Best Practices for Setting Public Sector Pension Fund Discount Rates

What are the accrued liabilities of America’s public sector pension systems? This is far from a straightforward, arithmetic matter. Depending on whom you ask, the 50 states have accumulated pension debt of anywhere between $500 billion and $5 trillion. To put it mildly, that is a massive range of opinion about the level of unfunded state pension liability.

The principal cause of this variation between estimates is the discount rate being used in the forecast of pension finances. The discount rate is a critical factor for determining how much gets saved today to pay pensions in the future. The higher the discount rate employed, the lower will be the net present value of anticipated pension benefits, which are also known as accrued pension liabilities. The lower the present value of the accrued pension liabilities (i.e. the value of all future pension benefits measured in today’s dollars), the less the government and employees will need to pay into pension coffers today to cover those promised benefits when they come due. Thus, the higher the discount rate, the lower the rate of contributions flowing into a pension fund (all else equal). Conversely, the lower the discount rate, the higher annual contributions will need to be to ensure a fully funded system.

Accurately identifying the present value of pension liabilities isn’t just important for understanding the current level of unfunded liabilities. It is also critical to ensuring that state and local officials make sufficiently large annual contributions to their pension funds now and in the future, thereby avoiding unfunded liabilities in the first place. Yet financial economists, actuaries and public officials disagree sharply over how the future benefit payments promised to public workers should be discounted, and therefore what the present value of pension liabilities is.

This policy brief lays out a case for how state and local officials should go about setting their discount rate. It begins with an outline of best practices for setting the discount rate. It then tackles several myths and misnomers about the discount rate that are prevalent in discussions about public sector pension reform nationwide. It concludes with recommendations.

The status quo approach to selecting discount rates for state and local governments is completely backward. Using the expected rate of return on assets to discount liabilities considers the risk of assets instead of the risk of liabilities. Public officials should instead adjust their discount rates to reflect the time value of money to their employees and the risks of the streams of cash payments to retirees.

• Platinum Standard: The ideal approach for a state or local government discounting liabilities would be to use an average yield on its own issued bonds. This would provide a proxy that includes a risk premium associated with the employer government’s capacity to pay its debts.
• Gold Standard: For public employee pensions that are not protected by constitutional guarantees, an alternate approach would be to identify assets whose distributions closely match the liabilities of a pension fund and use the expected rate of return on those assets as the discount rate. The closest characteristics of such assets would likely be found in a portfolio of corporate bonds. Given the cost of constructing such a portfolio, an appropriate proxy would be a high-grade corporate bond index. However, for the majority of pensions that are protected by a constitutional guarantee, a better proxy would be the municipal bond index, which has a much lower risk premium- and one that is far closer to the risk of a government defaulting on its pension payments.
• Silver Standard: Finally, pension funds could use a combination of a risk-free rate, such as that implied by the yield on Treasury bonds, plus a fixed basis point risk premium (e.g. 100 or 200 basis, with the specific rate being determined on a case-by-case analysis). For municipalities where certain stakeholders are fiercely opposed to changing the discount rate, this might be a good first step toward better practices because it would disconnect analysis of the discount rate from the expected return.
• Phasing-In Changes: Whatever approach an employer government takes will likely mean a substantial near-term increase in employer contributions. As pension financial statements more accurately reflect the accrued liabilities of a plan as higher than stated under a lower discount rate, amortization payments on unfunded liabilities (pension debt) will rise unless the amortization schedule is reset. To avoid a budget shock it would be appropriate to phase in changes to a discount rate over a three- to five-year period. Thus, a municipality with an 8 percent discount rate that is moving to a 5 percent discount rate as measured by the three-year average yield on its municipal bonds could lower the discount rate one percentage point a year for three years, instead of making the move all in one year.

Public officials should look carefully at their discount rate policies, and adopt the best practice that is politically and fiscally possible. Ultimately, the main objective should be for pension plans to shift toward having discount rates reflect the risks of liabilities-not the potential performance of assets. This is one of the most critical remedies to protect the future of public sector pension plan solvency.

Attachments

• Why Discount Rates Should Reflect Liabilities: Best Practices for Setting Public Sector Pension Fund Discount Rates