Défis du vieillissement

La retraite garantie est-elle chose du passé ?

Rapporteur's Report

Bruce Little 4 juin 2010

The recent financial crisis and recession have brought the viability and security of employer-sponsored pension plans, the sustainability of public pension programs and the adequacy of individual retirement saving into the spotlight since late 2008. In this context, and at a time when governments in Canada are actively consulting about the issue, the Institute for Research on Public Policy held a symposium to explore the retirement income prospects of today’s middle-income workers and possible reform options. This is in contrast with earlier pension debates, when then-current retirees were the focus. The event consisted of expert panels and presentations based on research commissioned by the IRPP and other groups; it brought together 60 invited experts and practitioners from government, business, labour and academia. The following highlights some of the issues raised and conclusions reached during the event.

  • Two complementary pieces of research pointed to a potential problem for future retirees. Grant Schellenberg and Yuri Ostrovsky’s research found differences of 7 to 9 percentage points between the earnings replacement rate of current 70-years-old male retirees who were covered by a registered pension plan (RPP) at age 55 and those who were not covered. The age of retirement was highlighted as a way workers adjust their replacement rate.
  • Michael Wolfson produced a sophisticated forecast of the incomes of future retirees under the current retirement income system using LifePaths, a unique Statistics Canada micro-simulation model. He found that “about half of individuals born between 1945 and 1970 who are in the middle 50% of the earnings distribution in their prime working years can expect a decline in their standard of living after retirement of at least 25%.” Further research is warranted to better incorporate items like housing and saving that are not explicitly tagged for retirement.
  • Insolvency law could be improved to better protect workers’ and pensioners’ claims in pension funds sponsored by private employers. However, more promising avenues involve reforming the funding and governance requirements of defined-benefit (DB) plans, and possibly setting up a national pension guarantee fund. The latter would require federal-provincial coordination to address “moral hazard” and “adverse selection” issues. Pension fund surplus ownership remains a core issue that needs to be resolved.
  • Workers without access to collective saving vehicles must face large financial and longevity risks on their own, which is costly for reasons such as a lack of expertise and the absence of economies of scale. The most efficient way to share and manage risks likely does not involve traditional DB structures, which fared well during the past three decades because of favourable demographic and market conditions that no longer hold, or traditional defined-contribution (DC) structures, which do not allow individuals to share risk efficiently under current regulations. In the spirit of reforms in the Netherlands, discussions should include alternatives like plans where the payment of a target level of benefits is conditional on financial outcomes. Today’s pension arrangements are often hedged by such conditions, and, according to rapporteur Bruce Little, absolute pension commitments – many of which promised more than they could deliver – might be a thing of the past.
  • What role can financial literacy play in improving the outcomes of individuals who must make complex decisions about retirement? US research indicates that employer-based financial education improves employees’ decision-making, and that this is the most efficient means of teaching retirement planning. However, well-crafted default options (in which workers are automatically enrolled, for example) are also important, and there is need to explore further the relative value of all available instruments in improving retirement outcomes.
  • Many programs targeted at smalland medium-sized enterprises (SMEs) in Quebec and the US, which were discussed by a panel, have resulted in a limited take-up, even though they address SME owners’ concerns regarding simplicity and cost. Some participants argued that a large percentage of SME workers do not want or need further pension coverage, and that it might be more efficient to simply improve on the options already available – including expanding access to individualized financial advice. Several participants noted that a lot of employers “want out of the pension business.”
  • Bob Baldwin compared reform proposals that alter the boundaries between voluntary and compulsory or private and public activity, and noted that preserving the parts of the current system that work well makes choices difficult; it might be easier to deal with a total failure that demands a greenfield approach. Two major issues are declining pension coverage and the ineffectiveness and inefficiency of existing retirement savings options, and he noted that the reform proposals differed on how to address these in their degree of compulsion, the treatment of the self-employed and the relationship to existing RPPs.
  • Baldwin also noted that with changing demographics, “stabilizing the relationship between contributions and benefits will require some mix of higher contributions, lower benefit amounts, and later retirement,” and that it is impossible for different cohorts to insure each other so that each cohort receives a fair benefit. This has a bearing on intergenerational fairness, which is a criterion by which to assess reform proposals. In contrast with past reforms, all current proposals assume that any new benefits would be fully funded; this is consistent with the current focus on avoiding imposing undue costs on future workers.
  • Based on his simulations, Keith Horner favoured an expansion of the Canada Pension Plan (CPP) over the creation of a new large DC plan. This is mainly because of the greater new savings it might generate; this would offset the future cost to government of the Guaranteed Income Supplement (GIS), which would be very large because tax-free savings accounts benefits are excluded from GIS eligibility calculations and because of the CPP’s efficiency on the labour and financial markets. His proposal would increase CPP benefits from 25% to 40% of earnings up to the year’s maximum pensionable earnings (YMPE) and from 0% to 25% of earnings between one and two times the YMPE, leaving room for additional discretionary saving, and it would encourage supplemental voluntary DC plans similar to the Alberta-BC proposal or Quebec’s SiPPs.
  • Jonathan Kesselman argued for further expanding the CPP and possibly phasing out old-age security in the future. In his view, we should displace inefficient employer-based plans, and expanding the CPP would address concerns about the current system and some of the reform proposals, such as costs, labour mobility problems and the need for risk pooling. Wolfson, based on his simulations, also favoured a doubling of the CPP, which would partially solve the problem identified in the status quo simulations he presented earlier.
  • Although there was some agreement on the advantages of auto-enrolment compared with purely voluntary plans, there was less agreement on the role of the financial industry and the degree of compulsion needed in any reform. Andrew Jackson defended the Canadian Labour Congress’s CPP-based proposal, highlighting its low cost and the fact that private sector employers and the financial sector had not been able to deliver sufficient coverage since the 1960s. Many participants expressed strong support for expanding the CPP in some fashion.
  • Some panellists argued that instead of expanding the CPP the focus should be on improving the current system. Bill Robson presented his assessment grid, and he emphasized the importance of avoiding forced savings or unrealistic promises, and increasing flexibility in terms of the options available to workers/savers and employers/sponsors. He favoured increasing tax-deferred saving room and expanding privately delivered and better regulated DC plans.
  • Flexibility was at the core of the Canadian Life and Health Insurance Association (CLHIA)’s proposal to further develop multi-employer pension plans, perhaps with mandatory pension coverage in firms with 20+ employees. Some participants opposed mandatory coverage, while Horner noted that the scale and targeting of mandatory plans involves a trade-off between increased retirement income security and higher welfare costs of forced saving.
  • The CLHIA’s proposal would address the concern many participants expressed with the prospect of putting large sums into the hands of a single organization; the Caisse de dépôt et placement du Québec was cited as an example to avoid.
  • Several participants argued that the outlook for retirees will keep worsening, as many Canadians want to retire earlier than previous generations did. Others pointed out that current generations are staying in school longer and taking on more debt, resulting in a longer retirement supported by a shorter working life. Some see this trend as unsustainable in the long run and thus likely to reverse itself (it has already begun to).
  • With increasing life expectancy and the changing nature of work, the age at which people retire is likely to rise naturally over the coming decades. Some believe it may be unnecessary to raise entitlement ages, as others suggested; but it will be crucial to take into account changing life cycle dynamics and the health of future retirees in analyses and policies.

Pension politics are difficult, and participants raised concerns that some proposals had little political appeal. But in his address Ontario Finance Minister Dwight Duncan said that “doing nothing is not an option,” pointing out that all reform options might not be mutually exclusive and that the optimal policy outcome – the one meeting agreed upon principles – may require a mix of different elements. He said political decisions will determine the balance among the criteria.

The rapporteur noted there is a danger that for all the consultation by government, the debate will now move behind closed doors, leaving little room for nongovernment voices to weigh in on the various proposals. At any rate, there is no quick fix and the recent political momentum could wane, so progress might be slower than expected. For all the recent urgency in public discourse, he noticed an implicit acceptance at the symposium that this round of talks could take years to come to fruition. Politicians – possibly excepting British Columbia – seem concerned that quick solutions will be suboptimal, and appear inclined toward caution. This was certainly true for Federal Finance Minister James Flaherty, who said that all options were still on the table, but he did not want to unduly increase the future burden on his triplet sons. The federal government, he said, would not avoid making difficult decisions, but it would not make them hastily.


Canada is in the midst of yet another debate over pensions, the latest in a long series of such discussions that have occurred in the postwar era. The financial crisis of 2008-09 exposed weaknesses when companies that failed – or faced bankruptcy – showed shortages in their pension plans that undermined the expectations of many employees and retirees for a financially secure retirement. Indeed, private sector employers have been moving for many years now to reduce their pension risk, mainly by shifting workers’ pension arrangements to defined contribution plans from defined benefit plans. Governments have also been active in reviewing pension issues. Four provinces – Ontario, Nova Scotia and (jointly) Alberta and British Columbia – have produced substantial reports on pension matters in recent years.1 The federal government has become more active with a public consultation paper of its own on federally regulated pensions, followed by cross-country consultations and changes to federal legislation. Together, the federal, provincial and territorial finance ministers established the Research Working Group on Retirement Income Adequacy, which released a summary report in December 2009 (Mintz 2009), which has been the subject of further consultations.

