Comparing OTPP to PSPP?

Frederick Vettese, the former chief actuary of a large actuarial firm and the author of Retirement Income for Life: Getting More without Saving More, wrote an interesting opinion column recently, How does the Ontario Teachers’ pension plan stack up against the federal public servants version?:
As secure workplace pensions continue to die out across the Canada, the Ontario Teachers’ Pension Plan (OTPP) and the federal Public Service Pension Plan (PSPP) are two gold-plated retirement plans that remain intact. So just how good are they – and how do they compare with one another?


These plans are large by any measure. At last count, the OTPP has 323,000 active members and pensioners while the corresponding figure for the PSPP is 590,000. Based on 2017 asset figures, that comes out to $587,000 per person in the OTPP and $327,000 per person in the PSPP.

Retirement age

Most Canadians think of age 65 as “normal retirement age.” In the public sector, there is nothing normal about retiring that late. The median retirement age in the public sector has been as low as 57.2 in the past, though it has since risen to 61.3.

In the case of the OTPP, teachers can retire without penalty once their age plus “qualifying years” total 85 or more. For example, a teacher with 32 qualifying years can retire without penalty at age 53.

The early retirement rules within the PSPP are also generous though they don’t quite match the OTPP. Civil servants used to be able to retire without penalty as early as 55 if they had 30 years of pensionable service. During the government of Stephen Harper, the age 55 condition was changed to 60 for new hires.

In both plans, members can retire even earlier subject to a modest penalty.

Pension benefits

When combined with CPP, the PSPP provides a pension of 70 per cent of the final five years’ average earnings if the member contributed for 35 years. After 65, the pension (including CPP) is actually a little more than 70 per cent because integration with the CPP is not perfect. In addition, members receive OAS pension.

The pension for OTPP members is virtually the same except that the pension payable after 65 is even higher than it is under the PSPP.

Members with 35 years of service under either plan will be very comfortable in retirement. For workers who were raising families or paying off mortgages, which means almost everyone, a 70-per-cent pension (plus OAS) translates into a standard of living in retirement that is significantly higher than what they enjoyed while they were working. With the CPP being enhanced, this can only go up.

Inflation protection

Both plans strive to provide 100-per-cent inflation protection of pensions, meaning that pensions rise each year in line with the annual change in the consumer price index (CPI). For convenience, this feature is referred to as indexation. The difference between the two plans is that indexation is guaranteed in the PSPP but conditional in the OTPP, depending on the funding level. The full, uncapped indexation within the PSPP is a rarity, even in the public sector.

Employee contributions

Cost-sharing in the public sector has been evolving in recent years. Most plans now split pension costs 50-50. This is one reason why employee contributions have been rising over the past decade. Another reason is to pay for the investment losses that occurred during the global financial crisis.

In the OTPP, employees contribute 10.4 per cent on their earnings below $57,400 and 12 per cent on earnings over $57,400 (which we refer to as 10.4/12). Contributions under the PSPP are a little lower at 9.49/11.67 (same threshold of $57,400).

When we add in the employer’s share, total contributions in both plans exceed 20 per cent of pay. In addition, employees can still contribute another 4 per cent or 5 per cent of pay to a registered retirement savings plan. Note that the federal government restricts tax-assisted saving by workers who are not in defined benefit plans (which describes 90 per cent of private sector workers) to only 18 per cent of pay. Another instance of “Do as I say, not as I do." The federal government should be called upon to explain this double-standard or better still, to level the playing field.

Even though the contribution formulas are similar, the 50-50 cost-sharing mentioned above is applied very differently in the two plans. In the OTPP, the active members and the employers share all costs 50-50. In the PSPP, the cost of deficits is borne purely by the government (i.e., the taxpayer). This can be a big deal since a federal plan deficit arising of $20-billion or more is not out of the question.

Approach to funding

In the OTPP, the actuary assumes that the fund will earn a nominal return of just 4.8 per cent. Under the PSPP, the assumed fund return is 6 per cent. The PSPP is thus taking a far more aggressive stance, which can become a problem if future returns are lower than past returns. (There are demographic reasons why this will actually be the case; aging societies – Japan, for example – tend to have lower interest rates.) This is largely why the PSPP’s assets per member are so much lower than under the OTPP.

