“People make mistakes. Even administrators of ERISA plans.”
Conkright v. Frommert,
— U.S. —, —,
Verizon’s pension plan contains erroneous language that, if enforced literally, would give Verizon pensioners like plaintiff Cynthia Young greater benefits than they expected. Young nonetheless seeks these additional benefits based on ERISA’s strict rules for enforcing plan terms as written. Although Young raises some forceful arguments, we conclude that ERISA’s rules are not so strict as to deny an employer equitable relief from the type of “scrivener’s error” that occurred here. We will accordingly affirm the district court’s judgment granting Verizon equitable reformation of its plan to correct the scrivener’s error.
I. Background
A. Bell Atlantic’s Pension Plans
Bell Atlantic, the predecessor of Verizon, operated the Bell Atlantic Management Pension Plan (“BAMPP”) until 1996. The BAMPP expressed an employee’s retirement benefit as a defined annuity, but employees also had the option of receiving a lump sum if they retired during specified “cashout windows.” For certain employees who retired during the 1994-1995 cashout window, the BAMPP provided a lump sum equal to the “actuarial equivalent present value” of the employee’s pension benefit, but calculated using an enhanced discount rate. Specifically, section 4.19 of the BAMPP required the use of a discount rate of “120% of the applicable ... PBGC [Public Benefit Guarantee Corporation] interest rate in effect” at the time of severance.
In 1996, Bell Atlantic adopted the Bell Atlantic Cash Balance Plan to replace the BAMPP. The new Plan expressed an employee’s benefit as a cash balance that grew steadily with the employee’s age and *813 years of service. Under the Cash Balance Plan, employees still had the option of receiving their retirement benefit as either an annuity or a lump sum.
Key to this transition to the Cash Balance Plan was converting the value of employees’ benefits under the old BAMPP to cash balances under the new Plan. The Plan used “transition factors,” a series of multipliers that increased with employees’ age and years of service, to make the conversion. The Plan language describing this conversion is critical, so we reproduce it in some detail (the emphasis is ours):
16.5 Opening Balance
16.5.1 Pension Conversions as of the Transition Date
Where a present value must be determined under this Section 16.k [sic, should read “Section 16.5”], the present value shall be determined as follows: (a) using the PBGC interest rates which were in effect for September of 1995 ....
16.5.1(a) 1995 Active Participants and 1995 Former Active Participants
... the opening balance of the Participant’s Cash Balance Account on January 1, 1996 shall be the amount described in subsection (1) or (2) below, as applicable:
16.5.1(a)(1) If Eligible for Service Pension
16.5.1(a)(2) Not Eligible for Service Pension
In the case of a Participant who is not eligible for a Service Pension under the 1995 BAMPP Plan as of the Transition Date, the amount described in this paragraph (2) is the product of multiplying (A) the Participant’s applicable Transition Factor described
in Table 1 of this Section, times (B) the lump-sum cashout value of the Accrued Benefit payable at age 65 under the 1995 BAMPP Plan, determined as if the Participant had a Severance From Service Date on December 31, 1995, based on Compensation paid through December 31, 1995, multiplied by the applicable transition factor described in Table 1 of this Section ....
B. Young’s Administrative Claim
Cynthia Young worked for Bell Atlantic from 1965 to 1997. When the Cash Balance Plan took effect in 1996, Young was not eligible for a service pension under the BAMPP — that is, her age and service level did not qualify her for full retirement benefits — so her opening cash balance was calculated using § 16.5.1(a)(2), for a resulting balance of $240,127. By the time Young retired in 1997, her cash balance had grown to the point that she received a lump-sum benefit of $286,095.
Several years later, in 2004, Young filed a claim with the Claims Review Unit of Verizon (which by then had taken over Plan administration as Bell Atlantic’s successor). Young claimed that Bell Atlantic made two errors in calculating her opening cash balance, and hence her ultimate pension benefit, under the Cash Balance Plan. First, Young read the language of § 16.5.1(a)(2) to require that the “applicable transition factor” be multiplied twice to convert her lump-sum cashout under the BAMPP to her opening cash balance under the new Plan. Bell Atlantic, however, multiplied the transition factor only once when making the conversion. Second, Young claimed that Bell Atlantic improperly applied the 120% PBGC discount rate used in the 1995 BAMPP to determine the “lump-sum cashout value” under § 16.5.1(a)(2). Young contended that Bell *814 Atlantic should have used a discount rate of simply 100% of the PBGC rate.