This rapporteur’s report is a summary of a symposium entitled “Avenues for Reforming the Canadian Retirement Income System” held by the Institute for Research on Public Policy on May 4-5, 2010, in Toronto. The symposium attracted about 60 invited experts and practitioners from government, business, labour and academia. It did not deal with the broader issue of income adequacy for the elderly, though the subject did arise occasionally. The very term “retirement” in the title of the event implies a life of work from which people have retired, and this was the focus of the presentations and the discussions that followed. There was a sense of history at work here, since many of the older participants recalled vividly (and some cut their teeth on) the “great pension debate” of the late 1970s and early 1980s. One participant drew the parallel immediately, adding: “But now we have data, unlike in 1980.” We also have considerably more experience with pensions themselves, in both the public and private sectors, and this history informed the debate as well.

Record of Proceedings

Session 1 – Income Replacement under the Current Retirement Income System

The session began with presentations by Grant Schellenberg of Statistics Canada and Michael Wolfson of the University of Ottawa. Rob Brown of the University of Waterloo commented on both papers.

Schellenberg, presenting work he did with Yuri Ostrovsky, compared the financial outcomes in retirement for men who had been members of a workplace registered pension plan (RPP) and those who had not, something that had not been done before. His central question was: Do individuals with and without RPP coverage at the age of 55 achieve comparable earnings replacement rates at the age of 70? His answer – which Brown called a bombshell – was no, earnings replacement rates did not differ significantly between members and nonmembers of an RPP. RPP members had more income from their pension plan, to be sure, but those with no RPP had more nonpension income – from investments and employment earnings.

That the no-RPP group had employment earnings at the age of 70 (the aggregates include both the retired and the nonretired) reflected another finding of the study: men with an RPP were more likely to be retired than men with no RPP. In income quintiles 2 through 5, over 90% of the group with an RPP were retired in the years 2003 through 2006, and that proportion rose each year. The probability of retirement also increased during this period for the group with no RPP, but significantly fewer of them were actually retired. In the three middle quintiles (2, 3 and 4), the retirement rates ranged from about 75% to 88%, while the rates for those in the top quintile increased from 69% to 79% over this period. Whether the group without RPPs kept working by necessity (having no workplace pension) or choice is not known.

An examination of men who have retired shows that in the middle of the income distribution (i.e., excluding those in the top and bottom quintiles) the median earnings replacement rate of men who do not have an RPP is 7 to 9 percentage points lower than that of those who do. However, the range of earnings replacement rates for the former is wider: some in the no-RPP group replace more of their income than RPP members, while others replace less. The same patterns are evident among couples: those with no RPP are less likely to be retired, their earnings replacement range is wider and their median earnings replacement rate is 3 to 6 percentage points lower than those with an RPP.

Wolfson’s presentation dealt with projections under the current retirement income system. Wolfson has been developing new data from Statistics Canada’s LifePaths simulation model, which will allow analysts to project the retirement income outcomes for future pensioners. He distinguished between income replacement rates that focus only on total (or gross) incomes before and after retirement, and “consumption possibilities” (or net incomes), which account for variations in taxes, saving and dissaving over the course of a person’s life. The popular norm for gross income replacement rates is 70%, although some think it should be closer to 60%; for example, the Mintz Report (2009). A better concept is “continuity of consumption,” where the natural norm is 100% – there is no sharp drop (or sharp increase) in a person’s consumption possibilities after retirement. Wolfson has tried to estimate, in his words, “this more fundamental concept of consumption possibilities over individuals’ lifetimes.”

Averages often hide important differences, he noted. Mintz (2009) found that one-fifth of Canadians may not have sufficient RPPs and RRSPs to replace at least 90% of their pre-retirement consumption, and modestand middle-income Canadians are most likely to be affected. Wolfson said his analysis using the LifePaths model, which is more sophisticated than Mintz’s analysis, suggests that roughly half of Canadian seniors who had middlelevel incomes in their pre-retirement years will face a decrease in their living standards (i.e., their consumption possibilities) of at least 25% after retirement. These results appear to show that the problem of retirement income inadequacy is greater than Mintz estimated. According to Wolfson, Mintz’s conclusion that the problem is primarily “inadequate saving discipline” is “inappropriate.” What some see as a lack of willingness on the part of Canadians to save for their longer term future, others will see as inadequacy in other parts of the retirement income system, such as the markets for private saving and annuities, the regulation of workplace pensions, the structure of tax incentives for retirement saving, and the limitations of the collective approaches embodied in the public pension system.

In his commentary on the two papers, Brown noted that while today’s retired people may not be doing too badly, the outcomes for future retirees may be more of a problem. Wolfson’s forward-looking approach is important, but is also risky, because projections depend on changeable factors such as life expectancy, fertility, inflation, real wage growth and market yields. Canadians can be split roughly into three groups: those who should not save more because they are suffiently protected by Old Age Security (OAS), the Guaranteed Income Supplement (GIS) and the Canada Pension Plan/Quebec Pension Plan (CPP/QPP); those who need not save more because they are already saving enough; and those who are failing to meet savings targets. The most important group is the last group, which is mainly in the middle-income range. How big is it? Are there enough data to make governments want to act?

In the discussion that followed, several participants focused on the declining retirement incomes of successive cohorts of the population. One noted that recent surveys show growing inequality in wealth and guessed that future retirees will not be as well off as those in the group Schellenberg and Ostrovsky analyzed, in part because of changes to RPPs. One person noted that we are already seeing declines in retirement incomes for successive cohorts, but suggested that this may relate to changes in OAS benefits. Wolfson confirmed that changes to OAS benefits have had the largest impact. In response to a question, Schellenberg said cohort effects may be difficult to extract from the data.

Many participants were keen to explore the differences and the (sometimes surprising) similarities in the retirement incomes of people with and without RPPs. One said the Canadian results were consistent with data from other industrialized countries. Another suggested that differences in retirement income may stem from differences in the age of retirement rather than income replacement rates. Some participants underlined Wolfson’s distinction between income and consumption replacement. One person had tracked retired individuals in the data and found that replacement rates did not diminish over time; indeed, they were more likely to increase, perhaps because of inheritances.

Schellenberg agreed that the age of retirement is an important factor in determining retirement income. He noted that retirement behaviours vary a lot, but we do not know whether these differences are driven by choice or necessity. A person’s position on the income scale is probably an important factor, he added. High earners tend to stay in the workforce longer, either because they are professionals who get intrinsic rewards from work and are still in demand, or because they need to replace a high level of income, or for both reasons, which is more likely.

The should-save-more group also received some attention in the discussion. One person said this group, as usually defined, includes people who really do not need to save more. Another said the numbers understate the problem because if they are saving privately, that group may still have unduly high costs (mutual fund fees), and they may face financial risks linked to longevity, inflation, or weak stock markets when they annuitize their RRSPs. Wolfson objected to the term “save more,” because he thought it biases the debate toward fixing private sector vehicles for saving; he preferred the term “provide more,” which includes other options.

Session 2 – Individual and Market Failures: The Role of Government in Retirement Saving

(2a) The Role of Government in Protecting Collective Retirement Savings: Pension Guarantees, Bankruptcy Laws and Creditor Protection

Speaker Ronald Davis of the University of British Columbia was unable to attend, so Murray Gold from Koskie Minsky LLP summarized his paper.

Pension funds have increasingly shifted to equities, which promise higher returns than fixed income investments, but also reduce the immediate cost of funding a pension. Various incentives have encouraged this trend. A corporate manager’s ability to manipulate the pension fund allows him or her to manipulate the company’s earnings and thereby affect his or her own compensation. Third-party managers have a vested interest in equity because their fees are higher. But with more equity, there is more risk, and risk is the enemy.

There are three areas where changes could reduce risk.

The first is funding, which is under provincial jurisdiction. In recent years, funding rules have been weakened and longer amortization periods allowed, to return a troubled pension fund to health. Davis advocates the opposite approach: The more risk a pension portfolio takes on, the greater the funding buffer should be; pensions should be funded to 110%, not 100%.

The second area for change is insolvency law, which is under federal jurisdiction. Because federal law trumps provincial law, provincial attempts to give priority to pension plan members in cases where companies are insolvent run into competing claims under insolvency law. In July, 2008, insolvency law amendments gave priority to arrears in normal cost contributions to the plan, but any shortfall in the pension fund remains at the end of the queue, including the arrears of special payment contributions. Davis argues that there is now no compelling reason in insolvency law why all arrears, including those for special payments, should not be given priority, and the law should be amended to accomplish this end. However, the same rationale does not apply to the shortfall remaining after all contribution arrears have been paid.

