Berkshire Hathaway, Inc., Plaintiff, Appellant, v. Textile Workers Pension Fund, Defendant, Appellee, 874 F.2d 53 (1st Cir. 1989)Annotate this Case
Heard Dec. 7, 1988. Decided May 9, 1989
William L. Patton with whom James B. Levy and Ropes & Gray, Boston, Mass., were on brief, for plaintiff, appellant.
Ronald E. Richman with whom Mark E. Brossman, Mary Ellen Koscs-Fleming, Chadbourne & Parke, New York City, N.Y., William F. Joy, Jr., Laurence J. Donoghue and Morgan, Brown & Joy, Boston, Mass., were on brief, for defendant, appellee.
Before COFFIN, BOWNES and SELYA, Circuit Judges.
COFFIN, Circuit Judge.
This case presents the issue of whether withdrawal liability may be assessed under the Multiemployer Pension Plan Amendments Act of 1980, 29 U.S.C. §§ 1381 et seq. (MPPAA or "the Act"), against an employer who withdraws from a fully funded plan. The district court held that such employers may incur withdrawal liability. We believe the district court failed to give adequate deference to the implementing agency's interpretation, which we view as reasonable, supported by the statutory language, and consistent with the Act's purpose as revealed by its legislative history. Accordingly, we vacate.
This is an appeal from summary judgment for the Textile Workers Pension Fund (the Fund) against Berkshire Hathaway, Inc. The Fund sponsors a multiemployer defined benefit pension plan under the Employee Retirement Income Security Act of 1974 (ERISA), 29 U.S.C. §§ 1001, et seq. See 29 U.S.C. §§ 1002(35), 1002(37), 1301(a) (3). Berkshire began contributing to the plan in 1970 pursuant to collective bargaining agreements with the Textile Workers Union and its successor, the Amalgamated Clothing and Textile Workers Union. Berkshire withdrew from the plan when it closed down textile operations at plants that employed participating workers. Berkshire challenged in the district court the Fund trustees' allocation of liability against Berkshire under the plan's withdrawal liability provisions. These provisions are mandated by statute, and regulated by the Pension Benefit Guaranty Corporation (PBGC).
Berkshire argues that section 4211 of ERISA, 29 U.S.C. § 1391, properly construed, does not authorize withdrawal liability where a pension plan's vested benefits are fully funded. The district court, without opinion, rejected Berkshire's position by granting the Fund's motion for summary judgment, based on the analysis in Ben Hur Construction Co. v. A.S. Goodwin, 784 F.2d 876 (8th Cir. 1986). Subsequent to the Eighth Circuit's consideration of the issue, the PBGC issued a "Notice of Interpretation," concluding that under the statute withdrawal liability should not attach where a plan's vested benefits are fully funded.1
The Supreme Court has recently reviewed the legislative history of the MPPAA in some detail. We quote liberally from its opinion in the appendix to familiarize the reader with the events leading to the passage of the MPPAA. Pension Benefit Guaranty Corp. v. R.A. Gray & Co., 467 U.S. 717, 720-23, 104 S. Ct. 2709, 2713-15, 81 L. Ed. 2d 601 (1984). The progress of the legislation makes clear that the MPPAA was designed to address two related problems. Before the MPPAA, employers in multiemployer pension plans had an incentive to withdraw from financially troubled plans in order to avoid financial liability should the plan have to be terminated. Congress was aware that many multiemployer plans were associated with financially troubled industries. Congress was concerned that this perverse incentive would hasten the demise of many multiemployer plans, and overburden the resources of the PBGC, whose insurance of vested benefits of single-employer plans it sought to extend to multiemployer plans.
The PBGC's recommendations were explained by its Executive Director:
To deal with this problem, our report considers an approach under which an employer withdrawing from a multiemployer plan would be required to complete funding its fair share of the plan's unfunded vested liabilities. In other words, the plan would have a claim against the employer for the inherited liabilities which would otherwise fall upon the remaining employers as a result of the withdrawal....
We think that such withdrawal liability would, first of all, discourage voluntary withdrawals and curtail the current incentives to flee the plan.
Pension Plan Termination Insurance Issues: Hearings before the Subcommittee on Oversight of the House Committee on Ways and Means, 95th Cong., 2nd Sess., 23 (1978) (statement of Matthew M. Lind) (emphasis added).
