Policies, Preferences and Perversions in the
Tax-Assisted Retirement Savings System
BarbaraAustin”
Reform of the retirement savings system has been
the subject of intense speculation and comment in recent
years. The system has been criticized for reasons rang-
ing from cost to equity. This paper attempts to evaluate
the current tax-assisted retirement savings system in the
context of stated federal government objectives of ensur-
ing that all Canadians have adequate retirement income
and encouraging increased private saving now to meet
pension needs in the future. In particular, the paper ad-
dresses three questions; (1) Are these policy objectives
being met in the most efficient way possible?; (2) Are
there better alternatives?; and (3) Is the system fair? The
paper concludes that further changes to the tax-assisted
retirement savings system are advisable in order to in-
crease both fairness and government revenue and to
meet the government’s stated policy objectives more ef-
ficiently and effectively than the current system seems
able to do.
Au cours des demires ann~es, la r6forme du r6-
gime d’alde fiscale 4 l’dpargne-retraite a donn6 lieu A
d’intenses conjectures et critiques qui portent autant sur
le cofit qu’il reprsente que sur son manque d’quit&
Cet article tente d’6valuer le present r6gime d’alde fis-
caleh l’Apargne-retraite dans le contexte des objectifs
6nonc6s par le gouvemement f&ldral, soit que chaque
citoyen canadien ait un revenu de retraite ad~quat et que
l’6pargne priv6e actuelle soit encourag6e afin de pour-
voir aux besoins futurs. En particulier, l’auteure se pen-
che sur trois questions, A savoir : (1) ces mesures sont-
elles appliqu(es de la manibre la plus efficace ; (2)
existe-t-Hi de meilleures alternatives ; et (3) le systime
est-il 6quitable ? L’article conclut que des modifications
suppl~mentaires au rgime d’aide fiscale A l’6pargne-
retraite sont n6cessaires afin d’augmenter l’6quit6 ainsi
que les recettes fiscales et de remplir les objectifs du
gouvemement de faqon plus efficace que ne semble le
faire le present rgime.
. The author wishes to acknowledge the assistance of Professor Neil Brooks, Osgoode Hall Law
School, who made very useful and insightful comments with respect to an earlier version of this pa-
per. The errors remaining and opinions expressed are those of the author.
McGill Law Journal 1996
Revue de droit de McGill
To be cited as: (1996) 41 McGill L.J. 571
Mode de r6fdrence : (1996) 41 R.D. McGill 571
MCGILL LAW JOURNAL/REVUE DE DROITDE MCGILL
[Vol. 41
Synopsis
Introduction
L The Present System
II. The Policy Objectives of Tax-Assisted Retirement Savings
m. Does the Tax-Assisted Retirement Savings System Meet its Policy Objectives?
A. Promoting Saving for Retirement
1.
2.
3.
Do Tax Incentives Increase Private Retirement Savings?
Other Tax Incentives for Saving
Should Private Saving Be Encouraged?
B. Ensuring Adequate Retirement Income and Saving the Public Purse
IV. Redesigning the System
A. Reducing the Contribution Limits
B. EliminatingAilDeductions for Contributions
C. Conversion to a Tax Credit
D. Taxing Fund Investment Income
E. Other Assorted Proposals
Conclusion
19961
Introduction
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVINGS
There are important questions related to tax expenditures. Do they ac-
complish their policy purpose in the most efficient way possible? Are
there alternatives that are better? Are the individual tax expenditures in
place fair? Or do they result in a situation where one part of society is
bearing more or less of the burden than it should? In that sense, as Ca-
nadians examine the choices ahead, we would suggest that they look at
tax expenditures in addition to program spending.’
The Honourable Paul Martin, Minister of Finance
The Minister of Finance, in his address to the House of Commons Standing
Committee on Finance (“Finance Committee”) prior to its first ever pre-budget
public consultation hearings, suggested that tax expenditures as well as programme
spending be considered as possible avenues toward deficit reduction. Many of the
presenters at the hearings recommended paying particular attention to reducing the
largest federal tax expenditure: tax-assisted retirement savings. A multitude of pos-
sible changes to the tax-assisted retirement savings system were suggested includ-
ing reducing the contribution limits, converting the deduction to a tax credit and
taxing the retirement fund. At the same time, there was a groundswell of support
for the present system. Many of those who made presentations to the Finance
Committee advanced dire predictions regarding the government’s ability to meet its
future financial obligations of providing Canadians with an adequate retirement in-
come should the present system of tax-assisted retirement savings be altered. The
Finance Committee made two recommendations. The first called for further study
of the issues; the second suggested that no future changes be made to the retirement
system without considering: (1) the need to encourage people to save more in order
to decrease future dependence on government; (2) the desirability of equality in the
tax treatment of different types of retirement plans; and (3) the desirability of
equality in the tax treatment between public-sector and private-sector retirement
plans.2 Despite pre-budget emphasis on retirement saving, both at the Finance
Committee level and in the media, only relatively minor changes were made in this
area in the 1995 and 1996 federal budgets.
This paper will first attempt to apply the Minister of Finance’s pre-budget
questions to the tax-assisted retirement savings system in order to determine
‘Minutes of Proceedings and Evidence of the House of Commons Standing Committee on Finance
(Ottawa: Queen’s Printer, 1994) at 58:14 [hereinafter Minutes of Proceedings and Evidence].
2 See Canada, Confronting Canada’s Deficit Crisis – Building Our Next Budget through Consulta-
tion: Tenth Report of the Standing Committee on Finance (Ottawa: Supply & Services Canada, 1994)
at 38 [hereinafter Canada’s Deficit Crisis]. Thus, the Liberal-dominated Finance Committee is clearly
committed to retaining tax incentives for retirement savings. However, this was not enough for the
Bloc Qudbicois members of the Committee who, in their dissenting opinion, criticized the Liberals
for refusing to eliminate the possibility of taxing pension and Registered Retirement Savings Plan
(“R.R.S.P”) funds (Canada’s Deficit Crisis, ibid. at 92). The Reform members did not specifically
comment on retirement savings in their minority report but stated that they are against all tax in-
creases and would only support tax changes if they were revenue neutral (ibid. at 104).
MCGILL LAW JOURNAL/REVUE DE DROITDE MCGILL
[Vol. 41
whether further changes are advisable with a view to increasing both fairness and
government revenue and more efficiently meeting the government’s stated policy
objectives. Some of the options before the Minister of Finance for reforming tax-
assisted retirement savings will then be reviewed. While the Minister did not effect
significant reforms in this area in the current budget, future cost-cutting measures
may well necessitate a closer look. This is particularly true given the government’s
commitment to “take action to reform Canada’s retirement income system on a
fairer and sustainable basis”.’ For the moment, suggestions for reform appear to be
focused on the Canada Pension Plan (“C.P.P.”), Old Age Security (“O.A.S.”) and
the Guaranteed Income Supplement (“G.I.S.”). The 1996 federal budget provides
for a new Seniors Benefit which will replace the latter two programmes commenc-
ing in 2001.”
As a preliminary matter, I will adopt the position of the Department of Finance
which regards tax-assisted retirement savings plans as a tax expenditure5 and will
not discuss the difficulties in defining a tax expenditure.” My limited objective is to
evaluate the current system in the context of stated federal government objectives.
I. The Present System
Canada’s retirement income system has three main components: (1) publicly
provided minimum-standards pensions;’ (2) publicly provided earnings-related
pensions;’ and (3) privately provided earnings-related plans which include em-
ployer-sponsored registered pension plans (“R.P.P.”s), deferred profit-sharing plans
(“D.P.S.P.”s) and R.R.S.P.s.
3 Department of Finance, “Budget in Brief” in Budget Papers (Ottawa: Department of Finance
Canada, 1995) at 13 [hereinafter Budget Papers 1995].
‘See Department of Finance, Budget Papers (Ottawa: Department of Finance Canada, 1996)
[hereinafter Budget Papers 1996].