The third area where change is warranted is that of pension guarantees. The US and the UK have guarantees; we too should have some form of pension insurance. There are two concerns with guarantees: moral hazard, in that the guarantee against loss might increase the incentives to engage in riskier behavior, and adverse selection, in that mispriced premiums might not reflect the actual insolvency risk of individual employers. Solvent employers would subsidize weaker ones through premium increases, and the former would eventually find a way to exit the insurance pool, leaving behind only financially distressed employers. To be most effective a pension insurance scheme would have to be federal and national; funding controls should be provincial. This poses challenges because the federal government would lack the necessary control over the policy instruments – the ability to regulate pension funding and benefit improvement rules – used to counter moral hazard.

Commentator Norma Nielson of the University of Calgary said Davis’s paper helped her articulate why special payment arrears should be treated differently than they are now in bankruptcy – i.e., why they should receive higher priority. But she stated she was biased in that she was not as frightened as some are of funding ratios in the 80% to 90% range, which she called acceptable, even impressive. Only a small proportion of plans need to hover around 100%. What is missing from the discussion, however, is the issue of who owns a pension surplus. This is a key reason for funding ratios of less than 100%; employers worry they may not get back surpluses that are not needed to fund the plan. This incentive to underfund plans needs to be fixed.

Protection for pensioners in cases of insolvency drew the most attention in the discussion that followed. One participant suggested that if governments have to choose between providing more protection for pensions under bankruptcy laws or providing more support for workers in the event of a bankruptcy, they will opt for the former to avoid moral hazard. Another person argued that moving pensioners nearer the front of the queue in an insolvency might undermine efforts to revive a troubled company. The company would need new capital to get back on its feet, but potential lenders would demand higher loan rates because their money would be less secure. (One person noted that the cost of credit in such situations rose by 30% in one jurisdiction that gave priority to pensioners.) An offsetting potential benefit, however, is that new lenders, because they are behind pensioners in the queue, would ensure that pension costs are covered. Still, the imbalance of power favours new investors. If the investors do not like the terms, they can simply not get involved; employees cannot walk away without serious consequences. Current retirees are especially powerless, because they represent legacy costs and will not be productive in the future. Governments must protect them; as one participant said, “insolvency is a brutal process with severe human consequences.”

Another person said Davis had ignored “the most obvious thing – making sure plans don’t get into trouble.” He also said that moral hazard is not a serious issue; Ontario got through the 1980s and 1990s with no serious problems. He asked why the existence of the Canada Deposit Insurance Corporation had not tempted the banks into foolish behaviour. Another participant said moral hazard cannot be ruled out; companies do not deliberately avoid full funding of their pension plans, but when conditions get tight, it is easier for them to become delinquent if the Pension Benefits Guarantee Fund is there to help.

2 (b) The Role of Financial Literacy in Retirement Saving and Individual Savers’ Decision-Making
Robert Clark of North Carolina State University reported on his research into financial literacy through employer-provided retirement-planning programs. Financial literacy is related to income replacement rates on retirement. If workers are going to make optimal retirement decisions, they need to understand basic financial economics (risk-return properties of investments), basic financial mathematics (compounding, discounting, real versus nominal values), and the properties of and incentives in company and public retirement plans.

In fact, US workers know surprisingly little about these matters and, as a result, make suboptimal choices, the results of which stay with them for the rest of their lives. The best place to learn financial literacy is in the workplace; companies have an interest in better outcomes. Clark partnered with large employers, and evaluated their retirement-planning programs for employees approaching retirement and their orientation programs for new hires. Key retirement decisions include when to leave a career job, whether to take a lump-sum distribution of a defined benefit (DB) retirement plan, whether to annuitize their 401(k) accounts, when to start drawing Social Security (SS); and whether to change investment strategies.

Employees are invited to participate in these programs, and most of those approaching retirement do so. Survey results show that participants’ financial knowledge increased, and that participants used that new knowledge to alter their retirement plans. Almost one-quarter of respondents changed their planned retirement age, and almost one-third changed the age at which they would begin taking SS benefits. The proportion of those expecting to work after retiring increased to almost 57% from just under 47%, and many altered their 401(k) and lump-sum distribution plans.

Higher levels of financial literacy can lead to altered plans and better choices. People do learn, Clark said, adding: “Now we are trying to find out how long the learning lasts.” Workplace education is the best method of reaching workers, but more investment in education is needed. Companies like the programs, because employees like them and give their employers credit for running the programs.

Commentator David Boisclair from the IRPP presented what little is known of financial literacy among Canadians. A 2008 survey for the Canadian Foundation for Economic Education found fairly low levels of financial literacy; fewer than half the respondents answered 9 of 13 multiple choice questions correctly, and only 7% answered 2 of 9 open-ended questions correctly. Although two-thirds said they were somewhat confident they can save effectively for retirement, only 60% were confident they knew where and when to invest their savings. A larger 2009 Statistics Canada survey found low general financial knowledge. Over two-thirds of the 15,000 people surveyed were planning for retirement, but only one-third knew that investments in stocks are not insured.

There are no Canadian studies on the effectiveness of financial education, although there are many tools and much material is available. As well, the public strongly supports including economic and financial education in the school curriculum. But while Ontario, Manitoba and British Columbia have added programs, Quebec has cut back its programs. The Canadian context differs from that of the United States – there are fewer large employers, fewer DC plans, fewer financial institutions and fewer governments. We should investigate the effectiveness of employer-provided retirement planning programs in Canada relative to those provided by governments or nongovernmental organizations. If the former are indeed better, should employers be required to run programs? How, then, can small employers be induced to do so? But mostly, do such programs improve financial outcomes more than do alternative measures, such as framing? Should financial education be viewed as a necessary, rather than a sufficient tool for optimal retirement planning?

The discussion was animated, perhaps because many participants drew from their own experiences in their observations.

One participant said the issue goes beyond one-time decisions, and discussions of financial literacy necessarily turn into the issue of retirement planning. Any forecast over a 30-year period is difficult because over time, governments will change programs that affect retirement planning. “If people start getting poor, governments will respond,” the participant said. Another person observed that financial planning is different for those who don’t have an employment-based pension plan. How those people save for retirement will have a huge impact on outcomes. They usually seek financial advice from their local bank, where they are in the hands of a young person who has taken a weekend course and wants to sell them a mutual fund with a 2.5% management expense ratio. His remarks were underlined by another participant, who asked how we can talk about financial literacy without talking about the financial advice industry.

One participant stressed the need for financial literacy to begin in the schools, because it involves more than retirement planning; kids should also be taught about paying off student debt, buying a house and starting a family. Another said it is difficult to engage young people in retirement planning, saying “People live in the now when they are young.”

One person noted that default options are an important element of retirement planning; that is, should people be offered a retirement savings plan that they can join if they wish, or should they be enrolled automatically unless they choose to opt out? Clark responded that default options matter; automatic enrolment is better, but people should know what they are doing.

2 (c) Individual Risk-Taking: Health and Longevity Risks, Investment Costs and Legal Risks
Malcolm Hamilton of Mercer began his presentation with the observation that risk avoidance is not a viable strategy when real long-term interest rates are at 1.5%, and nor is it a desirable strategy for the retirement savings pool of a country. The current round of reform is about overcoming previous denial of the problems. Canada needs to begin with realistic expectations: we can strive for either affordable adequate guaranteed pensions at age 65, or affordable adequate target pensions at 65 where participants are expected to adjust their savings, consumption and retirement ages depending on how the investments perform. The Netherlands decided on the latter, rationalizing that if guarantees are expensive, let’s stop guaranteeing; we’ll try for good returns, but you’ll have to accept risk to get them.

In Canada, we say we can’t deliver cheap guarantees, but if we can reconfigure or change tax rules, we can change the economics. Hamilton hoped we get not just new rules, but institutions that can deal with the new rules. Nostalgia about how wonderful DB plans were is really just nostalgia for the 1980s and 1990s, when there was no moral hazard because there was a bull market. With DB plans, the risks are large. They are manageable (but only to a point), and they are concealable and ignorable (but only to a point). Ultimately, the risks are borne by individuals. And contrary to popular belief, governments do not take risks; they just pass them on to taxpayers.

There are three broad types of risk – markets and investment, longevity, and other types like population aging and the maturity of pension plans. Investment risk is the biggest: Real interest rates have ranged from 1.5% to 5% in the past 20 years, while real returns have ranged from 0% to 9% in the past 50 years. At age 65, an individual may live for another 10 to 30 years.

Retirement plans cope with risk in one of the following three ways:

  • Risk concealment and denial: US public sector plans do it; CalPERS is a good example. It stretched its amortization period to 30 years from 15 years and assumed a 4.75% real return; even so, the results were bad.
  • Risk management/avoidance: This involves pooling and insuring diversifiable risks. It works reasonably well, but in the end, we are left with risk sharing.
  • Risk sharing/bearing: We have to figure out how to allocate risk to individuals, who then have to adjust their savings, consumption, retirement ages, workloads, bequests, and so on.

