There is a lot of momentum lately around the comeback of DB plans. We see continuous media coverage on how unprepared Canadians are for retirement and how hard it will be to save in the future, ongoing discussions about pushing the retirement age back for CPP/QPP and OAS to 67 or some other age beyond the current age 65, and the excitement around CAAT’s new DBplus arrangement for companies that want DB benefits and DC costs. Could all of this mean that DB is on the way back as I wrote last October?
A new report just published by the Canadian Public Pension Leadership Council (CPPLC) adds to the momentum in favour of DB. The Social Implications of Pensions is authored by Robert Brown, Professor Emeritus at the University of Waterloo. But before we just put the report on the ‘in favour’ side of the DB vs DC scale, I took the time to read the report and I want to share my thoughts.
The report reaches many conclusions including:
Let’s take a closer look at each of these:
DB plans promote worker wellbeing
To get to the bottom of this you have to dig back to media reports that Canadians are worried about insufficient savings for retirement and then draw the link that those with generous DB benefits worry less.
The root of the conclusion seems to be a report from Towers Watson from 2017 which shows that the ‘long-term financial confidence’ of workers has eroded over the past few decades and draws the causal link that new school DC plans are not generating the same level of retirement income as the old school DB.
But wait, there doesn’t seem to be clear evidence that the road to worker wellbeing is a DB pension as much as the road to worker wellbeing is in fact an adequate pension – DB or DC. There is no doubt that comparing DC pensions which cost one-half or one-quarter of their DB cousins is an apples to oranges game. The question we need to better understand in the DC context is how did the money not spent on pensions contribute to worker wellbeing? Did the employer spend the funds on better health care or paid time off? Did the employee spend it paying off a mortgage or going to school part-time? If the money went to fast cars or shareholder dividends that is probably not the answer we want.
Canadians want DB plans – and they are willing to pay for them
This conclusion is drawn from a CPPLC survey from 2017 titled The Pensions Canadians Want.
Dig a little deeper and here are four plan design features to which respondents replied positively:
Let’s be honest, none of these priorities say DB. What they say is employer sponsored retirement plan. DC plans can have matching employer contributions, contributions off every paycheck, and can be managed professionally by individuals with ‘aligned interest’. In fact, some would argue that the interests of the DC plan advisor community to see greater funds saved and higher investment returns earned is better aligned with the interests of plan members than DB managers that might be rewarded for reducing contributions to a DB plan beyond a prudent level – see Sears Canada.
I will address the ‘managed at a low cost’ in the next section.
The survey also directly asks what features in a plan would justify higher contributions by members. The answers:
On the surface this sure sounds like DB. But the first three can also be managed through an annuity at retirement in a DC plan. And the last one, is mostly just ‘more money in’ equals ‘more money out’, subject to reasonable parallels in the cost of management which is the issue I tackle next.
DB Plans deliver higher and more efficient income security
What actually triggered me to write this commentary was the statement:
“The analysis shows a huge advantage for workers in a DB pension plan compared to those saving for retirement using individual savings programs such as RRSPs. A worker in a DB plan contributes the same amount as a worker in an RRSP savings program but receives a pension amount in retirement that can be 2.2 times greater.”
I have heard the ‘DB gives you twice as much pension’ before. As a macro economist I am skeptical of this claim. If DB was really twice as efficient then we would not have had the mass migration to DC that we have seen in the last 30 years. Even people that are not very good at math recognize a super-good deal intuitively. I am not reluctant to concede that ‘large DB’ can be more efficient than ‘small DC’, but I am completely unwilling to accept the 2.2x multiple at virtually any size. We also know that at the under $100 million size plan – the cost advantage of DB vs DC is even more suspect.
Before I ask you to agree that I am right – let me offer a few facts and some math. First, how much does it cost to run a large DB plan? Costs will vary but 30 basis points is a reasonable proxy to the results you can find in publicly available annual reports. Some plans will be lower – others higher. Second, how much does it cost to run a DC plan. Not an easy answer since many factors including number of members, size of assets, and robustness of recordkeeper services are all factors. In addition, there are very few DC plans that exceed a billion dollars so when making comparisons it is hard to break out the difference in costs between the type of the plan versus the sheer size of the assets that have been accumulated long term. All in, a DC plan that has accumulated a reasonable level of assets per member will have a cost less than 100 basis points. For sake of argument let’s say a DC plan costs 80 basis points.
What is the impact of having DC dollars earn 50 basis points less each and every year over an entire career as compared to DB? About 12%. Now my math is back of the envelope here – but there is no way that a 12% advantage becomes a 120% advantage by fine tuning my numbers a little. So how does my work and the work of Professor Brown differ?
First, when I look at DC plans, I think about employer sponsored programs – which can be Group RRSPs or DC registered pension plans. Either way the costs are not too different than noted above. Second, I am focused on how much capital is saved at retirement whereas Professor Brown goes further to assume a DC plan member will annuitize at retirement (essentially invest in low risk bonds and mortgages) while a DB plan member will continue to enjoy the superior returns generated by higher risk equity investments.
In Professor Brown’s paper – he assumes that the cost difference between a DB plan and an individual RRSP is 276 basis points. This differential has no bearing in the employer sponsored DC environment. Secondly, very few DC plan members annuitize at retirement and increasingly, the DC industry is offering Target Date funds that continue to invest in equities ‘through retirement’ as a recognition with a possible 30- or 40-year lifespan after retirement, taking no investment risk at age 65 might not be the most prudent choice.
What have we learned?
Here is what I think we can really learn from the large DB plans and the small DC plans.