concluding that § 1054(c)'s actuarial equivalence requirement "means that plans must use the kind of actuarial assumptions that a reasonable actuary would use at the time of the benefit determination"Summary of this case from Belknap v. Partners Healthcare Sys.
Case No. 19-C-0505
DECISION AND ORDER
Plaintiff Rickie K. Smith alleges that the Rockwell Automation Pension Plan violates the Employee Retirement Income Security Act of 1974 ("ERISA") because it incorporates outdated actuarial assumptions that result in certain alternative pension payments being less than the actuarial equivalent of a normal pension. He seeks relief against the plan, the plan sponsor (Rockwell Automation, Inc.), and the plan's employee benefits committee. Before me now is the defendants' motion to dismiss the complaint for failure to state a claim upon which relief can be granted. See Fed. R. Civ. P. 12(b)(6).
Rickie K. Smith worked at Rockwell Automation for approximately 42 years and participated in its pension plan. The plan is a defined benefit plan within the meaning of ERISA. Under the plan, when a participant retires, he or she receives a pension calculated according to the plan's terms. The amount of the pension generally depends on the participant's years of service and compensation. Compl. ¶ 32.
ERISA requires that benefits from a defined benefit plan be paid to married participants in the form of a qualified joint and survivor annuity unless the participant, with the consent of his or her spouse, elects an alternative form of payment. See 29 U.S.C. § 1055(a) and (b). A qualified joint and survivor annuity is an annuity for the life of the participant with a survivor benefit for the life of the spouse that is not less than 50%, and not greater than 100%, of the annuity payable during the joint lives of the participant and the spouse. Id. § 1055(d)(1). For an unmarried participant, a qualified joint and survivor annuity is a "single life annuity." See 26 C.F.R. § 1.401(a)-20, Q&A 25. A single life annuity is a payment stream that starts when the participant retires and ends at death. See Compl. ¶ 2.
As indicated, a plan may offer participants alternative annuity forms, which are defined as "qualified optional survivor annuities." See 29 U.S.C. § 1055(d)(2); Compl. ¶ 18. The Rockwell plan allows participants to choose such annuities. One such optional annuity is a "certain and life" annuity, under which payments are made for the life of the participant or for at least a specified number of years. If the participant dies before receiving payments for the specified period, the remaining payments are made to the participant's beneficiary. When he retired, Smith elected to receive his pension in the form of a 10-year certain-and-life annuity, with his son as the beneficiary.
Because Smith proposes to represent a class of participants, he also includes allegations in his complaint about the Rockwell plan's qualified joint and survivor annuity. However, because the issue of class certification is not presently before me, I will not further discuss pension forms other than the one Smith elected, the 10-year certain-and-life annuity.
ERISA requires that a pension paid in the form of a certain-and-life annuity (or any alternative form of annuity) be "the actuarial equivalent of" a single life annuity. See 29 U.S.C. §§ 1054(c)(3); 1055(d)(1)(B) and (2)(A)(ii); see also Compl. ¶ 36. ERISA does not define "actuarial equivalent." However, in ordinary usage, benefits are actuarially equivalent when they are paid in ways that make them equally valuable to each other after factoring in the time value of money and the annuitant's life expectancy. See Call v. Ameritech Management Pension Plan, 475 F.3d 816, 817 (7th Cir. 2007). To determine actuarial equivalence, the plan must first use morality tables to predict how long the annuitant will live and thus determine how many payments he is likely to receive. The plan must then apply an interest rate to the payments to adjust them for the time value of money. As alleged in the complaint, "[t]he mortality table and the interest rate together are used to calculate a 'conversion factor' which determines the amount of the benefit that would be equivalent to the [single life annuity] the participant accrued." Compl. ¶ 3.
The Rockwell plan's governing documents specify that, to calculate actuarial equivalence for the annuity that Smith elected to receive, the plan must use a specific mortality table, known as the 1971 Group Annuity Mortality Table for Males (the "1971 GAM" table). Compl. ¶ 49. The governing documents further provide that the applicable interest rate is 7%. Id. According to the complaint, the 1971 GAM is "more than 40 years old" and reflects life expectancies of retirees in 1970. Compl. ¶ 51. In 1970, a 65-year-old had a life expectancy of 15.2 years. Id. However, in 2010, a 65-year-old had a life expectancy of 19.1 years, a 26% increase. Id. Thus, in 2010, the average retiree receiving a single life annuity would have expected to receive more payments than a retiree in 1970. The plaintiff alleges that, by using the 1971 GAM to calculate actuarial equivalence, the plan's optional annuities assume that the annuitant will die sooner than average and thus receive fewer payments than is likely. This, in turn, causes the value of the optional annuity to be less than the actuarial equivalent of a single life annuity, in violation of ERISA.