For individual savings plans, risk minimization is expensive, ineffective and, not surprisingly, unpopular. For example, one risk minimization investment strategy is to buy fully indexed deferred annuities, but the price is high and very sensitive to real interest rates, which have moved considerably in the past 20 years. In the 1990s, when a typical real interest rate was 4.5%, an annuity might cost 15% of pay; in the 2000s, with real rates at about 2%, the cost would be 30% of pay. Since there is no demand for such products, there is no supply. Individuals are more likely to invest in balanced portfolios before retirement, and annuities or low-risk products after retirement. The major risk is volatility near retirement, so it is important that people be flexible about the age at which they retire. Many individuals can ignore longevity risk: Low-income seniors already get government pensions; public sector workers have DB plans; people who save too much will be fine, as will those who live frugally.

Hamilton then turned to the question of large groups – can they deal with risk better? DB plans can transfer both cost and risk within and between generations. Under pay-as-you-go funding, each generation pays the benefits of the preceding generation. Fully funded DB plans have usually shifted risk to future generations; each generation bears the risk for the previous generation. Both funding systems work well for the first generation and after that until the population ages and the system matures. The time diversification hypothesis – i.e., that investment gains and losses will correct themselves if permanently ignored – proved to be wrong. The fact that this hypothesis is wrong is not well understood, but recent problems with the GM and Falconbridge pension funds proved it.

DB plans have a tendency to poorly represent the interests of future generations by pushing costs and risks forward, and by reacting quickly to surpluses and slowly to deficits. They tend to leave taxpayers or shareholders with the risk and plan members with the rewards of risk-bearing until taxpayers and shareholders figure things out, which does not happen until financial disclosure reaches an advanced state. In effect, members were paid more than they would otherwise have been paid if they had borne the full cost of the pension. But neither traditional DB nor DC plans, with their weaknesses, have lived up to expectations for large groups.

The key message is that the risks are large. Risk management cannot deliver safe, affordable, adequate pensions in a hostile economic environment. Individuals can bear risk and cope with disappointment if they are aware of their exposures and of the need to adapt to changing circumstances. A resilient retirement system requires a resilient population. Risk sharing is important, and we should expect retired members to bear some of the risk. Nevertheless, collective risk management mechanisms must better incorporate the interests of future generations.

Commentator Steven James of the Canada Pension Plan Investment Board took issue with Hamilton’s opening either/or characterization of what are realistic expectations for the system. Canada’s retirement income system, with its three pillars, does not involve a binary choice but is really a continuum. James said Hamilton was right to focus on risks, but noted that he did not talk much about longevity risk, which is a big reason why pensions exist. He suggested that people gain the most consumption potential in retirement if they annuitize all their wealth at retirement, but that option is not popular because it does not allows for legacies on death. In fact, partial annuitization is the optimal choice in the presence of a bequest motive. Market annuities and DB plans pool longevity risk; James was more optimistic than Hamilton was that we can pool risks, but he agreed with Hamilton that pension promises should be priced appropriately.

In the discussion, one participant said the risks of old age have always been borne by the younger generation. In the days before pensions existed, children looked after their parents when they were too old to work. The real winners when pensions were introduced were the kids, when their parents did not move in with them anymore. Another person wondered about how certain key variables would unfold over the next three or four decades. He mentioned long-term interest rates, the possibility that real wages will rise as labour becomes relatively scarce, and the possibility that the supply of savings will dwindle as seniors draw down their wealth. One person commented that financial market risk is swamped by a cultural shift toward a shorter working life as people stay in school longer and retire earlier. Yet another wondered how we can move from the current situation as Hamilton sees it to a more Dutch-style system. Hamilton responded it will be very slow. He said the Dutch seem to have been able to change their system retrospectively; we can’t. He long ago gave up predicting interest rates for distant years, but James said long-term rates are cyclically low now and should recover in due course, and that the high real interest rates of the early 1990s reflected disinflation engineered by central banks.

Guest Speaker: Ontario Finance Minister Dwight Duncan

Lunch speaker Ontario Finance Minister Dwight Duncan summarized the Ontario government’s approach to retirement and pension issues. Preserving seniors’ quality of life is “crucial to the health of our economy,” he said. The government moved to deal with issues affecting the employment pension system with temporary funding solvency relief after the 2008 market downtown, simplifying pension division when a marriage ends, and adding $500 million to the Pension Benefits Guarantee Fund. It is still trying to resolve what to do with the fund on a more permanent basis, he said, noting that this was not a partisan issue. As he said, “We’ve had five governments of three stripes who failed to fund it properly.” Amendments to the Pension Benefits Act have already been introduced (Bill 236), and a second pension bill will be introduced later this year. The first bill deals with the least contentious issues.

He summarized the main options facing the federal, provincial and territorial governments, along with the pros and cons involved. The options fall into four categories: expanding the Canada Pension Plan, launching supplementary DC plans for those without employment-based pensions, expanding coverage under existing employment pension plans and RRSPs, and tax reform to encourage more retirement savings and innovations in pension delivery. Changes must be made in a way that respects the principles of transparency, affordability, equity, accessibility and balance. Governments must be prudent, because changes to the design of the retirement income system could affect household consumption, savings behaviour, debt levels and business competitiveness. As he reeled off the long list of issues governments face, he interjected: “Jeez, I’m glad I’m not going to be finance minister forever.”

He said the Ontario government is committed to a prudent approach, but added that the diversity of the three-pillar retirement income system is one of the system’s strengths, and that each pillar may need to be changed in both size and scope. The optimal policy option – and the one on which consensus is most achievable – may include a mix of measures across the four categories he defined earlier. “Doing nothing is not an option.”

Responding to questions from the audience, Duncan argued that there is a connection between the issues of long-term care and retirement income; for example, Ontario’s drug costs will be affected if an employer stops health benefits at retirement. Finding a balance among the principles he mentioned will come down to political decisions, he said. The current system is not in crisis, but he has heard compelling examples of problems from the Canadian Labour Congress and private organizations.

Session 3 – Panel on Retirement Saving Instruments for Workers in Small and Medium-Sized Enterprises

Peter Drake of Fidelity Investments Canada said individuals need help in finding the right balance between consumption and investment, which he called one of the oldest problems in economic analysis. Only about one-quarter of pre-retirees – people over 45 – have a financial plan for retirement. People need advice because there are so many decisions to be made, and those who do get advice report feeling better off. Approximately 25% in the private sector have a workplace pension plan. Data for small and medium-sized enterprises (SMEs) are hard to come by, but companies with fewer than 100 workers account for 48% of the workforce, while plans with fewer than 100 members account for only 7% of private plan members.

The better plans are administratively simple, they balance investment returns and risk, and they are available at a reasonable cost. Given the obstacles presented by other options, which include strengthening the role of DB plans, DC and RRSP plans may be the best option for workers in SMEs. Various stakeholders have made recommendations, including increasing enrolment and participation; increasing the availability of workplace RRSPs (which could include the self-employed); and harmonizing federal and provincial retirement savings and pension policies. These changes are not trivial, and some have already been carried out in the US and Canada. None are perfect, but all would be a big improvement.

Doug Bruce of the Canadian Federation of Independent Business (CFIB) reported on an ongoing survey of its members on the issue of pensions. Questionnaires were sent to 105,000 members; there have been 8,000 responses to date, and 10,000 were expected by the end of the survey.

Business owners reported that their own retirement income will come from – in declining order of importance – the lifetime capital gains exemption, proceeds from the sale of their business, other personal savings, RRSPs, CPP/QPP, OAS/GIS, tax free savings accounts (TFSAs) and RPPs. Among very small companies (0-4 employees), only 11% have a retirement savings plan of some kind, but among companies with more than 40 employees, 45% have a plan. Those with no plan were asked why; 50% said it was too expensive; 37% said it was too complex; 32% said it was not common in their sector; 24% said they don’t know where to start; and 15% said their employees have no interest. (The latter response is most common in the hospitality sector.) Of those that do offer plans, the plans are usually RRSPs, and 76% match contributions at least partly.

As for policy changes they would like to see made, 78% said people should be able to contribute additional sums to the CPP on a voluntary basis, 73% opposed a doubling of the CPP, and 38% said we should require all companies with 20 or more employees to set up a plan without employer contributions. Among the CFIB’s recommendations: Don’t increase CPP premiums; don’t mandate employer contributions in any new system; bring in fairer tax treatment for RRSPs; index the lifetime capital gains exemption to inflation; and change the capital gains tax rules to promote business succession planning.

Pierre Plamondon, chief actuary of the Régie des rentes du Québec, outlined Quebec’s experience with the simplified pension plan (SiPP), which was introduced in Quebec in the mid-1990s. When SMEs were asked why they do not offer a pension plan or group RRSP to employees, 26% said it was too costly, 26% said their company was too small and not enough people would benefit, 19% said they had not been asked to offer a retirement saving vehicle, and 29% gave other reasons.

