In 2006, US regulators, reacting to the 2000 stock market crash and the so-called “perfect storm”, implemented broad rules for corporate defined benefit pension plans on liability valuation and financial reporting (SFAS 158), and on funding (the Pension Protection Act).
Today, there is a similar focus on the epic underfunding of defined benefit plans at US state and local governments, estimated at $1,000 billion to $3,000 billion, as accounting regulators have proposed more prominent disclosure of shortfalls on sponsors’ annual financial statements. Some observers caution, however, that the new principles will fall short and merely perpetuate current methods which greatly understate governments’ pension liabilities
The US public sector, excluding the federal government, employs more than 19 million people – 15 percent of the labor force – and their pension plan assets, according to the Federal Reserve, were $2,557 billion in June 2010, down from $3,198 billion at year-end 2007.
By any measure that total is not enough. Wilshire Associates estimates that as of June 2009, state plans were just 65 percent funded, while local government plan funding stood at 74 percent. Wilshire’s estimates do not reflect the 30 percent rise in the broad US stock market since then, but few dispute that plans are still badly underfunded.
In recent years many states have moved to shore up their DB plans by changing benefits for current and future employees, and a few have even dared to scale back payments to retirees. Others have issued dedicated pension bonds. Government pensions are emotional issues in a number of the state political races decided this week.
Adding to the tension, in June 2010 the Government Accounting Standards Board proposed a reporting and valuation framework to highlight governments’ pension disclosures. Many of the proposals echo the requirements imposed on corporate DB plans in 2006.
One essential change in Gasb’s 50-page proposal is to report the liability for unfunded pension benefits on the government’s balance sheet – a promotion from their current footnote status. “As a result of the employment exchange each year, the employer incurs an obligation to its employees for pension benefits… [and] that net pension liability is measurable with sufficient reliability for recognition,” states Gasb’s Preliminary Views.
Among the 190 comments on the proposals are many objections: some governments contend that footnote disclosure is adequate; others that because pension liabilities are so large and volatile, adding them would create more distortion than clarity. Others argue that the extent of responsibility to meet all estimated current and future liabilities is not clear, offering as evidence legal battles in Colorado and New Mexico that may grant governments a way to reduce their obligations for future benefits.
But Gasb left essentially untouched the method by which governmental plans value their liabilities – based on actuarial assumptions for asset returns, rather than a discount rate that reflects the near certainty that pension obligations will have to be paid.
Most government plans assume an 8 percent rate of return on investment; corporate plans, by contrast, are required to value liabilities at a high-grade corporate bond rate, currently below 5 percent. (The Gasb proposal would, however, require the portion of public plans’ liabilities that are chronically underfunded to be valued at municipal bond rates – currently between 3.5 and 4 percent.)
“State and local governments are being granted a special privilege, to use discount rates that understate the true economic value of their liabilities,” says Joshua Rauh, associate professor of finance at the Kellogg School of Management, Northwestern University, and a pensions scholar.
“The discount rate should reflect the certainty that the governments will have to pay them, and that is certainly no less risky than the rate on a state or local government bond.” Under both current and proposed rules, Prof Rauh explains, the value of the liability is based on the risk a pension plan takes with its assets, when the two rates are in fact independent.
In a recent paper written with Robert Novy-Marx, associate professor at the University of Rochester, Prof Rauh estimates that state pension plan liabilities alone, valued at treasury bond rates, came to $5,000 billion as of June 2009, while assets were just $1,940 billion. “This ‘pension debt’ dwarfs the states’ publicly traded debt of $940 billion,” he writes. City pension plans, he estimates, add another $383 billion aggregate underfunding.
One large public plan, the $35 billion New York City retirement system, supplements its disclosures based on conventional actuarial returns with several alternative calculations, including one based on a risk-free discount rate, in an effort to represent the certainty that the pension liabilities will be paid.
For June 2008, citing the most recent report, the plan’s funding ratio was 85 percent using actuarial discounting, but just 65 percent with Treasury-based valuation; since 1999 the difference between the two has ranged from 10 percent to 20 percent of assets at market value. While deficits under current methods have been publicized, “people have not come to realize how important this government accounting issue is”, says Prof Rauh.
“It traces back to the idea that there are limits on how much the governments can borrow. But by not measuring the pension liabilities realistically, it creates a huge loophole and exposure for taxpayers.”