A gold-plated burden
Hard-pressed American states face a crushing pensions bill
CHUCK REED is the Democratic mayor of San Jose, California. You might expect him to be an ally of public-sector workers, a powerful lobby in the Golden State. But last month, at a hearing on pension reform held by the Little Hoover Commission, which monitors the state’s government, Mr Reed lamented his crippling public-pensions bill. “City payments for retirement benefits have tripled over the last ten years even though our workforce has declined dramatically, and we have billions of dollars in unfunded liabilities that the taxpayers must pay,” he said.
Mr Reed estimated that the average cost to his city of employing a police officer or firefighter was $180,000 a year. Not only can such workers retire at 50, but some enjoy annual pension payments greater than their salaries. They are also entitled to cost-of-living increases of 3% a year, health and dental insurance for life and lump-sum payments for unused sick leave that could reach hundreds of thousands of dollars.
Plenty of similar bills are looming in America’s public sector: in municipalities, in the federal government, and especially at state level. Defined-benefit pensions, which link retirement income to salary, are expensive promises to keep. The private sector has been switching to defined-contribution plans, in which employees bear the investment risk. But the public sector has barely begun to adjust, and has built up a huge liability to its staff. Worse, it has not funded the promises properly.
Joshua Rauh, of the Kellogg School of Management at Northwestern University, and Robert Novy-Marx, of the University of Rochester, estimate that the states’ pension shortfall may be as much as $3.4 trillion and that municipalities have a hole of $574 billion. Mr Rauh calculates that seven states will have exhausted their pension assets by 2020—even if they make a return of 8%, a common assumption that looks wildly optimistic. Half will run out of money by 2027. If pension promises are to be kept, this will place immense strain on taxes. Several have promised annual payments that will absorb more than 30% of their tax revenues after their pension funds are exhausted (see chart 1).
The severity of states’ pension woes was disguised for years, because asset markets were so strong and because of the way states accounted for the cost of pension provision. But the 21st century has been dismal for stockmarkets, where most pension money has been put. State budgets came under huge pressure as a result of the 2008-09 recession, which caused tax revenues to plunge. Meredith Whitney, an analyst who made her name forecasting the banking crisis, believes the states could be the next source of systemic financial risk.
Now the problem is making headlines, especially in California, where taxpayer groups have been highlighting the generous pensions of some former employees. More than 9,000 beneficiaries of CalPERS, the largest state retirement plan, receive more than $100,000 a year.
The stage is set for conflict between public-sector workers and taxpayers. Because almost all states are required to balance their budgets, any extra pension contributions they make to mend a deficit will come at the expense of other citizens. Utah has calculated it will have to commit 10% of its general fund for 25 years to pay for the effects of the 2008 stockmarket crash. But attempts to reduce the cost of pensions are being challenged in court and will be opposed by trade unions, which still have plenty of members in the public sector.
A pension plan is a promise to pay employees after they retire. Most liabilities fall due well into the future, once those now working retire and receive payments until they die perhaps 20 or 30 years later. Such future liabilities have to be valued, using a discount rate to reflect what they are worth in today’s money. The higher the discount rate, the lower the present value.
States use the expected return on the assets in their pension funds as a discount rate. This is often around 8%, and reflects the performance of the past 20-30 years. However, such returns will be hard to come by in future. As Bill Gross of Pimco, a giant fund-management firm, pointed out recently, given current bond yields of 2% and a typical portfolio with 60% in equities and 40% in bonds, a total return of 8% requires a return of 12% on equities. And with American equities yielding just 2-2.5%, that in turn would require dividends to grow by 9-10% a year. Dividends grow roughly in line with the whole economy—and 9-10% is just not plausible.
This reliance on returns as the basis of the discount rate is extraordinary, when you stop to consider it. The more risk the pension fund takes (for example, by buying high-yielding bonds of companies with poor credit ratings), the lower its liabilities appear to be.
“Funding the liability with risky assets doesn’t make the liability any smaller,” says Andrew Biggs, of the American Enterprise Institute, a conservative think-tank. A state pension fund may achieve the desired returns by investing in the stockmarket. But if that does not work out, the state must still pay its pensioners.
David Crane, an adviser to Arnold Schwarzenegger, the governor of California, describes the treatment of state pension funds as “Alice-in-Wonderland accounting”. Suppose, he says, that a state had to pay a bondholder $30,000 a year for 25 years and to pay a pensioner the same sum for the same period. The bond obligation would have a present value of $425,000 in its accounts but the pension liability, with the same cashflows, would be valued at just $320,000.