An SiPP is a DC plan that is offered and administered by a financial institution, with contributions going into two accounts – one that is locked-in and the other that is not. Advantages of SiPPs for SME employers are they can set flexible contributions, they can avoid payroll taxes (as they would with an RPP), and the administration is simple (like that of an RRSP). Employees can be flexible with respect to locking in their contributions, they can make additional contributions, and they get the protections offered under the Supplemental Pension Plans Act. The employer sets the plan’s provisions and chooses the financial institution that will administer the plan (note that the plan is not administered by a pension committee). The financial institution is registered with the province and provides information both to members and government authorities; it also offers at least three diversified investment portfolios, reflecting different risk/return profiles. In 2009, there were 14 such plans involving 1,415 employers and covering 58,000 members. There are drawbacks to the SiPP. It is available only in Quebec and Manitoba, and for pension plans under federal legislation. It is not available for companies with employees in more than one province. It is a DC plan in which the risk falls on the employee, and it includes no retirement planning advice. Other alternatives for SMEs are group RRSPs and deferred profit sharing plans (DPSPs). Companies with 50+ employees can set up a retirement information committee to inform members and act as an intermediary with the financial institution.

John A. Turner of the Pension Policy Center in Washington, DC, noted that in the US, as in Canada, pension plan coverage is lower among SMEs. SMEs have access to the same pension plans as large employers – DB and 401(k) plans. Also, employees can contribute to an individual retirement account (IRA). In addition, any employer can set up a simplified employee pension plan, through which the employer contributes directly into a worker’s IRA; no employee contribution is allowed. Employers are not compelled to contribute every year and need not file documents with the government.

Some types of plans are available only to SMEs. One is a simple IRA, for any employer with 100 or fewer workers. Both the employer and the employees can make tax-deductible contributions, and financial institutions handle most of the paperwork. The employee can decide how much to contribute, and the employer must match this or contribute 2% of each employee’s compensation. The plan must be offered to any employee who has earned $5,000 in any two prior years and is likely to earn at least $5,000 in the current year.

A new plan being launched in 2010 is the DB(k) plan; it combines features of a DB plan and a 401(k), which is the most popular form of retirement saving. The DB(k) has less paperwork than separate DB and 401(k) plans and is designed to encourage a resurgence of DB plans that are smaller and less costly than earlier DB plans. The defined benefit equals 1% of final average pay for each year of service up to 20 years. Employees are automatically enrolled in the 401(k) portion of the plan unless they opt out; 4% of pay is automatically contributed to the 401(k) unless the employee changes the contribution level. The employer matches 50% of the employee’s contribution to the 401(k).

Much of the discussion consisted of questions seeking specific details about the Quebec and US systems, but one participant raised the issue of who benefits from certain savings plans. When one talks about RRSPs there is a danger of losing the focus on the middle income group. In the late 1970s when retirement income was under review, it was a dumb idea to put money into an RRSP unless you were in the upper income quintile, and it’s still a dumb idea, he said. Raising the upper limit on RRSPs would create big revenue losses for governments, but do nothing to help the middle income group. Turner said the US plans are designed so as not to disproportionately benefit upper income people.

There was also some talk about public service pensions. One person argued that we should not just say taxpayers are on the hook for public service pensions; public sector employees are very much on the hook. Another responded that unfunded liabilities are overlooked during collective bargain in the federal public service. Doug Bruce responded that in the early 1990s, federal public servants paid 50% of the cost of their pensions, but that has since fallen to 32% and the government is proposing to raise it again to 38% by 2012. Another participant noted that the real advantage to the federal government of a wage freeze would be the effect on pension liabilities; this “plays a real role” in the Minister of Finance’s actions.

Session 4 – Assessing Recent Reform Proposals: Criteria and Approaches

This was a particularly meaty session with three speakers who got into considerable detail about some of the major proposals now being floated for reforms to the retirement income system, and some new proposals as well.

Pension consultant Bob Baldwin reviewed the main reform options and what they tell us about the unresolved issues. Some proposals do not alter the boundaries between voluntary versus compulsory or private versus public activity. He focused on proposals that do alter the boundaries. The major ones are the provincial reform proposals (Ontario’s, Nova Scotia’s and the Alberta-British Columbia [ABC] proposal), the creation of a Canada supplementary pension plan (CSPP), and the expansion of the CPP. Ontario is looking at setting up an Ontario pension agency that could receive contributions from employers and employees for group pension plans; this adds options to the opportunity set.

Two major issues are declining coverage and future retirement income adequacy; and the effectiveness and efficiency of RPPs and individual retirement savings options. He noted the difference between the present situation and that during the “great pension debate” of the 1970s; today the agenda is different. Minimum income protection has been addressed, so it is no longer an issue. The concern now is for the future elderly, whereas in the 1970s, the debate was animated by the situation of the current elderly.

There are several points of agreement among the proponents of the different reform options. There is consensus that a problem exists and it will not be solved by voluntary activity (no voluntary system in the OECD has more than 50% participation); and the focus is on earnings replacement, mainly to maintain living standards. All the major proposals deal with future service – so full benefits would not begin for 40 years. We now accept there has to be full funding, though politicians might think this would not have political traction. There is also agreement that single employers should not be the delivery platforms, because small employers cannot offer the scale needed.

There are more points of disagreement than of agreement. One is the earnings levels at which participation/saving would be mandatory . At the low end, the exempt earnings would range from $3,500 to $30,000, while at the high end the ceiling would range from $47,200 to $120,000 – the limit defined in the Income Tax Act. Benefit design is another area of contention. The ABC proposal is a DC plan. Keith Ambachtsheer’s CSPP is a modified DC plan. Nova Scotia’s proposal is a DC plan, but it would allow groups to establish a target. The CPP expansion proposals are enhanced DB plans with a target benefit. There is also disagreement over the degree of compulsion (the CPP proposals are compulsory; the others are auto-enrolment), the participation of the self-employed (the CPP proposals are compulsory; the others are opt-in), and the relationship with existing RPPs (the CPP proposals would displace them; the others involve wrap-arounds).

Baldwin raised the issue of benefit design and contribution rate volatility. If benefits are defined and there is an assets-liabilities mismatch, the contribution rate will not be stable under the rules of full funding. Proposals need to balance the desired certainty of the benefit rate with the tolerable uncertainty of the contribution rate. In closing, he made several points. The CPP, he said, is a flexible tool, whose organizational structure is compatible with a wide range of policy decisions. The main analytic conclusion from the finance ministers’ Whitehorse meeting in December, 2009, remains valid – the current retirement income system works for many, but not for a significant chunk of the population, and the trends are worrisome. The sheer variety of need makes choices difficult; it might be easier to deal with a total failure that demands a greenfield approach. Lastly, he said, the reform debate launched in the mid-1970s lasted a decade; this one might too.

Commentator Byron G. Spencer of McMaster University described Baldwin’s presentation as a very thorough review of the major alternatives. He argued that it is difficult to measure plan coverage, because group RRSPs are not included in Statistics Canada’s measures (he wondered why not), and plans are too often misclassified, with plans often counted as being DB plans when they actually contain significant elements of DC plans. Also, the dynamic aspects of pensions are often lost. When individuals change jobs, they may change between DB and DC plans and no coverage at all, so snapshots might not be very informative. We need more dynamic information, he said, adding that in this respect Wolfson’s approach seems to be interesting.

His big concern is the extent and nature of private pension coverage, especially given the differences in regulatory arrangements across the country. He presented several quotes from Baldwin’s paper on margins for adjustment: Reduced mortality gives successive cohorts in a DB plan better benefits than previous cohorts; it is impossible to stabilize both a benefit and a contribution rate; it is impossible for different cohorts to insure each other in an equitable fashion; and, finally, he added, “stabilizing the relationship between contributions and benefits will require some mix of higher contributions, lower benefit amounts, and later retirement.”

Spencer noted that there is very little talk of the age of retirement – which was fixed at 65 in 1966. Since then, there has been a nine-year increase in life expectancy for men, but no change in age of eligibility. The age of entitlement has increased in the US, the UK and in other OECD countries. Why is that not part of the public debate in Canada?

Pension consultant Keith Horner began his presentation by noting that the case for reform has been well made: middle-income earners are not saving enough; pension coverage is declining; the RRSP savings rate has declined; retail savings plans have low investment returns; and Canadians will need to save more because of higher life expectancy, lower rates of return and the declining value of OAS (because it is indexed to prices, not wages).

Many people do not realize that the TFSA is a game-changer. The take-up has been extremely strong; in the first year, 4.7 million Canadians put $15.8 billion into TFSAs, a level RRSPs did not reach for 16 years after their inception. But because TFSA benefits are excluded from income in determining eligibility for benefits that are income-tested, people with large TFSA benefits could qualify for the GIS, which is supposed to target those with low incomes. The result is that lifetime TFSA savers will have higher consumption in retirement. Among households with earnings, the 25% with the lowest-income should not save, and of the remaining 75%, over half will be better off saving with TFSAs than RRSPs. TFSAs will bring about large increases in the cost of GIS and other income-tested benefits over the next 20 to 30 years, especially if governments continue to rely on GIS increases to maintain the effectiveness of the OAS/GIS income guarantee in combatting poverty.

