Funding Public Pensions

Potential Solutions: Different Accounting For Different Questions

Potential Solutions: Different Accounting For Different Questions 

If this article were merely a plea for less demanding pension funding requirements, that would be a difficult argument to sustain in the face of the facts. But though a central argument here is that 100 percent funding is hardly necessary to keep pension checks from bouncing, it is equally troubling that the abstract nature of pension funds also permits governments to ignore already existing funding requirements. The impacts on the current budget are always years away, making it easy to shave a little in the current year, which becomes a little more the following year, more the next, and so on. In a similar fashion, accounting rules that make clear a system’s strengths include the ongoing stream of payments from future employees would be preferable to rules that envision a system that might close tomorrow. Finding a form of funding and accounting practice that encourages adequate funding is a valuable goal—and is precisely the goal sought by the GASB in their 2012 revision—but one where the current rules fall short due to the many problems outlined here.

Alternate accounting rules are possible. Unfunded liabilities and funding ratios are not the only kind of planning values available. For example, instead of estimating a funding ratio, the Social Security trustees predict the year in which that fund will run out of money, given current trends. If, each year, the date advances, the system is in good shape. If it does not, there is cause for concern. This is simply another way to do the accounting for a pension plan, with advantages and disadvantages compared to the traditional method. It is worse at allocating the cost to any individual plan member, but it is arguably much better at respecting the fundamental philosophy of aggregated security behind a traditional pension plan.69

A pay-as-you-go system is also a different method of accounting. Like the depletion-year method it has positive and negative features. It does a poor job of predicting how much money ought to be put aside for future years, the original reason this practice was largely abandoned. However, a pay-as-you-go system will provide instantaneous budgetary feedback to pension changes. That is, changes enacted to make a system more generous will be immediately reflected in increased costs to the current budget (rather than the pension system budget) and a reduction in payments to the system will immediately be reflected in pension checks bouncing. These are the sorts of effects that, in a practical sense, constrain the actions of politicians, unlike vague promises that a bond rating might be threatened, or that a big tax increase will be necessary a decade or two hence. 

A hybrid system, combining the better aspects of pay-as-you-go and the GASB-approved systems would appear to be possible, perhaps by developing a formula that would keep a certain amount of the annual retiree payments for a pension system as part of the sponsoring government’s annual budget. This must be done not simply by recalculating the necessary contribution to the pension fund, but by arranging government finances so the pension checks to retirees will not clear unless such an appropriation is made. The linkage between policy and outcome must be immediate and clear in order to have an effect on policy. Under the status quo, the linkage is neither. 

For example, a government could seek to “monetize” the stream of payments made by its employees into the pension fund as a way to make clear that these payments are an important asset of the fund. A revenue bond backed by the premiums paid into the fund in future years could be bought annually from the government in exchange for a portion of the benefits paid to retirees in the current year. The bond itself would be an asset of the pension fund, offsetting its liability. The bond could be rated or have some small piece of it sold to another party to establish its value. Under such an arrangement, the government, rather than the pension fund, would be responsible for issuing benefit checks. The fund would pay the government for the next year’s bond, retaining enough to maintain or increase its funding level. The government would use the funds received to pay retirees, along with its employer share premiums. The debt to the retirees is of the government, while the pension plan incurs a debt to that government.

Under such a system, an abrupt change in benefits would have a direct impact on the government’s current budget. Were a mayor to promise large pension increases to the firefighters, part of the first installment would be paid directly from the city budget the very next year. Similarly, a reduction in premium payments would affect the value of the bonds, with an impact on the balance sheet of both the pension fund and whatever other agency might hold them. If the state owned some of the bonds in its cash pool, a governor who cut pension payments would see the value of the bonds drop, creating a loss to the state’s own balance, as well as that of the pension fund. Unlike the current system, the consequences of these decisions would be immediate. 

Such an accounting change could not remove political considerations from pension management. Whatever forces exist to keep the compensation of government employees from being cut—applied by organized labor or the job market—would remain, and a change in accounting could hardly do away with political pressure to hold down or cut budgets. But the consequences of policy decisions would be clearer and sooner, with much less opportunity for decision makers to shrug away responsibility for highly contingent events years hence. 

There is much detail to be added here, and this suggestion is only one possibility. It is merely a sketch of a possible system, intended to demonstrate that there are other ways to configure the relationship between pension fund, retirees, employees, and the government that may create clearer and better feedback and therefore better incentives for the various parties. 

  • 69. In truth, there is no reason a system using GASB standards cannot also provide a depletion date estimate as well, beyond the fact that the GASB rules provide no guidance for calculating it and that ratings agencies and bond buyers are not looking for it.