For certain other annuities, the specified mortality table is the 1984 Unisex Pension table. Compl. ¶¶ 5, 35. However, because Smith's benefits were calculated using the 1971 GAM table, I will not further discuss the 1984 Unisex Pension table.
The plaintiff brings three claims for relief based on this alleged violation on behalf of himself and other similarly situated plan participants. First, he brings a claim against the plan for declaratory and equitable relief under 29 U.S.C. § 1132(a)(3), which permits a plan participant to bring a civil action to redress ERISA violations. Second, he brings a claim under § 1132(a)(3) for "reformation" of the plan, by which he means an order requiring the plan to update its mortality tables so that it provides actuarially equivalent benefits when required by ERISA. Finally, he brings a claim for breach of fiduciary duty against the plan committee for failure to update the plan's actuarial assumptions and a related claim for breach of the duty to supervise and monitor the committee against Rockwell.
The defendants now move to dismiss all claims for failure to state a claim upon which relief can be granted.
To avoid dismissal under Rule 12(b)(6), a complaint must "state a claim to relief that is plausible on its face." Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570 (2007). "A claim has facial plausibility when the plaintiff pleads factual content that allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged." Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009). The complaint must, at a minimum, "give the defendant fair notice of what the claim is and the grounds upon which it rests." Twombly, 550 U.S. at 555. In construing a plaintiff's complaint, I assume that all factual allegations are true but disregard statements that are conclusory. Iqbal, 556 U.S. at 678.
In the present case, the outcome of the motion to dismiss turns on the meaning of the term "actuarial equivalent" in the parts of ERISA that require certain annuity forms to be actuarially equivalent to a single life annuity. See 29 U.S.C. §§ 1054(c)(3), 1055(d)(1)(B) and (2)(A)(ii). ERISA does not define "actuarial equivalent." In general, however, the parties agree that this term requires the use of mortality tables and an interest rate to adjust the benefits for the time value of money and the annuitant's life expectancy. But they disagree about how the term applies to a plan's selection of mortality tables and interest rates. The plaintiff alleges that a plan may not use a mortality table from more than 40 years ago and claim that the calculated benefit is actuarially equivalent to a single life annuity today. The plaintiff essentially argues that no reasonable actuary who was asked to calculate the actuarial equivalent of an annuity at the time of his retirement would use a 40-year-old mortality table. Thus, the plaintiff argues, his annuity cannot be deemed to be the actuarial equivalent of a single life annuity.
The defendants do not suggest that a reasonable actuary asked today to calculate the present value of future retirement benefits would use a 40-year-old mortality table. Instead, they contend that a plan pays actuarially equivalent benefits so long as it calculates actuarial equivalence using actuarial assumptions that were reasonable at the time they were written into the plan. See Br. in Supp. at 1; Reply Br. at 7. Under the defendants' approach, if a plan calls for use of a mortality table from 1970, and if the use of such a table was reasonable at the time the plan provision adopting this table became effective (such as in 1974), then the use of the table will always be reasonable. Even if 100 years have passed and retirees are living 20 years longer than they were in 1970, a retiree who receives an annuity based on the 1970 mortality table must, under the defendants' view of ERISA, be deemed to be receiving the actuarial equivalent of a single life annuity commencing in the year 2070.
The defendants make no attempt to ground their interpretation of the statute in the ordinary meaning of "actuarial equivalent," as that term is used in § 1054(c)(3) and § 1055(d). Instead, they point to other provisions that appear in ERISA, the Internal Revenue Code, and the regulations interpreting each of these statutes. The conclusion that the defendants would like me to draw from these provisions is that Congress could not have intended to require that plans periodically review their actuarial assumptions to ensure that they are reasonable at the time benefit calculations are made. However, as discussed below, nothing in these other provisions suggests that the term "actuarial equivalent" means "actuarial equivalent as of the date the plan adopted its actuarial assumptions."
A. The Requirement that Actuarial Assumptions be Specified in the Plan
The defendants first appeal to a provision of the Internal Revenue Code that requires a pension plan to specify its actuarial assumptions in the plan itself in a way that precludes employer discretion. See 26 U.S.C. § 401(a)(25). This provision arose out of earlier Treasury Regulations and Revenue Rulings, and so I begin by discussing these earlier sources.
