In 2010, then-Utah State Sen. Dan Liljenquist led a successful state pension reform effort that dramatically reduced the risk of a meltdown of the underfunded system.

The reforms gave new employees a choice: They could switch to a defined contribution retirement system or continue in a defined benefit system in which employees would bear the risks associated with underfunding.

Specifically, the state would cap its annual contributions to the pension system at 10 percent of payroll. If contributions were required to increase above that rate, those who chose to remain in the system would be responsible for the difference. (A similar arrangement has been proposed for new Detroit workers in its Plan of Adjustment.) Liljenquist spoke recently at an Issues & Ideas forum sponsored by the Mackinac Center for Public Policy. A video replay of the event can be watched here.

The reforms offer some important lessons for Michigan.

Contain the problem

The 2008 market crash cost Utah 22.3 percent of its retirement assets, creating a gap between how much the state had saved to pay for retiree benefits and the amount needed to meet those promises. Michigan's losses were comparable. The payments necessary to alleviate the shortfall would strain the budget and take away from other vital needs. Specifically, it required 10 percent of Utah's general fund revenue for 25 years.

By closing the pension funds or shifting the underfunding risks to pension fund members, the state contained its ability to rack up future budget-busting unfunded liabilities. Doing so, the state also made it more likely that it would be able to keep its promises to retirees.

Legislators need good information on the current system's risks

Stay Engaged

Receive our weekly emails!

Traditional defined benefit pension systems require their managers to make assumptions about the future, while also recognizing the likelihood that even their best guesses won't come true. In other words, prudent management means making conservative estimates about the future.

The difference between getting an 8 percent return on pension fund investments and a 6 percent return can be catastrophic. Legislators should insist that the managers develop projections based on a variety of assumptions.

In Utah, lawmakers found that keeping the annual pension system contribution rates at the same pre-recession levels would drain the system even if its future return assumptions came true. Further meetings with the system's actuaries mapped out the likelihood of different outcomes.

Transition costs are avoidable

Politicians are concerned about the alleged transition costs from no longer enrolling new hires into a defined benefit pension system. These "costs" are triggered by an accounting rule that requires the state to record a more front-loaded method for paying off unfunded liabilities. But these accounting rules do not mandate state funding, they only change how the state records its pensions in its financial statements.

Moreover, this change in accounting treatment is avoidable. By offering a hybrid system as well as a defined contribution plan, Utah left this area unchanged.

Opposition from current employees can be overcome

People are afraid of losing their pensions. But they misidentify the real threat, which is an underfunded system in an insolvent government. This was the lesson from Pritchard, Ala., and Central Falls, Rhode Island.

Transitioning to systems that eliminate the underfunding risk protects members' benefits.

When the Utah Legislature debated the reform legislation, public employees protested. Their concerns were mitigated, however, when the underfunding risk was fully explained. Government workers did not cause the underfunding problem and are the primary beneficiaries of reforms that eliminate the risks.

These lessons should be carefully considered as Michigan legislators discuss bills to close its underfunded teacher retirement system.


James Hohman is assistant director of fiscal policy at the Mackinac Center for Public Policy, a research and educational institute headquartered in Midland, Mich. Permission to reprint in whole or in part is hereby granted, provided that the author and the Center are properly cited.