Keith Ambachtsheer chimes in on pension reform
“Traditional problem-solving follows a linear model: gather data, analyze it thoroughly, formulate a solution, implement it. Designers follow a decidedly nonlinear model. While they also begin by trying to understand the problem, they almost immediately move to formulating potential solutions, and then jump back to refining their understanding of the problem. Far from orderly, the line plotting their thinking looks more like a seismograph for a major earthquake. Designers know that understanding of a problem can only arise from creating possible solutions, and that this understanding continues to evolve until the very end of the process.”
— Roger Martin
Rotman School of Management
University of Toronto
From debates to decisions
Over the course of the last five years, Canadians have engaged in (or have been observers to) an increasingly intense debate about pension reform. Three major themes have emerged:
1. Pension adequacy: It is a fact that ¾ of the private sector workforce does not have an employment-based pension plan, and that proportion is rising. At the same time, the need for retirement income is growing, as we continue to live longer. The debate is about whether this constitutes a problem, and if it does, how the problem is best addressed.
2. Cost-effectiveness: It is a fact that every percentage- point of additional annual costs charged to manage retirement savings (i.e., related to investment, administration, advice, or distribution) reduces the ultimate pension by some twenty percent. It is also a fact that the total cost of pension/retirement savings management and administration varies greatly (i.e., there can be a two percentage-point differential between the annual cost of managing a large-scale collective pension plan and the fees individuals pay when they place their RRSPs in actively-managed retail mutual funds). The debate is about whether this is a problem, and if it is, what should be done about it.
3. Defined Benefit pension plan sustainability: some 4.5 million workers are still members of DB plans (2.5 million in the public sector, 2 million in the private sector). The debate is about whether these plans are sustainable in their current forms, and if they are not, what should be done about it.
At their pre-Christmas meeting in Whitehorse, the federal and provincial finance ministers and their officials announced that they would engage in intensive consultations on these challenges, leading up to their next meeting sometime late May of this year, which in turn is a precursor to a fullfledged Federal-Provincial meeting on pension reform with the Prime Minister and Premiers in August. The Federal Government formally joined this consultation process on March 24, announcing it would conduct a series of coast-to-coast town-hall and roundtable meetings in April. A recent survey by TNS Canadian Facts indicates that ⅔ of Canadians agree that “the current system does not meet the needs of the average Canadian and should be reformed”.
In short, 2010 could see Canada moving from debates about pension reform to making actual decisions.
We have had the good fortune to be an active participant in this process since the beginning both as a contributor, and as a listener. Here, we offer our updated thoughts on each of the three major issues set out above, and how they might best be resolved. In doing so, we are mindful of Roger Martin’s observation that the discovery of better answers should continue until the very end of the design process.
4 recommendations
Before commenting on the three major challenges set out above in some detail, we summarize our recommendations on how they can best be addressed immediately below:
1. Simplify and modernize the current rules and regulations governing workplace and individual pensions to create greater flexibility.
2. Graft a national supplementary pension plan unto the current CPP/QPP in a way that enhances pension adequacy while maintaining employer and employee flexibility at the same time. This could be combined with some modest, targeted CPP/QPP expansion.
3. Create a task force mandated to find the best way for all Canadians to have the opportunity to access low-cost, but still expert pension management services.
4. Require ‘fair value’ accounting and positive risk buffers in defined benefit plans in order bring their sustainability and transparency up to the same prudential standards OSFI sets for insurance companies and other financial institutions.
Together with their public pensions (CPP/QPP/ OAS/GIS), the adoption of these four measures would give Canadians the best retirement income system in the world.
Pension adequacy
While there is broad agreement that ideal pension systems allow people to maintain their standard of living after they stop working, the debate about how to translate that ideal into a target earnings replacement rate continues. Is it the traditional seventy percent embedded in many DB plans and in the sales literature of the financial services industry? Or is fifty percent a better norm in the presence of a mortgage-free family residence, and the absence of work- and child-raising expenses? This is not a trivial question. For example, a new C.D.Howe Institute paper by Dodge, Laurin, and Busby calculates that a worker aged thirty who earns an inflation-adjusted $60K per year over thirty-five years will have to save about fourteen percent of pay to achieve a seventy percent earnings replacement rate (including the public OAS/ CPP pensions). The required savings rate drops to eleven percent for a sixty percent earnings replacement rate, and further to nine percent if the worker retires at age sixty-seven rather than sixty-five. There is a growing understanding that seventy percent is unnecessarily high for many people, and that sixty, or even fifty percent may be more appropriate in many cases.
