Unfunded Pension Liabilities and GDP

While pleased to be recognized in the Wall Street Journal (page A7) this morning under “People to Watch,” the feature also hammered home to me how entrenched the use of flawed official statistics still is.

The article includes a graphic, “How the States Measure Up,” including unfunded pension liabilities, sourcing Moody’s for the pension data. Moody’s covers itself by using official statistics provided in the annual financial reports of the systems — although according to another WSJ article from Tuesday Feb 22, Moody’s has “voiced support” for legislation that would require states to furnish annual reports to the Treasury using the correct methodology.

According to the graphic, unfunded liabilities to GDP maxes out at around 16% for the states Hawaii and Mississippi. The reality is that when proper financial discounting is applied to pension liabilities, the average unfunded liability across the 50 states represents 18% of GDP. For eight states it is over 25% of GDP. For three it is over 30% of GDP, and for one (Ohio) it is 35% of GDP. Here is a table. This includes only unfunded liabilities from systems sponsored at the state level, and only benefits promised based on service and salary up until today (legacy liabilities). Local liabilities increase the unfunded liability by about 20%.

If you are new to this and wondering what the difference is between my statistics and the official methodology, the issue is the standards established for pension measurement by the Government Accounting Standards Board (GASB). GASB allows state and local governments to disclose their debts by assuming returns on risky assets. The governments assume that the actual return will be identical to the targeted return, most commonly 8%, ignoring the fact that if the assets do not return 8%, the taxpayers are on the hook for the downside. GASB confounds the measurement of the amount of the debt with the government’s risky plans for repaying that debt.

As I wrote in my congressional testimony last week (click here for video):

Consider … how this methodology would work in the realm of personal finance. Imagine you personally have a $10,000 repayment of a loan due in 5 years. If you applied for another loan, any reasonable lender would ask you to record your other debts on the loan application. Are you allowed to assume a return on your stocks and bonds in recording the value of this debt? Of course not. Does the amount of debt you report to the bank depend on whether the savings you have is invested in stock or bonds? Once again, the answer is obviously not. But under GASB accounting procedures, state and local governments can treat the higher returns they hope to achieve as a sure thing. The higher the return the government assumes, the lower the pension debt that it reports to the public and to the rating agencies.

This procedure contrasts sharply with private sector accounting methods, as well as financial logic. It ignores the role of risk – the reality that a wide range of investment outcomes are possible – and, more specifically, the fact that future taxpayers will have to make up any shortfalls if the fund’s assets fail to generate the expected 8 percent return. If a taxpayer’s personal portfolio underperforms, he or she can cut back on their expenditures. But if the government’s pension portfolio underperforms, the taxpayer will be asked to pay to the government the difference between what the government promised to public employees and the resources that are left to meet those obligations.

Eight states have unfunded liabilities of over 25% of GDP. Is that a problem? For example, it is often pointed out that Greece has a debt to GDP ratio of about 144%.

The important point to note is that state governments collect about 5% of GDP in taxes on average. If one includes all local-level taxes as well, we are up to an average of around 8.5% of GDP. Greece, for all its troubles, collects 31% of GDP in taxes.

It is disturbing to realize that official statistics are failing to reflect reality by such a large magnitude. Borrowing spirals are like drinking problems – the first step to address the problem is acknowledging it. Unfortunately the states have a long way to go.

Additional reading:

Public Pension Promises: How Big Are They and What Are They Worth? (with Robert Novy-Marx), October 2010, Journal of Finance, forthcoming

The Crisis in Local Government Pensions in the United States (with Robert Novy-Marx), October 2010

Return Assumptions: It’s All Just Borrowing in Disguise, posted on this blog

Are State Public Pensions Sustainable? Why the Federal Government Should Worry About State Pension Liabilities, National Tax Journal 63(3), 2010 (Forum)