The Manhattan Institute and Economics 21 hosted a debate on pension accounting and reform last week, at which I was one of the panelists. The discussion was wide-ranging, but most of the debate boiled down to a key question: Does the government have a special advantage in bearing investment risk, thus justifying the use of defined-benefit pensions?
Taxpayers are not indifferent to investment risk assumed by governments, because investment losses and gains will be passed through to taxpayers and recipients of government services. While these parties also receive the excess gains when investments do well, that’s not a wash; the gains come when they are least needed and the losses when they are least affordable.
There were a couple of key discussion points at the event that are worth hashing out in writing. One is a response to a claim often heard from defenders of the status quo: Pension costs are not that high and shortfalls can be managed with little impact over a 30-year window.
It’s important to remember that a pension fund shortfall is akin to a 30-year mortgage. Pension funds depend on income generated by assets (typically, an 8 percent annual return) to cover the cost of benefits. If a pension fund is underfunded, taxpayers must effectively pay interest at that 8 percent rate on the amount of the underfunding, in addition to the “normal cost” of pension benefits that continue to accrue, or otherwise the funding ratio will continue to deteriorate.
Given the performance of the stock market over the last five years, it is likely that a typical pension fund that was fully funded in mid-2006 is now only 75 percent funded; as a result, doubling of annually required pension contributions should be considered typical, not abnormal.
So while pension costs have historically consumed about 4 percent of state and local budgets, that figure is likely to be far higher in the next several years, and therefore pensions are a major driver of state and local fiscal crises.
We also discussed whether defined benefit pensions are an efficient tool for attracting talent to work in the public sector. In my view, they are not; I discussed a Stanford study of a pension benefit increase offered in Illinois in the late 1990s, which found that Illinois teachers would not pay for an increased pension benefit unless offered a steep discount from the real cost. The fact that large corporations are increasingly abandoning the defined benefit model is also evidence that defined benefits are not especially useful for recruiting.
I think people on the left generally understand this, deep down, and actually support defined benefit pensions because of the retirement security that they provide. The problem with this justification (aside from the need to subject it to cost-benefit analysis) is that it is not a reason to offer a program specifically for the benefit of public workers.
In my view, the government ought to be in the business of providing some sort of guaranteed minimum income for the elderly so that people who are too old to work do not end up destitute. This is why I oppose the conversion of Social Security to a system of private accounts. We can discuss whether the floor set by Social Security is too low, but if so — then it should be expanded for the whole population. A special benefit just for government workers is not a good solution to a general problem.
Josh Barro is a senior fellow at the Manhattan Institute and editor of PublicSectorInc.org, which tracks public-sector issues in state and city governments.