Private-sector companies are no longer allowed to use assumed returns when calculating their pension-fund liabilities on their balance-sheets. They have to use corporate-bond yields. The contrast makes it appear as if public-sector pensions can be delivered on the cheap. “The accounting suggests that governments can provide pension benefits at half the cost of a private-sector fund,” says Mr Biggs.
A more prudent way of measuring the liability is to regard a pension as a debt that the state owes its employees. So one possible discount rate is the state’s cost of borrowing, the yield on its municipal bonds. Some argue that pensioners have even greater rights than bondholders and that points to using a “risk-free” rate like the Treasury-bond yield. Both rates make the present value of pensions liabilities much higher than that declared by the states.
Using Treasury bond yields as the basis for discounting, Mr Rauh and Mr Novy-Marx calculate that states’ pension liabilities are as much as $5.3 trillion. That is 68% more than reported by the states, and produces the authors’ figure of $3.4 trillion for the gap between liabilities and assets.
In their defence, the states say that they are following the Governmental Accounting Standards Board (GASB), which recommends discounting liabilities by assumed returns. Even on that optimistic basis, states are not putting enough aside. According to the Centre for Retirement Research, their average funding ratio, using the GASB approach, fell from 103% in 2000 to 78% last year (see chart 2); with a risk-free rate underfunding would be much worse. Despite this shortfall, 21 states failed to make their full contribution to their pension funds over the past five years, according to Eileen Norcross of George Mason University in Washington, DC.
Trouble in Trenton
New Jersey provides a prime example of America’s pension difficulties. In August the Securities and Exchange Commission (SEC) charged the state with fraud for misrepresenting the underfunding of its pension plans to municipal-bond investors. This was the first time a state had been charged with violating federal securities laws. It settled the case without admitting or denying the SEC’s findings.
A study by Ms Norcross and Mr Biggs outlines, using the Treasury yield as a discount rate, how New Jersey has run up a pensions deficit of $174 billion. That is equivalent to 44% of the state’s GDP, or more than three times its official debt.
The problems started in 1992 when the then governor, Jim Florio, increased the assumed return on pension assets from 7% to 8.75%. That allowed contributions to be reduced and helped the state balance its budget. Further reforms in 1994 and 1995 eased the accounting assumptions, allowing Mr Florio’s successor, Christie Whitman, both to cut taxes and to balance the budget. In the late 1990s the fund bet heavily on technology stocks, giving a brief boost to asset values. Employees’ contributions were cut from 5% of payroll to 3%. New Jersey also increased benefits, giving pension rights to surviving spouses in 1999 and a boost of 9.1%, in effect, to scheme members in 2001, just as the dotcom bubble was bursting and the fund’s assets were falling in value. The effect of this chronic underfunding on the pension scheme for the police and firefighters is shown in chart 3.
In March Chris Christie, the Republican governor elected in November 2009, reduced the pension benefits of new state employees. Last month he unveiled a more ambitious plan with several measures that affect existing workers. These include an increase in their contributions to 8.5%, raising the qualification for early retirement from 25 to 30 years of service, moving the normal retirement age to 65 and ending future inflation adjustments. It is too soon to tell whether Mr Christie will get this plan through. Not surprisingly, the unions oppose it. “Once again, it’s an attack on the middle class,” said Hetty Rosenstein, who heads the state chapter of the Communication Workers of America.
Other states have also started on reform, but have focused mainly on restricting the benefits of new employees. Michigan closed its defined-benefit scheme to entrants back in 1997 and Alaska moved to a defined-contribution plan for new staff in 2006. Utah is closing its defined-benefit plan to new employees next June.
But this makes only a small dent in a huge problem. The bulk of liabilities consist of promises to people already working or retired, which are often legally protected. Reducing this bill will take a much bigger reform. Mr Rauh and Mr Novy-Marx estimate that raising the retirement age by a year would trim the cost by 2-4%; a cut of a percentage point in inflation-linking would slash it by 9-11%. But states that have tried to adjust existing promises, such as Colorado, Minnesota and South Dakota, which have frozen cost-of-living adjustments, have faced challenges in court.
States will have to make difficult choices. A change to the rights of existing scheme members is sure to have an adverse effect on people on low pay, nearing the end of their careers or already in retirement. A cap on the cost-of-living adjustment, for example, would be a nightmare for pensioners were inflation to flare up, because they would have no way of making up the loss in their purchasing power.
But after years of neglecting the problem, such changes will be hard to avoid. In the words of Dan Liljenquist, a state senator in Utah and an architect of its reforms: “This is not a conservative-versus-liberal issue, this is a reality issue.”