This Web page has been archived on the Web.

1994 Report of the Auditor General of Canada

Assistant Auditor General: Shahid Minto
Responsible Auditors: Barry Elkin and Michael Adibe

Main Points

33.2 In 1992, approximately 7 million Canadian tax filers received tax assistance for retirement savings as contributors to, or members of, either registered pension plans (RPPs) or registered retirement savings plans (RRSPs), or both. About 5.2 million people and their employers contributed $19.7 billion to RPPs, while about 4.8 million individuals contributed $16 billion to RRSPs. In 1992, withdrawals amounted to $19 billion from RPPs and $4.7 billion from RRSPs, and the assets accumulated in plans were $417 billion.

33.3 Accountability for the costs and results of tax assistance for retirement savings needs to be improved. The Department of Finance needs to state clearly the objectives of the TARS program, and to improve monitoring and reporting of its costs, workings and results annually in order to foster scrutiny and evaluation. It also needs to develop alternative cost estimates to assist policy makers in assessing the long-term fiscal implications of TARS. The Department should ensure that its evaluations address key results and other issues of major concern.

33.4 While Revenue Canada has adequate procedures for reviewing and approving contributions to defined benefit plans and for verifying pension adjustments, its monitoring of registered plans' compliance with the Income Tax Act needs to be strengthened. The current compliance strategy needs to be reviewed and revised, as appropriate. The Department needs to develop, keep, report and use summary information on the nature, extent and revenue implications of identified non-compliance. It needs to document compliance profiles for registered plans. It also needs to increase the level of audit coverage of registered plans, and implement procedures to verify RRSP contributions, transfers between plans, and transactions of deferred profit sharing plans. Finally, in collaboration with the Department of Finance, Revenue Canada needs to establish effective penalties for cases of non-compliance where revocation of registration is considered inappropriate.

33.5 Both Finance and Revenue Canada need to improve the information provided to Parliament on TARS. We are concerned about inadequate documentation to support the cost estimates that Finance provided to a parliamentary committee in 1990. Finance needs to ensure that the results of its evaluation studies are provided to Parliament.

Introduction

Using the tax system to provide assistance

Tax assistance for retirement savings costs the federal government about $15 billion annually in deferred tax revenue

33.8 The TARS program is delivered through deferral of tax on savings in four types of plans that are registered by Revenue Canada: registered pension plans (RPPs), registered retirement savings plans (RRSPs), deferred profit sharing plans (DPSPs) and registered retirement income funds (RRIFs). Tax assistance for retirement savings is provided through provisions of the Income Tax Act that make contributions to these plans by individuals and employers deductible from taxable income, within defined limits. The investment earnings on assets in the plans are also tax-exempt until funds are withdrawn from the plans, but funds withdrawn from the plans are taxed. Individuals benefit because the tax on money contributed to the plans and accumulating in them is deferred until that money is withdrawn from the plans.

33.9 Tax assistance for retirement savings is a long-standing feature of the Canadian income tax system. Over the years, related tax legislation and regulations have undergone many revisions. The changes culminated in 1990 with the passage of Bill C-52, an Act that amended the Income Tax Act and brought in a uniform, comprehensive system of limits for all types of pension and retirement savings plans (see Exhibit 33.1 for historical background).

33.10 A key objective of that reform was to correct serious inequities in the system, which provided taxpayers at the same level of earned income with quite different opportunities to save for retirement. As reported by the Department of Finance, those inequities left many Canadians without adequate opportunities to build tax-assisted pensions, while others could obtain excessive tax benefits. The aim of the reform was to provide all Canadians at the same level of earned income with equal opportunities to provide for their retirement, regardless of the type of retirement savings plan they participate in, and whether they are employed or self-employed.

33.11 The government has estimated that the amount of deferred federal tax revenue it did not collect in 1991 because of the TARS program was about $15 billion, or 2.2 percent of gross domestic product. This figure excludes provincial tax expenditures for retirement savings.

33.12 Tax assistance for retirement savings is the single largest tax expenditure, and is a major component of Canada's retirement income system. It represented over 30 percent of federal direct expenditures and tax expenditures on retirement income security programs in 1991 (see Exhibits 33.2 and 33.3 ).

33.13 In 1992, approximately 7 million Canadian tax filers received tax assistance for retirement savings as contributors to, or members of, either RPPs or RRSPs, or both. About 5.2 million people (38 percent of the labour force or 47 percent of paid employees) and their employers on their behalf contributed $19.7 billion to registered pension plans, while about 4.8 million individuals (26 percent of all tax filers) contributed $16 billion to registered retirement savings plans. In 1992, withdrawals amounted to $19 billion from RPPs and $4.7 billion from RRSPs. The assets accumulated in those plans, including assets and reserves held by insurance companies for pension plans, amounted to $417 billion: $270 billion in RPPs and $147 billion in RRSPs. These exclude some assets in self-directed RRSPs, such as investments in bonds and stocks.

33.14 Canada's system of tax assistance for retirement savings is similar to those of other countries in the Organization for Economic Co-operation and Development (OECD), notably France and the United Kingdom. However, the Canadian pension system differs from most OECD countries in the strong role it gives to individual retirement savings arrangements. As well, Canada's system is unique in the way it integrates contribution limits for all types of pension and retirement savings plans (see Exhibit 33.4 ).