The key choices for pension reform are: DB versus DC plans; mandatory versus voluntary plans; and the scale and targeting of pensions. We can narrow down the range of options by insisting that DB plans be mandatory in order to avoid adverse selection, and that any mandatory plan must be national. The scale and targeting of mandatory plans involves a trade-off between increased retirement income security and higher welfare costs caused by inflexible forced saving.

One DB option could be to enrich the CPP with fully funded benefits that would be contingent on the funding status of the plan. Benefits would increase to 40% (from 25%) of earnings up to the year’s maximum pensionable earnings (YMPE) plus 25% (from 0%) of earnings between one and two times the YMPE. Such a plan would still leave room for discretionary saving, since income replacement rates would still fall below 50% for most earners.

Horner’s overall assessment included the conclusion that expanding the CPP would be positive:

  • Mandatory options are most effective at boosting retirement incomes, though there are some welfare costs.
  • A DB plan should outperform a mandatory DC plan, as intercohort risk pooling allows for a less liquid and somewhat riskier portfolio that raises expected returns, though this should not be overstated.
  • Some of the risks associated with DB plans – like job change and employer bankruptcy – do not apply to a national plan.
  • The history of the CPP suggests that investment risks can be managed.
  • Mandatory national plans have the best labour supply effects, partly because the benefits are portable for job changers. Labour supply effects are also better under a DB plan; the latter would replace some existing RPPs that have poor labour supply effects.
  • Mandatory new payroll taxes might have negative effects on the labour supply, but these could be alleviated by a restructuring of the financing of the COPP’s legacy costs, which account for about 4 percentage points of the current 9.9% rate. One way to do that would be to spread the legacy costs over all covered earnings – up to double the current YMPE.
  • A DB plan could reduce the future fiscal costs of TFSAs.

Horner’s explicit reform package would enrich the CPP and encourage supplemental voluntary multiemployer-defined contribution plans along the lines of the ABC proposals, Quebec’s SiPP-type plans and the Ontario Teachers’ Pension Plan. CPP premiums would be restructured to spread the legacy costs across a bigger earnings pool, and all CPP contributions would be made tax-deductible. Rules governing TFSAs would be changed so that for income-testing purposes (for the GIS especially), half of contributions would be excluded and half of benefits would be included. Governments could also reconsider the extent and organization of benefits regulation.

Commentator Jonathan Kesselman of Simon Fraser University strongly favoured an expansion of the CPP and was willing to go even further than Horner does – from a big CPP to a bigger big CPP. How big an expansion is a judgment call, he argued. Other big-CPP plans would replace 50% to 70% of earnings. Where Horner had referred to the welfare costs of oversaving, Kesselman asked what they were and how they could be measured. Limiting factors on the size of a bigger CPP would include constraints on individual choice, which would affect the timing and amount of saving, home purchases, child costs and preferences to consume while younger. There are opposing considerations that support greater enlargement of the CPP, however, and Kesselman noted reasons for believing a bigger CPP that displaces some RPPs would « not necessarily be undesirable.” A big CPP would have lower costs and risk, a higher return, and would avoid portability/mobility problems. Universal coverage would address adverse selection, annuitization costs and issues of risk pooling. Kesselman went even further, suggesting that a bigger CPP could one day lead to a phasing-out of OAS over the 47-year period of phasing-in a bigger CPP, with an expanded GIS benefit to offset the removal of OAS for low-income earners. The question of expanding the CPP is not whether we do it, but by how much.

The first day’s final paper was from Michael Wolfson, who built on his morning presentation to discuss three options for reforming the retirement income system. The first two were different ways of expanding mandatory DB plans such as the CPP/QPP. The third would index OAS, GIS and personal income taxes (PIT). His options include an intermediate phase with a new national opt-out DC plan.

The first option would double the nominal CPP, with no change to the year’s basic exemption (YBE) or YMPE, as proposed by the Canadian Labour Congress (CLC). The second would increase the YMPE, but leave the CPP replacement rate (25%) unchanged for those with low incomes (since they are already replacing more than 100% of income). Above the half-YMPE mark, however, the replacement rate would rise to 40%, creating a “wedge.” The third option would leave the CPP/QPP unchanged, but index the OAS, GIS and PIT to the increase in average wages rather than the Consumer Price Index.

The income replacement rates for all three options are little changed across the entire income spectrum. However, when you look at the fraction of the population with a net income replacement rate below 75%, the “wedge” option is targeted not to the middle income group, but the upper-middle groups, those with more than $80,000 in income. It would be better to target a doubling of the CPP/QPP at the middle group.

Wolfson then digressed into questions about seniors’ health and leisure time. He noted three claims that are often made about seniors: because they keep saving, they do not need more income; because they are less healthy, they cannot make much use of extra income in old age; and they have more free time, which can offset their income needs. Data for 2001 and a projection for 2021 show that population aging will bring with it somewhat of an increase in the incidence of disability, but the good news is that most seniors, even the “older old,” can expect to be healthy.

He introduced the concept of good life time (GLT), which is defined as the portion of an individual’s lifetime in which they have adequate time, money and health. They have three hours a day of leisure, either active (some form of physical activity), passive (watching TV, reading) or socializing (dining out, for example). They have over two-thirds of the median family disposable income, adjusted for household size. And they attain a specific score on the McMaster Health Utility Index. Discussions of retirement income are pertinent to the money part of this trio. Data show that health diminishes with age and leisure time increases, but the two-thirds cut-off for income puts many seniors below the mark to qualify as having GLT. How should income needs after retirement be judged if we do not take health and time into account? To conclude, he offered this piece of information: seniors get more satisfaction from work than they do from any of their leisure time activities, perhaps because they feel they are still contributing to society, perhaps because they enjoy the social interaction that work brings.

His conclusions: (1) a mandatory expansion of DB plans would reduce the problems with the retirement income system, though it is not clear how total savings would be affected by more prefunding, which would create a much larger explicit asset pool; and (2) pension policy analysis should be placed in a broader context, that is, to include health and leisure.

The ensuing discussion was particularly lively. Several participants wanted to talk about the politics of a major CPP expansion. “I want odds,” one said. There is probably another room like this one in which people are debating health care budgets, for which we are also asking the kids to pay the bills. What, he asked, are the tradeoffs? Another person argued that the topic of pension reform has gained political momentum, because we are dealing with a perceived crisis of recent vintage, and if we don’t act, we’ll lose that momentum.

One person worried that we would be putting too many eggs into one basket if we expand the CPP, which set off a spirited exchange with a second participant, who argued that the CPP has established institutions and governance rules; we should draw confidence from our experience with the 1997 reform, which created the bigger CPP fund, now in the hands of the CPP Investment Board. My concern, replied the first participant, is with the concentration of investment in one organization, not with governance. When the second participant then suggested that we could have a new organization, the first objected that this would increase costs. To which another participant interjected, “You cannot have it both ways.”

One person said we were ignoring the fundamental political problem we are trying to solve – which includes reducing poverty and increasing income equality, goals that go beyond income replacement. We talk about sharing risk across generations, but no one has discussed changing family structures or immigration (which implies bigger families). We should also be concerned with simplicity, which argues for an expansion of the CPP; in the UK, people are being forced to think about 100 pension parameters in making choices.

A participant took issue with Spencer’s focus on 65 as the age of retirement. Because people can begin to draw CPP as early as 60 or as late as 70, 65 is now irrelevant. Proposals to increase the retirement age are actually code for reducing benefits, he said. And for all the talk of people wanting to work when they are old, he added, look at what happens in Europe when anyone proposes an increase in the retirement age.

Another participant had some advice for anyone who thinks of reducing or eliminating OAS: “Don’t forget Charlie Brown,” a reference to the 1985 episode in which a near retiree publicly snapped at Brian Mulroney because of his stated intention to deindex the OAS. And, the person added, don’t forget that young people live in a different world than “we gray-hairs” do; they face a very different workplace.

Guest Speaker: Federal Finance Minister James Flaherty

The dinner speaker, Federal Finance Minister James Flaherty, said that the government is consulting widely on pensions and has not yet come to any conclusions, though he noted that he was not going to impose on his 19-year-old triplet sons costs they should not have to bear. “These are generational issues,” he said. The federal and provincial governments are working closely together on the issue, and they have commissioned a lot of research. Governments have to be careful not to do harm and not to jump too quickly to conclusions about the right solutions. Pension reform will be at the top of the agenda when the finance ministers meet again on June 1314 in Prince Edward Island.

The federal government will not avoid choices or tough decisions, he said. Research indicates that middle-income Canadians are not saving enough, and “we know from experience” that if some people cannot afford a reasonable standard of living in retirement, “other Canadians will be called upon to support them through their taxes and through reallocations of funds federally.” This makes retirement income an issue for all Canadians.

