Public Pensions are Underfunded

Featured in print Digest

There is a two-thirds chance that state pension plans will realize a shortfall in 15 years. The expected conditional shortfall is almost $1.5 trillion in 2005 dollars.

The extent to which public pensions are underfunded has been obscured by governmental accounting rules, which allow pension liabilities to be discounted at expected rates of return on pension assets, according to NBER researchers Robert Novy-Marx and Joshua Rauh. In The Intergenerational Transfer of Public Pension Promises (NBER Working Paper No. 14343), they report that over the next 15 years, state pensions are expected to grow to a total of about $7.9 trillion. But Novy-Marx and Rauh conservatively estimate a 50 percent chance that the system will be underfunded by more than $750 billion at that time, and a 25 percent chance of a shortfall of at least $1.75 trillion (in 2005 dollars). "Adjusting for risk, the true intergenerational transfer is substantially larger," they write. "Insuring both taxpayers against funding deficits and plan participants against benefit reductions would cost almost $2 trillion today, even though governments portray state pensions as almost fully funded."

Novy-Marx and Rauh collected data on the largest defined benefit (DB) pension funds sponsored by U.S. state governments. In a typical DB pension plan, an employer pledges an annual pension payment of an amount that is a function of the employee's final salary and years of employment. To assemble the list of plans, the authors began with data from the Census of Governments published by the U.S. Census Bureau. They studied all plans with more than $1 billion of assets. There were 112 such plans at the end of 2005. They then examined the Comprehensive Annual Financial Report (CAFR) for each pension plan and collected total actuarial liabilities for each, along with the discount rate used by state actuaries to calculate these liabilities.

States back pensions with stocks, bonds, cash, private equity, real estate, and hedge fund exposure. But the typical investment strategies, in conjunction with accounting rules, make the pension funding situation look much better than it actually is. Under the government accounting logic, states always could eliminate their underfunding, no matter how large, simply by investing in sufficiently risky assets. In fact, investing in riskier assets may raise expected returns, but it also increases the probability of a severe underfunding. Under current investment strategies and a standard equity premium of 6.5 percent, there is a two-thirds chance that state pension plans will realize a shortfall in 15 years. The expected conditional shortfall is almost $1.5 trillion in 2005 dollars.

Under any plausible discounting assumptions that reflect the true present value of state pension promises, the underfunding in state pension plans is larger than the total magnitude of outstanding state bonds if state pension promises were riskless, the underfunding would amount to $1.9 trillion. If the risk of state pension liabilities were roughly captured by municipal borrowing rates excluding the tax benefit, which allows for a possibility of defaulting on these pension liabilities, Novy-Marx and Rauh find that total state underfunding amounted to $862 billion as of late 2005.

All of the figures described above assume that all pension benefits that will be accrued in the future will be fully funded using appropriate discount rates, which Novy-Marx and Rauh demonstrate has not been the case in the recent past. The figures also do not account for other post-retirement employee benefits (OPEBs), which total $380 billion in present value. Therefore, the analysts' calculations probably understate the extent of the funding crisis.

-- Matt Nesvisky