James P. TEUFEL, Plaintiff-Appellant, v. The NORTHERN TRUST COMPANY, et al., Defendants-Appellees.
In 2012 Northern Trust changed its pension plan. Until then it had a defined-benefit plan under which retirement income depended on years worked, times an average of each employee's five highest-earning consecutive years, times a constant. Example: 30 years worked, times an average high-five salary of $50,000, times 0.018, produces a pension of $27,000. (We ignore several wrinkles, including an offset for Social Security benefits, a limit on the number of credited years, and a limit on the maximum credited earnings.) The parties call this the Traditional formula. As amended, however, the plan multiplies the years worked and the high average compensation not by a constant but by a formula that depends on the number of years worked after 2012. The parties call this arrangement the new PEP formula, and they agree that it reduces the pension-accrual rate. (There is also an old PEP formula, in place between 2002 and 2012, for employees hired after 2001; we ignore that wrinkle too.) Recognizing that shifting everyone to the new PEP formula would unsettle the expectations of workers who had relied on the Traditional formula, Northern Trust provided people hired before 2002 a transitional benefit, treating them as if they were still under the Traditional formula except that it would deem their salaries as increasing at 1.5% per year, without regard to the actual rate of change in their compensation.
James Teufel contends in this suit that the 2012 amendment, even with the transitional benefit, violates the anti-cutback rule in ERISA, the Employee Retirement Income Security Act. 29 U.S.C. §§ 1001–1461. He also contends that the change harms older workers relative to younger ones, violating the ADEA, the Age Discrimination in Employment Act. 29 U.S.C. §§ 621–34. The district court dismissed the suit on the pleadings, 2017 U.S. Dist. LEXIS 31674 (N.D. Ill. Mar. 6, 2017), and Teufel appeals.
The anti-cutback rule provides:
The accrued benefit of a participant under a plan may not be decreased by an amendment of the plan, other than an amendment described in section 1082(d)(2) or 1441 of this title.
29 U.S.C. § 1054(g)(1). Neither § 1082(d)(2) nor § 1441 matters to this case; the anti-cutback rule has other provisos too, but none applies. So all that matters is the basic requirement: the “accrued benefit” of any participant may not be decreased. Teufel insists that the 2012 amendment reduced his “accrued benefit” because he expected his salary to continue increasing at more than 5% a year, as it had done since he was hired in 1998, while the 2012 amendment treats salaries as increasing at only 1.5% a year.
To analyze this contention we need to be precise about how pension benefits are calculated for employees, such as Teufel, hired before 2002 and still covered by the Traditional formula until 2012. The plan first calculates an employee's accrued benefit as of March 31, 2012. That process starts with the number of years of credited service, multiplies that by the consecutive-high-five average salary, and multiplies by 0.018. The plan adjusts that result in following years by treating the high-five average (before 2012) as if that figure had continued to increase by 1.5% a year for each year worked after 2012. Finally, the plan adds benefits calculated under the new PEP formula for service after March 31, 2012.
This statement of the new formula shows why Teufel cannot succeed. If, instead of amending the plan in March 2012, Northern Trust had terminated the plan, calculated Teufel's accrued benefit, and deposited that sum in a new plan with additions to come under the new PEP formula, then Teufel would not have had any complaint. (He concedes that this is so.) What actually happened is more favorable to him: he gets the vested benefit as of March 2012 plus an increase in the (imputed) average compensation of 1.5% a year (for pre-2012 work) for as long as he continues working.
Teufel wants us to treat the expectation of future salary increases as an “accrued benefit,” but on March 31, 2012, when the transition occurred, the only benefit that had “accrued” was the sum due for work already performed. What a participant hopes will happen tomorrow has not accrued in the past.
Suppose the Traditional formula had remained unchanged but that in March 2012, as part of an austerity plan, Northern Trust had resolved that no employee's salary could increase at a rate of more than 1.5% a year. That would have had the same effect on the pre-2012 component of Teufel's pension as the actual amendment, but a reduction in the rate of salary increases could not violate ERISA, which does not require employers to increase anyone's salary. Curtailing the rate at which salaries change would not affect anyone's “accrued benefit.” Since that is so, the actual amendment also must be valid.
Teufel relies on decisions such as Hickey v. Chicago Truck Drivers Union, 980 F.2d 465 (7th Cir. 1992); Ruppert v. Alliant Energy Cash Balance Pension Plan, 726 F.3d 936 (7th Cir. 2013); and Shaw v. Machinists & Aerospace Workers Pension Plan, 750 F.2d 1458 (9th Cir. 1985). In these cases the language of the pension plan itself promised an increase in pension benefits—in one, a cost-of-living adjustment, in another a rate of interest added to the pension if the worker quit before retirement age, and in the third an adjustment in light of the salary earned by the current holder of the retiree's old job. The decisions all hold that these adjustments are part of the “accrued benefit” because they are among the pension plans' terms. See also Central Laborers' Pension Fund v. Heinz, 541 U.S. 739, 124 S.Ct. 2230, 159 L.Ed.2d 46 (2004) (plan cannot attach new conditions to benefits already accrued). But nothing in the Northern Trust plan's Traditional formula guarantees that any worker's salary will increase in future years. Teufel and others like him have a hope that it will, maybe even an expectation that it will, but not an entitlement that it will—and for the purpose of identifying the “accrued benefit” that's a vital difference. ERISA protects all entitlements that make up the “accrued benefit” but does not protect anyone's hope that the future will improve on the past. See Cinotto v. Delta Air Lines Inc., 674 F.3d 1285, 1296–97 (11th Cir. 2012).
