Pension Reform: Stop Billing the Grandkids
Intergenerational equity in retirement plans is long overdue.
Most pension reform discussions begin with two issues: the huge unfunded liabilities of public sector retirement plans (pension debt) and the abuses in the system such as pension spiking and enviable early retirements. The abuses are the easiest to fix, at least on a prospective basis. Unfunded liabilities are a "sunk cost." Changing the system for new hires won't eliminate the actuarial deficiencies of pension and retiree medical benefits (OPEB) plans. And in many states, the burden for those deficits falls entirely on governmental employers, because most public employees only pay for part of the normal cost of the plan. The question now is which generation(s) of taxpayers and employees should pay for the mop-up.
In its ongoing pension accounting project, the Governmental Accounting Standards Board (GASB) has squarely addressed the issue of intergenerational equity. Accountants call it "inter-period equity" because they focus on fiscal years more than generations, but the concepts are twin sisters. For those unfamiliar with the concept, it was first articulated in the academic literature 40 years ago by Richard and Peggy Musgrave in their classic collegiate textbook Public Finance in Theory and Practice. Intergenerational equity is the maxim that today's taxpayers should pay for today's services, so (1) we don't pay for current operations with long-term bonds, (2) debt repayments such as school building bonds and highway bonds are aligned chronologically with the benefits derived from the users, and (3) pension funds and other deferred benefits (such as OPEB) are actuarially funded rather than pay-as-you-go. In short, one generation should not burden the next generation for the public services it receives today.
Intergenerational equity is readily addressed by actuaries when they calculate the "normal cost" of pension and OPEB benefits. That's the amount of money that must be put aside each year to assure sufficient accumulated benefits for the employees when they retire. To make this calculation, actuaries study the plan design to see what is the normal or expected retirement age, then calculate the likelihood that employees will work a full career, and then use the expected return on investment assets to calculate the normal cost. The normal cost of a pension plan is typically shared between the employer and the employee. Most employers pay 5 to 10 percent of payroll for the normal cost of general employees, and between 10 and 15 percent of payroll for costlier public safety workers who retire at earlier ages because of the hazards of their work. That's just the normal cost, not the total cost. The plans with lower pension multipliers and higher retirement ages tend toward the low end of those ranges, and plans with earlier retirement ages and richer benefits formulas tend toward the higher end. And of course there are plans that fall outside those general boundaries for various reasons.
What the GASB has undertaken to evaluate is how to handle the situation employers presently face when the actuarial assumptions, especially those pertaining to investment returns, don't work out. When investments underperform, as they have in the past decade, the plan becomes underfunded. Now we have a pension debt, an unfunded liability. The question then becomes: how do we pay for it?
Historically, the philosophy of many pension plan officials and public finance professionals was that pension debt (and I include here OPEB debt which is a similar animal) could be paid off over very long periods of time. The old-school idea was that the government employer exists in perpetuity, and therefore we could amortize these unfunded liabilities over any period we chose. Likewise, the investment horizon of the pension plan was infinite, because the plan's life is assumed to be perpetual. So we could use very long-term investment strategies and investment return expectations, and we could amortize unfunded liabilities over very long periods. When the GASB's predecessor (the National Council on Governmental Accounting) wrote the rules before 1984, the accounting standards then permitted pension funds to amortize their unfunded liabilities for 40 years, and some states actually built that naïve presumption into their pension funding laws in what has proved to be a disastrous legacy of kicking the can. Show me a 40-year amortization plan and I'll show you a distressed pension system.
As GASB took over the job of promulgating accounting standards, it tightened up the amortization period to 30 years. But it left open a back door for some can-kickers to amortize over 30-year "open" actuarial periods — which means that the plan essentially resets the mortgage clock every year. I call this the "credit card amortization method" because the debt is never repaid. If you pay off 3 or 4 percent of an outstanding obligation every year but constantly "refinance," you never eliminate the debt. And herein lies the problem facing many public pension plans: they failed to match the amortization periods with the lives of either the employees working toward retirement, or the retirees who have already earned their benefits. And when the amortization period exceeds both the average remaining life expectancy of retirees as well as the average remaining service lives of the current workers, you've got yourself an intergenerational equity problem.
"Extended Smoothing." Some public pension plans presently take great liberties to smooth out the budgetary impact of stock market fluctuations. Bear in mind that the average stock market and business cycle in the U.S. since 1926 is 6 years: There have been 14 recessionary bear markets in 86 years. So any smoothing process that extends beyond 6 or 7 years is statistically suspect. Yet some plans have used smoothing periods as long as 15 years, and others employ so-called "double-smoothing" processes that punt the investment losses even further into the future. An even more dubious practice is now used in New York State, where legislation permits local governments to borrow from the pension fund to make their contributions.
Imagine that: a bank that will loan you money to make your minimum credit card payment. It doesn't take a genius to figure out why that is unsustainable public policy. Besides the obvious intergenerational inequities this creates, these deferral tactics are even more senseless from an investment standpoint: The pension fund thereby receives less new money to make up for losses while the markets are down, and then receives more money to re-invest when stock prices have recovered. "Let's buy more at higher prices" is a doomed investment strategy that betrays the dollar-cost-averaging concept that history has rewarded. In fact, it undermines the presumed long-term investment rates of return that many of these plans use today. You won't achieve 8 percent annual investment returns when you systematically invest new money at the high end of stock market cycles.