He stressed that the government is open to considering all options for reform. Canada has a sound financial sector, which means there is potential for new private involvement; “we shouldn’t be glib about turning our back on the private sector,” he said. The TFSA “has taken on a life of its own” and has become very popular in a very short time; in fact it is the most important tax change for savings since RRSPs were introduced in 1957. The government will also bring in a financial literacy strategy.

Session 5 – Perspectives on Recent Reform Proposals: Panel Discussion

Bill Robson of the C.D. Howe Institute led off by referring participants to a C.D. Howe Institute Backgrounder on pension complexity published in February (Robson 2010). He looked at principles for third-pillar pension reform, including what he called “broad bins” of solutions drawn from the Finance Canada discussion paper (2010). In some cases, he agreed with that paper, in others, he did not.

The first principle enumerated in the Finance paper is affordability for people and companies. Since many people should save less or not save at all, he argued, there should be no forced saving. The second is affordability for governments, and Robson noted that if this meant avoiding large increases in tax-deferred saving, he did not agree. However, he thought there should be ample savings room for those who need it. Federal public servants accrue pension benefits worth one-third of covered pay, and MPs accrue benefits worth one-half of their pay; Robson said these limits could be extended to all. Third, there must be intergenerational fairness, which requires people to fund DB pensions properly. And with real incomes likely to grow by only 1.6% annually over the next generation, pension reformers should remember that the projected health care costs will substantially draw on Canada’s modest income growth.

Fourth, reforms should strike a balance between individual and state responsibility, which Robson interpreted as meaning that they should not provide government backstops that introduce moral hazard to the system. Fifth, individual choice is vital, and sixth, any reform should increase accessibility to user-friendly savings vehicles. Sixth, and finally, another element of affordability for people and companies is striking the right balance between quality and cost. People should remember that while the CPP Investment Board’s operations cost only 0.2% of assets, the CPP costs the federal government another 0.4% for administration of the plan, and the Canada Pension Plan Investment Board (CPPIB) pays a further 0.4% for outside advisers. An additional factor that reformers should explicitly take into consideration but that is not included in the Finance paper is robustness; single-employer DB plans are brittle and sometimes break.

Robson then displayed an extremely useful slide – a matrix with reform options across the top and his list of principles down the left side, with check marks, x-marks and question marks in each box indicating how each option stacks up on each principle.

Robson is no fan of classic DB plans, which got no check marks on his matrix, for reasons of moral hazard, intergenerational fairness, and forced savings (which he opposes). He gave the most check marks to expanded flexible, privately delivered DC plans, combined with something Finance Canada also did not mention, more tax-deferred saving room.

Scott Sweatman of Spectrum HR Law LLP was co-chair of the Alberta-British Columbia Joint Expert Panel on Pension Standards (JEPPS), which reported in November 2008. He skipped lightly through his “too many slides” and summarized the JEPPS report and what has happened since. Contrary to a widely held view, Alberta and British Columbia are well on their way to producing “nearly identical legislation,” but it is unlikely there will be a joint regulator, as the JEPPS has recommended.

The objectives of the legislation will be to ensure the pension “deal” is kept, and to maintain flexibility. It will be based on principles, supported by rules where necessary. The key principles are eligibility, vesting, lockingin, portability, segregation of assets, governance and defining the pension deal. On the issue of governance standards, he said plan trustees and administrators need trustee/fiduciary education because “a lot did not know what they were doing”.

Fiduciaries need the protection of a “pension judgement rule” that would be parallel to the legal protection for company directors in the business judgement rule if they have to defend themselves against claims in respect of their decisions. “You can be wrong, but not stupid about being wrong.”

Companies with “legacy” surplus issues should be able to ring-fence old, i.e. DB, plans, and then wrap a new plan around it; it’s a very practical solution, for example, for companies that want to switch to a DC plan for new employees.

Sweatman stressed that the ABC plan is not dead, as reported. It was designed to meet the problem of people who are not saving enough and have big debts, plus the fact that many companies want out of the pension business. It is similar to plans in Saskatchewan, Quebec, Britain, Australia, and New Zealand, and to Ambachtsheer’s CSPP proposal (2008). It will be a private-sector-delivered plan, with help from government to get it going. The management expense ratio plus administrative costs should be no more than 0.5% of assets. It would be at arm’s length from governments, but this will not work if government backstops it.

Panellist Andrew Jackson of the CLC said the CLC is actively advocating the CPP be doubled on a fully funded basis, plus a 15% increase in the GIS. The CLC’s proposal is not correct in every respect, he said, but they wanted it on the agenda. The CLC would double CPP benefits to 50% of pre-retirement income and double the YMPE. Benefits would be phased in over a period of about 40 years and the YBE would be left alone. This is not aimed at the current elderly, but at people who are now in their 40s.

He argued that the CPP works; it is portable and it is a sound base for retirement income, but it is a barebones plan. That was deliberate on the part of its architects, because they wanted to give a larger role to the private sector than did other countries. “Maybe it’s time to come to a judgement on that,” he said.

The financial meltdown seized people’s attention, Jackson said. The average retiree gets only one-third of his or her preretirement income from public sources, and average CPP benefits are $472 per month, about half of maximum benefits (now $934). The average income from the CPP and OAS combined is $989 per month, which amounts to just under $12,000 a year. There are a lot of not-very-good plans in the private sector; there is no indexing. Expanding the CPP would take the burden off company pension plans. RRSPs do not work; the median RRSP for those aged 55 to 65 consists of only $60,000, enough to buy an annuity of less than $300 a month. The picture is not grim for today’s seniors, but what about those now in their 20s to 50s? They are less likely to find secure, well paid jobs until later in life, they carry more student debt and bigger mortgages, and they can expect less retirement income.

The expanded CPP would mean a premium increase of just under 6 percentage points on covered earnings, split between employers and employees. Workers who contribute the added 3% would get an additional $934, fully indexed to inflation, much more than they could get from a similar contribution to an RRSP. This is a very low-cost benefit. Many employers would see this as a good alternative to their own pension plans. He had an answer for those who make the “nanny state” argument: Many companies have compulsory pension plans, and no one screams about that. Ultimately, governments will be on the hook if seniors’ incomes are too low.

He said he takes seriously the argument that we should not put all one’s eggs in one basket, because he does not want to see a repeat of the problems with the Caisse de depot et placement du Québec.

Panellist Frank Swedlove of the Canadian Life and Health Insurance Association noted that his industry operates over 70% of private pension plans, where most of the growth has been in DC RPPs and group RRSPs. Still, almost half of private sector workers have no workplace savings plan; those with incomes between $30,000 and $100,000 are underserved. This is a missed opportunity. He set out four principles for new reform measures: provide for basic needs; offer incentives to save more; ensure private sector participation; and provide choice.

He reviewed two of the main proposals on the table. He outlined several concerns with them. One concern is that doubling the CPP would involve a 5 percentage point increase in payroll taxes, and benefits would not be funded for 40 years. Another concern is whether this would fill the gaps that need filling. Other concerns relate to government-sponsored plans like the ABC proposal or adding a DC component to the CPP. Under these proposals, there would be only one supplier, and infrastructure would be needed; the plans would take a long time to get rolling. If we were to add a DC component to the CPP, it would cause confusion with the CPP’s DB portion. There are also possible fiscal implications, and governments might get involved more than they planned to; in general, governments feel a need to step in when there are problems.

A better way, he suggested, is to encourage multi-employer pension plans (MEPPs), in which the employer is not a sponsor and there would be little administrative burden on employers. There would be automatic enrolment, with an opt-out provision. In the US, auto-enrolment led to a rapid increase in enrolment to 87% from 66%. Companies with 20 or more workers would be required to offer a savings vehicle.

For self-employed individuals, governments should expand the definition of income eligible for RRSP savings to include income from office, business and property, and they should defer the date for rolling over an RRSP into a RRIF to 73 from 71.

The system is not broken, he argued. The three pillars work, but we need to expand coverage. Improved access to workplace plans through the private sector could make the biggest difference, fast and at minimal cost.

In the discussion, the issue of one major investment board came up again when one participant agreed with Jackson that there should be other options; the recent experience in Quebec shows that having one big investor can go off the rails.

The same person drew a parallel with the original thinking behind medicare – that it was there to cover catastrophic illness. Perhaps the compulsory CPP should be regarded as a socially reasonable minimum base, and anything above it should be left to the private sector through workers and employers. He thought the ABC plans might fit better here than an expanded CPP.

Another participant said work needs to be done on quantifying the policy options; we do not know what would happen to existing pension plans under the various options, so we do not know what the net impact would be on retirement savings. A second person took up this theme; an expanded CPP would require additional contributions from employers, and we do not know what would be the effects of that. Also, an expanded CPP might encourage people to reduce other savings, because they think the government will look after them in the end.

Asked what the fees would be for MEPPs, Swedlove said medium plans tend to cost about 0.5% to 0.7% of assets; some might go as high as 1%.

Another person suggested that because group RRSPs are regulated by the federal government, they are easier for the insurance industry to administer because it does not have to worry about satisfying many different regulatory regimes.