Just as there is still a debate about the ‘best’ earnings replacement target to use for planning purposes, so is there a debate about whether Canadians are saving enough to achieve it. Everyone agrees that some are, and others are not. The debate is about the respective proportions of the ‘are’ and ‘are not’ groups. In an important sense, the debate is moot. Yesterday’s retirement savings behaviour, what ever it may have been, is not necessarily a good reflection of tomorrow’s. Similarly, the almost ideal investment environment of the 1980s and 1990s did not continue in the decade just ended, and may not in the years ahead. In this context, a more relevant policy question is: what can we do today to ensure adequate savings rates tomorrow? Or more specifically, what can we do today to give all workers a reasonable chance to achieve a post-work income target in the fifty to seventy percent range of working years earnings?
3 possible responses
The public pension combination of OAS/GIS/CPP by itself achieves high earnings replacement rates for low-income Canadians. For example, a couple, both aged sixty-five, with maximum government pension benefits, receives an inflation-indexed annuity of $34,218 today. This implies ‘adequacy’ policies should focus on Canada’s middle-income workers without pension plans. This group of some 3.5 million Canadians is likely to work in the private sector for medium- and small employers, or be self-employed. A supplementary pension arrangement that is automatic and low-cost would likely be a useful tool to help these workers achieve postwork income adequacy.
Three proposals have emerged to achieve this goal:
1. Allow ‘the market’ to work: Many in the financial services industry believe that the adequacy problem can be solved by removing the legal restrictions that prevent providers from offering the large-scale, multi-employer, multiprovincial plans that currently do not exist in Canada. More education and advice would lead to more informed participant decisions, and ‘auto-enrolment’ into these types of plans would materially enhance coverage.
2. Create a national supplementary pension plan: Advocates see this plan operating under the CPP/QPP legal umbrella, on an ‘arms-length’ basis from government. It would be low-cost, set a ‘default’ earnings replacement rate target, a consistent ‘default’ contribution rate, and a ‘default’ age-based investment policy. All workers without a pension plan would be autoenrolled to enhance coverage. Workers own their pension accounts. The plan is voluntary in the sense that workers and employers could override any of the ‘defaults’ if they wish, including opting-out altogether.
3. Expand the CPP/QPP: Advocates see this is the simplest way to enhance adequacy. Participation in the CPP (or QPP) is already mandatory, and all of the relevant low-cost investment and administration machinery already exists. There is still debate about what form any expansion should take beyond the current target to replace twenty-five percent of earnings up to the YMPE ($47K in 2010). Some believe the replacement rate up to the YMPE should be doubled; others believe the replacement should stay at twenty-five percent, but that the YMPE should be doubled; still others believe both should be doubled. There does seem to be consensus that, whatever the formula, it should be phased in over time on an actuarially-fair basis.
We assess the pros and cons of each proposal next.
Weighing the pros and cons of the 3 proposals
We see the pros and cons as follows:
1. Allow ‘the market’ to work: Letting ‘the market’ sort it out sounds like a strong ‘pro’ in principle. However, it requires motivated, knowledgeable consumers for it to work in practice. With its ‘Free Choice’ legislation, the Howard government set Australia on an ‘allow the market to work’ route in 2004. A 2008 study titled “Choosing Not to Choose” (Fear and Pace, The Australia Institute) found the legislation is not producing the desired results (e.g., it has not lowered member costs). The current Rudd government has acknowledged this market failure, and has launched a commission to study how the mandatory Australian superannuation system can become more costeffective. The Super System Review report is due mid-2010.