Why we chose to audit now

• the substantial amount of tax expenditures involved;
• Parliament's interest in the program as reflected, for example, in the 1990 recommendation of the House of Commons Standing Committee on Finance that a review of the program be undertaken after the new system (effective January 1991) had been in place for three years, and the Committee's request that Finance study certain issues related to the program and report back before the end of 1990. The Standing Committee on Public Accounts also asked Finance to evaluate the reform of TARS;
• the relative novelty and the complexity of the present regime, which requires an effective compliance strategy to manage the increase in compliance monitoring workload that resulted from the change to the new system; and
• the concerns about the program expressed by some employer groups, the pension industry and others. Those concerns include remaining and new inequities in the system, the results, costs and continued affordability of the program, the burden of compliance cost on plan sponsors, and the substantial and growing use of RRSP funds for purposes other than retirement income.

33.17 Our audit was directed at Finance and Revenue Canada, the two departments directly involved.

How the report is structured

DEPARTMENT OF FINANCE

The Role of the Department of Finance in TARS

Audit Scope and Objectives

• examining the manner in which Finance monitors, evaluates and reports on the results and costs of this significant tax expenditure program; and
• providing objective information to help Parliament scrutinize the costs and results of the program.

Observations and Recommendations

Need for Clear Statement of TARS Objectives

Objectives are not specific enough to allow their achievement to be assessed

33.23 There is no compendium statement of objectives for the TARS program at the disposal of the Department of Finance, other policy makers or the public. To identify the objectives, intent and rationale of the program, and the public perception of those objectives, we reviewed many documents, including academic journals and newspaper articles.

33.24 The objectives are not stated clearly. We are concerned that the objectives are not specific enough to allow their achievement to be assessed. Following are two examples that illustrate the reasons for our concern.

33.25 A common objective. Some of the objectives of tax assistance for retirement savings complement those of other retirement income programs -- Canada Pension Plan/Quebec Pension Plan, Old Age Security and Guaranteed Income Supplement. For example, the goals of income replacement and of helping Canadians to avoid serious disruption in their standard of living upon retirement are not unique to the TARS program, but are pursued as well by those public pension plans and retirement income security programs and by other tax-favoured investments (capital gains exemption on principal residences). For proper evaluation, it is necessary that the program objectives for TARS clarify its role relative to other retirement income programs.

33.26 As well, because the TARS program is used in combination with other programs to achieve those objectives, evaluation of their achievement needs to be conducted interdepartmentally, for example, with the Department of Human Resources Development. In our 1993 Report we observed that providing parliamentarians with the information they need on the $30 billion spent annually on seniors' pensions and the billions spent on related tax measures would require a sectoral analysis of the programs as a whole.

33.27 The target population is not evident. Public statements of the objectives suggest that the TARS program is available to those with taxable earnings who choose to contribute to plans. However, the target population is unclear. If the target population of the TARS program were specified clearly, it would be easier to assess whether the program indeed benefits that population.

Need for Better Estimates and Reporting of Costs

The cost of tax assistance for retirement savings has not been reported on a regular, consistent and timely basis

33.29 Since 1979, Finance has published estimates of tax expenditures for retirement savings intermittently along with estimates of other tax expenditures. This has left large gaps between publication dates: the four years between 1980 and 1985, and the six years between 1985 and 1992. During those periods, no updated estimates of tax expenditures for retirement savings were presented to Parliament or the public.

33.30 As well, the continuity of cost estimates was interrupted because estimates were not calculated for the period 1984 to 1987 but were resumed for the period 1988 to 1991.

33.31 The consistency and comparability of the historical estimates prepared by Finance were affected significantly by a major change in methodology, introduced in 1985. The present series were not linked to the earlier ones, nor were the previous estimates revised using the current methodology to permit comparability and trend analysis.

33.32 The timeliness of Finance's estimates has also not been adequate or consistent. There have been substantial lags in the calculation of estimates, caused partly by delays in the availability of data. Except for the earlier series, estimates have usually been published two or three years after the year to which the latest numbers relate. Thus, estimates for 1979 and 1980 were published in those years, whereas the latest year presented in the 1985 publication was 1983. In the 1992 publication, the latest year was 1989; in the 1993 publication, the latest year was 1991.

Cash-flow cost estimates need to be supplemented with longer-term, present value cost estimates

33.34 Since 1985, Finance has used an annual cash flow method to estimate the tax expenditure costs for TARS. Beginning with the 1992 Government of Canada Personal Income Tax Expenditures report, the basis of the cash flow estimates was revised. The cash flow approach estimates the tax revenue that the government did not collect in a given year because of the TARS provision in the income tax system. The approach has a number of advantages and disadvantages. Among the advantages, the estimates are easier to calculate because past data on contributions, investment income and withdrawals that are used to make the calculations are readily available. The estimates do not require projections of future trends in such variables as interest rate, inflation rate and take-up rates. The estimates are comparable to non-deferral tax expenditure costs. The fiscal impact of the tax expenditure in any specific year is readily discernible; and estimates are responsive in a specific year to changes in flow variables, such as contribution limits, that have significant impacts on tax revenues.