A Treasury Regulation promulgated under the Internal Revenue Code has long provided that a pension plan will qualify for tax-advantaged status only if it provides "definitely determinable benefits." See 26 C.F.R. § 1.401-1(b)(1)(i). Generally, this means that the plan document must specify how benefits are calculated in a way that is not subject to the discretion of the employer. See Rev. Rul. 74-385, 1974-2 C.B. 130. In 1979, the Internal Revenue Service issued a revenue ruling in which it concluded that a plan's stating that a form of benefit is "actuarially equivalent" to another form of benefit does not provide definitely determinable benefits unless it also identifies the actuarial assumptions that will be used to compute the benefit. See Rev. Rul. 79-90, 1979-1 C.B. 155. The revenue ruling states: "Whenever the amount of a benefit in a defined benefit plan is to be determined by some procedure (such as 'actuarial equivalent', 'actuarial reserve', or 'actuarial reduction') which requires the use of actuarial assumptions (interest, mortality, etc.) the assumptions to be used must be specified within the plan in a manner which precludes employer discretion." Id.
The revenue ruling also provides that the requirement to state actuarial assumptions in the plan may be satisfied by incorporating either "fixed" or "variable" standards. Id. One example of a "fixed" standard is the one Rockwell chose: specifying the exact mortality table and interest rate that would be used to compute benefits. See id. However, the revenue ruling makes clear that Rockwell was not required to specify an exact table and interest rate. The ruling provides that, "[a]s an alternative to these fixed standards, the plan may specify a variable standard which provides for self- adjusting changes which are independent of employer discretion." Id. It then sets out "[t]wo acceptable variable standards":
(1) specifying that the procedure will be performed by reference to a specified insurance or annuity contract available at the time of benefit determination from a specified insurance company, orId.
(2) specifying that the interest rate will be a designated percentage of the prime interest rate of a specified bank or banks at the time of benefit determination (while all other assumptions are also specified).
In 1984, Congress amended the Internal Revenue Code to codify the requirement that actuarial assumptions be stated in the plan. See Retirement Equity Act of 1984, Pub. L. No. 98-397, § 301, 98 Stat. 1426 (1984). It did so by amending § 401(a) of the Internal Revenue Code, which enumerates the requirements that a plan must satisfy in order to be "qualified" for tax-advantaged status. The amendment reads:
A defined benefit plan shall not be treated as providing definitely determinable benefits unless, whenever the amount of any benefit is to be determined on the basis of actuarial assumptions, such assumptions are specified in the plan in a way which precludes employer discretion.26 U.S.C. § 401(a)(25).
The defendants cite 26 U.S.C. § 401(a)(25) in their brief, but it does not support their position. This provision merely conditions a plan's tax-advantaged status on the plan's not granting its administrators the ability to manipulate actuarial assumptions. See Brengettsy v. LTV Steel (Republic) Hourly Pension Plan, 241 F.3d 609, 612 (7th Cir. 2001). Accepting the plaintiff's interpretation of ERISA would not in any way result in a plan's having the ability to manipulate actuarial assumptions. The plaintiff does not contend that all standards governing actuarial assumptions must be removed from the plan and that employers must be given discretion to choose which actuarial standards to use to perform a benefit calculation. Instead, he contends, a plan does not comply with ERISA if it incorporates outdated assumptions. Section 401(a)(25) does not prohibit employers from amending a plan's actuarial assumptions to bring them up to date. Indeed, § 401(a)(25) places no constraint whatsoever on an employer's discretion to amend the plan for any reason. And it is easy to draft an amendment that incorporates updated actuarial assumptions but does not also grant the employer discretion to manipulate those assumptions. The plan amendment merely needs to incorporate more accurate mortality tables and/or interest rates. Alternatively, the plan could adopt a variable standard that is self-updating, such as one of the variable standards identified in Revenue Ruling 79-90. Accordingly, § 401(a)(25) does not support the defendants' interpretation of "actuarial equivalent."
B. Provisions Specifying Interest Rates for Lump-Sum Distributions
Next, the defendants note that the Retirement Equity Act of 1984 added a requirement to ERISA that the interest rate used to calculate the present value of lump-sum distributions be no greater than the interest rate used by the Pension Benefit Guaranty Corporation to calculate lump sums. See Retirement Equity Act of 1984, Pub. L. No. 98-397, § 103, 98 Stat. 1426 (1984). The plaintiff's claim does not involve lump-sum distributions—it involves annuities—and thus this provision is inapplicable here. However, the defendants contend that because Congress chose to regulate the interest rate used to calculate the present value of lump-sum benefits but did not also regulate the actuarial assumptions used to convert single life annuities into alternative annuities, it must not have intended to require plans to periodically update their actuarial assumptions.