2. Create a national supplementary pension plan: On paper, this approach promises extended coverage, targeting, flexibility, and low cost in one package (see our 2008 CSPP proposal). In fact, the above-cited survey by TNS Canadian Facts indicated that eighty percent of respondents liked the idea of supplementing CPP/ QPP benefits with additional contributions on a voluntary basis. A potential ‘con’ is that this kind of arrangement has never existed in Canada on a large scale. Fortunately, there are operational models to learn from. For example, Saskatchewan’s Cooperative Superannuation Society Pension Plan (CSSPP) has successfully operated a similar structure since 1939. The Turner Commission in the United Kingdom came to the conclusion in 2004 that this was the right approach to extending pension coverage to seven million British workers without a pension plan. Legislation was passed in 2007. The Pension Accounts Delivery Authority (PADA) began implementation work in 2008, which continues as we write this. Operational testing is scheduled for 2011, followed by a full system rollout in 2012. If Canada decides to go this route, it could learn a great deal from the experience in the United Kingdom. At the same time, the CSSPP already has decades of successful operating experience in Canada. Finally, there is no reason why some of the proposed national supplementary pension plan functions could not be outsourced to the private sector.
3. Expand the CPP/QPP: This approach also promises increased adequacy and low operating cost in one package. But as it requires mandatory participation, it cannot offer flexibility; whatever specific ‘expand’ strategy is chosen will have to apply to all. However, some targeting is possible. If middle-income, private sector workers without pension plans are the right target group, expanding the current CPP / QPP benefit (i.e., twenty-five percent earnings replacement) beyond the current YMPE ceiling (e.g., doubling it from $47K to $94K) would serve them best, and would require a six percent of pay contribution rate on income between the old YMPE and the new one. Of course, middle-income earners with pension plans would be equally impacted, requiring equivalent benefit and contribution rate reductions in those plans to ensure overall pension benefits and costs remain the same.
All considered, it seems to us that: (a) simplifying and modernizing pension rules and regulations is a ‘no lose’ proposition, (b) the most direct, targeted but flexible route to enhancing future pension adequacy in Canada is to graft a national supplementary pension plan to the CPP/QPP foundation, and (c) step (b) could be combined with some modest, targeted CPP/QPP expansion. Neither (a), (b), nor (c) requires any major upfront expenditures.
The cost of managing Canada’s retirement savings
Just as enhancing pension adequacy in a targeted manner is an important pension reform goal for Canada, so is ensuring that retirement savings are converted into pension payments in a cost-effective manner. Recall that every percentage point of additional annual cost in managing retirement savings reduces the ultimate pension by some twenty percent. So what is the actual cost structure for managing and delivering pensions in Canada today? Table 1 provides some benchmarks. We divide total accumulated Canadian retirement savings into ‘wholesale’ and ‘retail’ channels. The ‘wholesale’ channel holds the accumulated assets of the CPP and QPP, and those of public and private sector employment-based pension plans. The ‘retail’ channel holds the retirement savings of individuals in such categories as RRSPs, RRIFs, LIFs, and LRIFs. These savings can be invested through a variety of sub-channels ranging from segregated funds, mutual funds, deposits, and self-directed accounts.
What does participation in each of these two channels cost? There is surprising cost variability within each channel. For example, the CEM Benchmarking Inc. database confirms that some pension funds operate at very low cost (e.g., under 0.25 percent per year), while others incur higher costs (e.g., over 0.5 percent per year.). Data from other sources confirm that the range of cost experience within the retail channel is even greater. At one end, Canadian ETFs and index funds may charge fees in the 0.3 percent-1.0 percent area, while at the other end, the fees of actively-managed mutual funds can exceed 3 percent. Having made this important point, it suffices here to focus on the average cost differential between the two channels. In the Table 1 calculations, we assume a 0.4 percent average cost for the ‘wholesale’ channel, and a 1.6 percent average cost for the ‘retail’ channel. Based on available information, we believe these are reasonable assumptions (see for example, a recent paper by Jog for the Research Working Group on Retirement Income Adequacy).
Using 2007 Statistics Canada estimates for accumulated retirement savings and our own average cost estimates, Table 1 indicates Canadians are spending almost $15.6B per year (0.9 percent of assets) to have their retirement savings managed, the bulk of it spent for retail channel services. What if the millions of retail channel savers could pay wholesale rates? Table 1 indicates their costs would decline by $8.4B per year., or equivalently, their retirement savings would grow by an additional $8.4B per year. Stated still differently, the resulting 1.2 percentage point reduction in annual costs is equivalent to a 24 percent boost in the ultimate pension the retirement savings can purchase.