Session 6 – Wrap-up Session

In the closing session, the rapporteur mainly summarized the messages heard during the previous day and a half, but he also jumped into the debate over whether creating one big investment fund was worrisome, because it meant putting all our eggs in one basket (the unfortunate example of the Caisse de dépoÌ‚t was on everyone’s minds). The money goes into one pool, he noted, but the investment managers, as they seek diversification and the best allocation of funds, then put the money into many baskets – bonds, stocks (public and private equity), infrastructure, and real estate; and any of these might be domestic or foreign. It is also worth noting, he said, that in the market meltdown and financial crisis of 2008 and 2009, all big pension funds lost money, but none (the CPPIB, Ontario Teachers’ Pension Plan, OMERS, HOOPP, bcIMC, AIMCo) ran into the same kind of trouble as the Caisse did. One reason for this is governance: the Quebec government exercises considerable influence over the decisions of the Caisse, while the others are arm’s length organizations.

Key Messages

The symposium produced a thorough overview of the problems in the retirement income system that need solving, and the challenges involved in coming up with solid and sustainable reforms. It generated many useful observations about the current debate and where it might go from here.

  • We have better data to assess the nature of the problem, not only as it is today, but how it is likely to evolve in the years or even decades ahead. Schellenberg and Ostrovsky’s research deepened our understanding of the role RPPs play in determining the income of seniors (not by as much as we might have thought), while Wolfson’s work offered a better idea of the outlook for seniors’ income in the future. This is a major advance from the pension debate of the 1970s and 1980s. Several participants argued that the outlook for retirees is worsening, cohort by cohort, and will continue to do so. While many Canadians now want to retire earlier than did previous generations, they are staying in school longer and take on more debt (student loans and mortgages) over their lifetime than their elders had; in addition, the nature of work has changed, it is less stable and offers less in the way of secure pensions. The result is that people expect to support a longer retirement with less work time and more financial stress than previous generations experienced. New evidence would be very helpful.
  • Today’s debate is different from its predecessors in several ways. Today, the debate focuses on the future retired; 30 years ago, it focused on the then-current retired – in the past, no one talked about big fixes for today’s retirees. And all the current proposals to repair the retirement income system assume that future pensions will be fully funded. This is in sharp contrast to the first three decades of the Canada Pension Plan (it was reformed in 1997) and the decades in which DB plans were inadequately funded, which had consequences that have become obvious in the past few years. A greater sense of humility was evident at this symposium than there was 30 years ago; absolute pension commitments seem now to be a thing of the past (many “guarantees” promsed more than they could deliver), so today’s arrangements are more often hedged by the ability-to-pay caveat. The notion discussed here that retirees might have to assume some of the risk – and the financial penalties implied – is quite new; seniors are no longer sacred cows, at least to nonpoliticians.
  • In this symposium there was no discussion of the health of the CPP or even the OAS/GIS, the main public pillars of retirement income. Rather, attention fell on the private sector pension system and on the inadequate savings – for whatever reason – of middle and upper-middle income earners. This helped to focus the debate.
  • We had two other very useful discussions. One involved the role of financial literacy and how it might be imparted both to those approaching retirement, and to the young, for whom pension issues seem far too distant a concern to bother thinking about. The concerns of small business also got considerable attention – more than in previous debates over retirement income.
  • The range of policy options is beginning to take shape, but it is clear that this debate will be very complex, simply because there are so many possibilities. Robson’s matrix – or variations on it – should be part of any future conference on pension policy. Everyone has principles (or criteria or goals) they want to advance in this debate, but the discussion would be much easier if everyone clarified their assessments and their preferred options with such a tool.
  • In terms of policy proposals, there was debate over not just the design of reforms, but also who will deliver pensions in the future. The longstanding divide over public versus private roles was the most voluble, with some supporting a greater role for the private sector and others advocating an expansion of public pensions. Whether retirement saving should be voluntary, mandatory or somewhere in between (automatic enrolment with an opt-out) was another key divide. Those elements can be mixed and matched in myriad ways, for example, mandatory or voluntary, through public or private vehicles.
  • The most surprising aspect of the policy proposals was the strong support for a substantial expansion of the CPP, which many participants probably regarded as a distant also-ran before this symposium. Governments seem disinclined to advance this solution, so it may go nowhere, but it is clearly on the table, along with the key provincial proposals from Alberta and BC, Ontario, and Nova Scotia.
  • There are no quick fixes. For all the recent urgency in the public discourse, there was an implicit acceptance that this round of the debate over retirement income could take a decade to work out. As the presentations by Duncan and Flaherty suggested, governments worry that too-quick solutions will be suboptimal solutions; they appear inclined to move carefully. There is one exception here: Alberta and British Columbia appear to be on a faster track than many people outside those two provinces have realized, and this may force the whole debate onto a faster track.
  • A final point, though it should not be regarded as an afterthought: Wolfson argued that consideration of retirees (or seniors) should be broadened to take account of not only their income, but also their health and leisure time activity. This is very difficult for policy-makers to do (especially since pension reform seems largely to be the domain of finance ministers), but it is worth exploring in future conferences of this type.

Where do we go from here? The politics of pension reform are difficult, and several participants raised concerns that some of the proposals on the table had very little political appeal. In every prolonged policy debate, there is a time for blue-sky thinking, where everything goes, and then there is a time for settling down to focus on solutions that are both sound and politically feasible. This debate has probably reached the transition point between the two.

The danger is that for all the public consultation by government, too much of the discussion will take place behind the closed doors of federal-provincial bargaining, leaving too little room for nongovernment voices – business, labour and academia – to weigh in on the various proposals as the decision-making progresses. Followup conferences, organized by the IRPP or any other organization that wants to get involved, would be welcome. And everyone should bring their matrix!

  1. See Joint Expert Panel on Pension Standards (2008); Ontario Expert Commission on Pensions (2008); and Pension Review Panel ( 2009).

Ambachtsheer, Keith, 2008. “The Canada Supplemental Pension Plan (CSPP): Towards an Adequate, Affordable Pension for All Canadians.” C.D. Howe Institute Commentary 265.

Finance Canada. 2010. Ensuring the Ongoing Strength of Canada’s Retirement Income System. Consultation document, March. Available at

Joint Expert Panel on Pension Standards. 2008. Getting Our Acts Together. A Report Prepared Jointly for the Governments of Alberta and British Columbia. Ministries of Finance, Edmonton and Victoria. Available at

Mintz, Jack M., 2009. Summary Report on Retirement Income Adequacy Research. Ottawa: Department of Finance). Available at

Ontario Expert Commission on Pensions. 2008. A Fine Balance: Safe Pensions, Affordable Plans, Fair Rules. Toronto: Queen’s Printer.

Pension Review Panel. 2009. Promises to Keep. Halifax: Ministry of Finance.

Robson, William T. “Cutting through Pension Complexity: Easy Steps Forward for the 2010 Federal Budget.” C.D. Howe Institute Backgrounder 126.

Glossary of Terms

AIMCo Alberta Investment Management Corporation

BcIMC British Columbia Investment Management Corporation

CPP Canada Pension Plan

CPPIB Canada Pension Plan Investment Board

CPP/QPP Canada Pension Plan and Quebec Pension Plan

CRA Canada Revenue Agency

CSPP Canada supplementary pension plan (as proposed by Keith Ambachtsheer)

DB  defined benefitDC  defined contribution

DPSP deferred profit sharing plan

GIS Guaranteed Income Supplement

GRRSP group registered retirement savings plan

HOOPP Healthcare of Ontario Pension Plan

IRA US individual retirement account

JEPPS Joint Expert Panel on Pension Standards (to advise the governments of Alberta and BC on workplace pension issues)

JSPP jointly sponsored pension plan MEPP multi-employer pension plan

OAS Old Age Security

OECP Ontario Expert Commission on Pensions

OMERS Ontario Municipal Employees Retirement System

PPIC Pension Plans in Canada (Statistics Canada survey)

PRP Pension Review Panel (to advise the government of Nova Scotia on workplace pension issues)

QPP Quebec Pension Plan

RIS retirement income system

RPP registered pension pla

RRIF registered retirement income fund

SiPP simplified pension plan

RRSP registered retirement savings plan

TFSA tax free savings account

YBE year’s basic exemption (under the CPP/QPP)

YMPE year’s maximum pensionable earnings (under the CPP/QPP)

Autres publications

Aînés Économie Régime de pensions

Articles connexes d’Options politiques

La supercherie des retombées économiques

par Yvan Stringer 24 janvier 2023

Le modèle québécois pour résoudre la pénurie de main-d’œuvre

par Yves Blanchet 29 août 2022

Recherches connexes du Centre d’excellence

The Federal Spending Power in the Trudeau Era: Back to the Future?

par Peter Graefe et Nicole Fiorillo 7 juin 2023

Provincial/Territorial Governments and the Negotiation of International Trade Agreements

par Christopher J. Kukucha 18 octobre 2016