The defendants read too much into Congressional inaction. As the defendants note, when Congress passed the Retirement Equity Act of 1984, it was concerned with a specific problem: employers using unrealistically high interest rates to compute lump-sum distributions. Br. in Supp. at 9 n.10 (quoting H.R. Rep. No. 103-632, pt. 2 at 57 (1994) and Mathews v. Sears Pension Plan, No. 95-C-1988, 1997 U.S. Dist. LEXIS 9892, *10 (N.D. Ill. Jul7 8, 1997)). Thus, Congress legislated with that problem in mind: it adopted a cap on interest rates for lump-sum distributions. Congress's adopting this cap without also addressing the actuarial assumptions used to calculate annuities does not suggest that Congress intended that plans be permitted to use outdated actuarial assumptions in connection with annuities. Rather, the most this suggests is that, in 1984, Congress did not think that legislation was needed to prevent plans from using unrealistic actuarial assumptions when converting single life annuities into alternative annuities. Accordingly, this provision of the Retirement Equity Act is not relevant to the meaning of the term "actuarial equivalent."
C. Anti-Cutback Rule
Next, the defendants contend that the plaintiff's interpretation of "actuarial equivalent" creates a conflict with what is known as the "anti-cutback" rule—a rule that prohibits any amendment of a pension plan that would reduce a participant's "accrued benefit." See Central Laborers' Pension Fund v. Heinz, 541 U.S. 739, 741 (2004). This rule appears in both ERISA and the Internal Revenue Code. See 29 U.S.C. § 1054(g) & 26 U.S.C. § 411(d)(6). In the Retirement Equity Act of 1984, Congress amended the anti-cutback rule to make clear that it applies to early-retirement benefits and optional forms of benefits, not only to normal retirement benefits. See Pub. L. No. 98-397, § 301(a)(1)-(2), 98 Stat. 1426 (1984); Central Laborers', 541 U.S. at 744. The alternative annuity form at issue in this case is an "optional form of benefit" under § 1054(g)(2)(B).
The alleged conflict between the plaintiff's interpretation of "actuarial equivalent" and the anti-cutback rule arises because a plan amendment that changes actuarial assumptions could reduce a participant's optional form of benefit on the date of the amendment. A revenue ruling notes that, under Treasury Regulation 1.411(d)-(3)(b), "plan provisions indirectly affecting a participant's accrued benefit [for purposes of the anti-cutback rule] include actuarial factors for determining optional or early retirement benefits." Revenue Ruling 81-12, 1981-1 C.B. 228. It then states that a plan amendment changing a plan's actuarial assumptions is subject to the anti-cutback rule. However, it notes that the anti-cutback rule does not prohibit an amendment that changes a plan's actuarial assumptions; it prohibits only amendments that have the effect of decreasing a participant's accrued benefit. But even if the amendment to actuarial assumptions decreases a participant's accrued benefits, the amendment will still be allowed if it is accompanied by other plan language that preserves the accrued benefit. The revenue ruling then outlines several methods by which the plan could preserve the accrued benefit. One method is for the plan to provide that the actuarial equivalent of the accrued benefit on or after the date of the change will be the greater of (1) the actuarial equivalent of the accrued benefit as of the date of change calculated under the old actuarial assumptions or (2) the actuarial equivalent of the total accrued benefit calculated under the new assumptions.
The Secretary of the Treasury has ultimate authority to issue regulations governing the anti-cutback rule under both the Internal Revenue Code and ERISA. See Central Laborers', 541 U.S. at 746-47.
The defendants contend that the anti-cutback rule effectively prohibits plans from amending actuarial assumptions during times when the use of a reasonable mortality table and a reasonable interest rate would decrease the dollar amount of the optional benefit. However, as just explained, the plan could amend its assumptions so long as it also adopted a provision guaranteeing that the participant would never receive less than the actuarial equivalent of his accrued benefit on the date of the amendment as calculated under the old assumptions. Moreover, unless the participant happens to retire or stop accruing benefits shortly after the amendment is adopted, it is likely that he will eventually receive greater benefits under the new assumptions because, as new accrued benefits are added to his old accrued benefits, the total payout on the date of retirement under the new assumptions will exceed his accrued benefit as of the date of the plan amendment under the old assumptions. (In this scenario, the amount of the accrued benefit would essentially be frozen between the date of the amendment and the date on which the addition of new accrued benefits causes the value of the total accrued benefit (as calculated under the new assumptions) to exceed the value of the old accrued benefits as of the date of the amendment (as calculated under the old assumptions). The anti-cutback rule does not prohibit this result, because the freezing of benefits would not reduce the benefits the participant already accrued. See 26 C.F.R. § 1.411(d)-3(a) (Example 2).)
