Closing DB Plans Increases Sponsor Costs

For the Week of December 10th from the Nation Institute on Retirement Security:

A new study by the National Institute on Retirement Security (NIRS) finds costs have risen markedly for plan sponsors in five states since they closed their defined benefit (DB) pension plans. NIRS’ new report, “No Quick Fix,” documents consistent patterns across these five states showing that, as the demographics within the plans change, cash flow is affected and negative cash flow increases. Workforce impacts are also a notable consequence, with each state examined experiencing challenges with employee turnover in recent years. Finally, NIRS says the retirement security of public employees has also been put at risk due to the high degree of “leakage” found in the defined contribution (DC) plans that replaced DB plans.

The states and plans featured in the NIRS report include:

• Michigan: the State Employees’ Retirement System (SERS) plan;
• Alaska: the Public Employees Retirement System (PERS) and Teachers Retirement System (TRS) plans;
• Kentucky: the five plans under the Kentucky Public Pensions Authority (KPPA), which are:
  • Kentucky Employees Retirement System (KERS) non-hazardous;
  • Kentucky Employees Retirement System (KERS) hazardous;
  • County Employees Retirement System (CERS) non-hazardous;
  • County Employees Retirement System (CERS) hazardous;
  • State Police Retirement System (SPRS);

• Oklahoma: the Public Employees Retirement System (PERS);
• West Virginia: the Teachers Retirement System (TRS);

NIRS describes these plans as “outliers” from the typical public plan. “Just as there are outlier plans that have a long history of poor funding and consequently find themselves in challenging financial situations, these plans represent another type of outlier— those that have moved sharply away from the DB pension model.”

The NIRS report’s analysis of the consequences of those decisions makes the following key findings:

• Employer costs increased significantly after closing a pension plan in the states studied. In Michigan, for example – which closed its State Employees’ Retirement System (SERS) plan in 1996 and replaced it with a DC plan –retirement costs have increased from $169 million in 1998 to $1.3 billion in 2022, or 7.8 times higher, while the funding ratio has fallen from 109 percent to 69 percent.
• In the closed plans studied, cash flows have become more negative over time as demographics shift and the plan begins to spend down its assets. In 2001, Michigan SERS was paying benefits equal to 4.5 percent of assets, with minimal contributions of 1.1 percent of assets. As NIRS explains, this meant the plan was responsible for making payments equal to 3.4 percent of assets if investment markets were flat throughout the year. However, by 2022, despite much higher contribution levels representing 5.2 percent of assets, negative cash flow increased to 5.7 percent because, with thousands of workers still accruing benefits in the DB plan 26 years after it was closed, benefit payments are now equal to 10.9 percent of assets. Complicating this is the fact that even a modest decline in investment returns during the year could necessitate selling a significant portion of plan assets while asset prices are low, NIRS underscores.
• Despite representations that younger workers will be attracted to savings-based plans such as cash balance and DC plans, NIRS finds the available retention data shows poor employee retention in the new DC/cash balance plans or tiers. In Alaska, for example, which closed its two statewide DB pension plans in 2006, replacing them with DC plans, the teacher turnover rate is significantly higher among those in the DC plan, higher than it was in the past, and higher than in other states. The Alaska Department of Public Safety (DPS) also reported to the state legislature that the lack of a DB pension was one of the primary obstacles to recruiting and retaining new state troopers. “Workforce management has become a challenge in many of these states with closed plans,” NIRS concludes.
• DC plans are failing to help many workers accumulate sufficient retirement savings, NIRS notes, pointing to a recent study which found, nationally, 41.4 percent of DC accounts are cashed out when a worker separates from a job. This holds true whether in the public or private sector. Furthermore, NIRS says “the evidence indicates those dollars are unlikely to be used for producing retirement income.” And the indications are that in the five states NIRS examined in this study, there is serious “leakage” from DC plans of savings intended to be used for retirement security. For example, the dollars leaving the Michigan SERS DC plan (pre-retirement) during the past 11 years have been between 40–50 percent of contributions in all years, except 2020.

One positive finding in the new NIRS report involves West Virginia, which closed its Teachers’ Retirement System, a DB pension plan, in 1991, placing new teachers in a DC plan. However, policymakers determined that teachers participating in the new DC plan were not saving enough to produce adequate income in retirement. Meanwhile, the funded status of the then-closed DB plan continued to decline. Upon closer examination, the legislature found if it returned to the DB plan, it could provide equivalent benefits at half the cost of the DC plan. The DB plan was then reopened in 2005 – initially only to new hires, but eventually to those participating in the DC plan who chose to switch back to the DB (and 78 percent did).

NIRS stresses this demonstrates reopening a closed pension plan is a viable option that can reverse many of the harmful trends documented in its new report, if reopening is combined with contribution discipline. For example, in West Virginia, pension plan costs have stabilized, and the plan’s funded status continues to climb.

Finally, NIRS comments on North Dakota, which passed legislation in the spring of 2023 to close one of its statewide pension plans. While this legislation has not yet taken effect and the plan remains open to new hires as of the new NIRS report’s drafting, NIRS warns that “North Dakota is likely to experience the same funding, workforce, and retirement security challenges that have plagued other states that have closed DB plans.”

NIRS also suggests that since the North Dakota legislature did amend its pension funding policy to improve the funding of the plan when they acted to close it, this “may make it appear, in the near term, that the funded status of the plan is improving.” However, NIRS points out that, as in the case of states like Michigan and Alaska, “this short-term improvement may be short-lived if a prolonged market downturn hits once cash flows begin trending more negative over time and the demographics within the plan become unbalanced.”

The new NIRS report concludes by stressing, the “available evidence shows over and over again” that closing a public pension DB plan presents unexpected and long-lasting challenges. “From plan funding to workforce management, a closed pension plan restricts the capacity of plan sponsors to operate in providing well-functioning public services,” NIRS underscores.

Moreover, “the do-it-yourself nature of DC plans and the high rates of cashing out at separation in those plans reduces the retirement preparedness of public employees.”

Dan Doonan, NIRS Executive Director, summed up the report by noting although there “often are claims that closing an existing pension to new hires will improve matters in a variety of ways,” he said “it’s hard to find proof among the public pension plans that have closed in the past 27 years.” Instead, Doonan says, switching away from a pension “starves the plan of employee contributions while the liabilities remain, creates higher negative cash flow, and leaves taxpayers supporting the costs of two plans for many decades.”

In short, “[c}losing a public DB plan offers no quick fix to the ongoing challenges of maintaining a robust and thriving public workforce and managing existing financial obligations,” NIRS warns.

One last point. NIRS points out “one important takeaway from the partial data” provided concerning cash outs of DC plans “is that it would be helpful for providers to track and report these data in a way that helps to clarify these impacts on retirement security, not just the plan’s financial results.” NIRS notes the term “disbursements” clarifies what is happening to the DC plan, but it does not show whether these savings might be used to provide retirement income through a rollover to another tax-preferred retirement savings arrangement, or whether they are otherwise “spent” at the time of withdrawal.

NCTR agrees with this observation. While administratively this may present certain challenges to providers, the results could potentially be very useful in addressing the major leakage associated with DC plans.

NCTR is proud to be a founding member of NIRS and encourages our members to familiarize themselves with the excellent research NIRS produces which can be very important in defending the DB pension model used by so many public pension plans. Thank you, NIRS!

• NIRS: “No Quick Fix: Closing a Public Pension Plan Leads to Unexpected Challenges”