Verizon’s Claims Review Unit denied Young’s claims, and on appeal, Verizon’s Cláims Review Committee affirmed. The Committee concluded that the intended meaning of § 16.5.1(a)(2) was to use only a single transition factor to calculate opening cash balances; the section’s second reference to the “applicable transition factor” was a drafting mistake. As for Young’s discount rate claim, the Committee concluded that § 16.5.1(a)(2) incorporated the 120% PBGC rate used in the 1995 BAMPP by referring to “the lump-sum cashout value ... under the 1995 BAMPP Plan.”
C. Young’s Federal Court Class Action
In 2005, Young brought a federal court action under ERISA § 502(a), 29 U.S.C. § 1132(a), against Verizon and its Cash Balance Plan (collectively “Verizon”). Young asserted the same claims she raised in Verizon’s administrative process, arguing that Verizon improperly applied only a single transition factor and the 120% PBGC discount rate to calculate her opening cash balance. The parties agreed to treat the case as a class action, and the district court certified a class of some 14,-000 Bell Atlantic/Verizon pensioners similarly situated to Young.
Young’s class action presented the district court, acting through Magistrate Judge Denlow, with a challenge. The court was confronted with a convoluted ERISA plan that seemed to contain a costly drafting error, but an uncertain state of law on the scope of the court’s review of such an error. So the court decided to bifurcate the trial into two phases and apply alternative standards of review. In the first phase, the court assumed that it was limited to examining the administrative record and reviewing the Verizon Review Committee’s denial of benefits under a deferential standard. (The Cash Balance Plan granted Verizon, as plan administrator, broad discretion to interpret the Plan, so judicial review was constrained to an “arbitrary and capricious” standard.
Black v. Long Term Disability Ins.,
Taking the district court’s cue, Verizon counterclaimed for equitable reformation of the Plan to remove the second transition factor in § 16.5.1(a)(2) as a “scrivener’s error.” The court took up Verizon’s counterclaim in the second phase of the trial, in which the court conducted a de novo review of the Plan and allowed the parties to introduce extrinsic evidence on the intended meaning of § 16.5.1(a)(2). And that evidence overwhelmingly showed that the inclusion of the second transition factor was indeed a scrivener’s error.
The drafting history of the 1996 Plan revealed how the second, erroneous transition factor came to be. Six drafts of the Plan were prepared prior to the final version. The first three drafts were prepared by Mercer Human Resources Consulting, an outside firm hired by Bell Atlantic, and contained no mention of a second transition factor. It was not until one of Bell Atlantic’s in-house attorneys, Barry Peters, took over drafting responsibility that the second transition factor appeared. In working on the fourth draft, Peters restructured the conversion formula under *815 § 16.5.1(a)(2) into a more readable “A times B” format, but in doing so, neglected to delete a trailing clause from the previous draft that referred to “the applicable Transition Factor.” Testifying in the district court, Peters admitted that he made this mistake in failing to delete the trailing clause in § 16.5.1(a)(2), thereby duplicating the transition factor. Peters’s mistake survived unnoticed in the fifth, sixth, and final drafts of the Plan.
In addition to the drafting history, the correspondence between Bell Atlantic and plan participants showed an expectation that only a single transition factor would be used to calculate opening cash balances. In October 1995, Bell Atlantic sent participants a brochure entitled, “Introducing Your Cash Balance Plan,” which clearly depicted opening cash balances as the product of an employee’s lump-sum value under the 1995 BAMPP and a single transition factor. In November 1995, Bell Atlantic sent participants personalized statements of their estimated opening account balances, which also illustrated the use of a single transition factor. Following the implementation of the Plan, Bell Atlantic sent participants personalized statements of their actual opening balances, and thereafter quarterly cash balance statements, which, again, reflected the use of only one transition factor. Notably, though, these Plan-related communications contained “plan trumps” provisions cautioning that, in the event of discrepancies between those communications and the Plan, the Plan would govern.