5See Department of Finance, Personal and Corporate Income Tax Expenditures (Ottawa: Depart-
ment of Finance Canada, 1993) at 52 [hereinafter Income Tax Expenditures]. According to Bernard
Fortin, this is also the view of most specialists:
The view that RRSPs form an intrinsic part of the tax structure is not likely to be shared
by most specialists. For many of them, the main goals of RPPs and RRSPs are to pro-
vide an adequate level of income for people upon retirement and to stimulate retire-
ment savings. In this perspective, RRSPs should be considered as true tax expenditures
and the incidence of their net benefit becomes a relevant policy issue (B. Fortin,
“Comment” in N. Bruce, ed., Tax Expenditures and Government Policy: Seventh John
Deutsch Roundtable on Economic Policy (Kingston: Queen’s University Press, 1988)
368 at 371).
6 See N. Bruce, “Pathways to Tax Expenditures: A Survey of Conceptual Issues and Controversies”
in Bruce, ed., ibid, 21.
‘These are publicly provided pensions which are not related to earnings and currently consist of the
G.LS. and the O.A.S.
‘ These pensions are provided through the C.P.P.
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVNGS
The tax-assisted portions of the retirement income system (R.PPs, D.PS.Ps
and R.R.S.P.s) were reformed in 1991 after more than ten years of task forces,
studies, policy statements, press releases and draft legislation. Under the new rules,
the tax deductibility of contributions to R.P.P.s, D.P.S.P.s and R.R.S.P.s has been
integrated and standardized so that a comprehensive limit of eighteen percent of
earnings or a dollar maximum (whichever is lower) now applies to all forms of
registered retirement income plans.’
The limit on contributions is based on a conception of what an adequate retire-
ment income would be. It has generally been accepted that to maintain the same
pre-retirement standard of living after retirement an individual would have to re-
ceive approximately sixty to seventy percent of his or her pre-retirement income.
The eighteen percent limit is based on the “rule of nine” –
the assumption that, on
average, it is necessary to contribute nine dollars in the current year in order to ob-
tain one dollar of annual pension. Therefore, if one sets aside eighteen percent of
one’s income annually, one will receive a pension equal to two percent of earnings
for every year of service. A two percent pension earned over a career of thirty to
thirty-five years will replace an additional sixty to seventy percent of pre-retirement
earnings in addition to what the public system delivers.
Individuals who are not members of R.P.P.s or D.P.S.P.s can contribute eighteen
percent of the prior year’s earned income up to a specified dollar maximum to an
R.R.S.P. For R.P.P. and D.P.S.P. members the limit is reduced by the retirement
benefits accruing to their credit under those plans.'” The benefit accrual is used to
determine a “pension adjustment” figure (“P.A.”). The P.A. is subtracted from the
individual’s potential R.R.S.P. contribution room, as otherwise determined, in order
to calculate what he or she may contribute to an R.R.S.P.” The P.A. is also used to
limit the benefits that may be provided under R.P.P.s and D.P.S.P.s.’2
The standardization of contribution limits for the various tax-assisted retire-
ment plans has greatly decreased the disparity in access to tax assistance among
those in different employment situations. One of the main criticisms of the prior
system was the preferential tax treatment afforded to R.P.P.s which are available
only to those employed by others but not to the self-employed.’3 Furthermore,
R.P.P.s, in practice, were primarily offered only to those employees who worked in
the public sector or for large corporations. As discussed below, however, the cur-
rent system has not served to equalize access to tax assistance between low- and
high-income earners or between members of private-sector and public-sector plans.
D.P.S.Ps, s. 146(5) for R.R.S.Es [hereinafter I.T.A.].
9 See Income TaxAct, R.S.C. 1985 (5th Supp.), c. 1, ss. 147.1(8),147.1(9) for R.PPs, s. 147(5.1) for
“See ibidL at s. 146(5).
“See ibid. at s. 146(1) for the definition of “unused RRSP deduction room”.
,2See ibiL ats. 147(5.1) for D.P.S.Es, s. 147.1(8) for R.P.Ps.
“The self-employed generally include farmers, professionals and business-persons.
MCGILL LAW JOURNAL/REVUE DE DROITDE McGILL
[Vol. 41
The 1995 budget reduced the dollar limit on deductible R.R.S.P. contributions
from 14,500 dollars in 1995 to 13,500 dollars for 1996 and 1997. In the 1996
budget this limit was frozen at 13,500 dollars until 2003. The dollar limit on contri-
butions to defined contribution R.P.P.s was also reduced to 13,500 dollars for 1996
and then frozen. The dollar limit on contributions to D.P.S.P.s will continue to be
one-half the contribution limit for defined-contribution pension plans. The maxi-
mum pension limit for defined-benefit pension plans will be frozen at its current
level through 2004. The pension and D.P.S.P. limits will be indexed beginning in
2005, and the R.R.S.P. limit will be indexed beginning in 2006.”
Essentially, the government has retained the integration of the contribution lim-
its for the various tax-assisted retirement plans. This is consistent with the recom-
mendations of the Finance Committee.” The government, however, has also re-
tained its commitment to providing tax assistance for contributions based on earn-
ings up to two and one-half times the average wage. Therefore, while there may be
a delay in reaching the target maximum limits, the same targets remain. The result
is that the current changes to contribution limits amount to no more than short-term
tinkering that will not significantly increase tax revenue or fairness.
The government has also introduced certain adjustments that will make the tax-
assisted retirement savings system marginally more equitable. Pursuant to the 1995
federal budget, the rollover of retiring allowances to R.R.S.P.s will be gradually
eliminated. Perhaps more importantly, the over-contribution allowance of 8,000
dollars will be reduced to 2,000 dollars in 1996. This is the amount that may be
contributed to an R.R.S.P. in excess of the limits for deductible contributions before
a penalty tax becomes exigible.” In the past, financial advisors have suggested us-
ing this over-contribution room because of the ability to tax-shelter earnings, even
though the taxpayer receives no deduction for the contribution and will be required
to include the amount in income when it is withdrawn. The 1996 federal budget
proposes to eliminate the seven-year limit on the carry-forward of unused R.R.S.P.
room in recognition of the fact that many taxpayers find it difficult to contribute to
an R.R.S.P. in the early stages of their working lives. In addition, the age at which
individuals must mature their retirement savings is being reduced from seventy-one
to sixty-nine years of age.
” See: “Tax Assistance for Retirement Savings” in Budget Papers 1995, supra note 3; “Measures
Relating to Retirement Saving” in Budget Papers 1996, supra note 4 [hereinafter “Retirement Sav-
ing”].” See Canada’s Deficit Crisis, supra note 2 at 38.
16 See LT.A., supra note 9 at s. 204.2(1.1), which defines the cumulative excess amount in respect
of R.R.S.P.s. Where a taxpayer has an excess amount in his or her R.R.S.P, a penalty tax is payable.
A taxpayer who is at least 18 years of age, however, is permitted to contribute up to 8,000 dollars
more than his or her contribution room before it is considered that there is an excess amount in the
I-R.S.P.
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVNGS
The estimated cost of tax assistance through R.P.P.s and R.R.S.P.s in 1991 was
14.915 billion dollars at the federal level.’7 No estimates are available for the cost of
the tax expenditure on D.P.S.P.s. It should be noted that estimates for each tax ex-
penditure assume that there are no changes in other income-tax provisions, gov-
ernment policy, taxpayer behaviour or aggregate economic activity. Eliminating the
tax assistance to R.P.P.s and R.R.S.P.s, therefore, may not result in an increase in tax
revenue of approximately fifteen billion dollars, particularly if taxpayers could re-
direct their savings to another tax-assisted form. Furthermore, the R.R.S.P. tax ex-
penditure estimates may exhibit an upward bias because, due to the immaturity of
the system, contributions currently exceed withdrawals.’ On the other hand, be-
cause the contribution limits were raised significantly in 1991 and since the partici-
pation rates have been increasing, one might expect that the 1991 figures may un-
derstate the current tax expenditure.’9
It should also be noted that in comparing the cost of the tax expenditure esti-
mates to direct spending, a dollar of tax preference is often worth substantially
more to the individual than a dollar of direct spending since direct spending is
usually included in the recipient’s taxable income. Since all contributions and
earnings are eventually taxed, the value of the tax assistance is sometimes doubted.
