A Simple (But Hard) Way to Improve Public Pensions

Retirement systems could go a long way toward addressing their problems by getting real about their investment returns.
by , | November 7, 2012 AT 5:00 PM

Public employee pensions have long been a hot topic for Better, Faster, Cheaper. I have more than once made the case for switching from the traditional defined-benefit approach to a defined-contribution system with a cash-balance option that would insulate governments from short-term market fluctuations. For risk-averse workers, the cash-balance component would guarantee employees an annual interest rate on their contributions and the employer's match.

I stand by the recommendation, but for those unwilling to go quite that far, there is a simpler reform that can help address many of the problems that currently afflict public pension systems: more realistic assumptions about their investment returns.

As of 2010, state pension systems had an estimated unfunded liability of $1.38 trillion, a number that almost certainly has grown since then. And that doesn't include hundreds of teacher, firefighter and other local government systems. One thing almost all of them do have in common, however, is that their unfunded-liability estimates are based on overly optimistic assumptions about the annual returns the funds will earn.

One recent study looked at the impact of investment-return variations on Massachusetts' state pension fund, which assumes annual returns of 8.25 percent. The study looks at three scenarios. Under the most optimistic, which assumes annual returns of 7.5 percent, state payments for unfunded liabilities alone would rise to $3.6 billion by 2040, when state law requires the unfunded liability to be retired. That is about $400 million more than the current projected 2040 payment of around $3.2 billion.

This column appears in our monthly Better, Faster, Cheaper email newsletter. Click to subscribe.

Under a scenario that assumes a 5 percent annual return, the 2040 payment would increase by more than $2 billion, from $3.2 billion to about $5.25 billion. The worst-case scenario assumes annual returns of just 2 percent. In that case, the 2040 payment to the state fund would more than double, to about $7.9 billion.

Massachusetts' assumed returns are on the high side, but they are hardly unique. In a National Association of State Retirement Administrators survey, nearly four out of five funds assumed returns of between 7.5 and 8 percent.

The bottom line is that unfunded liabilities are even worse than the official estimates, and public pension funds need to adopt more realistic investment-return assumptions. It is indeed a simple reform, but I never said it would be easy.

Dialing back investment assumptions means admitting to higher unfunded liabilities and either paying a little more now or much more later on to retire the liabilities. Politically, paying more now is a tall mountain to climb, but Rhode Island Treasurer Gina Raimondo recently scaled it, reducing her state's assumed annual rate of return from 8.25 to 7.5 percent. As Raimondo told The Wall Street Journal, "I could have cranked up our discount rate to 10 percent and dropped our unfunded liability by half--except in 10 years someone wouldn't get a pension check."

Public pensions are clearly one of those issues that make you want to avert your glance; every time you look closely, the situation is worse than previously thought. Even if they choose not to adopt more-comprehensive reforms, state and other public pension systems should adopt more-realistic assumptions about their investment earnings.

If not, as Gina Raimondo said, the time will quickly be upon us when retirees won't get their checks. The political fallout from that would dwarf the pain of finding more money to kick in now.