Also convincing was the course of dealing between Bell Atlantic/Verizon and plan participants. Bell Atlantic consistently calculated opening cash balances using a single transition factor and paid benefits accordingly. Taking Young’s case as an example, her transition factor was 2.659. The estimated opening balance statement that Young received illustrated the multiplication of this 2.659 transition factor by her BAMPP lump-sum cashout value of $90,027, for an estimated opening balance of $90,027 x 2.659 = $289,381. The actual opening balance statement that Young received in 1996 applied the same, single-transition-factor formula to slightly different numbers: $90,307 x 2.659 = $240,127. Prior to Young’s lawsuit, no employee complained that opening balances should have been increased by an additional transition factor. For her part, Young admitted that she never relied on the transition factor language in § 16.5.1(a)(2) prior to this litigation.
Based on this evidence of the intended meaning of the Plan, the district court found that the second transition factor in § 16.5.1(a)(2) was a scrivener’s error and granted Verizon’s counterclaim for equitable reformation. The court also resolved a host of other arguments raised by the parties, many of which we discuss below. But suffice it to say, the district court’s treatment of the issues presented by this case was exhaustive. Over the course of a four-year, multi-phase litigation, the court built a complete record, fully explored alternative bases of decision, and sharply honed the issues for appellate review. These commendable efforts by the district court, as well as the fine advocacy by both sides, have greatly assisted this court in deciding this complex ERISA case.
II. Analysis
A. Statute of Limitations
Before reaching the merits, we must address each side’s argument that the other’s claims are barred by the statute of limitations. ERISA does not provide a limitations period for actions brought under § 502, 29 U.S.C. § 1132, so we borrow the most analogous statute of limitations from state law.
Berger v. AXA
*816
Network LLC,
The parties agree that Pennsylvania’s four-year statute of limitations for breach of contract actions, 42 Pa. Cons.Stat. § 5525, should apply to this ERISA case. Pennsylvania has the most significant connection to this dispute, since Bell Atlantic was headquartered and drafted the Cash Balance Plan there. Also, more class members currently live in Pennsylvania than any other state, and while a few class members live in the forum state of Illinois, Young has never lived or worked there. We further note that the Plan contains a choice of law provision stating that Pennsylvania law will fill any gaps left by federal ERISA law.
See Berger,
The real point of contention is the accrual date of the parties’ claims, that is, when Pennsylvania’s four-year limitations period started to run. Although federal courts borrow state limitations periods for certain ERISA claims, the accrual of those claims is governed by federal common law.
Daill,
Beginning with Young’s ERISA claim, we have held that a claim to recover benefits under § 502(a) accrues “upon a clear and unequivocal repudiation of rights under the pension plan which has been made known to the beneficiary.” Id. at 66. In this case, Young did not receive a clear repudiation of her claim for additional benefits until 2005, when Verizon’s Review Committee resolved her administrative appeal. (Actually, the Committee denied Young’s claim with respect to the discount rate issue in 2005 but took until 2007 to deny her claim with respect to the transition factor issue. Since it is obvious that Young’s entire federal court action, filed in 2005, would be timely using a 2005 accrual date, this distinction is immaterial.) Prior to denying Young’s administrative claim, Verizon did not inform Young that it rejected her interpretation of the Plan calling for two transition factors and a 100% PBGC discount rate. Cf. id. at 66 (claim accrued upon correspondence from plan disagreeing with participant’s understanding of benefits).
Verizon argues that Young’s claim accrued in February 1998, when she received her lump-sum benefit computed under Verizon’s interpretation of the Cash Balance Plan. At that time, however, the parties’ dispute over the correct interpretation of the Plan had not developed. And nothing suggests that the $286,095 payment that Young received should have been a red flag that she was underpaid.
Cf. Redmon v. Sudr-Chemie Inc. Ret. Plan for Union Employees,
Moving to Verizon’s counterclaim, Seventh Circuit precedent provides
*817
less guidance on the accrual of a claim for equitable reformation under ERISA § 502(a) — understandably so, since the cognizance of such a claim is an issue of first impression for this court. The general federal common law rule is that an ERISA claim accrues when the plaintiff knows or should know of conduct that interferes with the plaintiffs ERISA rights.
