In a paper released earlier this month by the Competitive Enterprise Institute, former Virginia Libertarian gubernatorial candidate Robert Sarvis explains how state governments have understated the underfunding of their pension systems for years through the use of dubious accounting methods. Namely the use of a discount rate-the interest rate used to determine the present value of future cash flows-that is too high. As a result, state pension systems are likely in even worse shape than government data suggests and without reform, state government debt could burden labor markets and worsen state’s business climate.
In his paper, Sarvis aggregates several estimates of states’ pension debts and ranks the states from best to worst in terms of unfunded liability compared to state GDP. Included in his composite ranking are unfunded liability estimates calculated using fair market value techniques by Naughton, Petacchi, and Weber in their recent paper “Economic Consequences of Public Pension Accounting Rules”, by Novy-Marx and Rauh in their 2011 paper “Public Pension Promises: How Big are They and What are They Worth”, by Cory Eucalitto in “Promises Made, Promises Broken-The Betrayal of Pensioners and Taxpayers”, by Moody’s Investor Service, and unfunded liability estimates based on each states official assumed rate of return. Below are the five best funded states and the five most underfunded based on Sarvis’ aggregation of the rankings:
| The Five Best Funded | The Five Worst Funded |
| Nebraska | New Mexico |
| North Carolina | Illinois |
| Tennessee | Mississippi |
| Delaware | Kentucky |
| South Dakota | Ohio |
In determining whether a pension fund’s assets are sufficient to meet future liabilities, one must compare the valuation of assets in the pension fund with the calculated net present value of future payments to retirees. The net present value of future liabilities is calculated using a discount rate that represents that represents the risk and timing of those liabilities. Sarvis argues that the discount rate used by public pension systems should be a low risk rate, ideally as low as the rate on Treasury bonds (around 2 percent), but at least as low as other government bonds (municipal bonds return about 5 percent) or high-quality corporate bonds.
However, state pension system discount rates average around 7.7 percent, and are often as high as 8 percent or 8.5 percent. As a result of these high actuarial assumptions, states understate their unfunded pension liabilities. For example from 1990-2009, as a percentage of state GDP, New Mexico officially reported underfunding their pension system by 5.3 percent over 20 years, but when one reestimates the numbers using the fair market discount rate used by Naughton, Petacchi, and Weber, it turns out New Mexico’s pension system has averaged being underfunded by 20.9 percent of state GDP from 1990-2009.
Sarvis explains that under defined benefit pension systems (the most common type of public pensions) the amount of benefits to be paid out to future retirees is determined by a formula and legally guaranteed, so public pension systems that are underfunded may require further infusions of cash to remain solvent. Often times, it is politically and legally easier to shore up a pension system by either raising taxes or shifting spending from somewhere else in the budget than by increasing contributions from public employees. According to Sarvis, the fact that so many pension systems are underfunded beyond what the government data reveals creates the expectation of future policy changes (like raising taxes or budget cuts) which affects the business climate and labor market within a state.
For a plain language guide on unfunded pension liabilities see Reason’s piece from June, “The Public Employee Pension Crisis Explained.”