33.35 The major limitation is that such estimates do not represent the "true cost". They do not reflect the long-term costs of the program, since they account for only the immediate costs of annual contributions, tax deferral of interest earnings of plan assets, and the tax recovered on withdrawals in the current year.

33.36 Another accounting basis that Finance could use to make its cost estimates is the present value (discounted) approach. The advantage of this approach is that all current year and future costs (calculated in constant dollar terms) arising from an activity undertaken in any given year can be provided to help evaluate the long-term impact on tax revenue. One drawback is that the present value approach requires making various discounting assumptions and projections of future flows and parameters, such as the average length of time the contributions will remain in the pension plan, their future earnings rate, the future tax rate and the discount rate. These projections introduce a substantial element of uncertainty into the estimates.

33.37 To assist policy makers in assessing both the current and long-term fiscal implications of the TARS program, Finance needs to develop and publish, in addition to the cash flow estimates, cost estimates based on the present value approach or on some other method that would measure long-term costs or aggregate lifetime costs.

Need to Improve Performance Measurement and Program Evaluation

Performance of the TARS program is not monitored and reported regularly

33.39 In our opinion, starting with that information, the Department of Finance can develop a set of key performance indicators that it could monitor and report, ideally, on an annual basis. Exhibit 33.5 presents examples of performance indicators that Finance could consider in selecting those for ongoing monitoring and regular reporting. Some of the indicators are already being reported by, or are available from, Revenue Canada and Statistics Canada. We have also developed some questions to aid in developing and refining the indicators and for evaluating the results and impacts of the program. These are presented in Exhibit 33.6 .

Monitoring of results needs to be supplemented by periodic evaluations

Evaluation studies do not address key results and other issues of major concern

33.42 The Department has done no additional formal evaluation studies of tax assistance for retirement savings. However, it has done other studies, such as a 1979 task force report, The Retirement Income System in Canada: Problems and Alternative Policies for Reform (the Lazar Report); the 1982 Green Paper, Better Pensions for Canadians ; and, in 1984, Building Better Pensions for Canadians: Improved TARS . Other federal studies include the 1966 Royal Commission on Taxation , the 1979 Report of the Special Senate Committee on Retirement Age Policies and the 1983 Parliamentary Task Force on Pension Reform. The studies made several recommendations that helped to shape the present system of tax assistance for retirement savings.

33.43 In response to a request by the Public Accounts Committee, Finance prepared, in October 1992, an evaluation assessment of the reform of tax assistance for retirement savings. The evaluation assessment report included a framework for evaluating the reform of the limits on tax assistance for retirement savings, and identified nine issues for further study.

33.44 We found that progress in conducting the evaluations has been very slow. The summary plan prepared in 1992 scheduled four studies for completion in 1993 and the remaining five in 1993-1994. At the time of our audit, Finance officials informed us that work had started on three of the studies and that at least two should be completed by the end of 1994. They expect all of the nine studies to be completed by late 1996. A revised plan for the studies was to be reviewed and approved by the Department's Tax Evaluation Advisory Committee by the fall of 1994.

33.45 Our review of the plan shows that, although the evaluation studies are directed at evaluating the reform - as requested by the Public Accounts Committee - they will not address several key effects of the system. For example, the program's impact on pension income will not be examined, because it will take some time before changes in levels of pension income become apparent. While it is true that the impacts of reform will take some time to be manifest, this should not be an excuse for not evaluating the program's past performance.

33.46 Both the 1989 evaluation of the incidence of pension plans and public assistance to the elderly and the evaluation studies planned by Finance leave many fundamental concerns unaddressed. We noted that the aim of one of the studies planned by Finance is to identify areas of concern among employer and employee groups and, at a later stage, to provide more systematic analysis of those areas.

33.47 Exhibit 33.7 presents examples of issues of concern and related questions that we believe need to be addressed in future evaluation studies. Two of those issues, affordability and the burden of compliance cost, are discussed below.

33.48 Burden of compliance cost on plan sponsors and administrators needs to be assessed. A common concern among many plan sponsors is the cost to them of

compliance. Finance has not assessed this cost burden since the 1990 reform. Plan sponsors continue to express concern about rising costs imposed on them, not only by the Income Tax Act and Regulations but also by the provincial Pension Benefit Standards Acts . This escalating compliance cost is cited as a major reason for some plan sponsors shifting from defined benefit plans to defined contribution plans, and to group RRSPs as an alternative to traditional RPPs.

33.49 An objective identified in Finance's 1992 evaluation assessment was to examine compliance costs to employers. The Department intends to cover the compliance cost issue in one of its planned nine evaluation studies, through consultations with employer and employee groups.

33.50 Affordability and fiscal implications need to be examined. We would expect Finance to assess the costs of the existing system as well as the potential costs of alternative systems of tax assistance for retirement savings, and their implications for the deficit and the public debt in the context of the government's overall fiscal objectives. We saw no evidence to suggest that this has been done or is planned.

