For some states that have curbed their COLAs, including Colorado and South Dakota, it has brought billions of dollars in much-needed relief to their state retirement systems and avoided more painful choices on the investment side. “It can put the plan in a more positive cash flow near term, and with less money flowing out, there is more money to be invested,” said Ron Snell, senior fellow at the National Conference of State Legislatures in Denver.
The saving that comes from changing COLAs depends on how dramatic the changes are, and how long the cuts will last. The $10.6 billion Augusta-based Maine retirement system canceled its COLA for three years so far, while the $58.8 billion Olympia-based Washington state fund eliminated it for some employees' retirement benefits and limited it for others. Virginia will cap COLA increases in retirement benefits at 3% for non-vested participants and 5% for vested participants in the $53.6 billion Richmond-based fund.
Several pension funds, including the $6.5 billion Oklahoma City-based Oklahoma Public Employees and $72.1 billion Trenton-based New Jersey Division of Investment, tie any resumption of COLAs to specific levels of prefunding or investment returns. Some funds, like the $6.9 billion Providence-based Rhode Island Employees, tie it to both.
“It's very tempting,” said Alicia H. Munnell, director of the Center for Retirement Research at Boston College, in an interview. “You get a big reduction in your liability immediately. There are very few (other) things you can do. I think that COLAs are vulnerable.”
The concept of a COLA is straightforward, but the design is another matter. “We were impressed with the wide variety of them,” said Keith Brainard, research director for NASRA in Georgetown, Texas, in an interview. NASRA identified a dozen basic COLA configurations, including adjustments that are automatic or provided ad hoc by a governing board. Some state and local governments base COLAs on a fixed rate, typically 3%, or the consumer price index, while others tie it all or in part to the investment performance or funding level of the plan.
And plan administrators are getting ever more creative, Mr. Brainard noted, with some applying COLAs to a limited portion of a retiree's annual benefit — for example, the first $35,000 — or making retirees wait longer for it. One variation pegged to investment returns involves the creation of a separate reserve account that is funded and distributed only when the main fund has excess earnings.
One of the richest COLAs, offered by 10 states, is an automatic 3% boost compounded on the accrued benefit. That can add 26% to a plan's benefits cost, according to an analysis by Gabriel Roeder Smith & Co., an actuarial and pension consulting firm. Even a modest 1% COLA can add 7% to total costs, according to the analysis.
That makes COLAs ripe for cost-cutting. That was the case in South Dakota, where a previously automatic 3.1% COLA that had to be prefunded accounted for 25% of present benefit costs. “That's $2 billion that we knew were out there,” said Robert A. Wylie, executive director/administrator of the $8 billion South Dakota Retirement System, Pierre, in an interview. “It is a huge leverage on the overall cost of the plan, and people don't recognize that.”
It helped that state statute required retirement system officials to make changes when the funding threshold fell below 80%. When that happened in 2010, they calculated that getting back to 80% would take $400 million. They got three-fourths of the way by simply changing the COLA to 2.8% from 3.1%. “You can't get that kind of savings with other benefits,” Mr. Wylie said. The fiscally conservative state didn't want to lower anyone's benefit, “but we wanted to slow the growth,” he said.
So far, the change “has worked out amazingly well,” Mr. Wylie said. The plan went to 103% funded in the fiscal year ended June 30, 2011, from 88% in fiscal 2010 and 76% in fiscal 2009. He attributes 8% of that growth to investment performance, and 4% from the COLA change, which also saved on opportunity costs by having more money to invest, he noted. “It's a compounding effect.”