Wholesale or retail? A Question of ‘Value for Dollars’
Table 1 raises the important question of why all retirement savers should not have the opportunity to pay wholesale fees. Addressing it requires recognizing the financial services industry is the beneficiary of the current annual $11.2B ‘retail’ channel cash-flow, which calculations show would fall by $8.4B to $2.8B with ‘wholesale’ pricing. The financial services industry argues that surveys show their clients are satisfied with current arrangements, and that the additional $8.4B per year represents unavoidable costs (e.g., daily valuation and liquidity). More importantly, the money buys their clients valuable advice and almost unlimited choice.
In response, we question why long term retirement savers should have to pay for services they do not need (e.g., daily valuation and liquidity). The ‘valuable advice’ argument would be more persuasive if there was evidence that this advice is actually producing higher risk-adjusted returns for clients. In fact, the evidence points the other way. Empirical studies in Canada, the United States, and Australia all confirm what theory predicts: the higher the average costs of investing, the lower average net returns. Finally, behavioural studies confirm what common sense tells us: retail investors have far too much choice. In fact, most do not want to choose at all.
So from a design perspective the question is clear: how can we best help millions of Canadian retirement savers who want adequate pensions at affordable savings rates, but who don’t want to get mired in the complexities of investing? There are two broad possibilities:
1. A retail approach involving hundreds of investment choices and tens of thousands of financial advisors.
2. A wholesale approach based on a series of automatic ‘default’ settings (e.g., for autoenrolment, auto-escalation of contribution rates, and age-based auto-investment policies).
Too often, we frame choices in ‘either or’ space when the right question is ‘why not both?’ People should have the option to go the retail route. Indeed, it is possible to go this route and pay wholesale fees (e.g. 0.4 percent per year). However, it is equally true that many retail savers are steered to high-cost options without understanding the consequences. These people would benefit from an automatic, low-cost, collective alternative.
Designing the wholesale option
Canada is fortunate to be home to some of the bestmanaged pension plans in the world. By ‘bestmanaged’ we mean that by global standards, they score well on a ‘value-for-dollars’ basis. These plans have five common characteristics:
1. Arms-length: close alignment of interests between the plan and plan participants.
2. Well-governed: plan oversight is a provided by boards of trustees/directors who (a) have appropriate skill/experience sets, and (b) are motivated to represent the interests of plan participants.
3. Sensible investment beliefs: they do not engage in value-reducing trading strategies, but instead focus on acquiring sustainable cash flows at reasonable prices.
5. Scale: are large enough to assemble the requisite resources at low unit costs.
5. Competitive: have compensation structures that permit them to attract top talent.
In addition to producing value for plan participants, organizations with these characteristics also produce an important positive ‘externality’. They are both motivated and well-equipped to act as responsible shareowner stewards with respect to the corporations they invest in. Research confirms that responsible corporate governance action by knowledgeable long-term investors improves corporate behavior and long-term profitability.
The current pension reform dynamic creates an opportunity to examine how millions of Canadians needing a low-cost, expert management option for their retirement savings could be connected to Canadian pension plan organizations already providing such a service. At the same time, we should recognize that commercial organizations also have valuable experience and expertise to bring to the table. The challenge is designing a structure that delivers that experience and expertise at wholesale pricing. All this leads us back to the British decision some years ago to create a working group (PADA) to create an expert, low-cost personal pension account delivery design suited to the United Kingdom context. Canada should undertake a similar initiative to examine how this could best be done in a Canadian context, in concert with the decision to create a national supplementary pension plan.
Defined benefit plan sustainability problems
Over the course of the last 50 years or so, the pension promises in defined benefit plans have evolved from voluntary gratuities without advance funding, to enforceable financial contracts with advance funding. Defined benefit plan regulation has lagged in responding to this reality. This lag has had two types of consequences:
1. Solvency problems in the private sector: the regulation and supervision of defined benefit (DB) plans in Canada continues to be fundamentally different from the regulation and supervision of insurance companies and banks. Our regulation/supervision processes ensure that these financial intermediaries can meet their payment obligations by requiring them to carry prescribed levels of risk capital on their balance sheets. We do not hold the sponsors of DB plans to the same standard. First, our regulation/supervision processes allow sponsors to understate the ‘fair value’ of accrued pension payment obligations. Second, we allow sponsors to undertake material balance sheet mismatch risk. And third, we do not require that sponsors hold prescribed levels of risk capital on their balance sheets as buffers against adverse events. These realities created the recent Nortel and General Motors Canada debacles, and will likely create similar debacles in the years ahead.
