 2019).
 Author’s estimate based on a custom tabulation provided by Statistics Canada.
 See J. Kesselman, “Tax-Free Savings Accounts: Expanding, Restricting, or Refining?” Canadian Tax Journal 63, no. 4 (2015): 504-32; and J. Kesselman, “Behind the Headlines: Who’s Really Benefiting from Higher TFSA Limits” (Toronto: Broadbent Institute, June 2015).
 One could either deal solely with the GIS and limit the amount of TFSA withdrawals that the GIS ignores or, more broadly, include some TFSA withdrawals in net income (line 236 of the tax return), which is used for the GIS and other income-tested benefits such as OAS. Another approach to limit the extent to which TFSAs can be used to avoid income-tested clawbacks would be to modify the income tax code so that a tax filer could include a portion — say, the first $5,000 — of TFSA withdrawals in net income (line 236), but be given a corresponding deduction so that it is not taxable. This way, the included portion of TFSA withdrawals would affect income-tested benefits (most of which use net income) but TFSAs’ tax-free status would be retained. This mechanism is currently used for certain tax-free income sources, such as workers’ compensation and welfare income.
 Kesselman, “Tax-Free Savings Accounts.”
 D. Macdonald, “The Number Games: Are the TFSA Odds Ever in Your Favour?” Behind the Numbers (Ottawa: Canadian Centre for Policy Alternatives, May 2015).
This paper was published as part of the Faces of Aging research program, under the direction of Colin Busby. The manuscript was copy-edited by Barry Norris, proofreading was by Barbara Czarnecki, editorial coordination was by Francesca Worrall, production was by Chantal Létourneau and art direction was by Anne Tremblay.
Richard Shillington is an Ottawa-based statistician whose research interests include poverty measurement, tax policy and the design of effective supports for low-income Canadians, particularly seniors. He has conducted research for more than 30 years.
To cite this document:
Shillington, Richard. 2019. Are Low-Income Savers Still in the Lurch? TFSAs at 10 Years. IRPP Insight 27. Montreal: Institute for Research on Public Policy. DOI: https://doi.org/10.26070/114a-qf18
The opinions expressed in this study are those of the authors and do not necessarily reflect the views of the IRPP or its Board of Directors.
IRPP Insight is an occasional publication consisting of concise policy analyses or critiques on timely topics by experts in the field.
Cover photo: Shutterstock, by Wutzkohphoto.
ISSN 3392-7748 (Online)
Montréal – Le compte d’épargne libre d’impôt (CELI) avait notamment pour but d’aider les Canadiens à faible revenu à optimiser leur épargne-retraite, mais beaucoup d’entre eux laissent encore de l’argent sur la table, révèle une publication de l’Institut de recherche en politiques publiques.
Au fur et à mesure que vieillira la population, de plus en plus d’aînés dont l’épargne-retraite est insuffisante devront recourir à des programmes gouvernementaux comme le Supplément de revenu garanti (SRG) pour maintenir un niveau de vie raisonnable. Mais à l’approche de leur retraite, bon nombre de petits épargnants n’utilisent toujours pas le CELI de manière à profiter pleinement des programmes fédéraux, soutient le statisticien d’Ottawa Richard Shillington.
Selon le principal avantage du CELI, les revenus de placement et les retraits qu’on y effectue n’ont aucune incidence sur l’admissibilité aux prestations et crédits fondés sur le revenu. À l’inverse, les économies investies dans un REER sont contreproductives pour les futurs retraités moins nantis, puisqu’une grande partie de leur épargne enregistrée sera récupérée directement par l’impôt et indirectement par une réduction des prestations du SRG.
« Comme on conseille rarement aux futurs bénéficiaires du SRG de délaisser leur REER, note Shillington, beaucoup continuent d’y cotiser inutilement. » Depuis 2008, seuls 36 % des travailleurs sans régime de pension d’employeur ont ainsi ouvert un CELI. Vu ses avantages potentiels pour les aînés à faible revenu, les Canadiens qui sont admissibles au SRG devraient pourtant privilégier le CELI au lieu de conserver leurs REER. Au moment de sa création, les décideurs croyaient d’ailleurs qu’il serait nettement plus populaire.
