EASTERBROOK, Circuit Judge.
Carle Clinic Association claims to be one of the nation’s largest medical-practice groups. See <http://carle.com>. Together with the Carle Foundation, it operates several clinics plus the hospital affiliated with the University of Illinois in Urbana-Cham-paign. Carle promised employees (including physicians) a pension that could be as large as 50% of average earnings. Doctors Richard Helfrich and Daniel Nelson learned to their dismay, however, that the total annual pension under Carle’s defined-benefit plan cannot exceed $160,000. Paying more would cost the plan its “tax-qualified” status, which allows Carle to deduct pension expenses and employees to defer until retirement all taxes on the value of this fringe benefit. See 26 U.S.C. §§ 401(a), 415. The cap changes with the cost of living; $160,000 is close enough for current purposes. It is less than the pension that Helfrich and Nelson believed had been promised, and when the plan refused to pay more they filed this suit under § 502(a) of the Employee Retirement Income Security Act, 29 U.S.C. § 1182(a). Observing that the plan’s terms explicitly restrict payouts to a level consistent with retaining tax advantages, the district court granted summary judgment for Carle. Later it ordered plaintiffs to reimburse Carle for the legal expense it had incurred in defending the suit.
Plaintiffs’ principal contention on appeal is that summaries Carle handed out to its employees override the terms of the plan. Every pension plan must publish a summary plan description, and conflicts between this document and the plan itself are resolved in favor of the summary plan description (unless it alerts the reader to look for additional terms in the full plan). See, e.g.,
Mathews v. Sears Pension Plan,
144 F.3d 461, 466 (7th Cir.1998). Plaintiffs want us to extend this rule to other descriptive material. They did not make such an argument in the district court (relying there on estoppel, which we discuss below); and plaintiffs’ argument, like this whole suit, reflects confusion between the employer and the plan, which under ERISA is a separate trust. Plans can control the contents of summary plan descriptions, which they prepare (with, one hopes, a degree of care appropriate to the reliance they engender and the legal consequences they carry). Plans cannot control what miscellaneous recruiters and personnel managers may say, nor does even a large employer’s human-resources staff draft descriptions with the precision that
the plan itself will do — for the employer knows that its staff can refer to the summary plan description. Because ERISA requires plans to prepare summary plan descriptions, and because their content is within the plan’s control, it makes sense to give these documents legal effect when relied on. Employer-prepared summaries, by contrast, have no footing in ERISA and could not be enforced against the plan without disregarding the boundary between two distinct entities: the plan and the employer. Blurring that line is what plaintiffs hope to achieve, and they have taken their own view to heart by suing only the employer. Claims based on the plan (or the summary plan description) must be enforced against the plan, which is not a party. Yet Carle makes nothing of this omission, and § 1132(a) does not make federal jurisdiction depend on the plan’s inclusion as a defendant. It is enough if the plaintiff is a plan participant and makes a claim for benefits. Still, plaintiffs’ decision not to sue the plan reveals their working assumption that the plan and the employer are indistinguishable. ERISA does not operate that way.
One summary plan description is in the record. This document, dated October 1996, is 21 single-spaced pages. It fully describes the cap required by statute as a condition of tax deferral. The three documents that plaintiffs call “summaries” do not look remotely like summary plan descriptions and must have been prepared by Carle rather than by the plan. “Summary I” is a single-page handout; “Summary II” is a brochure that covers all of Carle’s fringe benefits in the equivalent of two letter-sized pages; and “Summary III,” a section of Carle’s employee handbook, though longer (at 15 pages), covers many topics in addition to the pension formula. None of the three summaries alerts employees to the $160,000 cap on tax-qualified plans, most likely because the latest of the three dates from 1981. Ever since ERISA’s enactment in 1976, the Carle plan has provided that benefits will not exceed what tax-qualified plans can provide. Congress amended the Internal Revenue Code in 1986 to create the limit (originally $90,000 but lifted to $160,000 under a formula enacted in 1991), and pre-1986 brochures were unlikely to describe future legislation.