The deferral of tax, however, does have a very real value to taxpayers. It also
causes governments to incur significant costs since tax deferral is equivalent to an
interest-free loan from the government in the amount of the tax otherwise payable.
In addition, further tax revenue is lost through deferrals to the extent that taxpayers
face lower tax rates in retirement than they do during their working lives.
II. The Policy Objectives of Tax-Assisted Retirement Savings
The government has established two basic policy objectives behind providing
tax assistance to retirement saving: (1) ensuring that all Canadians have adequate
retirement income,” and (2) encouraging increased private saving now to meet
pension needs in the future.2′ Ensuring adequate retirement income has generally
meant both guaranteeing a basic level of retirement income for all Canadians and
assisting them in avoiding serious disruption of their pre-retirement living stan-
dards. Maintaining pre-retirement living standards is generally considered to re-
quire sixty to seventy percent of earnings since income tax, personal savings needs
” See Department of Finance, Agenda: Jobs and Growth Creating a Healthy Fiscal Climate
(Ottawa: Department of Finance Canada, 1994) at 89 [hereinafter Jobs & Growth].
” See Income Tax Expenditures, supra note 5 at 53.
‘9 Empirical research has shown an increase of 14 percent in the number of R.R.S.P. contributors
and an increase of 30 percent in the amount of R.R.S.P. contributions in 1991 (see H. Frenken, “Note
on RRSP Contributions and Payouts” [1993] Persp. Lab. & Inc. 49).
” See Department of Finance, Budget Papers (Ottawa: Department of Finance Canada, 1984) at 18.
This objective was established under Finance Minister, Marc Lalonde.
21 See Department of Finance, News Release 89-132 (11 December 1989). This objective was es-
tablished under Finance Minister, Michael Wilson.
MCGILL LAWJOURNAL/REVUE DE DROITDE MCGILL
[Vol. 41
and job-related expenses decline at retirement.’ As discussed above, this is the ba-
sis for the current contribution limits for tax-assisted retirement saving.
The government’s policy objectives are premised on the following assump-
tions: (1) tax incentives encourage saving; (2) publicly assisted earnings-related re-
tirement plans are desirable; and (3) if the government fails to assist individuals to
save for their retirement, they will become a burden on the state. The last assump-
tion is based not only on the view that the state has an obligation to ensure its citi-
zens have a basic retirement income, but also on the view that those who are as-
sisted by the tax incentives are otherwise unlikely to have sufficient private retire-
ment income. In fact, however, those members of society who receive the most
benefit from tax-assisted retirement savings are those who are least likely to require
public support in retirement.’
Some have argued that a paternalistic government programme that encourages
or requires increased retirement savings is necessary to maximize individual wel-
fare because people often do not act in their own best interests in planning for their
retirement.’ This alleged inability to plan responsibly may be due, in part, to the
difficulties in determining one’s necessary level of current saving in order to meet
specific retirement goals –
a process that involves complex calculations and as-
sumptions as to inflation rates, interest rates, future income streams and needs. Psy-
chological factors including impulsiveness, death anxiety, inconsistent preferences
over time and situational preference changes may also affect the saving patterns of
individuals. While these factors may strongly militate in favour of both publicly run
compulsory savings plans and universal locking-in requirements,’
they do not, in
themselves, constitute a persuasive argument for earnings-related retirement subsi-
dies.
Another argument in favour of retaining the current system is that retirement
savings are a source of funds for capital investment, and that more saving generally
is needed in order to facilitate economic growth in Canada. This reasoning, how-
ever, rests on the questionable assumption that the increased savings would primar-
ily be invested to increase Canadian productive capacity.”6 Moreover, it is highly
‘ See K. Homer, “Policy Foundations of the New Tax Treatment of Retirement Savings” [1986]
‘ This is a design flaw of the current system. It is not fundamental to every possible tax-assisted re-
Can. Tax Found. 17:1 at 17:18.
tirement savings system.
2 See: D.M. Weiss, “Paternalistic Pension Policy: Psychological Evidence and Economic Theory”
(1991) 58 U. Chi. L. Rev. 1275; J. Bankman, “Tax Policy and Retirement Income: Are Pension Plan
Anti-Discrimination Provisions Desirable?” (1988) 55 U. Chi. L. Rev. 790.
25Funds held under R.PPs are, in general, locked-in until retirement pursuant to pension benefits
standards legislation. Generally, however, funds held under R.R.S.Rs and D.PS.P.s may be withdrawn
prior to retirement.
2 As discussed at text accompanying notes 48-49, below, the current foreign-content rules do not
prevent a large portion of retirement savings from being invested outside of Canada.
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVINGS
debatable whether tax incentives actually encourage saving or merely serve to
channel savings to tax-assisted vehicles.
M. Does the Tax-Assisted Retirement Savings System Meet its Policy Objectives?
A. Promoting Saving for Retirement
There are several arguments against using a tax-assisted retirement savings
system as a saving incentive: (1) tax incentives do not increase private retirement
savings; (2) there are sufficient other tax incentives for saving; (3) increased saving
is undesirable; and (4) fairness concerns. This last argument will be dealt with in
the next section.
1.
Do Tax Incentives Increase Private Retirement Savings?
There has been a great deal of debate among experts over whether tax incen-
tives increase savings or whether they merely serve to redirect savings to a tax-
assisted vehicle. There is no question that tax incentives for a certain type of saving
will attract funds to that type. The important measure for determining the success of
the incentive, however, is national savings which includes both public-sector and
private-sector savings. The question is: what impact does tax-assisted retirement
saving have on other forms of private saving and on government tax revenue.
Most economic studies in this area have focused solely on the effects of
R.R.S.P.s on saving. The conventional economic view is that these vehicles have
two conflicting effects on personal saving rates. The “substitution effect” dictates
that a tax incentive for saving makes current consumption more costly relative to
future consumption, with the result that an individual is likely to save more. The
“wealth (or income) effect”, however, recognizes that the tax incentive increases an
individual’s wealth.*7 The individual can now accumulate a target level of retire-
ment income at a lower rate of saving; he or she may set aside less money since the
government is also contributing. Increased wealth leads to increased consumption.
In addition, the loss of the tax revenue caused by the incentives leads to a reduction
in government savings (or an increase in the deficit).
According to Ragan, even the conventional view does not predict that R.R.S.P.s
will necessarily increase national saving.’ It predicts only that the substitution ef-
fect of the R.R.S.P. leads to more saving. Ragan argues that in a progressive in-
come-tax system, the substitution effect of R.R.S.P.s actually works to reduce the
“See generally C. Ragan, “Progressive Income Taxes and the Substitution Effect of RRSPs”
28See ibiL at 44.
(1994) 27 Can. J. Econ. 43.
MCGILL LAW JOURNAL/REVUE DEDROITDE MCGILL
[Vol. 41
level of personal saving. The result is that all the economic effects –
income and revenue- operate to reduce national saving.”
substitution,
In addition to the economic explanations, there are also psychological explana-
tions for why tax incentives will or should operate to increase saving. For instance,
individuals may save more because the advertising campaign around the annual
deadline for R.R.S.P. contributions may remind and encourage people to save. If
this is true, then increased saving may be the result not of the tax incentive, but,
rather, of advertising. Moreover, as Ragan has pointed out, psychological explana-
tions can go both ways. People may in fact save less because they fall to recognize
the future tax liabilities of their R.R.S.P. saving and, consequently, perceive them-
selves to be wealthier than they are.” Without any empirical evidence, such argu-
ments are not very persuasive.
