It's time to bust the myth that assumptions drive pension costs

There’s a lot of talk these days about public pension plans, the size of their unfunded liabilities and the level of their costs. It’s appropriate that policymakers and policy influencers focus their attention on, “What to do?”

But when these conversations turn to actuarial assumptions, you can bet there’s something amiss. The woefully misguided — and dangerous — idea that assumptions drive costs leads decision-makers down a perilous path. Cost management is not achieved by “managing” actuarial assumptions.

Costs are determined by what actually happens, not by what we assume or predict will happen.

That assumptions don’t drive pension costs goes against everything we’ve always thought, right?

Here’s the truth. The cost of a pension plan, paid by making contributions over time, is a function of the pensions paid out and of the investment earnings of the fund. Period. Costs are not a function of assumptions.

Real management of a pension plan involves setting the level of pensions and deciding how to invest the assets. The cost of the plan — again, paid by making contributions over time — is unaffected by assumptions. It is simply the sum total of the pension payments to plan members, offset by the pension fund’s investment earnings.

Raising or lowering the formula used to determine employees' pensions does drive costs. And, how long those pensions are paid also drives costs — so actual longevity matters.

But, here’s a trap. The recent mortality tables from the Society of Actuaries, reflecting significantly improved life expectancy, provoked this typical response: “The new tables will increase plan costs — perhaps by as much as 8%.”

That is, of course, not true. Because people in real life are living longer, pension costs already have increased. The new tables simply recognize that fact. Plans using “old” mortality assumptions that do not take into account recent improvements in mortality or projections of further improvements in the future are generally underestimating what recent research and wisdom tell us about life expectancy and trends in improving life expectancy.

The point is this: Assumptions don’t drive costs. Reality drives costs.

The same is true for investment returns. Pension costs are affected by the actual investment earnings on assets in pension funds, not by what we assume the earnings will be. Higher actual investment earnings means smaller contributions are necessary. However, merely assuming a higher investment return does not, unfortunately, translate to smaller contributions.

By the way, a stricter view of “cost” is that the cost of pensions is determined when the pensions are earned and does not change as investment earnings materialize. Any reduction or increase in contributions due to investment earnings comes from the risk/rewards of investing in a particular asset portfolio, not from the cost of the pension promise.

Well, of course they matter. But let’s be very clear about where they matter and where they don’t. Too often, discussions about how to deal with significant unfunded liabilities or about how to maintain a sustainable level of cost steer toward the topic of actuarial assumptions.

Although assumptions don’t drive costs, they do absolutely affect the allocation of that cost over time — the pattern of contributions through the years. This is an important distinction, one that too often clouds the discussion of important issues such as how to achieve a sustainable level of costs.

The whole topic of accelerating costs vs. deferring costs is a topic for another day. For now, let’s just say that an unaffordable plan doesn’t become more affordable just because you change assumptions in order to defer contributions.

How can pension costs be reduced? How should pension costs be managed?

The only ways to cure an underfunded situation are to contribute more, earn more or reduce pension payouts. And, we all know (we hope!) that earning your way out of an underfunded situation involves risk. It’s a crapshoot. And you certainly can’t earn your way out merely by managing the actuarial assumptions. That’s a failed strategy.

Over-attention to assumptions can lead — and has led — to bad decision-making. If we pay too much attention to assumptions, we’ll overlook what’s really driving costs and what can put a troubled pension plan back onto a path toward sustainability.

Let's be clear on what we’re talking about in order to move forward toward responsible management of public pensions. It’s important to use good actuarial assumptions, but “managing assumptions” as a means of managing true pension costs just doesn’t work. Let's make sure we listen with a discerning ear so we don't confuse the two.

Brian M. Septon is a principal at The Terry Group, Chicago.