The Institute for Research on Public Policy recently published a paper by economist Keith Horner entitled A New Pension Plan for Canadians: Assessing the Options, in which he concludes Canadians would be better served by proposals to expand the CPP, which delivers DB pensions, over other options such as the pooled registered pension plan (PRPP) that deliver pensions on a DC basis.
The DB versus DC debate drives me crazy in that it perpetuates a number of DB plan myths that seem opaque to even those who have a fair degree of financial sophistication, including many pension experts.
Here are some of the points that are ignored in the acceptance of these DB myths along with some very simple math that shows how DC benefits can be better than DB.
1. The best lesson I ever learned from a wise actuary is that “money in equals money out.” Although this sounds trivial, when it comes to DB pension funding it is anything but, as the dimensions of time and pooling of risks/rewards add a high degree of complexity. Thus regardless of whether a plan is DC or DB, the simple math is as follows:
2. DB plans and DC plans can be invested identically across a diversified portfolio. Offering investment choice is a weakness for most DC plans—the administrative complexity of offering investment choice drives up administration costs, and when averaged among a cohort with many unsophisticated investors, investment returns fall short of those in a pooled diversified portfolio.
Removing investment choice and mandating investment in an appropriately diversified portfolio removes investment results as a differentiator between DB and DC, except for the effects of timing of contributions/benefits relative to individuals. Thus In becomes a constant “k” for comparisons between pure DB and pure DC:
3. DC with no investment choice is simpler to administer than DB, thus lowering expense payments, which leaves more for benefits.
The formulae above do not account for the following:
- Longevity risk pooling, which is a feature of DB that is absent from DC (except to the extent that a DC plan may offer life annuities at retirement)
- Timing of contributions relative to delivery of benefits (e.g. risk pooling between age cohorts as articulated in the Horner paper), which can also be characterized as intergenerational funding
These are critical differentiators that result in requirements for intergenerational funding increases when the following population and economic factors are present.
- Increasing longevity
- Demographic bulge (e.g. baby boomers) moving into benefits phase
- Recessionary periods
Of course, intergenerational funding requirements can decline if any or all of the above factors reverse direction, but often the tendency is to instead increase benefits while holding funding steady or even to reduce funding through a “contribution holiday,” as has been widespread practice in the private sector.
Alternatively, excess funding reserves could be built up to weather periods of adverse population and economic factors, but few governments—Canada being a notable exception in respect of the CPP—and almost no private plan sponsors have perceived the wisdom of building such reserves. This is why there is worldwide concern, at crisis levels for many countries, about pension funding for all models.
The Horner paper is an excellent treatise. However, its principle failing is that the risks of the “critical differentiators” I describe above have not been assessed.
How can the significance of such risks be ignored in the context of riots in Europe and widespread media coverage of the state of public sector pension plans in the U.S. and elsewhere? The current policy direction being followed in Canada is to preserve the stability of our current government programs and to enhance them with stronger, more universal DC arrangements (in the form of the PRPP framework).
The former provides a strong, basic level of pension, while the latter can empower individuals to ensure their own retirement income security with far less exposure to the demographic and economic risks that have undermined the security of DB plans worldwide.