33.51 To date, Finance has not undertaken estimates of the long-term tax expenditure cost of the TARS program that take into account the fact that tax deferrals ultimately will be recoverable, at least in part, when funds are withdrawn from RPPs and RRSPs. Once the plans have reached the stage of maturity when contributions equal withdrawals, the ongoing tax expenditure will arise largely from deferral of tax on interest earned by the assets of the plans. As a result, long-term cost estimates may be significantly different from the costs implied by the short-term estimating techniques used by Finance. Officials of that Department informed us that they are examining the feasibility of developing a long-term model to capture the deferral aspects of the program, as part of the policy review promised in the February 1994 Budget, Preparing For An Aging Population .

Information to Parliament Needs to Be Improved

• its own responsibilities regarding TARS and the program's effectiveness;
• consolidated summary-level information on the cost and operation of the complete TARS program including that part carried out by Revenue Canada; and
• estimates for past years and projections for future years for TARS.

33.54 In our 1992 chapter on departmental reporting, we noted that there was little information in Finance's Part III on its stewardship responsibilities for major areas such as tax expenditures, or references to where it can be found. In its 1994 Part III, Finance included a reference to Personal and Corporate Income Tax Expenditures and listed other studies and analyses of existing tax measures undertaken over the previous year, but did not provide brief summaries of the findings.

33.55 We believe that Finance needs to include in its Part III, or in any other regularly published accountability document, an overview of TARS and other significant tax expenditures. This entails notifying Parliament when evaluations of specific tax expenditures are completed, providing a brief summary of the evaluation findings and disclosing where the detailed report can be obtained.

33.56 For example, in 1989 Finance did an evaluation on the incidence of pension plans and public assistance to the elderly. However, the report was not referred to in Finance's Part III or made available to Parliament and the public during the debate on the legislation for pension reform. Yet that report declares that it was intended to stimulate public and parliamentary discussion.

33.57 Finance officials have informed us that the Department has now established a standard procedure for disseminating information on evaluation studies it completes. These procedures include issuing a press release on the study and sending copies to the Standing Committee on Public Accounts and to each member of Parliament.

33.58 In 1990 the Standing Committee on Finance asked the Department of Finance to study the following issues and report back before the end of that year:

• the availability of alternative forms of retirement income;
• the possibility of changing tax deductions for retirement savings into tax credits; and
• the desirability of measures that would narrow the differences between registered pension plans and registered retirement savings plans (for example, RRSPs are not subject to many of the restrictions imposed on RPPs, such as locking in and survivor benefits).

Documentation was inadequate for a cost estimate provided to Parliament

33.61 In a 1990 presentation to the House of Commons Finance Committee, the Department of Finance said that the tax revenue it would not collect in 1989-90 because of the TARS program was $5.5 billion. Three years later, in December 1993, the Department published a significantly different estimate of $12.5 billion in the Government of Canada Personal Income Tax Expenditures . Finance officials have advised us that the difference between the two estimates was due to a change in methodology. However, the Department was unable to provide us with adequate documentation to support the earlier estimate. Consequently, we are unable to provide any assurance to Parliament on the validity of the cost estimate of $5.5 billion presented to the Committee.

33.62 Finance also presented to the Finance Committee the projected cost of the 1990 reform of the TARS program. The Department concluded that the reform's net impact on revenue would be neutral. It estimated that, for fiscal 1991-92, there would be additional foregone revenue of $300 million to $350 million. At the same time, it estimated that this cost would be offset by revenue savings resulting from measures that included stopping unlimited salary deferrals in employee benefit plans, removing the deductibility of past service additional voluntary contributions, restricting transfers between plans and the rollover of pension payments into RRSPs, and preventing a major increase in RRSP limits through a continuing shift to non-contributory plans.

33.63 Current estimates of tax expenditures suggest that net cost to revenue increased by nearly $1 billion in 1991. However, Finance has no information on the portion of that increase that is attributable to the reform, or on some variables, such as direct transfers between plans, that will be necessary to verify the offsets. The Department plans to evaluate its estimate that the net impact of reform on revenue would be neutral in one of the evaluation studies it is undertaking for completion in 1996. The objective of the study will be to obtain improved estimates of the impact of reform on tax revenues.

Differences with Provincial Pension Legislation Are Being Resolved

33.65 In the course of our audit, pension experts brought to our attention certain areas of difference between the federal legislation and provincial legislation. Under the auspices of CAPSA, some of the differences - for example, reductions in benefits and refund of contributions - have been resolved, and amendments made to the Income Tax Act . Other issues, such as the equal periodic payments of pension and the treatment of eligible contributions under money purchase plans, are being addressed respectively through amendments to the legislation being drafted by Finance and discussions at CAPSA meetings.

33.66 Finance officials advised us that the Department has been working with Revenue Canada and CAPSA members to identify, assess and resolve potential areas of conflict. We encourage both Finance and Revenue Canada to continue to find ways to resolve those and any other differences.

Conclusion

Accountability to Parliament for TARS is inadequate

33.68 We have concluded that the accountability for the costs and results of the substantial tax assistance for retirement savings program needs to be improved. The objectives of the TARS program need to be clearly specified. Consolidated information on the cost, workings and results of the program, and on the roles and responsibilities of departments involved, is not regularly provided to Parliament. Evaluation studies of TARS by Finance need to address key results and other issues of major concern.