Any impact of tax incentives on retirement saving appears to be dwarfed by the
non-tax reasons for retirement saving, particularly lifecycle influences. The analy-
sis of J.B. Burbidge and J.B. Davies indicates that participation in, and amounts
contributed to, R.R.S.P.s increased with age (up to normal retirement age) and with
income.” This is in line with the results of demographer David Foot’s research
which indicates that as Canada’s population ages, savings should increase: “Front-
end boomers entering their 40’s are now entering their prime savings years …
[I]ncreasingly, baby boomers are worried about savings, so they are accumulating
assets.” 2
In addition to the impact of lifecycle influences, there are other non-tax reasons
for rational employers and employees to establish pension plans. A defined-benefit
pension plan permits employers to reduce employee turnover and influence the age
of retirement. In defined-benefit plans, retirement income is generally a percentage
of earnings at the time of retirement or separation multiplied by years of service. If
an employee changes jobs, the benefits from the first employer will be based on the
salary at the time of separation. Thus, in calculating the employee’s retirement in-
come attributable to those years of employment, he or she will lose the benefit that
would have resulted from future wage increases. In a defined-benefit plan, there-
fore, the value of pension accruals typically increases steeply with age and years of
service.
Employees also receive certain advantages from pension plans independent of
tax savings. Some may have difficulty saving on their own and find the forced-
” See ibid at 56.
30See ibid
31See J.B. Burbidge & J.B. Davies, “Government Incentives and Household Saving in Canada” in
J.M. Poterba, ed., Public Policies and Household Savings (Chicago: University of Chicago Press,
1994)19.
N D. Marston, “The Great Boom and Bust Ahead –
an Interview with Demographer David Foot”
Cashing in … Tapped Out (Winter 1995) 5 at 6.
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVINGS
saving aspect of a pension plan attractive. Perhaps more importantly, employer-
sponsored plans generally earn a higher rate of return from investments due to their
professional management and because a higher percentage of their investments are
usually held in equities. Pension plans may also reduce the cost of annuities by
providing them at group rates. Furthermore, a defined-benefit plan gives the em-
ployee the security of a certain, known income stream during retirement.3
Theoretically, therefore, the effect of tax incentives on retirement saving is un-
certain. While empirical studies undertaken in the late 1980s seemed to indicate a
small increase in national savings, more recent work casts doubt on the proposition
that tax assistance for retirement savings results in any increase in national or per-
sonal savings?’ The net effect of tax assistance for retirement savings on total sav-
ings is, at best, ambiguous.
A recent U.S. study supports the view that tax incentives do not work to pro-
mote retirement saving?’ Similarly, an analysis of Canadian data suggests that very
few people save only in the form of an R.R.S.P. This seems to indicate that an in-
crease in contribution limits may have only wealth effects and no substitution ef-
” See D.I. Halperin, “Special Tax Treatment for Employee-Based Retirement Programs: Is it ‘Still’
Viable as a Means of Increasing Retirement Income? Should It Continue?” (1993) 49 Tex. L. Rev. 1
at 10.
‘4See e.g.: E. Duskin, “Changing the Mix of Public and Private Pensions: The Issues” in OECD,
Private Pensions and Public Policy (Social Policy Studies No. 9) (Paris: OECD, 1992) 1; Ragan, su-
pra note 27.
3 This is the conclusion that many scholars have reached, including: Duskin, who states, “Empirical
evidence on the savings effects of private pensions is scarce; what evidence exists is controversial”
(Duskin, ibid. at 19); R.M. Bird, D.B. Perry & T.A. Wilson, “Tax Reform in Canada: A Decade of
Change and Future Prospects” (Discussion Paper No. 1, International Centre for Tax Studies, Faculty
of Management, University of Toronto, 1994) [unpublished]:
It appears reasonable to conclude that the new system has generated greater contribu-
tions to RRSPs. Whether these contributions represent increased savings, or asset real-
locations cannot be readily determined, however. Published studies of the impact of the
RPP/RRSP system on savings show diverse results (Bird, Perry & Wilson, ibid. at 19).
M.E. Donnelly has also observed:
Despite the fact that the government appears to have wholeheartedly embraced the
point of view that increasing tax-assistance for retirement savings will encourage Ca-
nadians to save for their old age so that the government does not have to, the opinions
of experts are decidedly mixed as to whether or not any additional net savings will be
produced … there is sound empirical evidence to suggest that the presence of govern-
ment subsidies, whether delivered via the tax system or otherwise, likely brings about a
reduction in savings (M.E. Donnelly, ‘Tax-Assisted Retirement Savings for Women in
Canada – A Feminist Critique of Pension Reform” (LL.M. Thesis, University of To-
ronto, 1993) at 66 [hereinafter “Retirement Savings for Women”]).
Portions of the thesis were published in M.E. Donnelly, “The Disparate Impact of Pension Reform
on Women” (1993) 6 C.J.W.L. 419 [hereinafter “The Disparate Impact’].
6 See J.C. Gravelle, “Do Individual Retirement Accounts Increase Savings?” (1991) 5 J. Econ.
Pers. 133.
MCGILL LAW JOURNAL/REVUE DE DROITDE MCGILL
[Vol. 41
fects. While the same analysis revealed a strong gross correlation between tax in-
centives for saving and the personal saving rate, the authors warned that careful
empirical work would be required in order to accurately assess whether a causal
link exists between these two factors. 7
Furthermore, there appears to be a great deal of evidence to the effect that sav-
ings are generally unresponsive to tax rates. Over the last two decades there have
been significant changes in real interest rates, the rate of inflation and tax measures,
but “[t]he savings rates in most countries have remained relatively constant through
most of recent history.”38 This uncertainty over the effects of tax incentives on re-
tirement savings should, at the very least, lead the government to question the wis-
dom of spending such enormous sums on a programme with such uncertain results.
2.
Other Tax Incentives for Saving
The tax-assisted retirement savings system is only one of many tax incentives
for saving. Current contribution limits seem to reflect a belief that tax-assisted re-
tirement savings can only occur through R.P.P.s, R.R.S.P.s and D.P.S.P.s, whereas in
fact, any measure that subsidizes asset accumulation, in effect, subsidizes retire-
ment income.
The treatment of principal residences under the I.T.A. is a good case in point.
On disposition, principal residences are exempted from capital-gains tax and the
proceeds of disposition can be used to provide retirement income. 9 Moreover, the
imputed rent of owner-occupied homes is not income for I.T.A. purposes, The
value of both of these tax savings depends on the value of the home owned by the
taxpayer. Presumably, the greater one’s income, the more expensive a home one can
afford.
Burbidge and Davies estimate (using 1990 data) that one-half of personal net
worth is in the form of owner-occupied housing and consumer durables, neither of
which is taxable under the I.T.A. About one-third of personal financial assets are in
fully tax-sheltered forms such as R.R.S.P.s or private pensions. In all, the income on
approximately two-thirds of personal wealth is not taxable under the Canadian in-
come-tax system.”
Tax preferences are also given to capital gains, giving a greater benefit to those
who can purchase more expensive assets. Capital gains are taxed at the time of the
disposition of the asset, not as they accrue. The value of this deferral is equivalent
37 See Burbidge & Davies in Poterba, ed., supra note 31 at 55.
3 N. Brooks, The Canadian Goods and Services Tax: History, Policy, and Politics (Sydney: Aus-
tralian Tax Research Foundation, 1992) at 120. See also Brooks, ibl. at 121-24, for a general discus-
sion of the empirical research on the effect of taxes on private savings.
39 See I.T.A., supra note 9 at s. 40(2)(b).
, See Burbidge & Davies in Poterba, ed., supra note 31 at 25.
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVNGS
to that of an interest-free loan from the government in the amount of the tax that
would be payable. Upon disposition of the asset, only seventy-five percent of the
capital gain is included in income. Finally, until recently there was a 100,000 dollar
lifetime capital-gains exemption. There is still a 500,000 dollar capital-gains ex-
emption for “qualified small business corporation share[s]” and for “qualified farm
property”.”‘ While these tax preferences are not generally viewed as part of retire-
ment policy, their effect is similar to direct tax assistance for retirement saving in
that they act to subsidize asset accumulation which can be used to provide retire-
ment income.