Generally, a decrease in benefits is more likely to occur when the plan changes its interest rate rather than its mortality tables, because interest rates are more volatile than mortality tables. However, it is conceivable that life expectancy could decrease over a given period, and thus updating mortality tables could also result in a decrease in benefits. In any event, the plaintiff's interpretation of "actuarial equivalence" would have to be applied to both mortality tables and interest rates (because both always factor into the actuarial-equivalence calculation), and thus even if life expectancy always improved, it would still be possible for an amendment to actuarial factors to result in a decrease in benefits.
Still, the defendants contend, this approach creates a conflict with the plaintiff's interpretation of "actuarial equivalent" because the plan provision that preserves the calculation that existed under the old assumptions could result in a beneficiary receiving an annuity that is not the "actuarial equivalent" of a single life annuity. As I understand the defendants' argument, they contend that if a plan updates its actuarial assumptions to reflect current interest rates and mortality but, to comply with the anti-cutback rule, also provides that some participants are entitled to an annuity that is calculated using the old actuarial assumptions, then the plan will violate the requirement that the alternative annuity be the "actuarial equivalent" of a single life annuity. In this scenario, the value of the annuity calculated under the old assumptions will be greater than the actuarial equivalent of a single life annuity. The defendants contend that, to avoid this conflict, I should interpret ERISA to mean that plans are never required to update their actuarial assumptions, provided that their existing assumptions were reasonable when they were adopted.
I agree with the defendants that an amendment to a plan's actuarial assumptions could result in some conflict between the actuarial-equivalence requirement and the anti-cutback rule. But this conflict could also result if the defendants' interpretation of "actuarial equivalent" were adopted. Under the defendants' interpretation, an alternative annuity is actuarially equivalent to a single life annuity so long as the actuarial assumptions the plan uses to calculate the benefit were reasonable at the time they were adopted. But what if those assumptions were unreasonable at that time? Then the plan would violate the actuarial equivalence requirement under the defendants' interpretation, and its actuarial assumptions would have to be changed. It is conceivable that the change will result in a decrease in the accrued benefit of the participants who earned credit under the old, unreasonable actuarial assumptions. Thus, to comply with the anti-cutback rule, the plan would have to preserve the accrued benefit of those employees by guaranteeing that they will never receive less than the value of their benefit as calculated under the old, unreasonable assumptions. In this scenario, some retirees will receive a benefit that is greater than the actuarial equivalent of a single life annuity. Accordingly, the defendants' interpretation of "actuarial equivalent" does not resolve the potential conflict between the actuarial-equivalence requirement and the anti-cutback rule.
In any event, even if the defendants' interpretation resolved the conflict, it would not be a reasonable solution. This is so because there is simply no reasonable interpretation of the statutory language under which "actuarial equivalent" could mean "actuarial equivalent as of the date the plan adopted its actuarial assumptions." Moreover, it is possible to resolve this conflict without distorting the meaning of "actuarial equivalent." One reasonable approach is to interpret ERISA as requiring plans to achieve actuarial equivalence by increasing the amount of the single life annuity whenever the anti-cutback rule prohibits them from decreasing an optional benefit to achieve actuarial equivalence. Actuarial equivalence is essentially an equation with a single life annuity on one side and the optional form of benefit on the other. If the anti- cutback rule requires that the optional-annuity side of the equation to be raised above actuarial equivalence, a plan can restore equivalence by adding value to the single-life-annuity side. The defendants concede that this way of preserving actuarial equivalence is possible, but they contend that "[n]othing in ERISA requires plan sponsors to continually increase benefits in this manner." Reply Br. at 5. However, both the actuarial-equivalence requirement and the anti-cutback rule are "things" in ERISA, and it may be that these two provisions together require plan sponsors to occasionally increase benefits.
I also note that plans can minimize conflict between the actuarial-equivalence requirement and the anti-cutback rule by adopting variable actuarial assumptions that self-adjust to reflect changes in mortality and interest rates. Under Revenue Ruling 81-12, "in the case of a variable standard, any variation in accordance with the plan standard is not subject to" the anti-cutback rule. For example, if the plan incorporates a variable interest rate—such as by stating that the applicable interest rate will be a percentage of the prime interest rate of a specified bank or banks as of the date of the benefit determination, see Revenue Ruling 79-90—any change in the rate will not implicate the anti-cutback rule. Here again, the defendants contend that nothing in ERISA requires plans to use a variable standard, and I agree with them in that regard. However, the point is that a plan that is concerned about having to "continually increase benefits" (Def. Reply Br. at 5) has the option of adopting a variable standard. Once the variable standard is adopted, the plan will not have to continually increase benefits to comply with the anti-cutback rule—only those employees who accrued benefits under the old, fixed standard would potentially be entitled to increased benefits. A plan may eschew a variable standard if it wishes, but then the need to "continually increase benefits" will be a problem of its own making.