33.69 As the Department responsible for the policy and program design and the management of the TARS program, the Department of Finance should:

a) clearly state the objectives of the TARS program and clarify its role relative to other retirement income programs;

b) monitor and regularly account for the program by providing consolidated information on the estimated costs, results and workings in Part III of the Estimates or in any other regularly published accountability document;

c) develop, in addition to estimates of tax expenditures for retirement savings made on a cash flow basis, alternative cost estimates using a present value method and other approaches to assist policy makers in assessing the longer-term fiscal implications of the TARS program;

d) complete the nine evaluation studies planned, and carry out further studies to address other key effects and major concerns;

e) establish an interdepartmental evaluation framework to ensure that TARS is evaluated in conjunction with other related public pension and retirement income security programs;

f) estimate and assess, in its further evaluation studies, the long-term costs and related fiscal implications of alternative systems of TARS; and

g) make reports on its evaluation studies available to Parliament to contribute to informed discussion, scrutiny and decisions about the TARS program.

Department's response:

a0 The Department of Finance believes that the objectives of tax assistance for retirement savings, and its role relative to other retirement income programs, have been clearly outlined in public documents. A recent example is the booklet "Pension Reform: Improvements in Tax Assistance for Retirement Saving", which was tabled in Parliament by the Minister of Finance on December 11, 1989. The objectives and role of tax assistance for retirement savings are described on pages 5 and 6 of that booklet.

b) The Department will consider expanding the scope of its ``Personal and Corporate Income Tax Expenditures" document to meet this recommendation.

c) Cost estimates using the present value approach will be calculated and published in future editions of "Personal and Corporate Income Tax Expenditures".

d) Once the current evaluations of Tax Assistance for Retirement Savings are complete, consideration will be given to further evaluations such as those recommended by the Auditor General.

e) This approach will be examined in developing a possible evaluation framework for future evaluations of TARS.

f) This recommendation will be taken into account in developing any future evaluation framework for TARS.

g) For several years now, it has been the Department's policy that all tax evaluation studies are made public and, in particular, that they are provided to parliamentarians. This will continue to be the case.

REVENUE CANADA

The Role of Revenue Canada in TARS

33.71 Registration of plans. The registration of pension, retirement savings and deferred profit sharing plans is a prerequisite for claiming tax assistance for retirement savings. The registration process entails a front-end review of applications for plan registration and for amendments, to ensure that the plan texts comply with prescribed conditions and requirements. After an initial review of an application for registration, a plan is "deemed registered" if the application complies with the "prescribed manner" specified in the Income Tax Act . This proviso is intended to avoid holding up the operation of plans while they await full compliance review for final registration.

33.72 Registration of retirement savings plans is controlled by the Registered Plans Division of the Policy and Legislation Branch of Revenue Canada. The Division has a staff of about 115 persons. It provides a large number of publications and a national enquiries service to facilitate compliance by clients. It also approves eligible contributions to defined benefit pension plans, with assistance from the Office of the Superintendent of Financial Institutions.

33.73 About 75 of the Division's staff of 115, or 65 percent, are engaged in the front-end review work, the results of which are reported primarily in terms of submissions processed. The Division estimated that, as at 31 March 1994, a total of 33,143 plans were registered, broken down as shown in Exhibit 33.8 .

33.74 The Division's records show that in 1993-94 it processed submissions for initial registration of 2,845 RPPs, approval of 19,918 RPP amendments, and approval of eligible contributions to 6,194 defined benefit RPPs. It processed submissions for approval, amendment or termination of 4,961 RRSPs, 1,571 DPSPs and 1,159 RRIFs, and answered 29,648 telephone enquiries and 3,874 written enquiries. The Division has taken initiatives to promote and facilitate compliance by publishing forms, guides, information circulars and newsletters. It also meets semi-annually with pension industry representatives to consult on their needs and opinions.

33.75 Monitoring compliance. The monitoring process comprises primarily audits: to verify that the plans continue to qualify for registration, to determine the validity of income tax deductions for contributions claimed by taxpayers, and to determine that withdrawals from, and transfers between, registered plans have been dealt with properly. The Registered Plans Division manages compliance programs designed to identify plans that do not comply with the terms of registration. Five of its staff of 115 provide functional direction for back-end compliance work. The Verification, Enforcement and Compliance Research Branch is responsible for auditing deductions claimed by taxpayers for contributions to and withdrawals from registered plans, and for auditing transfers between plans.

Audit Scope and Objectives

• determining if the rules and procedures applied have been sufficient to assure an effective check on the right of registered plans to continue to qualify for registration; on the validity of the income tax deductions claimed by taxpayers; and on the proper tax treatment of withdrawals from, and transfers between, registered plans; and
• providing objective information to help Parliament scrutinize the administration of the program.