Les décideurs doivent donc s’efforcer d’accroître le taux d’adoption et d’utilisation du CELI chez les petits épargnants, conclut l’auteur. Et pour aider les Canadiens à épargner plus efficacement en vue d’optimiser leur revenu de retraite, on pourrait combiner des options d’épargne par défaut en milieu de travail, des incitations fiscales et même un « crédit à l’épargnant ».
On peut télécharger l’analyse Are Low-Income Savers Still in the Lurch? TFSAs at 10 Years, de Richard Shillington, sur le site de l’Institut (irpp.org/fr).
L’Institut de recherche en politiques publiques est un organisme canadien indépendant, bilingue et sans but lucratif, basé à Montréal. Pour vous tenir au courant de ses activités, veuillez vous abonner à son infolettre.
Renseignements : Shirley Cardenas tél. : 514 594-6877 firstname.lastname@example.org
The report card on Tax-Free Savings Accounts (TFSAs) at 10 years shows that, despite being designed to assist people in lower-income brackets make better retirement-savings decisions, TFSAs are still falling short of this goal. In a recent Institute for Research on Public Policy paper, I urge policy-makers to focus on improving TFSA use among low-income savers. With a combination of improved education, default rules for savings accounts and perhaps a “savings credit,” their retirement incomes could be boosted. Further, this would spread the benefits of TFSAs more evenly across the population; at present, benefits are slanted towards high-income savers.
The value of TFSAs for low-income seniors was trumpeted in the 2008 budget speech, which noted that TFSAs “will provide greater savings incentives for low- and modest-income individuals because neither the income earned in a TFSA nor withdrawals from it will affect eligibility for federal Guaranteed Income Supplement and tax credits.”
The fact is, low-income savers who choose to save in Registered Retirement Savings Plans (RRSPs) over TFSAs are leaving money on the table. This is because, in retirement, RRSP withdrawals are taxable, and they also reduce the government benefits their holders might have qualified for if they had saved in a TFSA. Too many future recipients of the federal Guaranteed Income Supplement (GIS) — the income transfer for low-income seniors — are not getting the advice they need to shed their RRSPs, and some are still, wastefully, saving in them.
While growing numbers of low-income Canadians are using TFSAs, RRSPs still dominate. The data are clear: although a substantial minority of low-income savers have opened up TFSAs — about 36 per cent of seniors without an employer pension plan — about 90 per cent of all their savings are still in RRSPs.
Why do low-income earners still seem to prefer RRSPs over TFSAs? To some degree, the transition to TFSAs will take several decades. But a major reason is likely that financial institutions are not tailoring appropriate financial advice for low-income savers. Anecdotal information suggests that the advice that low-income Canadians obtain from banks is incorrect. Banks, and even financial-advice periodicals, tell low-income seniors they would be foolish to cash in their RRSP before age 65. This is often bad advice.
The inability to improve TFSA usage among low-income Canadians underscores legitimate equity concerns about TFSAs, where most benefits still go to higher-income Canadians. Plus, economists and the Parliamentary Budget Office have shown that the long-run cost of TFSAs in terms of federal and provincial tax revenues, mainly from this high-income group, will be large.
The data also show that inequities in terms of who benefits from TFSAs are growing. Between 2014 and 2016, the assets in TFSAs worth over $500,000 more than doubled from $45 million to $108 million, and the assets in TFSAs worth over $1 million increased from $20 million to $49 million. These inequities will continue to grow as lifetime contribution limits expand: The 2019 TFSA lifetime contribution limit is $63,500 and it will reach $100,000 by 2025.
What to do? At the 10-year mark, the federal government should focus on boosting the benefits for low-income Canadians with options to encourage more effective saving decisions. This could include a “savings credit,” which — similar to RESPs — would provide matching grants for a certain amount of TFSA contributions for lower-income savers. As well, it should investigate ways to ensure that financial advisers, especially those in banks, give better advice to GIS-bound seniors.
Of course, these fixes cost money. But the amount of money required pales in comparison to the long-run government costs from TFSAs. Should they not be addressed soon, the federal government should consider capping lifetime contributions of TFSAs — at around $100,000 to $250,000 — to avoid creating more TFSA millionaires while shortcomings persist among lower-income Canadians.