In short, the alleged conflict with the anti-cutback rule does not support the defendants' interpretation of "actuarial equivalent."
D. Plaintiff's Interpretation Does Not Require a Court to "Legislate" Actuarial Assumptions
The defendants next contend that, to accept the plaintiff's interpretation of "actuarial equivalent," the court "would have to legislate a detailed set of rules . . . specifying when a plan must change the actuarial assumptions it used to determine its contractually promised annuity benefits, how a plan should decide which mortality tables and interest rates to use and for which plan participants." Br. in Supp. at 15. I disagree. ERISA already contains the relevant rule: plans must ensure that any optional annuity forms are actuarially equivalent to a single life annuity. See 29 U.S.C. §§ 1054(c)(3); 1055(d)(1)(B) and (2)(A)(ii). This means that plans must use the kind of actuarial assumptions that a reasonable actuary would use at the time of the benefit determination. A court does not have to specify further details to enable plans to comply with the rule. They may comply by periodically consulting with professional actuaries who will review the plan's actuarial assumptions for reasonableness and recommend whether changes to mortality tables or interest rates are needed. Here, I agree with the defendants that reasonableness is a zone, not a point, see Br. in Supp. at 18, and thus ERISA likely does not require that plans use any specific mortality table or any specific interest rate at any given time. Rather, they may choose from the options that fall within the range of reasonableness at the time of the benefit determination, as determined by professional actuaries.
However, I disagree with the defendants' related assertion that their current actuarial assumptions must be reasonable as a matter of law because those assumptions reduced the value of Smith's pension by only 3.5% from what he alleges reasonable actuarial assumptions would have produced. See Br. in Supp. at 18-19. For purposes of converting annuities, ERISA requires actuarial equivalence, not something that approximates actuarial equivalence. See 29 U.S.C. §§ 1054(c)(3); 1055(d)(1)(B) and (2)(A)(ii). Although actuarial equivalence may encompass a range of values, any value outside of the reasonable range will violate ERISA. Here, the plaintiff alleges that the defendants' assumptions produced a value that is not within the range of reasonableness. Compl. ¶¶ 49-67. The fact that the difference between a reasonable calculation and an unreasonable calculation is only 3.5% does not excuse the violation.
As for how often a plan must revisit its assumptions, that will depend on the totality of the circumstances, as does any legal standard based on reasonableness. Mortality tables do not become outdated overnight, so they likely do not have to be reexamined frequently. Interest rates are more volatile, so they might require more frequent review. But in the case of both mortality tables and interest rates, a plan may protect itself by adopting a variable standard, as discussed in Revenue Ruling 79-90. A variable standard is self-updating, and thus a plan that uses one will not have to wonder whether it is reexamining its actuarial assumptions often enough. If, however, the plan elects to stay with a fixed standard, it will have to decide when to reexamine its assumptions. At any given time, a plan participant will have the option to commence a lawsuit alleging that a fixed standard has become outdated, and then a court will determine whether the standard is actuarially reasonable. If a plan prefers to avoid the risk of having the reasonableness of its actuarial assumptions determined through case-by-case litigation, it should either review its assumptions frequently or choose a variable standard. E. The Second Circuit's Decision in McCarthy v. Dun & Bradstreet Corp.
The defendants contend that their interpretation of ERISA—that a plan is never required to update actuarial assumptions that were reasonable at the time they were written into the plan—was adopted by the Second Circuit in McCarthy v. Dun & Bradstreet Corp., 482 F.3d 184 (2d Cir. 2007). This is debatable. In McCarthy, the question presented was whether the district court properly granted summary judgment to a plan on the issue of whether its use of a 6.75% interest rate in the course of actuarially reducing certain benefits was reasonable. The court, after considering multiple factors—including expert testimony by an actuary on the reasonableness of the rate and evidence that the rate was consistent with current market rates—agreed that the plaintiffs had not presented evidence from which a factfinder could conclude that the rate was unreasonable. See id. at 202-07. The court did not hold that the rate was reasonable at the time of the benefit calculation solely because it was reasonable at the time it was written into the plan. However, the court did place significant weight on the fact that the rate was reasonable at the time it was adopted. See id. at 204 & 205-06 (noting that the 6.75% rate was comparable to the rate on thirty-year government securities "that existed around the time the plan was created"). It also seemed to reject the plaintiffs' argument that ERISA required plans to periodically adjust the rate used to determine actuarial equivalence. Id. at 206.