One additional ERISA contention calls for brief mention. Teufel maintains that the plan's administrator violated 29 U.S.C. § 1054(h)(2) because it did not furnish all participants with a writing that described the 2012 amendment “in a manner calculated to be understood by the average plan participant”. To the extent Teufel faults the description for failing to tell participants that the amendment eliminated an accrued benefit, this contention fails for the reasons we have already given. To the extent that Teufel finds the language too complex—well, it seems clear to us, and it isn't apparent how it could have been made much simpler (all of these pension formulas have complexities). True, what seems clear to a federal judge may not be clear to “the average plan participant”, but Northern Trust provided its staff with an online tool that showed each worker exactly what would happen to that worker's pension, under a number of different assumptions about future wages and retirement dates, and under both the pre-2012 approach and the amended plan. A precise participant-specific summation is hard to beat for clarity and complies with § 1054(h)(2). Teufel makes a few other arguments based on ERISA, but they do not require discussion.
Teufel's argument under the ADEA fares no better. He acknowledges that the plan as a whole, and the 2012 amendment, is age-neutral, for pension eligibility is distinct from age. See Kentucky Retirement Systems v. EEOC, 554 U.S. 135, 128 S.Ct. 2361, 171 L.Ed.2d 322 (2008); Hazen Paper Co. v. Biggins, 507 U.S. 604, 113 S.Ct. 1701, 123 L.Ed.2d 338 (1993). Still, he maintains, the correlation between pension eligibility and age—plus the fact that the high-five-average feature of the Traditional formula was most valuable to older workers approaching their highest-earning years—means that the 2012 amendment produces a disparate impact that violates the ADEA. (Smith v. Jackson, 544 U.S. 228, 125 S.Ct. 1536, 161 L.Ed.2d 410 (2005), holds that a form of disparate-impact analysis applies under the ADEA.) The Traditional formula treats older workers better than younger ones (the high-five-average feature is more valuable the older one gets); and from this it follows that the elimination of the formula (or its reduction to a 1.5% annual increase) harms older workers relative to younger ones. So the argument goes.
We are skeptical about the proposition that curtailing a benefit correlated with age, and so coming closer to eliminating the role of age in pension calculations, can be understood as discrimination against the old. Kentucky Retirement Systems holds that a pension benefit for older workers does not violate the ADEA, but not that any such benefit, once extended, must be continued for life. At all events, the Supreme Court has never held that the disparate impact of an age-neutral pension plan can violate the statute. To the contrary, Kentucky Retirement Systems tells us that the relation between the ADEA and pension plans should be understood through the language of 29 U.S.C. § 623(i), which directly addresses the topic.
Section 623 as a whole is the basic rule against age discrimination. Section 623(i)(2) provides that “[n]othing in this section” (that is, all of § 623) prohibits an employer from “observing any provision of an employee pension benefit plan to the extent that such provision imposes (without regard to age) a limitation on the amount of benefits that the plan provides or a limitation on the number of years of service or years of participation which are taken into account for purposes of determining benefit accrual under the plan.” Just to avoid any doubt, § 623(i)(4) adds: “Compliance with the requirements of this subsection with respect to an employee pension benefit plan shall constitute compliance with the requirements of this section relating to benefit accrual under such plan.” In other words, a pension plan that complies with § 623(i) does not violate the ADEA.
The Northern Trust pension plan, both before and after the 2012 amendment, complies with § 623(i). Benefits depend on the number of years of credited service and the employee's salary, not on age. Because salary generally rises with age, and an extra year of credited service goes with an extra year of age, the plan's criteria are correlated with age—but both Kentucky Retirement Systems and Hazen Paper hold that these pension criteria differ from age discrimination. An employer would fall outside the § 623(i) safe harbor if, for example, the amount of pension credit per year were a function of age rather than the years of credited service, or if pension accruals stopped or were reduced at a firm's normal retirement age. See 29 U.S.C. § 623(i)(1). Stopping pension accruals at age 65 used to be a common feature of defined-benefit plans. Under § 623(i)(1)(A) that is no longer lawful. The Northern Trust plan, however, allows accruals past the normal retirement date, and accruals do not otherwise depend on age. Because the plan complies with § 623(i), it satisfies the ADEA.
Easterbrook, Circuit Judge.