Coming soon: New GASB rules. GASB's latest exposure draft seeks to address the amortization problem by (1) accounting for investment gains and losses as "expenses" of the employer over five-year periods and (2) amortizing other actuarial changes over the average remaining service lives of employees. The latter standard aligns closely with how private corporations amortize pension debt under corporate rules of the Financial Accounting Standards Board (FASB). Their project timetable anticipates a new pension accounting standard this summer. The GASB standards will apply to future changes in financial condition, and they haven't really addressed what they will do with the $700 billion of outstanding pension debt and $1.5 trillion of unfunded OPEB liabilities. The industry will await guidance from the GASB on how this will be handled.
Funding vs Accounting. As I've reported previously the accounting and the funding for governmental retirement plans will probably be "divorced" as GASB sets accounting standards that will be virtually impossible for many employers to match in terms of their actuarial funding practices. Funding policies will have to be crafted by the pension community in consultation with the government finance community (i.e., the CFOs of the states and localities). And that's where we need a much more thoughtful discussion about intergenerational equity.
I've spoken with numerous public-sector actuaries and plan administrators whose unshakable mindset has been that their plans are perpetuities and therefore it's reasonable to defray accumulated pension deficits (unfunded liabilities) over extended periods that have no relationship to the lives of the retirees and current workers. As you might expect, there are very few public budget officers and even fewer elected officials who want to bite the bullet and pay off the investment losses of the past decade any sooner than they are required. But here's the problem with that thinking: Today's elongated amortization periods virtually guarantee that most retirees will die before their employers have paid for their benefits, and today's workers will already be collecting pension checks before the taxpayers finish funding their pensions and retiree medical benefits. In both cases, that is a blatant violation of intergenerational equity.
States and localities really need to establish shorter amortization periods for their unfunded retirement obligations, and begin working toward those in an orderly, feasible migration. In my written comments to the GASB, I suggested a transitional provision that would immediately shorten the amortization period for currently outstanding obligations to 20 years and then begin working downward toward the average remaining service lives (ARSLs) of current employees. (Explanation: If pension plans permit retirement upon 30 years of service, then today's average worker will retire in about 15 years. For aging public safety workers in 25 year careers, the average will be closer to 12 years. And after five years of government hiring freezes, the ARSLs are probably even lower.)
Investment horizons have similar implications. There's one more conventional assumption that pension professionals need to rethink: their investment horizons. For decades, we have assumed that the proper investment horizon for determining the discount rate used in actuarial projections should be perpetual, because the plan is perpetual. But that is flawed thinking, especially in times like this when bond yields are very low and when most investment professionals agree that stock-market returns in the shorter term are likely to be impaired by global debt problems and the deleveraging of the American economy after its debt binge of past decades. A bond portfolio starting today with Treasurys yielding 3 percent is not very likely to produce the same returns in the next 15 years that it will over the next 30 years after rates normalize at higher levels. And stock returns in the coming decade could well be lower than those which result in the much longer term. Thus, it is appropriate to use a somewhat lower discount rate for that portion of the liabilities which must be funded sooner than the traditional 30-year amortization assumptions and lifespan assumptions used by actuaries in many pension and OPEB plans. Otherwise, these plans are highly likely to experience continuing investment shortfalls which will only create new unfunded liabilities and even further burden the future taxpayer.
Unfortunately, this is not news that anybody working with public pension and OPEB plans wants to hear. Most policymakers want to believe that investment returns in the coming decade will match historical norms, but that's not what my research from the comparable "reconstruction" period following the 1930s shows. History suggests that equity (stock market) returns will likely meet or beat their historical averages over the next 30 years, but the next decade is far more risky and less assured. And it's mathematically almost impossible to produce bond returns of 5 percent in the next decade given today's low rates, which will result in capital losses for bondholders whenever rates do increase. It's hard to contrive a scenario in which bonds earn more than their current coupons before 2020 unless it's accompanied by runaway inflation that causes even worse problems for pension funds. I have no quarrel with discounting the liabilities of new hires over 30 years using long-term rates of expected investment returns, but I can't understand the logic of assuming improbable (higher) rates of return over the shorter periods when today's unfunded liabilities must actually be repaid. That just increases the odds that the next generation will pay higher costs because their elders kicked the can.
In fact, there is a growing risk that pension trustees clinging to ambitious investment return expectations in their actuarial assumptions will burden their children and grandchildren with more unfunded liabilities. What I fear now is that an improving economy will lull pension policymakers into believing that a few years of above-average stock market returns "in the teens" will continue indefinitely so they don't need to change their actuarial assumptions. After watching stock market indexes double since the last cycle-bottom, some folks are already engaged in a "willing suspension of disbelief" by assuming that investment returns in the next ten years will not include a recession down-cycle, despite 85 years of stock market history showing an average loss of 30 percent in equity prices when that happens. Unlike Lake Wobegon, every year cannot be above average.
It's a fair bet that there is another recession coming before the end of this decade, and investment expectations need to take that into account.
Wake-up call for policymakers. As I've written before, this suggests that pension and OPEB plan trustees should now be using discount rates closer to 7 percent, rather than their current levels between 7.5 and 8 percent. A naturally lower rate aligns with the likely outcome of what the GASB's proposed "blended" discount rate for underfunded pension plans may produce anyway. Thus, I would hope to see a "directional convergence" of policies, assumptions and methodologies in 2013, especially if GASB follows through on its previous intentions to invoke a lower discount rate when there are substantial unfunded liabilities. Every pension board, governor, budget office, county commission, school board and city council should be discussing the implications of these issues, and confronting their own demons by plotting a strategy to resolve their unfunded liabilities in a prudent way that does not burden future generations. Even as budgetary revenues begin to creep upward as our stagnant economy gathers steam, higher pension and OPEB costs will crowd out other spending requests — one way or another, sooner or later. Let's not turn a problem in our hands into a crisis for our grandchildren.
Join the Discussion
After you comment, click Post. You can enter an anonymous Display Name or connect to a social profile.