3.
Should Private Saving Be Encouraged?
It is far from universally accepted that an increase in private savings is neces-
sarily beneficial for the economy. While a link is often suggested between increased
saving and higher real investment, this assumes that, at least in the long run, eco-
nomic resources are fully employed.” If an economy is in a recession, an increase
in personal saving represents a reduction in aggregate demand which may prolong
the recession.’3 For example, David Foot has suggested that increased saving has
prolonged the current recession: “This saving trend is also a drag on the economy.
The consumers that led us out of the ’81/’82 recession are not around to lead us out
of the ’91/’92 recession and that’s why people continue to experience tough
times.”
In pre-budget consultations, Michael McCracken suggested to the Finance
Committee that in attempting to reduce spending, the government should aim to
minimize the harm to the economy by making cuts that lead to the fewest job losses
and cause the least damage to hiring decisions, business investment, consumer ex-
penditure and the like. He stated that with this aim in mind the government will fo-
cus cuts on people who save rather than on those with low incomes.”
In summary, increased personal saving results in decreased demand for goods
and services and increased unemployment. Unemployment creates additional de-
mands on government revenues and restricts personal saving opportunities.’ Con-
“Both of these terms are defined in the I.T.A., supra note 9 at s. 110.6(1).
‘2 There is also another implicit assumption. As Brooks has pointed out,
what matters for economic growth is not household savings but total national savings.
Both corporations and the government also contribute to national savings. If the rate of
national savings is too low, it is not clear that household savings should be targeted for
an increase (Brooks, supra note 38 at 116).
,3 See J.E. Pesando, “The Economic Effects of Private Pensions” in OECD, supra note 34, 126.
Marston, supra note 32 at 6. S. Ingerman & R. Rowley, “Tax Expenditures and Retirement Sav-
ings” (1994) 2 Can. Bus. Econ. 46 at 48, have suggested that the 1991 reforms of the tax-assisted re-
tirement savings system may have deepened the 1990-1991 recession and weakened the subsequent
recovery.
“3 See Minutes of Proceedings and Evidence, supra note 1 at 59:26.
46 See “Retirement Savings for Women”, supra note 35 at 70.
MCGILL LAW JOURNAL/REvUE DEDROITDE MCGILL
[Vol. 41
versely, high levels of demand raise employment levels and, thereby, directly lower
poverty levels: ‘They also improve the fiscal position of governments and improve
the opportunities for expanding the current system of redistributive measures.”‘
Another argument against using government policy instruments to encourage
private saving is that the underlying assumption that increased savings will result in
increased domestic investment is flawed. This is no less true of retirement savings.
Up to twenty percent of a fund’s assets may be invested in foreign property before
any penalty is imposed.’ Recently, there has been a proliferation of investment ve-
hicles created solely to avoid the foreign-content restrictions. There are certain hy-
brid schemes and segregated funds, widely advertised by insurance companies and
investment dealers, which result in foreign investments being treated as wholly Ca-
nadian for the purposes of the penalty tax. Even savings that are invested in the
shares or debt of active Canadian companies do not necessarily result in increased
domestic productive capacity since the Canadian company may then use the funds
raised to carry on business or invest abroad.”
B. Ensuring Adequate Retirement Income and Saving the Public Purse
Many of those who made presentations to the Finance Committee advanced
dire predictions regarding the future financial obligation of the government to pro-
vide Canadians with an adequate retirement income should the present system of
tax-assisted retirement savings be altered.” Such predictions are based on assump-
tions both about the cost of the tax expenditure relative to future costs and about
those who benefit from tax-assisted retirement saving. It is implied that those who
benefit now are those who otherwise would likely become eligible for direct gov-
ernment assistance during their retirement.
“‘Economic Council of Canada, The New Face of Poverty: Income Security Needs of Canadian
Families (Ottawa: Supply & Services Canada, 1992) at 11.
” See I.T.A., supra note 9 at s. 206(2). The 20 percent figure refers to the cost amount of the prop-
erty rather than to its current value.
9This is true only if certain conditions are met (see I.T.A., ibid. at s. 206(l)(d.1), proposed ss.
206(d.1), 206(1.1)).
50 For example, Christine Hollett, Executive Director of the Advisory Council on the Economy,
warned that by taxing pension plans the government may be gaining revenue today at the expense of
tomorrow. She stated that if the provisions are changed,
you’ll find a major reduction in the amount of moneys people are saving for their re-
tirement. When the working population of today approaches retirement tomorrow or
next year or twenty years down the road, you’ll face a major financial drain on your
purse at that point (Minutes of Proceedings and Evidence, supra note 1 at 63:39).
See further statements to the same effect from: Martin D. Salloum, General Manager of the Edmon-
ton Chamber of Commerce (see ibid at 66:22); Michael White, representing the Employer Commit-
tee on Health Care (see ibid at 66:86-87); Pierre Caron, representing the Retirement Savings Alliance
(composed of the four largest pension consulting firms) (see ibid at 73:35); Leo-Paul Landry, Chair
of the RRSP Alliance (see ibid at 85:5); George Anderson, President of the Insurance Bureau of Can-
ada (see ibid at 65:91); Robert Schultz, Chair of the Investment Dealers Association of Canada (see
ibid at 79:6).
19961
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVNGS
The design of the tax incentive leads to inefficient and inequitable results. A
deduction from taxable income gives the greatest incentive to save to those who
need it least, namely, high-income earners who have the ability to accumulate sig-
nificant assets before retirement and to take advantage of other tax incentives for
saving. They are probably not potential candidates for future income-tested gov-
emnment programmes. Those who are likely to need assistance in obtaining ade-
quate retirement income –
receive little benefit from this
costly government programme; a deduction is worthless to those who have no tax-
able income.’
low-income Canadians –
As with all deductions, the value of a deduction for a contribution to a tax-
assisted retirement savings plan increases as the taxpayer’s marginal tax rate in-
creases. Higher-income taxpayers receive proportionately more assistance for any
given amount of retirement saving than a taxpayer whose marginal tax rate is below
the maximum. For example, assume that there are three individuals who each con-
tribute one hundred dollars to an R.R.S.P. The high earner who has a marginal tax
rate of twenty-nine percent will receive a subsidy of twenty-nine dollars. The tax-
payer who has a marginal tax rate of seventeen percent will receive a subsidy of
seventeen dollars while the individual with no taxable income will receive no assis-
tance whatsoever. Furthermore, since the level of permissible contributions in-
creases with the level of earned income, the high-income eamer is also able to con-
tribute larger dollar amounts and, therefore, able to receive a larger benefit. In ad-
dition, the proportion of income that is available for saving or discretionary spend-
ing increases as income increases. High-income earners are, therefore, more likely
to be in a position to contribute.
It should be stressed that the major tax benefit arising from the retirement sav-
ing incentives is not the deduction, but the sheltering, of investment income. High-
income earners will also benefit to a much greater extent from the tax deferral on
investment earnings for the same reasons outlined above. Since they would other-
wise be paying tax on the income at a higher marginal rate, a greater amount of tax
is deferred. Furthermore, because they have the legal and financial capacity to con-
tribute a greater amount in absolute terms, they will also benefit from the deferral
of tax on a larger amount of investment income.
This theoretical analysis may be verified empirically by an analysis of Revenue
Canada data on annual contributions to R.R.S.P.s for the 1992 tax year. 2 Not sur-
” The regressivity of tax deductions and the failure of either deductions or credits (unless refund-
able) to help those without any taxable income are two common flaws with the tax-expenditures
method of programme delivery. A subsidy can be more easily and directly targeted.
*2 See Revenue Canada, Taxation Statistics (Ottawa: Supply & Services Canada, 1994) [hereinafter
Taxation Statistics 1994]. Some have argued that the lifetime of a taxpayer is a less arbitrary period
than a year for the purpose of measuring tax incidence including the incidence of tax expenditures.