In any event, McCarthy's reasoning on the issue of whether ERISA requires periodic updates to actuarial assumptions is not persuasive. The entirety of its reasoning is the following:
ERISA does not specifically require that retirement plans periodically adjust their actuarial interest rates. If a plan were required to do this, an employer potentially could manipulate the benefits provided to a participant, particularly in a year in which interest rates were extraordinarily high. The court recognizes the concern expressed in the relevant provisions of Title 26, Title 29, and the related regulations, that employers should not be able to manipulate actuarial assumptions to their benefit and to the detriment of employees. See, e.g., 26 U.S.C. §401(a)(25).Id. This reasoning is not persuasive for several reasons. First, while it is true that no provision of ERISA specifically states that plans must periodically adjust their actuarial interest rates, ERISA does state that plans must provide certain benefits in actuarially equivalent forms, see, e.g., 29 U.S.C. §§ 1054(c)(3); 1055(d)(1)(B) and (2)(A)(ii), which implies that the benefits will be calculated using reasonable actuarial assumptions. Thus, if actuarial assumptions that were reasonable at the time of their adoption become unreasonable due to changed circumstances, then ERISA requires an adjustment.
Second, while I agree that ERISA should be interpreted to prevent an employer from manipulating benefits, it is difficult to see how requiring plans to periodically review their actuarial assumptions for reasonableness could result in such manipulation. In McCarthy, the court stated that allowing periodic adjustments to actuarial assumptions might lead to an employer's manipulating a participant's benefits by changing rates in a year in which interest rates are "extraordinarily high." Id. However, the anti-cutback rule would prevent an employer from using a plan amendment to reduce the benefits that a participant had already accrued. See 29 U.S.C. § 1054(g). And a requirement to periodically review rates for reasonableness would prevent an employer from manipulating rates by amending the plan at a time when rates were extraordinarily high. Although the extraordinarily high rate might be reasonable and thus permissible at the time of the amendment, it would have to be reduced when interest rates returned to normal. Thus, it is the defendants' interpretation of ERISA that allows employers to manipulate benefits: if reasonableness is determined by conditions that existed when the rate was incorporated into the plan, then an employer could amend the plan at a time when rates were extraordinarily favorable to it and lock in the favorable rate for eternity. But under the plaintiff's interpretation, rates would have to be reexamined periodically, and thus an employer could not manipulate benefits by locking in a favorable rate.
For these reasons, I conclude that McCarthy does not support the defendants' interpretation of "actuarial equivalent."
F. The Complaint Adequately Alleges that the Rockwell Plan's Actuarial Assumptions Do Not Provide Actuarial Equivalence
Next, the defendants contend that the plaintiff has not adequately alleged that the plan's actuarial assumptions caused him to receive a benefit that was not the actuarial equivalent of a single life annuity. The defendants first argue that, as a matter of law, the plan's mortality tables must be reasonable because they are used by the Pension Benefit Guarantee Corporation and the Department of the Treasury. They then argue that, even if the tables are not reasonable as a matter of law, the plaintiff has failed to include allegations in his complaint that make his claim of unreasonableness plausible.
1. Reasonableness as a matter of law
The defendants argue that their continuing to use the 1971 GAM mortality table must be deemed reasonable because two federal regulations prescribe use of either the same table or a similar table.
First, the defendants point to a regulation that governs the Pension Benefit Guarantee Corporation ("PBGC"), 26 C.F.R. § 4022.8. This regulation states that when the PBGC pays an optional benefit form (such as a certain-and-life annuity), it will start by taking the amount of the normal retirement benefit and then "convert the benefit amount . . . to the optional form elected, using PBGC factors based on" mortality and interest. 26 C.F.R. § 4022.8(c)(7). The regulation then specifies that mortality will be based on "a fixed blend of 50 percent of the male mortality rates and 50 percent of the female mortality rates from the 1983 Group Annuity Mortality Table," id. § 4022.8(c)(7)(i), and that the interest rate will be six percent, id. § 4022.8(c)(7)(ii).
Notably, the regulation does not state that the PBGC's conversion will necessarily result in actuarial equivalence. --------
According to the defendants, if the PBGC uses the 1983 GAM table to convert benefit forms, then the plan's use of the 1971 GAM table must be reasonable. However, these are two entirely different mortality tables, and the defendants have not shown that the PBGC's use of the 1983 GAM table produces similar actuarial conversions as the plan's use of the 1971 GAM table. The plaintiff contends that the PBGC's mortality table and interest rate would produce conversions that are significantly more favorable to participants than the table and rate used by the plan. See Br. in Opp. at 21-22, n.23. Thus, at this stage of the case, I cannot say that the PBGC regulation is even evidence that the plan's use of the 1971 GAM table is reasonable, much less that it establishes that the plan's use of the 1971 GAM table is reasonable as a matter of law.