Overall assessment

The two plans are fairly similar in most respects except for cost-sharing. The fact that PSPP does not extend the 50-50 cost-sharing principle to the sharing of deficits makes the PSPP considerably more generous overall.

In both plans, the contributions being made exceed what private sector workers can contribute to RRSPs or defined contribution pension plans.

In both plans, the early retirement rules are quite generous. This might have made some sense in a previous era when the country had more potential workers than the economy could absorb. That is no longer the case, and in fact, we are on the brink of a time when workers will be scarce, even with high levels of immigration. For at least the next two or three decades, it is in the public interest to encourage people to work longer. The PSPP and the OTPP do exactly the opposite.
This is a great comment by Frederick Vettese and I'd like to discuss it further.

According to Vettese, "the two plans are fairly similar in most respects except for cost-sharing. The fact that PSPP does not extend the 50-50 cost-sharing principle to the sharing of deficits makes the PSPP considerably more generous overall."

No doubt, Ontario Teachers' Pension Plan is a jointly-sponsored plan where the cost of the plan is equally borne by teachers and the government of Ontario. It's not the only jointly-sponsored plan in Ontario. The Healthcare of Ontario Pension Plan (HOOPP) and CAAT Pension Plan are also jointly-sponsored but the employer in this case is not the government of Ontario.

Interestingly, all three plans are fully funded plans which argues for a model where the costs of the plan are shared equally among employees and employers.

One thing Ontario Teachers' adopted a while ago was conditional inflation protection, allowing it to become young again. In a nutshell, because Ontario teachers live longer than most Canadians, there will soon be more retired teachers than active ones (click on image below, from page 11 of OTPP's 2017 Annual Report):

As the demographics of the Ontario Teachers' Pension Plan shift to reflect an aging population, it makes sense to adjust inflation protection retired teachers receive for a period of time when the plan runs into trouble than asking active members to bear most of the cost of the plan (through higher contributions).

Typically, partial inflation protection is introduced when the plan runs into trouble, meaning instead of a guaranteed cost-of-living-adjustment (COLA), OTPP's pensioners receive 80, 70, 60% or less for a period until the plan's fully funded status is restored.

For pensioners, this is a small change to the benefits they receive but for the overall plan, it has a drastic effect on cutting down the pension deficit, and this will increasingly be the case as more Ontario teachers get set to retire, changing the ratio of active to retired teachers.

Conditional inflation protection is also a fair policy to adopt from an inter-generational view, sharing the risk of the plan between retired and active members. The more retired members there are, the more risk they can collectively assume relative to active members (without hurting them as much as active members).

None of this risk-sharing is present in the federal Public Service Pension Plan (PSPP). Federal public-sector employees have guaranteed inflation protection (like OMERS and OPTrust) and if the plan runs into trouble, the federal government (ie. Canadian taxpayers) is on the hook.

According to Vettese, "the fact that PSPP does not extend the 50-50 cost-sharing principle to the sharing of deficits makes the PSPP considerably more generous overall."

You'll recall, this is what Malcolm Hamilton and Philip Cross called the dirty secret behind Canada's pensions. That comment generated many more comments from my readers and ultimately led me to write a lengthy comment on retirement security in Canada.

After writing that last comment, Malcolm Hamilton came at me hard in an email exchange stating this:
So here's the problem with your position. You compare Target Benefit Plans to Defined Benefit Plans that cost twice as much. It's like comparing a Pontiac to a Porsche (I'm dating myself).

Properly priced, a public sector DB plan costs 40% of pay. By changing the guaranteed benefit to a target benefit that will be adjusted as circumstances require (by changing the level of the benefit or the indexing), we reduce the cost of the pension plan to 20% of pay. So yes, the DB pension is much more valuable. But it is much more valuable because it is much more expensive, not because it is a better plan design. I see no evidence that members will voluntarily pay the extra 20% of pay. I believe that most public sector employees would rather pay 20% of pay for the Target Benefit Plan than pay 40% of pay for the Defined Benefit Plan. In other words, the Target Benefit Plan is better value.