Davies’s study concludes that the incidence of net benefits from R.R.S.P. saving over a lifetime is ap-
proximately as regressive as the incidence reported in annual studies (see J.B. Davies, “Incidence of
MCGILL LAWJOURNAL/REVUE DE DROITDE MCGILL
[Vol. 41
prisingly, both the percentage of tax-filers who contribute to an R.R.S.P. and the
average contribution increase substantially as income rises. While only ten percent
of tax-filers reporting income of less than 25,000 dollars contributed to an R.R.S.P.
in 1992, over seventy-two percent of those earning over 100,000 dollars contrib-
uted to an R.R.S.P. Of those who did contribute, the average contribution for low-
income earners (under 25,000 dollars) was 1,565.86 dollars, whereas the average
contribution for high-income earners (over 100,000 dollars) was 9,560.97 dollars.
The average tax-filer earning over 100,000 dollars contributes forty-four times as
much to an R.R.S.P. as the average tax-filer earning under 25,000 dollars. When the
different marginal income-tax rates of these two groups are factored in, the average
benefit received by high-income earners is seventy-five times that received by low-
income earners. Furthermore, these numbers fail to take into account those indi-
viduals who do not file a tax return. 3
R.R.S.P. Contributions for 1992 by Income”
Income
Under $25,000
$25,000 – $50,000
$50,000 – $70,000
$70,000 – $100,000
Over $100,000
Percentage
of all
tax-filers
Percentage of tax-
filers who contributed
to an R.R.S.P
Average
contribution of
tax-filers
61.5
27.9
7.0
2.2
1.3
10.0
42.8
60.9
70.0
72.1
$156.39
$1,119.80
$2,384.19
$4,191.41
$6,890.06
Average
contribution of
contributors
$1,565.86
$2,613.96
$3,918.03
$5,990.45
$9,560.97
The Ontario Fair Tax Commission has pointed out that what is considered to be
an adequate retirement income for tax-assisted retirement savings purposes is much
higher than what is considered to be an adequate retirement income for the pur-
poses of other forms of government retirement assistance. The C.P.P. covers only
twenty-five percent of the Year’s Maximum Pensionable Earnings, which is
equivalent to the average wage. By contrast, the earnings corresponding to the
maximum contribution to tax-assisted retirement savings plans is approximately
two and one-half times the average wage.” O.A.S. benefits are clawed back if the
recipient has an income over, approximately, 55,000 dollars. The Ontario Fair Tax
Tax Expenditures in a Lifetime Framework: Theory and an Application to RRSPs” in Bruce, ed., su-
pra note 5, 339 at 366).
53 Pursuant to the I.T.A., supra note 9 at s. 150, an individual who has not disposed of capital prop-
erty need not file a return for a year in which no tax is payable. While many individuals with no in-
come would still file a return in order to claim the refundable tax credits, not all do. It has been esti-
mated that as many as 15 percent of the lowest-income families do not apply for refundable tax cred-
its (see Brooks, supra note 38 at 95).
‘ This chart is derived from data contained in Taxation Statistics 1994, supra note 52 at table 2, pp.
54-61.
“5 The federal government affirmed its commitment to maintaining this ratio in the 1995 Federal
Budget.
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVINGS
Commission recommended that the maximum retirement benefit eligible for tax
assistance be reduced to one and one-half times the industrial wage after which
contribution limits should be indexed to maintain the ratio.”
Tax assistance for retirement savings tends to reinforce current economic ine-
qualities. Therefore, it should not be surprising that women benefit to a much lesser
extent from such assistance.’ Women are less likely to have the jobs or the dispos-
able income that enables adequate retirement savings. In addition, women are less
likely to belong to a pension plan because they are disproportionately represented
in low-wage, intermittent and part-time work. A pension model that assumes a
thirty-five year stint in the paid labour force does not reflect the experience of most
women. The present system, coupled with women’s economic inequality, has
translated directly into lower subsidies for retirement saving for women even
though the poverty rate of elderly women in Canada is double that of elderly men.”
The Women and Taxation Working Group of the Ontario Fair Tax Commission
concluded that the current system of tax-assisted retirement savings results in sys-
temic discrimination against women. 9
A programme that gives no assistance to those most likely to rely on the gov-
ernment to provide them with a retirement income in the future and that, at great
cost, primarily benefits those who are highly unlikely to require government subsi-
dies to obtain a basic retirement income, can never be seen as a good policy. In
these days of fiscal restraint and overwhelming public concern for the size of the
government debt and social spending cuts, the continued existence of such a pro-
gramme is not only wasteful but obscene.’
56 See Ontario Fair Tax Commission, Fair Taxation in a Changing World: Report of the Ontario
Fair Tax Commission (Toronto: University of Toronto Press, 1993) at 331-32 [hereinafter Fair Taxa-
tion].57For instance, Taxation Statistics 1994, supra note 52 at table 4, p. 107, shows that in 1992,
women made up 42 percent of R.R.S.P contributors, but that only 34 percent of R.R.S.P contribu-
tions were made by women.
58 See “The Disparate Impact”, supra note 35 at 421.
59 See: Ontario Fair Tax Commission, Women and Taxation (Working Group Report) (Toronto: Fair
Tax Commission, 1992) at 22; Minutes of Proceedings and Evidence, supra note 1 at 91:34 (M.
Jackman, National Association of Women and the Law).
60 It should be noted that R.R.S.P funds can be withdrawn at any time. Therefore, there is no guar-
antee that they will be used for retirement. In fact,
[d]ata for 1990 indicate that two-fifths of total RRSP income was reported as cash
withdrawals by tax-filers under 55 years of age … another fifth was paid to those aged
55-64-of which 90 per cent was cash withdrawals! Fully three-quarters of RRSP in-
come was in the form of cash rather than payments from annuities or a Registered Re-
tirement Income Fund, … a proportion which is hardly suitable for any normal pension
programme (Ingerman & Rowley, supra note 44 at 52).
An obvious and simple way to improve the efficiency of the system would be to require that
R.R.S.P funds, like pension funds, be locked-in until retirement. Of course, such a requirement would
not detract from any of the criticisms mentioned above.
MCGILL LAW JOURNAL/REVUE DEDROITDE MCGILL
[Vol. 41
IV. Redesigning the System
Numerous possibilities have been suggested for redesigning the current system
with a view to making it more efficient, more equitable and less costly. In evaluat-
ing these alternatives the following general principles should be kept in mind: (1)
tax treatment should be equitable between private-sector and public-sector plans;
(2) tax treatment should be equitable among different types of plans; and (3) tax
treatment should be equitable between high- and low-income earners. In order to
meet the policy objective of saving the public purse from a future drain on revenue,
the system should be designed to provide a greater benefit to low-income earners.
There are three main transactions that provide possible occasions for the taxa-
tion of retirement income: (1) contributions; (2) income derived from the invest-
ment of contributions; and (3) the payment or withdrawal of retirement benefits
from the accumulated fund.
A. Reducing the Contribution Limits
Reducing contribution limits, either by reducing the maximum dollar amount,
the percentage of earnings that may be contributed or both, was the suggestion for
reform most often made to the Finance Committee” and was adopted to some ex-
tent by the government in the 1995 and 1996 federal budgets. As noted above, the
Ontario Fair Tax Commission made the same recommendation. ‘2 The fairness and
efficiency of the system could be improved somewhat by reducing the maximum
dollar amount to a level that most members of the middle class could afford,
thereby eliminating contribution room which is only available to those with very
high incomes. The Confederation of National Trade Unions (“C.N.T.U.”) has sug-
gested that an appropriate amount would be 7,000 dollars or 8,000 dollars.6′ Under
the current rules, these figures correspond to annual incomes of approximately
38,000 dollars to 45,000 dollars. An income of over 80,000 dollars would be re-
quired in order to reach the maximum contribution for 1995 of 14,500 dollars under
the present rules.