Observations and Recommendations

Need to Improve Monitoring of Compliance

Results of the front-end compliance review need to be summarized

33.79 Officials of the Division advised us that they discuss non-compliance at their management meetings. However, a review of the Division's quarterly and annual reports for 1993 and 1994 showed that it does not systematically maintain, analyze and report summary information on the results of its front-end compliance review, in terms of the nature, frequency and implications of non-compliance identified and forestalled. Such summary information would be useful for developing compliance profiles and appropriate compliance monitoring strategy.

There are delays and backlogs in registering plans and approving amendments

33.81 Delays and backlogs are increasing. As at 31 December 1993, 36 percent of the case files were more than six months old, and at 31 March 1994, 40 percent were more than six months old. Some 390 files received between 1986 and 1991 still had not been completed at 31 December 1993. The backlog of plans awaiting approval for registration or amendment increased by 48 percent from 1991-92 to 1992-93 and by 41 percent from 1992-93 to 1993-94.

33.82 The delays have frustrated plan sponsors and administrators. As well, the delays have increased the risk that plans not registered and those not yet amended might be operating while not in full compliance. At the time of our audit, there was no information on the age of some 1,500 plans deemed registered, although some plans have been operating under "deemed registered" status for two years or longer. Delays of two years in registering and amending plans cause difficulties and expense to employers and plan sponsors, and they create business risks for pension plan developers.

33.83 Factors contributing to the delays and backlog include the increases in workload and the complexity of the legislation that have led to a significant increase in requests for more information and explanations. Departmental officials advised us that delays are sometimes caused by clients themselves, who may take nine to 12 months to reply to initial letters of request for further information. The causes of delays also include a lack of adequate and timely information for clients to facilitate their compliance. As well, the Department lacks formal standards for assessing its service performance. The Department recognizes these problems and has initiated steps to address them. For example, it has adopted a modified review process and is working on the development of service standards.

33.84 Service standards. The Registered Plans Division proposed standards for processing applications, responding to enquiries, and providing other services to clients. It conducted a survey of clients, and consulted managers to obtain their views and suggestions on standards.

33.85 There was a big gap between client expectations and the proposed standards of service. For example, while the Division's proposed service standards are four to six months for registration of new plans, and six to nine months for amendments and for terminations, the majority of clients surveyed expect these transactions to be completed within one month - an expectation the Department considers unrealistic. In its continuing effort to improve services, the Division subsequently proposed goals for quicker service to narrow that gap. However, the standards have not yet been published. The Department has stated that it will publish the standards in February 1995. The establishment and publication of service standards, and the monitoring and reporting of service performance against published standards, would provide an important basis for planning and managing reduction in delays and backlog.

Post-registration audit of plans and verification of transactions need to be improved

33.87 Client compliance profiles are not formally maintained to target plans for audit. Revenue Canada does not maintain compliance profiles of registered plans to help it assess the risk of non-compliance and to target plans for audit. Officials of the Department informed us that they have personal knowledge of compliance profiles of registered plans and areas of risk. However, their knowledge is not systematically documented, maintained and summarized to help select plans for audit, to provide continuity in the Department's corporate memory, and to capture changes in risk profiles.

33.88 In addition, we note that annual information returns and tax forms do not provide sufficient information to select plans for audit. Officials advised us that they plan to redesign annual information returns to address the problem.

33.89 Summary information on the nature, frequency and revenue implications of non-compliance identified through front-end compliance review, as well as through post-registration regular audits, is not maintained to help develop compliance profiles.

33.90 Without compliance profiles and sufficient information on annual information returns and tax forms, the Department's ability to identify and audit cases of potentially significant non-compliance is hindered.

33.91 Audit coverage is inadequate to monitor pressure points in the system. We identified areas where significant non-compliance could occur (referred to as "pressure points") and reviewed Revenue Canada's audit program and other procedures to ascertain whether they include the necessary steps to assure compliance in those areas. Some of the pressure points are shown in Exhibit 33.9 .

33.92 Revenue Canada's audit programs have steps to monitor these pressure points and to monitor contributions to, and investments of, pension plans to ensure that they conform with the requirements of the legislation. However, the low level of audit coverage inhibits any effective and in-depth review of those areas (see paragraphs 33.93 to 33.97). In the circumstances, regular audits need to be supplemented with special projects to determine the nature and magnitude of non-compliance in those areas. The Registered Plans Division has carried out such special projects to examine terminated pension plans for excess transfers; it also examined executive pension plans to determine if contributions were contravening an anti-avoidance rule.

33.93 The number and coverage of audits concluded in 1992-93 and 1993-94 are summarized in Exhibit 33.10 .

33.94 Revenue Canada concluded 23 regular audits of registered pension plans in each of the years 1992-93 and 1993-94. On an annual basis, the regular audits represent about 0.1 percent of the 22,536 registered pension plans. In 1992-93, a total of 127 plans were audited, out of some 33,000 deferred income plans of all types. In 1993-94, a total of 82 regular audits were concluded plus 26 special project audits.

33.95 Of particular concern is the extremely low audit coverage of registered pension plans (RPPs). At 31 March 1994, these plans represented 68 percent of all plans registered with Revenue Canada. Yet the 23 regular audits of registered pension plans

represented only about 18 percent of all regular audits of plans in 1992-93, and 28 percent in 1993-94.