Berkshire argues that the district court erred in relying on Ben Hur, and in not deferring to PBGC's subsequent Notice of Interpretation. Berkshire further argues that Ben Hur misread the statute and its legislative history in determining that an employer may incur withdrawal liability where a plan's vested benefits are fully funded. The Fund disputes this, and argues that the "plain language" of the MPPAA supports withdrawal liability in such circumstances. Indeed, because the matter is so clear, the Fund argues, the PBGC's interpretation is entitled to no deference whatsoever. Permitting withdrawal liability for fully funded plans will further the Act's purpose of discouraging withdrawal and encourage additional employers to participate, it argues.2
We are persuaded that deference to the PBGC's interpretation is appropriate in this case. The PBGC, a product of the original ERISA legislation, has extensive experience in the area, and has had a central role in the implementation of ERISA from the outset. See Belland v. Pension Benefit Guaranty Corp., 726 F.2d 839 (D.C. Cir. 1984) (PBGC interpretation of ERISA entitled to "great deference"). The MPPAA was founded on a specific study and accompanying recommendations by the PBGC, as set forth in R.A. Gray (see appendix). It is evident that the withdrawal liability provisions of the MPPAA were specifically directed toward preserving the resources of the PBGC in anticipation of extending its insurance coverage to multiemployer plans. The PBGC, after a fifteen-month notice and comment period, promulgated its Notice of Interpretation, indicating that this concern is simply not implicated where a plan's vested benefits are fully funded. Of course the Ben Hur court did not have the benefit of the PBGC's Notice of Interpretation, and instead looked to a General Accounting Office report consistent with the PBGC's earlier position on the issue. 784 F.2d at 879 & n. 4.3
Perhaps most importantly, PBGC's interpretation is consistent with the statutory language, and Congress expressly delegated substantial regulatory authority to PBGC relating to withdrawal liability. Section 4201 of ERISA, 29 U.S.C. § 1381, provides that an employer's withdrawal liability is "the amount determined under section 4211 of this title [29 U.S.C. § 1391] to be the allocable amount of unfunded vested benefits." Section 4211 in turn authorizes various alternative schemes for determining withdrawal liability. This section permits plan trustees to choose one of the alternatives in advance, subject to approval by the PBGC.
The Fund trustees had selected a variation of the "direct allocation method." Under this method, the trustees assessed Berkshire withdrawal liability for:
(i) the plan's unfunded vested benefits which are attributable to participants' service with the employer
29 U.S.C. § 1391(c) (4) (A). As defined by the statute:
The plan's unfunded vested benefits which are attributable to participants' service with the employer is the amount equal to the value of nonforfeitable benefits under the plan which are attributable to participants' service with such employer (determined under plan rules not inconsistent with regulations of the [PBGC] decreased by the share of plan assets ... which is allocated to the employer....
29 U.S.C. § 1391(c) (4) (B).
The Fund argues that this language equates "the plan's unfunded vested benefits" with the amount by which an employer's negotiated contributions underfund the defined benefits of that employer's participating workers. But an equally plausible, and perhaps more intuitive, reading is that Sec. 1391(c) (4) (B) presumes that the plan has unfunded vested benefits--i.e., in the aggregate--before one reaches the allocation of that deficit on an employer-by-employer basis. This interpretation has been adopted by the PBGC:
The use of the term "allocable" [29 U.S.C. §§ 1381, 1391] at least suggests the prior existence of unfunded vested benefits to be allocated. Further, the discussion of withdrawal liability in the legislative history of the [MPPAA] strongly reinforces this suggestion. For example, Senator Javits, one of the Senate sponsor [sic] of MPPAA, stated:
"Under the bill, an employer who withdraws from a multiemployer plan would be liable to the plan for a proportionate share of the unamortized amount of the plan's unfunded vested benefits."