Reducing the percentage of earned income that may be contributed would make
it more difficult for low-income earners to reach the dollar maximum and, there-
fore, to receive a benefit similar to that accessible to high-income earners. While
raising the percentage of earned income that may be contributed may be worth
considering, it would be difficult, in reality, for a low-income earner to save more
6, It was suggested by more than a dozen groups including: the Canadian Real Estate Association
(see Minutes of Proceedings and Evidence, supra note 1 at 77:4); the Child Poverty Action Group
(see ibid. at 82:3); and the Confederation of National Trade Unions (see ibid, at 62:12).
6 See Fair Taxation, supra note 56 at 332. The Commission also noted that the limits have effec-
tively been reduced in real terms over the years due to the effects of inflation (see ibid. at 332).
6′ See Minutes of Proceedings and Evidence, supra note 1 at 62:26 (P. Paquette, Secretary General,
C.N.T.U.).
19961
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVINGS
than eighteen percent of earned income. Such a move, therefore, might only serve
to enhance the tax-assisted savings opportunities of those with significant invest-
ment income and low earned income. Essentially, the percentage should approxi-
mate that which low- or low- and middle-income earners can realistically contribute
taking into account lifecycle saving patterns.
One way of making the system more progressive and of increasing tax reve-
nues would be to restrict the deduction as income increases. The Carter Comnis-
sion suggested that limits should be raised for low-income earners and significantly
restricted for high-income earners.”
There is also the issue of maintaining equity between public-sector and private-
sector pension plans. Not all public-sector plans are R.P.P.s due both to the govern-
ment’s tax-exempt status and because a large number of government plans are un-
funded. In order to maintain equity it would be necessary to calculate a P.A. for
each member of a non-registered public-sector plan and use it to reduce his or her
R.R.S.P. contribution room. The last time the current system was reformed, the De-
partment of Finance proposed calculating pension credits in respect of certain un-
registered retirement plans maintained by tax-exempt employers so as to reduce the
R.R.S.P. contribution room of those who are entitled to pension benefits under these
plans.” While the proposed regulations are still not law, a P.A. is currently calcu-
lated in respect of certain non-registered public-sector arrangements.
B. Eliminating All Deductions for Contributions
Another option would be to eliminate the deduction for contributions to tax-
assisted retirement savings plans altogether.” The question then arises as to how to
tax employer contributions to R.P.P.s and D.P.S.P.s. For defined-contribution plans
and D.P.S.P.s it is easy and reasonable to include the amount of the employer’s
contributions in the employee’s income for the year as a taxable benefit. In a de-
fined-benefit plan, however, there may be little connection between current em-
ployer contributions and the increase in an employee’s wealth. To resolve this diffi-
culty, the P.A., which is a method of measuring the accrual of pension benefits,
could be used; the portion of the P.A. that relates to employer contributions could
be included in the employee’s income without adding significant complexity to the
current I.T.A. provisions. In order to maintain equity between public-sector and
&’See R. Krelove & S.A. Rea, Jr., “Private Pensions and the Tax System: Twenty Years After the
Carter Commission” in N. Brooks, ed., The Quest for Tax Reform (Toronto: Carswell, 1988) 121 at
127. The Carter Commission also suggested that the maximum pension limit be independent of
earnings and that withdrawals from R.S.P.s prior to retirement be limited. Excess withdrawals
would be subject to a penalty (see ibid. at 127).
“See proposed regulations 8308.3 ind 8308.4, which were part of the December 18, 1992 draft
regulations.
6I note that this would not result in the elimination of contribution limits as the taxation of the
funds’ income would still be deferred. Therefore, this could be combined with changes to the contri-
bution limits.
MCGILL LAW JOURNAL! REVUE DE DROIT DE MCGILL
[Vol. 41
private-sector pension plan members, P.A.s should be calculated for all members of
public-sector plans, and the portion relating to the government’s contribution
should, likewise, be included in each member’s income.
Liquidity may become a problem if the amount of an employer’s contribution is
included in an employee’s income. The employee will be taxed on an amount that is
not available to help pay the tax. In practice, this should not be a problem since
there would likely be a withholding requirement and since the income tax conse-
quences would probably be taken into account in the negotiation of future employ-
ment contracts.
Finally, supporters of the current system may argue that eliminating the deduc-
tion for contributions would eliminate or greatly reduce the savings incentive. As
discussed above, it is highly debatable whether tax incentives work to increase
savings at all. To the extent that people contribute to their R.R.S.Ps more for the
current deduction than for the sheltering of investment income, there may be some
argument for taxing the fund instead. Practically speaking, however, financial advi-
sors commonly suggest using the 8,000 dollars over-contribution room as a shelter,
and a large number of individuals with sufficient financial resources to access it
have taken their advice even though they not only receive no deduction for the
contribution, but will have to include the amount as income when it is withdrawn.
This seems to indicate that removing the deduction for contributions would not
greatly affect the savings incentive.
If contributions to tax-assisted retirement savings plans are not deductible, then
it is inappropriate to tax the entire amount of the pension benefit when it is re-
ceived, since tax will already have been paid with respect to a portion of that
amount. If the income of the funds is not taxed, it has been suggested that the full
exemption from tax of the pension benefit would result in a similar impact on tax
revenue as the current system, although the timing of the tax payments would be
advanced.’ In order to raise more revenue and improve equity there would have to
be a system that distinguished between underlying contributions and income. One
suggestion is to track the amount of contributions. If the entire amount in the
R.R.S.P. were withdrawn at one time, determining the amount of income would be
straightforward. If, as is more likely to happen, benefits are received in the form of
an annuity or similar partial payment or withdrawal in any one year, the income
portion could be determined using a formula similar to that used to determine the
capital element of annuity payments.’
67See A. Dilnot, “Taxation of Private Pensions: Costs and Consequences” in OECD, supra note 34,
72.
” See I.T.A., supra note 9 at s. 60(a). It is interesting to note the relatively new Australian system
where all tax-deductible contributions to pension plans are subject to a 15 percent tax when received
by a fund. Investment income and capital gains are taxed at 15 percent, but capital gains are taxed af-
ter adjustment for inflation. Pension benefits are taxed at the individual’s marginal rate less 15 per-
cent. While this system has the effect of reducing the size of the budget deficit by bringing tax reve-
1996]
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVINGS
C. Conversion to a Tax Credit
Converting the current deduction to a tax credit at the lowest marginal income-
tax rate (currently seventeen percent) would increase the progressivity of the sys-
tem since all taxpayers with sufficient taxable income would receive the same tax
benefit per dollar of contribution. Those with no taxable income would continue to
receive no benefit from the tax-assisted retirement savings system. Replacing the
deduction with a tax credit would also be less costly for government.
This approach was recommended by, among others, the Ontario Fair Tax
Commission, who also recommended that withdrawals should continue to be taxed
as ordinary income. ‘ It was also the second most popular suggestion for reforming
the tax-assisted retirement savings system which was presented to the Finance
Committee in its pre-budget consultations.”
If a tax credit were used, it would require including in income an amount with
respect to employer contributions to R.P.P.s and D.P.S.P.s, as discussed above. Fur-
thermore, a tax credit would give rise to an increased possibility of double taxation,
assuming that pension benefits and R.R.S.P. withdrawals would continue to be
taxed at the individual’s marginal rate. Such a risk would only arise, however, if the
individual’s marginal tax rate is higher than seventeen percent both at the time of
contribution and at the time of withdrawal or receipt of the pension benefit.7′ Since
individuals who fit this criterion are those who least need-the subsidy, it is perhaps
not unfair to reduce their benefits. The fear that the use of a tax credit might reduce
the savings incentive for high-income earners appears to be unfounded since many
individuals consider it to be in their interest to use the over-contribution room de-
spite the much larger penalty involved.
D. Taxing Fund Investment Income
It is somewhat interesting to note that only one of the hundreds of groups that
appeared before the Finance Committee suggested that the investment income of
retirement funds be taxed.’ It is unclear whether this proposal is unpopular, even
among groups that advocate reform of the system, due to perceived technical or
political problems. A superficial review of public opinion in the media indicates
nues forward, it is no more progressive than our current system (see D.M. Knox, A Review of the Op-
tions for Taxing Superannuation (Sydney: Australian Tax Research Foundation, 1990) at 43).