33.96 Moreover, although public sector plans accounted for 48 percent of pension plan membership at 1 January 1992, no audits of public sector plans have been undertaken to date.

33.97 In our opinion, audit coverage is not sufficient to detect and deter non-compliance. In our 1990 audit of Revenue Canada's enforcement of the Income Tax Act , we also expressed concern about the low level of audit coverage for various activities.

33.98 The review and approval of contributions to defined benefit plans is reasonable. For an employer's contribution to a defined benefit plan to be deductible from income, it must be ruled an eligible contribution or must be made on the recommendation of an actuary. The contribution must be approved by the Minister of National Revenue, in writing, on the advice of the Office of the Superintendent of Financial Institutions.

33.99 With the assistance of actuarial experts, we reviewed the procedures for assessing the eligibility of contributions to defined benefit plans. We found them to be reasonable. As no actuarial calculations are required for defined contribution plans, these plans are not subject to the same sort of review as defined benefit plans.

33.100 Verification of pension adjustment is adequate. Each deferred profit sharing plan and each registered pension plan produces a pension credit for the member. The pension credit is a measure of the value of the benefit earned or accrued during the calendar year. A member's pension adjustment (PA) is the total of that member's pension credits from all plans in which the member's employer participates in the year, excluding registered retirement savings plans.

33.101 It is the responsibility of the employer to calculate a pension adjustment for each plan member. The employer must report the pension adjustment for each employee to Revenue Canada on the T4 or T4A information slip by the last day of February each year.

33.102 We examined Revenue Canada's procedures for verifying employers' calculations of pension adjustments and found them to be reasonable.

33.103 Past service pension adjustments (PSPA) are not systematically verified. In addition to the benefit earned by a member for a current year (reflected in the member's pension adjustment), pension benefits may improve as a result of events related to past service. These "past service events" occur for periods of past service after 1989 where benefits are increased retroactively, where an additional period of past service is credited to the member, or where there is a retroactive change in the way the member's benefits are determined.

33.104 Administrators of defined benefit plans are responsible for calculating and reporting PSPAs when necessary. Reporting of PSPAs did not start until 1992 because of delays in promulgating the Income Tax Regulations. To June 1994, Revenue Canada had processed $187 million in PSPAs.

33.105 Officials informed us that the Department does not verify plan administrators' calculations of PSPAs systematically.

33.106 Because the calculation of PSPAs is not verified, it is not known if they are being calculated and reported in accordance with the Income Tax Regulations. Any understatement of PSPAs could result in contributions to RRSPs that exceed the limit, since PSPAs reduce the amount that taxpayers can deduct for contributions to an RRSP.

33.107 Lack of information to verify RRSP contributions. In 1992, $16 billion was contributed to RRSPs. However, there is no requirement for issuers of RRSPs, mainly financial institutions, to submit information to Revenue Canada showing contributions made to RRSPs by purchasers. Thus, no matching is possible as is done for pension plan payments and registered plan withdrawals, which require that information slips be submitted. Consequently, the Department has virtually no information from an independent source to determine the validity of deductions claimed by taxpayers, or to determine if there are significant contributions over the permissible cumulative limit of $8,000 on "over-contribution". In effect, the Department has limited means in place to ensure compliance with Income Tax Act rules for contributions to registered retirement savings plans.

33.108 The problem of lack of information for verifying RRSP contributions has been further affected by the introduction of electronic filing (EFILE). Taxpayers using EFILE are not required initially to submit any supporting documentation. In our 1993 audit of Revenue Canada's electronic filing of individual income tax returns, we observed that there could be significant potential for revenue loss if taxpayers' voluntary compliance deteriorates.

33.109 Deferred profit sharing plan contributions and withdrawals are not known. Revenue Canada does not have the information on the amounts of contributions to and withdrawals from DPSPs needed to effectively plan the audit of those transactions.

33.110 Validation of withdrawals is inadequate. In 1992, withdrawals in the form of periodic payments and transfers amounted to $19 billion from RPPs, $4.7 billion from RRSPs and $1.8 billion from registered retirement income funds.

33.111 The Verification, Enforcement and Compliance Research Branch audits withdrawals from registered plans. Information on tax slips (T4A, T4RSP, T4RIF) showing withdrawals from registered plans is taken as accurate in the initial assessment of returns. When doing its matching programs or verifying supporting information on tax returns, the Department does not question the amounts on those slips. Responsibility for reviewing their accuracy rests with the Registered Plans Division. The Division verifies validity as part of its audit of registered plans. However, as we have noted, audit coverage is extremely low.

33.112 Legislation no longer requires reporting of information on direct transfers between plans. Many annual direct transfers between plans are not reported to Revenue Canada because, under the new legislation, there is no longer a requirement to do so. The Income Tax Act sets limits on transfers from RPPs to other vehicles. Without a requirement to provide information on the amounts transferred, the Department has no information to verify if the limits on direct transfers have been exceeded and excess amounts transferred without being taxed. The Department did identify excess contributions in a review of terminated pension plans for direct transfers from RPPs to RRSPs.

