The Actuary Civil War of 2016 that broke out this past summer between the Society of Actuaries and American Academy of Actuaries is emblematic of the challenge facing public sector pension funds in America today. There is considerable consensus that public plans are grossly underfunded. According to their own actuarial reports, the largest pension plans run by the 50 states had a cumulative overfunded position in 2001 that collapsed into roughly $1 trillion in unfunded liabilities by 2014. There is profoundly less consensus on what factors have caused this underfunding to emerge and whether actuarial valuations are accurately reflecting the true health (or lack thereof) of American pension plans.
Consider that at the heart of the Actuary Civil War is a disagreement over discount rate policy. Historically public sector pension funds have used their assumed rate of return on assets to discount accrued liabilities. This flies in the face of how financial economics understands risk, which would suggest calculating the value of accrued pension liabilities should be relative to the risk of liabilities, not the risk of the plan assets. The debate is important because if public plans were to use market based liability values, total unfunded liabilities for the biggest state plans combined would likely be closer to $3 trillion or $5 trillion.
But there is a puzzle here: from the financial economics view of the world, pension funds explicitly choosing inappropriate discount rates have tacitly been underfunding their pension obligations for years by underreporting unfunded liabilities. Excessively high discount rates are therefore a proximate cause of at least part of the nation’s unfunded liability growth problem. Thus, we could argue that low funded ratios for public sector pension plans are in part caused by inappropriate discount rate policies.
However, should a given pension fund decide to lower its discount rate (whether or not they are embracing a financial economics approach to valuing liabilities), the result would be the fund reporting a higher value of accrued liabilities, a higher recognized amount of unfunded liabilities, and then a lower reported funded ratio. We know that as assumed rates of return have fallen over the past decade, so too have discount rates — the average discount rate used by public plans has fallen from more than 8% at the turn of the century to around 7.25% as of 2015. Thus, we could also argue that low funded ratios for public sector pension plans are in part caused by more appropriate discount rate policies.
This duality poses a problem for the academic (and policy) quest to understand exactly what is going on with America’s public sector pension plans: the metric we have for defining whether a plan is fiscally sound or not (the funded ratio) could be the product of either historically bad or recently good policy — or, in many cases, both. […]
A Pension Web with Many Strands
Looking past the funded ratio directly allows analytical focus to be put squarely on the pension funding behaviors themselves — and the factors that influence the behaviors in all of their interconnectivity. A standard model in epidemiology for why disease occurs and spreads is to take account of the interconnected relationship of multiple factors (Goldenfeld 2011). This “web of causality” model can be applied in understanding unfunded liabilities as a disease plaguing state and municipal finances across the country. First, there are a myriad of behaviors, i.e. explicit causes, which can drive of unfunded liability growth, such as:
- Underfunding actuarially determined contribution rates, either specifically in a given budget or systematically through statutory language;
- Adopting actuarial assumptions that are misrepresentative of reality, including using old mortality tables or salary growth assumptions that are based on historic trends but not recent demographic patterns;
- Adopting an open amortization policy that leads to negative amortization;
- Adopting a poorly designed asset allocation strategy;
- Increasing benefits without ensuring the change is fully funded, whether changes to benefit multiplier or ad-hoc COLAs; and
- Using actuarial cost methods that allow for systematic asset depletion, such as granting credits against normal cost when actuarially overfunded or allowing for experience account programs that skim assets off the top of investment returns in good years to fund benefit increase programs.
Second, there is a range of actuarial experiences, i.e. implicit causes, which result in growing unfunded liabilities, such as:
- Underperforming investment returns;
- Changes in mortality rates;
- Changes in inflation trends;
- Death or disability shocks, particularly for smaller public safety plans; and
- Salary experience being adjusted due to a sudden change in hiring policy or economic growth.
While each of these explicit and implicit causes might be analyzed independently, rarely do they occur independently. Paying 100% of employer contributions in a given year may be related to a spike in actuarial experiences; underperforming investment returns may be specifically related to the adoption of unrealistic assumed rates of return; benefits may be increased because the actuarial assumptions used by a plan report strong funded status, even if the long-term reality under different assumptions shows a poorly funded plan. Thus, when seeking to understand why a pension plan’s unfunded liabilities have grown, we should consider more than just individual actuarial experience (implicit causes) or funding behaviors (explicit causes). Instead those funding policy behaviors and their associated actuarial experience should be understood as an interconnected web that is being cohesively influenced by some set of factors.
Read the rest of the paper for a discussion about a three part model for refining our understanding of what causes inappropriate funding policy behavior, including: misaligned political incentives, misalignment between the groups that have “captured” policy decision making, and “rent seeking” by the various actors and organizations in pension funding policy.
This article republishes selections from “There is a Tangled Web of Factors Causing Inappropriate Pension Funding Behavior” by Anthony Randazzo, an academic article published in the The Journal of Law, Economics & Policy in 2017. A working paper version is available here.