See Berger,
The district court found, and Verizon does not dispute, that Verizon’s predecessor Bell Atlantic learned of the scrivener’s érror in 1997. Indeed, Bell Atlantic removed the second, erroneous transition factor from the 1998 plan that it adopted to replace the 1997 version of the Cash Balance Plan. Still, we conclude that this 1997 discovery did not give Verizon notice of the need to reform the scrivener’s error, given a course of dealing consistent with Verizon’s interpretation of the Plan.
Verizon always treated the Plan’s second transition factor as a drafting mistake, and through correspondence with plan participants, it communicated that only a single transition factor would be used to calculate opening cash balances. Verizon consistently paid benefits using this formula, and prior to Young’s administrative claim, no employee communicated a contrary understanding that Plan benefits should be calculated using two transition factors.
Cf. Tolle v. Carroll Touch, Inc.,
None of the parties’ claims accrued before 2005 when Young brought her federal court ERISA action, so these claims are timely under the applicable Pennsylvania four-year limitations period. We may proceed to the merits of Verizon’s claim for equitable reformation and Young’s claim for additional benefits under ERISA § 502(a).
B. Equitable Reformation Due to Scrivener’s Error
ERISA is a comprehensive statute designed to uniformly regulate employee benefit plans.
Aetna Health Inc. v. Davila,
While ERISA’s strict requirements “ensure[ ] fair and prompt enforcement of rights under a plan,” Congress was careful not to make those requirements so onerous “that administrative costs, or litigation expenses, unduly discourage employers from offering plans in the first place.”
Conkright v. Frommert,
— U.S. —, —,
Another ERISA provision that promotes equitable plan enforcement — and the statute important here — is § 502(a)(3), which allows a plan participant, beneficiary, or fiduciary to bring a civil action for “appropriate equitable relief.” 29 U.S.C. § 1132(a)(3)(B). The Supreme Court has explained that the statute authorizes “those categories of relief that were
typically
available in equity” during the days when common law courts were divided as courts of law or of equity.
Mertens v. Hewitt Assocs.,
We have never considered whether § 502(a)(3) authorizes equitable reformation of an ERISA plan due to a scrivener’s error, but our case law addressing the related problem of ambiguous plan language suggests that such relief may be appropriate.
In
Mathews v. Sears Pension Plan,
We reached a different result in
Grun v. Pneumo Abex Corp.,
Other circuits have directly addressed claims for equitable reformation of an ERISA plan. Using reasoning similar to that in Mathews and Grun, these courts have either concluded that ERISA authorizes such relief or does not foreclose the possibility.
Verizon’s strongest case is
Int’l Union v. Murata Erie N. Am., Inc.,
The Ninth Circuit distinguished
Murata
in
Cinelli v. Sec. Pac. Corp.,
From this authority, we conclude that ERISA § 502(a)(3) authorizes equitable reformation of a plan that is shown, by clear and convincing evidence, to contain a scrivener’s error that does not reflect participants’ reasonable expectations of benefits. Though complex in design, ERISA maintains the basic goal of “protecting employees’ justified expectations of receiving the benefits their employers promise them.”
Cent. Laborers’ Pension Fund v. Heinz,
We acknowledge, like the Third Circuit in
Murata,
Even so, since we interpret § 502(a)(3) to authorize the equitable reformation claim asserted here, we cannot simply reject such a claim based on the added litigation burden that it might represent. Moreover, we see little difference between the intent-based inquiry that took place in this reformation case and what must occur in the related case of an ambiguous ERISA plan. In each case, the court must look beyond the plan document to extrinsic evidence to determine the parties’ understanding of the plan.
See Mathews,
Also, other limitations on the equitable reformation claim that we recognize under § 502(a)(3) will mitigate its impact on the plan documents rule. Only those who can marshal “clear and convincing” evidence that plan language is contrary to the parties’ expectations will have a viable claim.
Murata,
In this case, though, we agree with the district court that Verizon presented enough objective, convincing evidence to show that the second reference to the transition factor in § 16.5.1(a)(2) of the Cash Balance Plan was a scrivener’s error inconsistent with participants’ expected benefits.