2. Cost problems in the public sector: a 2009 study by the British-North American Committee recalculated the net pension liabilities owed by Canada’s federal and provincial public sectors on a ‘fair value’ basis. Instead of the reported $190B in net liabilities (twelve percent of GDP), the amount jumped to $422B (27 percent of GDP) when the yields on inflation-linked government bonds were used to discount future pension obligations. On the same discount basis, the most recent actuarial report of the federal Public Service Plan estimates plan benefits cost 34 percent of pay, well above the current 18.5 percent contribution rate. This systemic understatement of pension liabilities and costs is problematic. It leads to a material overstatement of the fiscal soundness of Canada’s public finances, and to a material understatement of the true cost of public sector compensation.
Below we describe how the Dutch came to grips with these defined benefit plan sustainability problems.
A dose of Dutch logic
The Dutch Central Bank (DNB) supervises all financial institutions (i.e., banks, insurance companies, pension plans) in the Netherlands. Prior to the bursting of the equity market bubble in 2001, the DNB had been lax in its supervision of pension plans, as Canada continues to be to this day. The 2001 market crisis led DNB to the conclusion that there is no logical basis for treating promises to pay pensions differently from promises to pay insurance claims, or promises to pay interest and repay principal on term deposits. DNB issued a letter to Dutch pension plans to that effect in September 2002. It stated that it would commence regulating pension plans using the same principles as for insurance companies and banks (e.g., ‘fair value’ accounting and positive risk buffers) starting January 1, 2004.
For a while, the letter caused a great hue and cry in the Dutch pension management community. DNB was accused of everything from destroying the vaunted Dutch pension system, to sheer pigheadedness. In the end, DNB prevailed, although it did push back the start date of the new regulatory regime to January 1, 2005. The result has been a significant shift in Dutch pensions from ‘hard’ to ‘soft’ promises, contingent on ability to pay. The rethinking of the nature of the pension promise in the Netherlands continues to this day, with the recognition that fuzzy pension contracts lead to fuzzy property rights, which in turn continue to threaten pension plan sustainability in difficult economic environments.
What are the lessons in all this for Canada? We should start by accepting the reality of the ‘from gratuities to financial contracts’ story. Then we should discuss the implications of that reality. For example, hard guarantees are neither afforadable nor sustainable today. Once we accept that, then we can move on to the difficult but necessary discussion about how we transition DB plans from a hard guarantees regime to a softer, more sustainable one. We should not fool ourselves into believing that DB plan sustainability can be achieved with band-aid solutions such as tweaking the solvency funding rules, or the rules governing bankruptcy proceedings. Nor does the answer lie in creating a new pension insurance scheme, as some suggest. The financial difficulties of America’s Pension Benefit Guaranty Corporation are increasingly visible to all. Only the prudential requirements of ‘fair value’ accounting and the maintenance of positive risk buffers can create genuine defined benefit plan sustainability. Canada could achieve this sustainability goal after a transition period if it chose to follow in the prudent footsteps of the DNB.
The best pension system in the world
The research, discussions, and debate that have taken place over the course of the last five years have now placed Canada in a position to lead the world in pension system design. In the spirit of our Olympic ‘Own the Podium’ campaign, we should aim to move from having the world’s #4-ranked pension system in terms of quality, to #1. The reforms of the 1990s have already given us a strong public pensions component. We now know where Canada needs to do better in its supplementary pensions component. In our view, four steps would lead Canadians to the top of the pension podium:
1. Simplify and modernize the current rules and regulations governing workplace and individual pensions to create greater flexibility.
2. Graft a national supplementary pension plan unto the current CPP/QPP in a way that enhances pension adequacy while maintaining employer and employee flexibility at the same time. This could be combined with some modest, targeted CPP/QPP expansion.
3. Create a task force mandated to find the best way for all Canadians to access expert, low-cost pension management services.
4. Require ‘fair value’ accounting and positive risk buffers in defined benefit (DB) plans in order bring their sustainability and transparency up to the standards of insurance companies and other financial institutions.
The question now is: can we muster the collective will to actually do this? We could find out this year.
Keith Ambachtsheer is director of the Rotman International Centre for Pension Management at the University of Toronto and president of KPA Advisory Services, an international, strategic pension consulting firm. For more information, visit www.kpa-advisory.com.