Second, the defendants point to a Treasury Regulation that identifies the 1971 GAM table as a "standard mortality table." See 26 C.F.R. § 1.401(a)(4)-12. Here, the defendants contend that because the Treasury Regulation describes the 1971 GAM table as a standard mortality table, the plan's continuing to use that table to calculate actuarial equivalence must be reasonable. However, the regulation does not state that the 1971 GAM table is a standard mortality table for all purposes. Instead, the regulation applies only to the Internal Revenue Code's requirement that pension plans not discriminate in favor of highly compensated employees. See 26 U.S.C. § 401(a)(4) (nondiscrimination requirement); 26 C.F.R. § 1.401(a)(4)-12 (explaining that definitions apply to regulations relating to the nondiscrimination requirement). The regulation prescribes the use of standard mortality tables to ensure that actuarial equivalence is "determined in a uniform manner for all employees" during testing for nondiscrimination. See 26 C.F.R. § 1.401(a)(4)-3(f)(7). For this purpose, it does not matter whether the mortality table is up to date; all that matters is that the same table is used to compare benefits for all employees. Thus, the Treasury's listing the 1971 GAM table as a standard mortality table for purposes of testing for nondiscrimination does not imply that the Treasury deems the 1971 GAM table a reasonable table to use when paying benefits.
2. The complaint adequately alleges that the plan did not provide Smith with an actuarially equivalent annuity
The defendants contend that the complaint does not plead sufficient nonconclusory facts to give rise to a plausible claim that the plan violates ERISA by incorporating outdated mortality tables. However, the complaint alleges that the 1971 GAM table is more than 40 years old and does not reflect the significant improvements in life expectancy that have occurred since they were published. See Compl. ¶¶ 49-51. These allegations, by themselves, give rise to a reasonable inference that the plan's actuarial assumptions are unreasonable and thus do not produce actuarial equivalence. Therefore, the plaintiff's claim is plausible. See Iqbal, 556 U.S. at 678 ("A claim has facial plausibility when the plaintiff pleads factual content that allows the court to draw the reasonable inference that the defendant is liable for the misconduct alleged."). Whether the tables are in fact outside the range of reasonableness is a matter to be resolved after discovery, including, if necessary, expert actuarial testimony.
G. The Plaintiff Has Stated Claims for Relief Under ERISA
Finally, the defendants contend that the plaintiff's specific claims for relief under ERISA's civil-action provisions are defective. However, as explained above, the plaintiff has plausibly alleged that the plan violates ERISA by paying optional benefit forms that are not actuarially equivalent to a single life annuity. Thus, the plaintiff properly brought this action under 29 U.S.C. § 1132(a)(3), which allows a participant to bring a civil action to enjoin and redress ERISA violations.
The plaintiff alleges a separate count for "reformation" of the plan under § 1132(a)(3) and then makes a request for recovery of benefits under the reformed plan pursuant to § 1132(a)(1)(B). This count appears to be entirely duplicative of his count alleging ERISA violations, and so I am not sure what purpose it serves. In any event, at this stage of the case, I see no reason to "dismiss" this part of the complaint, which is likely only an alternative legal theory and not a distinct claim for relief, and thus is not required to appear in the complaint at all. See, e.g., Shea v. Winnebago County Sheriff's Department, 746 F. App'x 541, 545 (7th Cir. 2018) ("Plaintiffs need not plead legal theories, and must plead in 'counts' only if they are founded on different occurrences and doing so would promote clarity.").
The plaintiff also includes a count in the complaint for breach of fiduciary duty against the plan's committee (i.e., its administrators) for failing to update its actuarial standards and against Rockwell for failing to monitor the committee to ensure that it used reasonable actuarial assumptions. Again, these claims appear to be entirely duplicative of the plaintiff's claim that the plan violates ERISA. If the plaintiff prevails on that claim, then he will receive the relief that he requests in connection with his breach-of-fiduciary duty claims without needing to prove the additional elements of the latter claims, such as that the violation of ERISA was so plain that the plan administrators had a duty to override the plan's terms. See 29 U.S.C. § 1104(a)(1)(D). But while the claims are redundant, they are not implausible, and so I will not dismiss them. Still, I encourage the plaintiff to limit himself to a single claim under § 1132(a)(3) seeking redress for the plan's alleged violation of ERISA's actuarial-equivalence requirement, which should provide him with all the relief to which he (and any eventual class) could conceivably be entitled in this action. See Vodak v. City of Chicago, 639 F.3d 738, 750-51 (7th Cir. 2011) (encouraging plaintiffs to forgo largely duplicative claims).
For the reasons stated, IT IS ORDERED that the defendants' motion to dismiss the complaint is DENIED.
Dated at Milwaukee, Wisconsin, this 10th day of February, 2020.
United States District Judge