The drafting history left little doubt that the second transition factor in § 16.5.1(a)(2) was a mistake. It first appeared in the fourth draft of the Plan, the first draft prepared by Bell Atlantic attorney Barry Peters. This draft reformatted the multiplication formula in § 16.5.1(a)(2), but in doing so, failed to omit the prior draft’s trailing clause that referred to the transition factor, thereby duplicating the transition factor. We need not rely on Peters’s arguably self-serving testimony to conclude that this botched reformatting led *821 to the second transition factor; so much is clear by comparing the fourth draft with the prior version. And given the absence of any evidence contemporaneous to the fourth draft suggesting that Bell Atlantic was reworking the Plan to increase benefits, it is evident that duplicating the transition factor was a drafting mistake.
The communications and course of dealing between Bell Atlantic/Verizon and plan participants further illustrate that the parties intended a single-transition-factor formula. Young and other participants received a Plan brochure that described their opening cash balances as the product of their lump-sum values under the 1995 BAMPP and a single transition factor. Although the brochure did not explicitly state that a “single” transition factor would be used, the formula depicted in the brochure makes clear that only one multiplier would apply. That was confirmed in the personalized statements sent to participants of their estimated and actual opening cash balances, which reported values based on the use of a single transition factor. By way of illustration, Young received an estimated opening balance statement that reported her transition factor of 2.659 and her BAMPP lump-sum cashout value of $90,027, for an estimated opening balance of $239,381. Her actual opening balance reported in a later statement, $240,127, was calculated similarly. If a second 2.659 transition factor were applied to these figures, Young’s estimated and actual opening balances would have been $636,514 and $638,498, respectively. Bell Atlantic/Verizon never squared transition factors in this manner but instead calculated benefits using only a single transition factor, consistent with the Plan communications. Prior to Young’s claim, no employee complained that cash balances should have been increased by an additional transition factor.
Granted, many of the Plan communications, including the Plan brochure and opening balance statements, are less compelling because they contain what Young describes as “plan trumps” provisions, which stated that the communications were subordinate to any contrary language in the Plan. As Young points out, were the situation reversed and the employee-favorable language contained in a Plan communication rather than the Plan itself, Verizon no doubt would contend that these plan trumps provisions barred Young from relying on the communication.
See Kolentus v. Avco Corp.,
Based on this evidence of the intended meaning of the Plan, the district court correctly found that the second transition factor in § 16.5.1(a)(2) was a scrivener’s error inconsistent with plan participants’ expected benefits. Under these circumstances, equitable reformation of the Plan to remove the error is appropriate.
We close our discussion of Verizon’s reformation claim by considering additional *822 defenses to equitable relief. Because Verizon’s claim is one for “appropriate equitable relief’ under ERISA § 502(a)(3)(B), 29 U.S.C. § 1132(a)(3)(B), it is subject to the traditional equitable defenses at common law, provided that they are not inconsistent with ERISA.
Young raises the defense of “good faith” and “fair dealing,” under which a contracting party may be precluded from reforming a mistake caused by the party’s own “gross” negligence. Restatement (Second) of Contracts § 157 & cmt. a (1981). As the district court put it, Bell Atlantic/Verizon’s failure to prevent the drafting mistake in § 16.5.1(a)(2) was “profound” negligence. Bell Atlantic charged a single in-house attorney, Barry Peters, with revising a critical provision of a multibillion-dollar pension plan, apparently without critical review by another ERISA expert. It is baffling that a major corporation would not invest greater resources to ensure accuracy in the drafting of such an important document. Still, we cannot agree with Young that this institutional failure showed a lack of good faith. Verizon never misrepresented its intended meaning of the Cash Balance Plan, and indeed, based on the extrinsic evidence examined above, it made great efforts to accurately communicate how participants’ benefits would be calculated. Cfi id. cmt. a, illustration 2 (misrepresentation that party verified bid for accuracy was failure to act in good faith).
For similar reasons, we do not accept Young’s “unclean hands” defense, under which “equitable relief will be refused if it would give the plaintiff a wrongful gain.”
