Colorado’s pension debt may be worse than policymakers think

Despite two years of outstanding investment returns, Colorado’s public employee pension plan has continued to experience growing pension debt. Nearly all of Colorado’s state workers are enrolled in the Public Employees’ Retirement Association (PERA), a defined benefit retirement plan. PERA uses a 7.25% assumed rate of investment return to project its market gains and calculate sufficient annual contributions. Over the past two years, there’s been good news—PERA’s investment returns have greatly exceeded the assumed rate of return. In 2019, PERA achieved a return of 20.3%, and the plan recently reported a 17.4% return for 2020. Yet, despite those investment returns, PERA recently saw a $1 billion increase in its unfunded liabilities.

Defined benefit retirement plans depend on several assumptions to accurately price the cost of promised benefits. Assumptions on investment returns tend to be a major point of discussion. Long-term market gains need to exceed, or at least match, the expectations of a plan, or they will see growth in unfunded liabilities. But there are other assumptions that a plan uses that can also generate unexpected costs if they end up being inaccurate. This appears to be the case for Colorado’s public pension plan.

PERA’s ability to fully fund the benefits promised to Colorado’s workers will ultimately depend on being able to match or exceed the 7.25% assumed rate over several decades, so it is important not to read too much into one or two years of results. That said, investment returns being above assumptions should lead to a modest reduction in the plan’s unfunded liabilities.

Effective pension funding relies on more than just investment returns, however. Judging by PERA’s good 2020 return, one would expect to see a one-year reduction in pension debt for PERA, but the latest reporting shows the plan falling further behind by over $1 billion.

The cause of this added debt is a major adjustment the plan recently made to the assumptions used to predict mortality for PERA’s members. According to the plan’s reporting, the positive market gains reduced the fund’s unfunded liabilities by nearly $2.8 billion. But the change to its mortality assumptions, which more accurately project the length of time current members will be drawing benefits from the plan, added $3.1 billion to the calculation of pension liabilities.

Demographic factors like actual retirement rates, hiring rates, and mortality rates that differed from the plan’s assumptions about these factors also added about $300 million to liabilities in 2020. This has been a consistent issue for PERA, which justifies the plan’s move to adjust its mortality assumptions.

On top of that, contributions going into the plan were short of the amount suggested by actuarial standards by $280 million because the legislature decided to forgo a pension payment. Furthermore, interest on the nearly $31 billion in pension debt added another $233 million to the plan’s total debt in 2020. Colorado’s 2018 pension reform took steps to address the state’s chronic problem with insufficient contributions, but it will take time and further commitment to eliminate this annual drag on PERA’s funding.

Colorado’s experience is a good illustration of the wide range of factors that can affect a pension plan’s ability to reduce unfunded liabilities. As most US plans work their way back to full funding, it is clear that this process involves much more than just investment returns.

One of the largest challenges facing public pension plans is the fact that there are significant costs still unaccounted for and hidden in assumptions that plans make about everything from investment returns to how many new employees will be hired to how long retirees will live. As these hidden costs emerge, plans see their progress to full funding stagnate or even backslide.

Policymakers should be aware of these challenges and seek policies that can reduce the chances of future unexpected costs. Adopting a safer, more conservative approach to return assumptions would reduce a plan’s reliance on market outcomes to fulfill the retirement promises made to public workers. Policymakers should also confirm that plans are frequently and accurately evaluating their other assumptions about mortality, retirement rates, payroll growth, and more.

As experience has shown in Colorado, the actual price tag attached to pension benefits may be significantly higher than what the current balance sheet shows. This needs to be addressed to avoid a constant pension funding treadmill in which debts are not reduced despite good market results.