Fraser Institute report misses the mark on retirement security

The best measure of the political saliency of expanding public pensions might be the progression of conservative think tank studies purporting to show that there is no need for an expanded CPP or for the Ontario Retirement Pension Plan.

In the past few months, we have heard:

These studies suffer from significant methodological weaknesses and inconsistencies, and massively oversell their results.

Michael Wolfson clearly analyzed the limitations of the McKinsey and Hamilton studies in a CCPA-Ontario report earlier this month.[1]

The Fraser Institute study’s main conclusion—that CPP premium increases largely replace private retirement savings—is fundamentally irrelevant to the pension debate, even if it is true. It is irrelevant because the pension debate is about retirement income adequacy, not contribution rates.

However, the Fraser Institute study doesn’t even establish that past CPP premium increases have replaced private savings. It relies on a statistical model of retirement savings whose results fall far short of any reasonable standard of statistical significance. The authors themselves warn readers that their results “must be interpreted cautiously. CPP changes between 1996 and 2004 did not occur in a vacuum; they coincided with many factors that may have affected the savings behaviour of households in unpredictable ways.”

That may be the only understatement in the paper. The best result from their primary model of retirement savings explains only 25.9% of the variation in savings rates; depending on the age group, the statistical power of their study ranges from 21.0% to 25.9% of the savings variation. Their alternative model’s explanatory power peaks at 5.3% for the youngest age group, and is even lower for other age groups.

As noted above, the most troubling aspect of the Fraser Institute study is its explicit assumption that savings in the form of CPP contributions are equivalent to private RRSP savings. In making that critical assumption, the Fraser study inadvertently highlights the fundamental reason why CPP expansion is a better deal for Canadians.

RRSP contributions are not equivalent to retirement savings through contributions to a universal, defined-benefit pension plan like the CPP. CPP contributions are significantly more effective in generating retirement income than RRSP contributions.

The CPP advantage is attributable to four key factors: investment returns; fees; longevity risk; and inflation protection.

Investment returns

A large fund like the CPP is able to take advantage of investment opportunities that are simply not available to individuals investing privately through mutual funds. A study in the United States comparing the investment returns of large defined contribution funds (401(k) plans) with those of Individual Retirement Accounts (the American equivalent to RRSPs) have found that large fund returns exceed those of IRAs by 1.8% annually. That might not sound like a big difference. But, over a working lifetime, the same savings rate results in over 47% larger investments in the large funds than in individual funds.


The corrosive effect of investment management fees on retirement savings has recently been highlighted as an issue by the U.S. Securities and Exchange Commission. The significantly higher mutual fund fees paid by Canadians make their impact on retirement savings an even bigger issue in this country. In a 2014 CCPA report, I estimated that 36% of a Canadian’s savings over their working lifetime would be soaked up by mutual fund managers’ fees.

Longevity risk

A pension plan is better positioned to manage longevity risk—the risk that you will outlive your retirement savings—in two respects. The most obvious advantage is that it is far more expensive for an individual to guarantee a retirement income than it is for a pension plan. It is far more cost effective to pool longevity risk over a large population than to protect against that risk as an individual. In my 2014 CCPA report, I estimated that even to reduce the probability of running out of retirement savings to 25% would require 18% more retirement savings.

The other longevity risk management advantage for pension plans is more subtle, but equally significant. In general, prudent investment management requires that, the closer you are to retirement, the more conservative your investment portfolio should be. As they age, investors trade higher returns for greater certainty. Because a pension plan pools the retirement savings of people at all different life stages, on average pension plans age only very slowly, if at all. As a consequence, pension plans can take advantage of the benefits of investment risk pooling to seek higher investment returns.

A recent analysis by the Canadian pension performance measurement organization CEM found that large defined benefit plans earned 1.5% annually more than large defined contribution plans, and that most of that difference—about 1.3%—was attributable to differences in asset mix, or investment portfolio risk.[2]

Inflation Protection

In addition to these advantages, in the current Canadian annuity market, it is not possible to replicate through individual savings the inflation protection provided through the Canada Pension Plan. Indexed annuities are simply too expensive.

These key advantages for the CPP and by extension the ORPP relative to individual retirement savings are additive. Each contributes separately and independently to a massive efficiency advantage for the CPP/ORPP.

In my 2014 report, I compared the cost of a given level of retirement income through a defined benefit pension plan with the cost of providing the same income through RRSPs the cost advantage for the pension plan is dramatic.[3] The percentage of pay required as a contribution to fund the RRSP option would be between 2.4 times and 2.5 times the percentage of pay required to fund the pension option.

This is a conservative estimate of the advantage, given that the comparison was biased in favour of the RRSP option. The RRSP option took no account of the inflation protection cost advantage for the pension option and was based on a 25% probability of outliving retirement savings, not the 0% probability provided through the pension option.

The inefficiency disadvantage for the RRSP option is staggering, and begs for an explanation as to why we are having this debate at all.

The answer clearly lies in conservatives’ ideological opposition to public policy responses when a private alternative is available.

As a look behind the Fraser Institute study’s casual assumption that private savings and CPP contributions are equivalent in generating retirement income demonstrates, that ideological preference comes at an extremely high cost to individual retirement savers and to Ontarians and Canadians generally.

Hugh Mackenzie is a CCPA research associate. Follow him on Twitter @mackhugh.

[1] See also Michael Wolfson’s state-of-the-art study, “Projecting the Adequacy of Canadians’ Retirement Incomes: Current Prospects and Possible Reform Options.” no. 17. April 2011.

[2] Source: CEM Benchmarking Inc.

[3] Key assumptions: longevity based on the 2010 CPP Life Tables for Ontario males and females; inflation, 2.0%; annual employment income increase 3.0%; Defined Benefit returns 6.0%; RRSP/DC returns 4.5% until age 65; 3.5% thereafter. The lower RRSP return after age 65 reflects the greater degree of conservatism in investment strategy in the period during which retirement income is being paid out than in the period in which assets are being accumulated. The 1.5% and 2.5% differentials in returns pre-and post-65 between DB and RRSPs are extremely conservative, given the much larger differences in fees, gross returns and asset mix return potential observed empirically between large DB plans and individual retirement savings plans.

One comment

  1. Wow. I knew the advantage of a public plan was substantial, but I had no idea the difference was that huge.

Join the Discussion

Your email address will not be published. Required fields are marked *

Before commenting, please read our Comment Policy