Now if a summary plan description from 1981 had omitted mention of a rule (the plan’s provision that no benefit costing the plan its tax-qualified status would be paid) that had a potential to curtail retirement income, plaintiffs might have a better position. But the record does not contain any summary plan description preceding 1996. Plaintiffs complained at oral argument that the district court had blocked their discovery into this subject, but counsel admitted that he had not sought documents of this kind. Instead plaintiffs noticed several depositions of plan officials, and the district judge stopped this process as burdensome because nothing that any plan official could say about the handling of Helfrich’s or Nelson’s demands for larger pensions would affect the proper disposition of the suit, which depends on the validity of the plan itself. Plaintiffs’ counsel then earned the district judge’s enmity by noticing
another
set of depositions, showing that he would do what he could to evade the judge’s ruling. The judge did not abuse his discretion in curtailing plaintiffs’ efforts to take depositions; and plaintiffs did not seek production of the kind of documents that could assist them (or serve requests for admissions about the nature and contents of earlier summary plan descriptions). Maybe counsel knew that there is no helpful document waiting to-be found; Helfrich was for a time chairman of Carle’s board and a member of its pension committee, so his files may well contain whatever relevant documents the plan distributed.
So the rule that summary plan descriptions, if relied on, trump the plan itself does not assist plaintiffs. And, as we have concluded that this rule should not be extended to documents prepared by the employer, it follows that the same contention under the banner of “estoppel” fares no better. The doctrine that the summary plan description prevails over the plan is a
form
of estoppel; to establish the limits of this doctrine is to establish the limits of estoppel. No matter what label applies, documents prepared by an employer do not supersede those documents that establish the terms of a pension plan. Whether, and to what extent, estoppel is available with respect to welfare benefit plans under ERISA is an issue on which the judiciary has not come to rest, compare
Frahm v. Equitable Life Assurance Society,
Free access — add to your briefcase to read the full text and ask questions with AI
EASTERBROOK, Circuit Judge.
Carle Clinic Association claims to be one of the nation’s largest medical-practice groups. See <http://carle.com>. Together with the Carle Foundation, it operates several clinics plus the hospital affiliated with the University of Illinois in Urbana-Cham-paign. Carle promised employees (including physicians) a pension that could be as large as 50% of average earnings. Doctors Richard Helfrich and Daniel Nelson learned to their dismay, however, that the total annual pension under Carle’s defined-benefit plan cannot exceed $160,000. Paying more would cost the plan its “tax-qualified” status, which allows Carle to deduct pension expenses and employees to defer until retirement all taxes on the value of this fringe benefit. See 26 U.S.C. §§ 401(a), 415. The cap changes with the cost of living; $160,000 is close enough for current purposes. It is less than the pension that Helfrich and Nelson believed had been promised, and when the plan refused to pay more they filed this suit under § 502(a) of the Employee Retirement Income Security Act, 29 U.S.C. § 1182(a). Observing that the plan’s terms explicitly restrict payouts to a level consistent with retaining tax advantages, the district court granted summary judgment for Carle. Later it ordered plaintiffs to reimburse Carle for the legal expense it had incurred in defending the suit.
Plaintiffs’ principal contention on appeal is that summaries Carle handed out to its employees override the terms of the plan. Every pension plan must publish a summary plan description, and conflicts between this document and the plan itself are resolved in favor of the summary plan description (unless it alerts the reader to look for additional terms in the full plan). See, e.g.,
Mathews v. Sears Pension Plan,
144 F.3d 461, 466 (7th Cir.1998). Plaintiffs want us to extend this rule to other descriptive material. They did not make such an argument in the district court (relying there on estoppel, which we discuss below); and plaintiffs’ argument, like this whole suit, reflects confusion between the employer and the plan, which under ERISA is a separate trust. Plans can control the contents of summary plan descriptions, which they prepare (with, one hopes, a degree of care appropriate to the reliance they engender and the legal consequences they carry). Plans cannot control what miscellaneous recruiters and personnel managers may say, nor does even a large employer’s human-resources staff draft descriptions with the precision that
the plan itself will do — for the employer knows that its staff can refer to the summary plan description. Because ERISA requires plans to prepare summary plan descriptions, and because their content is within the plan’s control, it makes sense to give these documents legal effect when relied on. Employer-prepared summaries, by contrast, have no footing in ERISA and could not be enforced against the plan without disregarding the boundary between two distinct entities: the plan and the employer. Blurring that line is what plaintiffs hope to achieve, and they have taken their own view to heart by suing only the employer. Claims based on the plan (or the summary plan description) must be enforced against the plan, which is not a party. Yet Carle makes nothing of this omission, and § 1132(a) does not make federal jurisdiction depend on the plan’s inclusion as a defendant. It is enough if the plaintiff is a plan participant and makes a claim for benefits. Still, plaintiffs’ decision not to sue the plan reveals their working assumption that the plan and the employer are indistinguishable. ERISA does not operate that way.