Scheiber v. Dolby Labs., Inc.,
Finally, Young raises the equitable defense of laches, or unreasonable delay, by Verizon in seeking equitable reformation. Laches means “culpable delay in suing” and may apply if the plaintiff commits an unreasonable, prejudicial delay in bringing the suit.
Teamsters & Employers Welfare Trust of Ill. v. Gorman Bros. Ready Mix,
In sum, no equitable defenses bar Verizon’s equitable reformation claim under ERISA § 502(a)(3), and the district court properly granted that claim to remove the scrivener’s error from the Cash Balance Plan.
C. Discount Rate for Opening Cash Balances
In addition to her argument regarding the second transition factor in § 16.5.1(a)(2), Young claimed that Verizon improperly applied the enhanced, 120% PBGC discount rate used in the 1995 BAMPP to calculate her opening balance under the Cash Balance Plan. Verizon’s Review Committee denied Young’s discount rate claim, and because the Plan grants the administrator broad discretion to interpret Plan provisions, we review the Committee’s decision for an abuse of discretion.
See Black v. Long Term Disability Ins.,
The interpretation of ERISA plans is governed by federal common law, which draws on general principles of contract interpretation to the extent they are consistent with ERISA.
Mathews,
The use of a discount rate to calculate opening balances under the Cash Balance Plan occurs by operation of § 16.5.1(a)(2). That section defines opening cash balances as the product of two variables (assuming, of course, one ignores the second “transition factor” that we have disregarded as a scrivener’s error): “(A) the Participant’s applicable Transition Factor described in Table 1 of this Section, times (B) the lump-sum cashout value of the Accrued Benefit payable at age 65 under the 1995 BAMPP Plan----” Under § 4.19 of the BAMPP, which was attached to the Cash Balance Plan as an appendix, lump-sum payments for employees who retired during the 1994-1995 cashout window were calculated using a discount rate of 120% of “the applicable PBGC interest rate.”
Reading the language of § 16.5.1(a)(2) in the context of the entire Cash Balance Plan — including the attached 1995 BAMPP — the best interpretation is one that applies the 120% PBGC discount rate used in the 1995 BAMPP to calculate opening cash balances. The plain meaning of the “(B)” variable in § 16.5.1(a)(2) — “the lump-sum cashout value ... payable ... under the 1995 BAMPP Plan” — is the lump-sum value as calculated under the 1995 BAMPP. Since the BAMPP used a 120% PBGC discount rate, that same methodology carries over to calculating opening balances under the Cash Balance Plan.
Young points to the umbrella section 16.5.1, which provides that any “present value” that “must be determined under this Section 16.[5] shall be determined ... using the PBGC interest rates which were in effect for September of 1995.” Young would apply this present value definition, which uses a discount rate of simply 100% *824 of the PBGC rate, to determine the “lump-sum cashout value” in § 16.5.1(a)(2). Young’s interpretation ignores the explicit reference in § 16.5.1(a)(2) to the cashout value “under the 1995 BAMPP Plan.” Because § 16.5.1(a)(2) specifically uses the 1995 BAMPP formula for discounting lump-sum values, the more general present value formula in § 16.5.1 does not apply to that section.
We also disagree with Young that incorporating the 1995 BAMPP, 120% PBGC formula into § 16.5.1(a)(2) in this manner renders the 100% PBGC formula in § 16.5.1 superfluous. The latter formula applies broadly to calculate present values under “this Section 16.[5].” Notably, unlike § 16.5.1(a), provisions in § 16.5.2(a) use the “present value” term defined in § 16.5.1 to determine opening cash balances for employees covered by those sections. So it harmonizes all the language in § 16.5 to give effect to the 120% PBGC rate incorporated into § 16.5.1(a)(2) for that specific provision, while giving effect to the general 100% PBGC rate for other provisions in § 16.5.
The most reasonable reading of § 16.5.1(a)(2) is one that applies the 120% PBGC discount rate to calculate opening cash balances. At the very least, Verizon’s Review Committee did not abuse its discretion in adopting this interpretation.
III. Conclusion
ERISA’s rules for written plans are strictly enforced, but they are not so strict as to prevent equitable reformation of a plan that is shown, by clear and convincing evidence, to contain a scrivener’s error that is inconsistent with participants’ expected benefits.
Affirmed.