Court Holds That ERISA Does Not Require Actuarial Equivalence Factors to Be "Reasonable"
The U.S. District Court for the District of Massachusetts has dismissed an Employee Retirement Income Security Act (ERISA) class action lawsuit challenging a retirement plan's use of an outdated mortality table in calculating actuarially equivalent retirement benefits. Belknap v. Partners Healthcare System, Inc., No. 19-cv-11437, 2022 WL 658653 (D. Mass. March 4, 2022). In dismissing the case, the court held that there is no requirement under ERISA that the assumptions used to determine a participant's actuarially equivalent retirement benefits must be "reasonable." The decision is significant in that it diverges from a number of recent decisions addressing this issue.
Under ERISA § 204(c)(3), if a participant's accrued benefit under a defined benefit pension plan is determined at any time other than the participant's normal retirement age and/or in a form other than a single life annuity (SLA), the participant's benefit must be the "actuarial equivalent" of the SLA commencing at the participant's normal retirement age.
The lead plaintiff in Belknap was a former employee of Partners Healthcare System who retired in 2016 at age 62 and, under the company's defined benefit pension plan, began receiving a joint and survivor annuity (JSA) with his spouse as beneficiary. To calculate the JSA, the plan converted the SLA benefit beginning at the participant's normal retirement date using an adjusted 1951 mortality table.1 The plaintiff brought a class action suit under ERISA, alleging that using such an outdated mortality table was unreasonable (and violated ERISA) because it was based on outdated life expectancies, yielding a lower benefit that was not, in fact, actuarially equivalent to an SLA beginning at normal retirement age.
The court rejected the plaintiff's claims, holding that there is no requirement under ERISA that the assumptions used in calculating actuarial equivalence must be "reasonable." The court emphasized that ERISA does not define "actuarial equivalence" and does not specify — either by statute or under any regulation applicable to annuity benefits — that calculating actuarial equivalence requires using "reasonable" assumptions. Further, there did not appear to be any industry practice of using objectively "reasonable" criteria, as several expert witnesses testified that an actuary should follow the actuarial assumptions described in the plan document when calculating an actuarially equivalent benefit. Indeed, the only place where "actuarial equivalence" was defined in any relevant sense was within the plan itself, and the plan specifically stated that actuarially equivalent benefits are determined by using the adjusted 1951 mortality table.
The court therefore rejected the plaintiff's argument that ERISA requires the use of "reasonable" assumptions when calculating actuarial equivalence, particularly when the plan itself specifically requires the use of particular actuarial equivalence factors. The court also discussed the problems that would arise had it determined otherwise, such as the lack of any guidance regarding when actuarial equivalence factors would no longer be considered "reasonable" and how frequently such factors would be required to change.
The court's decision in Belknap is significant because it diverges from a number of recent decisions addressing this issue. Those decisions have typically concluded that the assumptions used to calculate actuarial equivalence must be "reasonable," and that when a plan uses an outdated mortality table, plan participants can at least plausibly state a claim under ERISA that they are not receiving actuarially equivalent benefits. See, e.g., Masten v. Metro. Life Ins. Co., 543 F. Supp. 3d 25, 33 (S.D.N.Y. 2021). The Belknap decision is therefore a bit of an outlier. Still, if the decision stands on any ensuing appeal, it could persuade other courts considering the issue to dismiss class action lawsuits alleging ERISA violations based on the use of outdated mortality tables.
1 1951 Group Annuity Mortality Table projected to the 1960 Mortality Table, set back two years for participants and three years for beneficiaries (Adjusted 1951 Group Annuity Mortality Table).