One summary plan description is in the record. This document, dated October 1996, is 21 single-spaced pages. It fully describes the cap required by statute as a condition of tax deferral. The three documents that plaintiffs call “summaries” do not look remotely like summary plan descriptions and must have been prepared by Carle rather than by the plan. “Summary I” is a single-page handout; “Summary II” is a brochure that covers all of Carle’s fringe benefits in the equivalent of two letter-sized pages; and “Summary III,” a section of Carle’s employee handbook, though longer (at 15 pages), covers many topics in addition to the pension formula. None of the three summaries alerts employees to the $160,000 cap on tax-qualified plans, most likely because the latest of the three dates from 1981. Ever since ERISA’s enactment in 1976, the Carle plan has provided that benefits will not exceed what tax-qualified plans can provide. Congress amended the Internal Revenue Code in 1986 to create the limit (originally $90,000 but lifted to $160,000 under a formula enacted in 1991), and pre-1986 brochures were unlikely to describe future legislation.
Now if a summary plan description from 1981 had omitted mention of a rule (the plan’s provision that no benefit costing the plan its tax-qualified status would be paid) that had a potential to curtail retirement income, plaintiffs might have a better position. But the record does not contain any summary plan description preceding 1996. Plaintiffs complained at oral argument that the district court had blocked their discovery into this subject, but counsel admitted that he had not sought documents of this kind. Instead plaintiffs noticed several depositions of plan officials, and the district judge stopped this process as burdensome because nothing that any plan official could say about the handling of Helfrich’s or Nelson’s demands for larger pensions would affect the proper disposition of the suit, which depends on the validity of the plan itself. Plaintiffs’ counsel then earned the district judge’s enmity by noticing
another
set of depositions, showing that he would do what he could to evade the judge’s ruling. The judge did not abuse his discretion in curtailing plaintiffs’ efforts to take depositions; and plaintiffs did not seek production of the kind of documents that could assist them (or serve requests for admissions about the nature and contents of earlier summary plan descriptions). Maybe counsel knew that there is no helpful document waiting to-be found; Helfrich was for a time chairman of Carle’s board and a member of its pension committee, so his files may well contain whatever relevant documents the plan distributed.
So the rule that summary plan descriptions, if relied on, trump the plan itself does not assist plaintiffs. And, as we have concluded that this rule should not be extended to documents prepared by the employer, it follows that the same contention under the banner of “estoppel” fares no better. The doctrine that the summary plan description prevails over the plan is a
form
of estoppel; to establish the limits of this doctrine is to establish the limits of estoppel. No matter what label applies, documents prepared by an employer do not supersede those documents that establish the terms of a pension plan. Whether, and to what extent, estoppel is available with respect to welfare benefit plans under ERISA is an issue on which the judiciary has not come to rest, compare
Frahm v. Equitable Life Assurance Society,
137 F.3d 955, 961 (7th Cir.1998), with
Shields v. Teamsters Pension Plan,
188 F.3d 895 (7th Cir.1999), but with respect to pension plans (which by statute must be very formal) the only variance that any court has allowed depends on the plan’s own writings, such as summary plan descriptions. See, e.g.,
Downs v. World Color Press,
214 F.3d 802, 805-06 (7th Cir.2000);
Krawczyk v. Harnischfeger Corp.,
41 F.3d 276, 280 (7th Cir.1994);
Shields,
188 F.3d at 900 n. 3. And we cannot imagine a good claim of estoppel that would require a plan to forfeit the tax benefits on which its actuarial soundness relies. Plaintiffs’ position, if accepted, would cost Carle, their co-workers, and themselves considerable sums in taxes. Helfrich estimated before the case began that, under his position, Carle would need to contribute about $20 million extra to the plan; Carle also would lose tax deductability of past and future contributions, costing it (and thus its physician-owners) millions more. It is unclear whether plaintiffs would be winners; that depends on whether the increase in pension benefits would be enough to cover their tax obligations. Most other current and former Carle employees — those not affected by the cap because their annual earnings were less than $320,000 — would be certain to lose because they would suffer tax consequences without offsetting gains. Overriding the plan’s terms to achieve that dubious end is not a proper function of an equitable doctrine.