33.113 Revocation needs to be supplemented with other sanctions for non-compliance. Sanctions and penalties for non-compliance by plans are stipulated in the Income Tax Act . They include monetary penalties, revocation of registration, summary conviction and fines. For example, the legislation imposes monetary penalties for failure to meet statutory deadlines, while revocation applies to registered plans that do not comply with the prescribed conditions for registration or that are not administered in accordance with their terms as registered, or for other violations set out in ss. 147.1 (11) of the Act.

33.114 Revocation is used, usually in rare cases that involve flagrant non-compliance and where revocation does not prejudice employee beneficiaries. Revenue Canada asked Finance in 1993 to develop alternative sanctions, primarily monetary penalties, against employers and plan administrators who fail to bring a plan into compliance with the conditions of registration for cases where revocation is considered inappropriate. Finance is still considering that request.

33.115 Compliance monitoring strategy has not been reviewed. Since the 1990 reform, the Registered Plans Division has not carried out a review of its approach to compliance monitoring. It has continued to apply its front-end compliance review strategy, with little emphasis either on back-end audits of registered plans or on determining the validity of transactions - contributions, withdrawals and transfers.

33.116 Officials of the Department informed us that, following the 1990 reform to the legislation, they made a conscious decision to continue using the same front-end compliance review strategy. They maintain that the legislation dictates the present approach. However, no study or review of that strategy has been carried out to compare it with alternative strategies or approaches for achieving the same results.

Information to Parliament Needs to Be Improved

33.118 However, it does not report the results of its compliance activities, such as the number of plans found to be non-compliant and the associated implications for revenue loss. Also, there are no cross-references to other departmental sources of information - for example, Taxation Statistics , which provides the number of contributors and the amount of contributions to pension and retirement savings plans.

Co-ordination with Provinces

33.120 For example, the administrative actions of the Régie des rentes du Québec, the Pension Commission of Ontario and those of Revenue Canada are similar in such areas as the registration of pension plans, examination of plan documents and review of actuarial valuation reports. Departmental officials informed us that they initiated a joint project to study the feasibility of harmonizing the federal and provincial annual information returns, so that there would be one return required instead of two. The project is nearing completion. Ontario intends to adopt registration numbers for pension plans that are the same as those of Revenue Canada to facilitate communication between the two levels of government, and other provinces have expressed interest in following suit.

33.121 Other potential areas for co-ordination of administrative actions include compliance reviews, monitoring of pension plans and the maintenance of a central filing system containing all official pension plan documents.

33.122 As Revenue Canada and the provincial pension authorities examine essentially the same pension documents, they could co-ordinate their activities so that one jurisdiction could have the other carry out the examination on its behalf, thereby improving overall efficiency and achieving cost savings. We encourage the Department to continue to seek, through the Canadian Association of Pension Supervisory Authorities, ways of making the administrative actions of the various jurisdictions compatible, to reduce duplication and overlapping of operations.

Conclusion

The current compliance strategy needs to be improved

33.124 The Department intends to deal with most of the areas requiring improvement noted in this chapter through the initiatives listed in Exhibit 33.11 . We would expect the Department to monitor and regularly report to senior management on their implementation. We will follow up and report on this in future audits.

33.125 In order to deal more fully with the required improvements, Revenue Canada should:

a) review its compliance strategy and explore ways to improve it. If the legislation precludes the Department from adopting a more effective compliance strategy, it should seek changes to the legislation;

b) implement a verification program for RRSP contributions and transfers between registered plans to ensure that those transactions are within the prescribed limits;

c) obtain information on the amounts of contributions to, and withdrawals from, deferred profit sharing plans, to enable it to effectively design and implement an audit program to verify the transactions;

d) continue to urge the Department of Finance, emphasizing the need, to formulate effective penalties for cases where revocation is considered inappropriate; and

e) provide to Parliament, in Part III of the Estimates, information on the results of its compliance activities with respect to TARS.

Department's response:

a) The Department will develop and implement a system to maintain and report summary information on the results of front-end compliance review as well as audit results. This information will assist in the assessment of the effectiveness of its current compliance strategy with regard to Tax Assistance for Retirement Savings. If changes to legislation are required, they will be discussed with officials of the Department of Finance.

b) The Department does not agree. There is currently a verification program in place.

Procedures are in place to verify RRSP deductions. While many RRSP claims are accepted on the initial assessment of the T1 return (both paper and EFILE returns), clients are required to substantiate their claims by sending in receipts upon request.

The Department is aware of the importance of monitoring clients' voluntary compliance with the Income Tax Act . The Department's upfront verification steps (error clues and confidence validities), as well as the post processing review program, provide the necessary safeguards to ensure that clients' RRSP claims are valid.

The Department's audit program includes the verification of transfers between registered plans.

c) The Department does not agree that by obtaining the cited information it could more effectively design and implement its audit program. The existing audit program for DPSPs addresses all of the pressure points listed in Exhibit 33.9.

d) The Department's objective is to ensure compliance with the requirements of the Income Tax Act and Regulations. We will continue to work with officials of the Department of Finance to secure penalties as appropriate.

e) The Department will provide information on its compliance activities with respect to TARS, provided that the information is significant and warrants inclusion in Part III of the Estimates.