9 See Fair Taxation, supra note 56 at 332.
70 The representatives of at least seven groups suggested this option including the Child Poverty
Action Group (see Minutes of Proceedings and Evidence, supra note 1 at 82:32-33), the Hamilton
Chamber of Commerce (see ibiL at 72:99) and the B.C. Federation of Labour (see ibid. at 64:66).
” Even under the current system theie is no guarantee that an individual’s marginal rate will not be
higher at the time of withdrawal or receipt of the pension benefit either due to increases in the indi-
vidual’s income or due to general increases in tax rates.
” The Ecumenical Coalition for Economic Justice suggested a modest tax on investment income
(see Minutes of Proceedings and Evidence, supra note I at 72:18).
McGL LAw JOURNAL/REVUE DE DROITDE McGILL[
[Vol. 41
that any taxation of pension funds would be even more unpopular than decreasing
the contribution limits or replacing the deduction with a tax credit.
One advantage of taxing the fund earnings rather than contributions is the reve-
nue potential. It has been estimated that in 1992 the value of pension and R.R.S.P.
funds in Canada was approximately 440 billion dollars.’ The Department of Fi-
nance estimated that in 1991 the non-taxation of investment income in R.R.S.P.s
and R.P.P.s cost the government almost twelve billion dollars. No numbers were
available with respect to the revenue lost through the non-taxation of investment
income in D.P.S.P.s. ‘
Furthermore, the tax exemption may lead to inefficiencies in the capital mar-
kets as different tax rates for different investors may affect their investment prefer-
ences. Particular transactions may become profitable due to the presence of tax-
exempt investors. Such transactions may merely represent a transfer of taxable in-
come to a tax-exempt investor. In addition, the tax structure offers an incentive for
tax-exempt investors to favour debt over equity since dividends are paid from after-
tax profits, whereas interest is paid from pre-tax profits. This preference can have a
significant effect on government revenue since the I.T.A. permits a deduction for
interest paid on the assumption that it will be taxable in the hands of the recipient.”
There are two basic ways of taxing investment earnings. Either the fund itself
could be taxed directly, or the earnings could be attributed to the relevant individu-
als and taxed at their marginal rates. Problems arise with each of these approaches.
It is impossible to apply a single tax rate to retirement funds without either main-
taining the increased tax benefits to high-income earners or penalizing low-income
earners for participating. In addition, it may be argued that a direct tax on retire-
ment funds would increase inequities between public-sector and private-sector
plans. The result of the tax on private-sector plans would likely be, at least in the
long run, a reduction of the benefit level provided. Since the public sector is tax-
exempt and since many of the plans are, in any event, unfunded, there would be no
similar pressure to reduce the benefits provided under public-sector plans.
There do not appear to be any significant conceptual or administrative prob-
lems in attributing investment earnings to individuals with respect to defined-
contribution R.P.P.s, D.P.S.P.s or R.R.S.P.s. For defined-benefit R.P.P.s, it would be
necessary to include an amount with respect to the benefit accrual rather than the
investment income, since the investment income, like a contribution, is not directly
related to the ultimate benefit received by any particular plan member. This would
be similar to the P.A. calculation. The issue of equity between public-sector and
‘ See Ingerman & Rowley, supra note 44 at 50. This figure includes the accumulated assets held
under R.P.s and R.R.S.Ps but excludes assets held under D.PS.Ps and in government-consolidated
revenue arrangements.
” See Budget Papers 1996, supra note 4 at 18.
7
5 See Dilnot in OECD, supra note 34.
19961
B. AUSTIN – TAX-ASSISTED RETIREMENT SAVNGS
private-sector plans could be resolved by requiring a similar calculation and inclu-
sion for members of public-sector plans. The major problem with taxing investment
income at the level of the individual is one of liquidity; quite simply, the individual
beneficiaries may not have sufficient funds to pay the taxes as they become due.
The problem is particularly acute with respect to R.P.P.s since pre-retirement with-
drawals are prohibited.
E. Other Assorted Proposals
It is beyond the scope of this paper to exhaustively explore every possible re-
form. For the sake of completeness, however, I will list some other potential re-
forms:
(1) A limit could be placed on the total amount of tax-assisted savings, that
is, the value of all R.R.S.P.s held and all pension benefits. Once this limit is
reached, the taxpayer would no longer be entitled to any deductions for
contributions, and the investment earnings would no longer be tax-
exempt.’
(2) The age at which R.R.S.P.s must be terminated was reduced from sev-
enty-one years of age to sixty-nine years of age in the 1996 federal
budget;’ it could be further reduced to sixty-five years of age so that tax
could be collected earlier. Other tax deferral systems, such as registered
retirement income funds, could be similarly restricted. This would be un-
likely to affect any saving incentive.’
(3) R.R.S.P.s could be subject to a locking-in requirement. This change
might not have much of a current effect on revenue but would ensure that
the funds are used to provide a retirement income.”
(4) The 8,000 dollar over-contribution limit that was reduced to 2,000 dol-
lars in the 1995 federal budget could be totally eliminated, especially since
the I.T.A. permits a deduction for withdrawals (to offset the income inclu-
sion) where an over-contribution has mistakenly occurred.’ It is unlikely
that an over-contribution would be made by mistake since Revenue Can-
ada informs each taxpayer of his or her personal contribution limit for the
year. This might not significantly affect revenue but would increase pro-
gressivity.
7See Minutes of Proceedings and Evidence, supra note 1 at 73:46. This was suggested by the
Chair of the Finance Committee who pointed out that currently one could have millions in an R.R.S.P.
(with good investment performance) and still have the earnings sheltered from taxation.
See “Retirement Saving”, supra note 14.
78See Minutes of Proceedings and Evidence, supra note 1 at 69:14 (D. Burrell).
“Among others, this was recommended by the Ontario Fair Tax Commission (see Fair Taxation,
supra note 56 at 329).
eo See I.T.A., supra note 9 at s. 146(8.2).
MCGILL LAW JOURNAL/REVUE DE DROIT DE MCGILL
[Vol. 41
(5) R.R.S.P.s and pension funds could be required to invest a certain per-
centage of their assets in special government bonds. Assuming that the
government bonds would not carry a market interest rate, this would, in ef-
fect, be a tax on the funds. For the reasons outlined above, this would not
be a progressive measure. Furthermore, restricting investment in this man-
ner might decrease the attractiveness of tax-assisted retirement savings
plans even more than a direct tax since it may increase uncertainty about
future government intentions.
(6) The current investment restrictions could be revised so as to permit less of
the funds to be invested outside Canada. At present, the foreign-content rules
permit twenty percent of a fund to be placed in foreign investments.” This per-
centage could be reduced so as to require greater investment in Canada.
Conclusion
The purpose of this paper was to determine whether further changes to the tax-
assisted retirement savings system are advisable in order both to increase fairness
and government revenue and to meet the government’s stated retirement policy
objectives more efficiently and effectively than the current system seems able to do.
My conclusion is a resounding “yes”. It is highly questionable whether the current
tax-assisted retirement savings system meets its policy objectives of helping to en-
sure that all Canadians have adequate retirement income and encouraging increased
private saving now to meet future pension needs. It is absolutely clear, however,
that the system does not accomplish these policy purposes in either the most effi-
cient or the most equitable way possible. Thus, the answers to the questions posed
by the Minister of Finance in the quote cited at the beginning of this paper would
suggest that major reforms of tax-assisted retirement savings should be pursued.
Given the minor tinkering to the system in the 1995 budget, which was, generally,
the toughest federal budget in decades, and the focus on publicly provided pensions
in the 1996 budget, however, it may be unrealistic to expect such reforms in the
foreseeable future.
, See ibkL at Part 11. As discussed at text accompanying notes 48-49, above, the current foreign-
content rules may be inadequate to encourage investment in Canada.