Notice of Interpretation, 51 Fed.Reg. 47,343 (1986) (citing 126 Cong.Rec. 20,178 (1980)) (emphasis added).4
This reading appears consistent with the two primary purposes behind the 1980 amendments: to protect the PBGC's limited resources, and to remove the incentive for employer withdrawal from shaky multiemployer pension plans. See Keith Fulton & Sons v. New England Teamsters & Trucking Indus. Pension Fund, 762 F.2d 1137, 1145 (1st Cir. 1985) (en banc). Where a plan is fully funded, the risk that the PBGC's resources will be called upon to fund those benefits is low. Likewise, when a plan is fully funded, employers have little incentive to withdraw. If the plan is not fully funded, withdrawal liability will attach, thus discouraging withdrawals.
We believe that the relevant statutory language supports the position that Congress did not intend to impose withdrawal liability on employers in a fully funded plan. But even viewed in the light most favorable to the Fund, the statute is at best ambiguous. The PBGC, authorized by Congress to implement the statute, has promulgated a reasonable interpretation, one which is consistent with the statute's purpose and overall structure. If we consider the precise statutory provisions at issue as ambiguous, then we believe the circumstances call for deference to the PBGC's interpretation. See Chevron U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837, 104 S. Ct. 2778, 81 L. Ed. 2d 694 (1984); cf. Wilcox v. Ives, 864 F.2d 915, 924-27 (1st Cir. 1988) (discussing this circuit's approach to the issue of agency deference).
The judgment of the district court is vacated, and the case is remanded for further proceedings consistent with this opinion.
The Notice reads, in part:
The fundamental purpose of withdrawal liability is to prevent the burden of unfunded benefits from falling on a shrinking contribution base and possibly accelerating the exodus of employers from the plan. In situations where this purpose is not served, the reason for withdrawal liability disappears.
* * *
ERISA does not permit the assessment of withdrawal liability under any statutory allocation method against employers that withdraw from a plan when, as of the end of the preceding plan year, the plan has no unfunded vested benefits.
Fed.Reg. 47,344 (1986)
This last suggestion is somewhat puzzling. Employer participation in multiemployer pension plans is a matter of bargaining between employers and unions. Likewise, the level of an employer's contribution is a matter of bargaining between the employer and the union. These plans are governed by trustees drawn equally from employers and unions. Benefits to covered employees, however, are set independently by plan trustees--drawn equally from employers and unions--and not by the employers or unions themselves. Thus, multiemployer pension plans are structured as "pooled" funds, such that some employers, in effect, "subsidize" the employees of other employers
In this case, Berkshire's contributions did not cover the level of benefits to be drawn by its employees. To the extent its employees received benefits in excess of the cost to the company, Berkshire would appear to profit from the pooled nature of the multiemployer plan; it had no apparent incentive to withdraw. Given that Berkshire only agreed to a level of contribution below that necessary to independently fund full benefits for its employees, it can be inferred that Berkshire would have further resisted participation in the Plan had it been required to provide the higher level of funding. Reading the MPPAA to permit withdrawal liability where an employer's contributions are less than its employees' benefits would appear to discourage participation by additional employers.
The GAO suggested that Congress amend the statute to exempt employers from withdrawal liability where a plan is fully funded, concluding: "Such an exemption would be consistent with withdrawal liability based on a share of the plan's unfunded vested benefits and should have little effect on the plan or its contributing employers." General Accounting Office, Effects of Liabilities Assessed Employers Withdrawing from Multiemployer Pension Plans 40, GAO/HRD-85-16 (Mar. 4, 1985)
It is true, as the Fund argues, that the Notice of Interpretation contradicts the PBGC's earlier view. See Opinion Letter 83-19, 50 Fed.Reg. 36,504 (1985). Inconsistency in agency interpretation of a statute is grounds for according less deference to the agency's interpretation. See Wilcox v. Ives, 864 F.2d 915, 924-25 (1st Cir. 1988). However this principle does not mandate a wholesale abandoning of deference whenever an agency changes its view. The PBGC's Notice of Interpretation followed a fifteen-month notice and comment period soliciting comments on Opinion Letter 83-19. As the subsequent Notice explained:
The principal basis for [Opinion Letter 83-19] was that the result seemed to be compelled by a literal reading of the statutory language.... This apparently straighforward [sic] reading does not, however, survive closer scrutiny.
Fed.Reg. 47,343 (1986). The Notice goes on to ground the new position on the structure, policy goals, and legislative history of the MPPAA. The agency's explanation for its change of position strikes us as credible. The agency's current interpretation should not be discounted in this case because it is the result of a more considered examination of the issue