One might imagine an argument that, although the existing plan must respect the statutory rules for maintaining tax benefits, Carle should be obliged to create another plan that is not tax-qualified and brings pensions up to the promised level. Cf.
Bartholet v. Reishauer A.G. (Zurich),
953 F.2d 1073 (7th Cir.1992). Yet although plaintiffs hint at such an argument they do not make it expressly — perhaps because it has no footing in the three “summaries” on which they rely. These documents describe two plans: the noncontributory defined-benefit plan at issue here, and a defined-contribution plan in which employees could elect to participate. An argument for creating still a third plan, limited to Carle’s best-paid physicians, has no support in the documentation. Moreover, because plaintiffs have not made this argument explicit, Carle was not put on notice of the need to discuss its effects. Would such a plan limited to well-paid employees jeopardize the tax benefits for the rest? Would it have other adverse consequences? These are subjects that we need not explore given plaintiffs’ sketchy argument. And if the contention were cast as a demand for cash from Carle without a plan, it is hard to see why employers should insure their employees against legislative change. But for the amendments to the tax code, even the unguarded language of the “summaries” would have held. Many people are disappointed, and their fortunes affected, by statutory change; this does not imply that some private party must step in to undo the effects of the legislation. The “sum
maries” described extant pension plans; they did not include promises to protect employees from the transition effects of statutory change.
As for attorneys’ fees: the district court did not abuse its discretion in concluding that Carle, the prevailing party, is entitled to recompense under 29 U.S.C. § 1132(g)(1), which says that “the court in its discretion may allow a reasonable attorney’s fee and costs of action to either party.” Nor did the district judge err in concluding that fees in the range of $160,000 are reasonable, given the stakes of the case: a loss would have cost Carle $20 million or more and had awful tax consequences for many current and former employees. Carle retained ERISA specialists to protect its interests; this was a prudent step. (Plaintiffs stress that they hired less expensive lawyers. The difference shows. You get what you pay for.)
Unfortunately, however, the district judge did not exercise the billing judgment that is essential in all fee-shifting cases. Carle sought compensation for all legal time and expenses racked up during the suit’s pendency. That included, for example, about $886 billed to Carle for the expense of preparing a press release describing the suit and $95 for the time one lawyer devoted to preparing an application for admission to the bar of the United States District Court for the Central District of Illinois. Press releases are not part of a legal defense (even if the client elects to have the drafting done by lawyers), and admission to the local bar would have a benefit outlasting this case. Time devoted to exploration of insurance issues also should not be shifted to the plaintiffs; it is no concern of this litigation how things happen behind the scenes in the event the plaintiffs prevail. The district court should eliminate these items on remand and review the remaining charges to see whether fee-shifting is appropriate with respect to each kind of service rendered. Carle is entitled to recover the cost of legal defense (including the cost of preparing a budget, a normal incident in any substantial suit), but not legal expenses that do not produce a defense against the plaintiffs’ claims. On balance, however, the tab will go up — for although a few items must be subtracted, the cost of defending Carle on this appeal must be added. When fees have been awarded in the district court on the authority of a fee-shifting statute, the costs of appellate work are automatically shifted.
Commissioner of INS v. Jean,
496 U.S. 154, 110 S.Ct. 2316, 110 L.Ed.2d 134 (1990). The district court should determine, and award, the legal fees that Carle incurred in defending its judgment.
The decision on the merits is affirmed. The award of attorneys’ fees is vacated, and that subject is remanded for further proceedings consistent with this opinion.