In support of my position, I draw your attention to the behavior of employees in DC plans that allow them to choose between safe investments (long term government bonds) and more risky ones. Virtually no retirement saver voluntarily invests all of their money, or even half of their money, in government bonds. When employees retire, few decide to buy annuities. They prefer to take investment risk and, by so doing, to strive for higher returns and, eventually, higher incomes. They could save twice as much and take less risk, but they choose not to do so.

I believe that the same will happen in Target Benefit Plans. The plans will invest in balanced portfolios that will typically deliver good returns and good pensions, but with no guarantees. The plans will take less risk than today's public sector DB plans because employees will no longer be able to pocket the reward for risks borne by the public - hence risk-taking will be less advantageous to plan members.

You may not understand this. You may not want to understand it. But this is how things work. If you feel differently, explain how the public is now rewarded for the investment risks it bears. Or perhaps you think that public employees deserve a free ride? As I wrote earlier, the federal government's DB plan is no different than telling federal employees that they can use 20% of their compensation to buy long term RRBs with 4% real interest rate while offering other Canadians a 1% real interest rate on the bonds they buy.
I have tremendous respect for Malcolm, think he's one of the best-retired actuaries in the country but I'm not in full agreement with his proposed Target Benefit Plan because it leaves retired members of a plan vulnerable to the vagaries of markets.

I shared his comment with HOOPP's CEO, Jim Keohane, who shared this with me:
This is a bit of a “glass half empty “ view. You could also view it that prudent risk-taking, scale, good governance and good management allow Canadian model plans to provide good pensions for 20% of pay which would otherwise cost 40% of pay thus saving taxpayers 20% of pay.

The Federal Public service plan is unique in Canada in that the employer assumes all of the downside risk. Virtually every other public sector plan is a shared risk plan with contingent benefits. In the case of HOOPP, neither the employers or the province guarantee the pension. The only obligation the employer has is to pay their share of the annual contributions as long as they remain members of the plan (which is voluntary). When you consider that COLA is not guaranteed and can be reduced or eliminated should plan funding be insufficient, employees accept most of the risk of underfunding. The notion that taxpayers are taking all the risk and that plan members are getting all the benefit is simply not true.

What we should be focusing on is the efficiency of the conversion of pension savings into pension payments. Our recent research paper “The Value of a Good Pension” shows the efficiency of moving from individual savings plans to collective plans. If more Canadians were fortunate enough to be members of Canada model plans there would be a much larger pool of savings to pay pensions which would benefit all of Canadian society and taxpayers.
I couldn't agree more with Jim on this last point. Jim was actually in Washington D.C. last week where he spoke at a conference held by the National Institute on Retirement Security. You can find details of that event here and download Jim's presentation here.

Below, I share the key findings (click on image):

Back to Fred Vettese's comment above. It's true that OTPP uses a lower discount rate than PSPP and pretty much all other large Canadian public plans but that's because it's a more mature plan and the demographics are different (more retired members).

Other key differences that Vettese doesn't highlight is OTPP is a pension plan that manages assets and liabilities of just one client, Ontario teachers, whereas the federal Public Service Pension Plan's (PSPP) assets are managed by an independent Crown corporation, PSP Investments which is the investment manager for the pension plans of the Public Service, the Canadian Armed Forces, the Royal Canadian Mounted Police and the Reserve Force.

What OTPP, PSP Investments and other large Canadian pensions have in common is world-class governance, allowing them to manage assets in the best interests of the members of their plan.

And that, in my humble opinion, is the most important point to get across when comparing large Canadian public-sector plans.

Below, last year, Ontario Teachers’ net assets reached $189.5 billion, up tenfold since the Plan’s inception; the Plan was 105% funded as at Jan. 1, 2018.

I expect OTPP's 2018 results which come out in early April will be more challenging but the funded status will remain largely unchanged and that's the most important measure of success.