Posted
Mar 2010
Pensions in Peril
The funding status of state-sponsored pension plans
Based on the research of Robert Novy-Marx And Joshua Rauh
As the races for governor heat up in thirty-six states, the question of how to fix troubled state-sponsored pension plans may be one of the most challenging that candidates will have to face. State pension plans are underfunded by $3.2 trillion when misguided accounting practices are corrected according to research by Joshua D. Rauh, an associate professor of finance at the Kellogg School of Management, and Robert Novy-Marx at the University of Chicago, published in the Journal of Economic Perspectives. Furthermore, because pension funds are highly exposed to market risks, there is only a 5 percent chance that they will perform well enough to meet the needs of retirees in fifteen years.
State governments in the United States had approximately $2 trillion set aside in pension funds and $3 trillion of “stated” pension liabilities in December 2008. By this measure, the funds seemed to be short nearly $1 trillion. But according to Rauh and Novy-Marx, the shortfall is more than three times larger, at $3.2 trillion. The lower estimate, they say, is the result of government accounting standards that require states to apply accounting procedures that severely understate their defined-benefit pension plan liabilities.
Many states seem to be in dire straits. Relative to its revenues, Ohio faces the largest burden. It would need to devote over eight years of tax revenue solely to retirement funding to cover already-acquired pension liabilities. Colorado, Rhode Island, and Illinois are close behind. “Vermont, the least underfunded as a percent of total tax revenue, would still need over twenty months of tax revenue to make up for its pension fund shortfall,” Rauh says.
Vermont, the least underfunded as a percent of total tax revenue, would still need over twenty months of tax revenue to make up for its pension fund shortfall.
Most U.S. state governments offer defined-benefit pension plans, retirement packages that differ from private sector defined-contribution plans such as the 401(k) or the 403(b). In a defined-benefit pension plan, employees accrue the right to an annual monetary benefit that begins upon retirement. Such payment is usually determined as a flat percentage of a worker’s average salary over the final years of employment multiplied by the worker’s years of service with the employer.
In a defined-benefit plan, the employer is responsible for determining how much money to set aside for its employees’ retirement fund and how to invest it—the risk is assumed by the employer rather than the employee. If a pension fund is short when workers retire, the state has three options: raise taxes, cut spending, or default on its obligation to retired employees. The good news is the minimum annual benefit promised by a defined-benefit pension plan is certain. The bad news is the pension fund’s solvency could be anything but.
Problematic Calculations
Estimating pension liabilities is by no means straightforward. One measure of pension liabilities is based on a grim assumption—that a state can immediately lay off all of its employees. At that point, the employees would cease accruing benefits. If a state’s fund were large enough to make all of its retirement payments given those layoffs, it would be considered fully funded. This is called the accumulated benefit obligation, or “termination liability.” In reality, this measure is quite conservative because states cannot simply stop accruals through mass layoffs. Furthermore, the minimum the states have already promised to their employees is generally protected by law. However, Rauh and Novy-Marx find that even if states could stop accruals at will—a big assumption—state pensions would still be severely underfunded.
Furthermore, any measure of liabilities requires a discount rate to convert the expected future pension payments into a present value. Government accounting standards instruct states to discount their liabilities at the expected return on their assets. Doing so creates a false equivalence between future pension payments, which almost certainly will be incurred, and the outcome generated by a risky investment portfolio, which by definition is highly uncertain.
Standard financial theory suggests that future payments should be discounted at a rate that reflects their risk. So what is the risk associated with state pension payments? Under many state constitutions, state pension obligations are the most senior types of unsecured state debt. Given their seniority, labeling pension obligations “risk-free” would be a reasonable starting point, one that would allow them to be discounted using a risk-free interest rate, such as one based on Treasury bills and bonds. However, most states use an unreasonably high and seemingly arbitrary 8 percent discount rate. As a result, the present value of future pension liabilities is often dramatically understated. Using the interest rate on Treasury securities as of January 2009, Rauh and Novy-Marx estimate the present value of pension liabilities to be $5.2 trillion. Thus, rather than being $1 trillion underfunded, state pension plans are actually short $3.2 trillion.
Risky Discounts
Because state pension liabilities are not completely risk-free, Rauh says, “some might argue it might be more appropriate to use rates on municipal bonds to discount future pension liabilities. States could potentially default on their pension obligations the same way they can default on their bonds.” For example, state constitutions could be changed to make it easier to default on currently owed pension benefits. Taxpayers employed by the private sector might prefer the default over increased taxes.
“This possibility, however, seems highly unlikely in most states,” Rauh points out. In the past, a few municipalities in the United States have defaulted on their debt while maintaining employee pensions at the same time. Still, using municipal bond rates may not be much of a solution. “Even when we used interest rates on municipal bonds as of January 2009, we calculated the present value of pension liabilities to be $3.3 trillion—which still implies an underfunding of $1.3 trillion,” Rauh notes.
“There is no doubt that the decline in asset markets in 2008 has worsened the problem,” explains Rauh, “but unfortunately the financial crisis is far from being the main cause of the problem.” Pension funds were vastly underfunded even before the financial crisis. Although state pensions appeared to be fully funded by government accounting standards in 2005, they were already short $800 billion by Rauh and Novy-Marx’s calculations.
The severity of the underfunding is not the only challenge plaguing the system. Many funds hold a portfolio of risky assets, a strategy that has a small probability of generating extremely high positive returns and a very high probability of relatively low or even negative returns. States will be able to meet pension obligations only in the unlikely event that these risky investments perform well. In all likelihood, tax increases or spending cuts will be needed to shore up yawning funding deficits.
Further reading:
Joshua Rauh blogs about this topic and related subjects in Everything Finance




4 Comments
Mar 3 2010
Great article and I am working on pension related projects as we speak. For Navy shipbuilders, they pass on pension costs in their contracts and with high levels of unfunded liabilities, increased likelihood of pension expense increases due to discount rate changes, asset allocations and pension expenses on the P&L, all this means that taxpayers will pay more for U.S. Navy ships into the future.
Great article and thanks for letting weigh in. A prospective Kellogg MBA applicant.
Oct 25 2010
This and other writings by Dr. Raul are not only severely flawed and based on inaccurate assumptions, but border on alarmist, perhaps pushing a conservative anti-government agenda.
A spot-on rebuttal to this alarmism can be found at http://www.nasra.org/resources/RauhNorvyMarxMuniStateCritique.pdf .
Oct 25 2010
It is “Rauh”, not Raul. At any rate, I beg to differ—fiscal responsibility is not a “conservative anti-government agenda.” In reading the two-page of the National Association of State Retirement Administrators (NASRA, hardly a neutral party), I see that it misinterprets the Rauh and Novy-Marx research. They do not assume rates of return on the assets, they are calculating the present value of already existing pension obligations. Most pension funds, they point out, use an 8% discount rate. Such a high discount rate understates the present value of future obligations—8% understates even compared to the 4.5% expected return for the assets mentioned in the NASRA two-pager. That is the main point, in my opinion, of the analysis.
Mar 18 2011
What about nuclear decommissioning liabilities on utilities balance sheets. These are discounted at GAAP defined accretion rate which may be higher than NRC/PUC mandated expected return on fund assets which again is higher than risk adjusted rate (presumabely ultimately the Feds would step in to decommission a nuclear plant - should it be a long-term treasury rate?) depending on the valuation method selected. When you add to this the Fed’s liability relating to spent fuel this may be astronomical.