Harrison Combs v. Classic Coal Corporation

931 F.2d 96, 289 U.S. App. D.C. 251, 13 Employee Benefits Cas. (BNA) 1993, 1991 U.S. App. LEXIS 7342, 1991 WL 62420

• Court: Court of Appeals for the D.C. Circuit
• Decided: April 26, 1991
• Docket: 90-7066
• Status: Published

Opinion

Opinion for the Court filed by Circuit Judge SENTELLE.

SENTELLE, Circuit Judge:

Classic Coal Corporation (“Classic” or “appellant”) appeals the district court’s grant of summary judgment in favor of trustees of the United Mine Workers of America 1950 and 1974 Pension Plans (“Trustees” or “appellees”). Upon Classic’s withdrawal from the Plans, the Trustees assessed almost $2 million in withdrawal liability against Classic. Classic demanded arbitration on the amount assessed. The arbitrator determined that the actuarial assumptions used in calculating Classic’s liability were unreasonable. The district court granted summary judgment, overturning the arbitration order and holding that the actuarial assumptions were reasonable in the aggregate within the meaning of the Employee Retirement Income Security Act (“ERISA”), 29 U.S.C. § 1393(a) (as amended by the Multiemployer Pension Plan Amendments Act of 1980 (“MPPAA”)). We affirm.

I. Background

A. Factual Summary

Appellees are trustees of the United Mine Workers of America (“UMWA”) 1950 and 1974 Pension Plans (“Plans”), which are multiemployer pension benefits plans maintained pursuant to collective bargaining agreements between UMWA and the Bituminous Coal Operators Association. Numerous employers contribute to the Plans on behalf of their employees, and the Plans, in turn, disburse benefits to eligible plan participants.

Appellant Classic Coal engaged in coal mining and other related activities from March 1978 until March 27, 1981, and, as a signatory to the collective bargaining agreements with UMWA, made contributions to the Plans during this period. In 1981, however, Classic ceased all covered operations and withdrew from the Plans. The MPPAA provides that, upon withdrawal, a participating employer must pay withdrawal liability. This liability represents the withdrawing employer’s proportionate share of the plan’s unfunded vested benefits (“UVB”), which are those benefits that are nonforfeitable by the plan’s participants, but are as yet unfunded by the plan.

The MPPAA prescribes statutory methods for calculating withdrawal liability. This calculation involves the allocation of a plan’s UVB among contributing employers. See 29 U.S.C. § 1391. To make this allocation, a plan’s trustees must first calculate the plan’s total UVB, defined as the difference between the actuarial present value of nonforfeitable pension benefits earned by employees and the value of plan assets actually on hand. 29 U.S.C. § 1393(c). Benefits in a plan vest after a participant completes a specified term of service covered by the plan. 29 U.S.C. § 1053. Vested benefits include amounts payable in the future to participants who have not yet retired as well as amounts currently being paid to retirees or their beneficiaries. Thus, the trustees’ actuary must estimate the present value of the plan’s liability for both vested and non-vested, accrued benefits. See 26 U.S.C. § 412(c)(9); 29 U.S.C. §§ 1082(c)(9), 1393(b)(1).

To determine the present value of a plan’s liability for vested benefits, the actuary employs certain assumptions. 29 U.S.C. § 1393(a). Most pertinently to the present case, because the Plans will be funded in part by investment earnings on current assets and in part by assets the Plans are expected to acquire in the future, their actuaries must select an interest rate assumption to apply in discounting the liability for future benefit payments. Increasing the interest rate assumption decreases the employer’s withdrawal liability. In calculating Classic’s withdrawal liability, the Trustees used a 5.5% interest rate assumption, the same rate they routinely used for the long-range funding determinations of the individual Plans. The Trustees fixed Classic’s withdrawal liability to the 1950 Plan at $877,406.76 and to the 1974 Plan at $1,092,094.73.

Classic protested the assessment and requested a reduction of its liability, asserting that the calculation used an unreasonable interest rate assumption. The Trustees refused to reevaluate the withdrawal liability, stating that Classic had failed to furnish evidence of the assumption’s unreasonableness. It appears to be undisputed that the 5.5% assumption is significantly lower than the Plans’ actual rates of return. After Classic’s withdrawal, in response to suggestions from its actuaries and assistant treasurer, the Plans’ Trustees increased the interest rate assumption for the 1950 Plan from 5.5% to 6.5%. This increase, however, applied retroactively to the Plan year following Classic’s withdrawal.

On August 23, 1983, Classic filed a demand for arbitration pursuant to 29 U.S.C. § 1401(a), which provides that disputes over withdrawal liability must be resolved initially in arbitration. After hearing, the arbitrator ruled for Classic, finding that the Plans’ actuarial assumptions and methods were unreasonable in the aggregate under 29 U.S.C. § 1393(a)(1).

B. The Arbitrator’s Decision

In arbitration proceedings, the trustees’ calculations “are presumed correct” unless the employer “shows by a preponderance of the evidence” that the determination was unreasonable or clearly erroneous. 29 U.S.C. § 1401(a)(3)(A), (B). Because of the inherent ambiguity of the term, the arbitrator in this case expressed his willingness to accept a fairly wide range of “reasonableness” in making his § 1393(a)(1) determination. Nevertheless, based on evidence from Classic’s actuarial witness and “statements made in connection with the Plans’ own deliberations,” he determined that the 5.5% interest rate assumption for withdrawal liability was, on balance, unreasonable. He found that this rate, adopted for funding purposes in 1975 when the Plans had few assets to invest, did not reflect either the substantially higher interest rates available in June 1980, or the actual return on assets of approximately 9.2% realized in 1980.

Next, the arbitrator examined the Trustees’ assertion that the 5.5% assumption was justified by the extended length of time over which returns for contributions would be received, but found that the evidence did not justify the length of the Trustees’ actuary’s projection. He further noted that the Trustees increased the interest rate assumption the year following Classic’s withdrawal. Consequently, he determined that the Trustees had calculated the interest assumption at an unreasonably low rate.

Based upon his belief that an unreasonable interest rate assumption would affect the entire calculation package, the arbitrator concluded that the Trustees’ actuarial assumptions, in the aggregate, were unreasonable under § 1393(a)(1). Accordingly, he ordered the Trustees to recompute Classic’s withdrawal liability using a higher interest rate — 6.5% for the 1950 Plan and a “blended” assumption for the 1974 Plan. The Trustees then filed the present action in district court seeking reversal of the arbitrator’s decision.

C. The District Court’s Opinion

The district court determined that “the arbitrator unjustifiably failed to apply the presumption in favor of the correctness of the Plans’ determinations, which were shown to be rationally supported, which is all the law requires.” Memorandum and Order, Civ. No. 84-1562, at 17, 1990 WL 66583, citing 29 U.S.C. § 1401(a)(3)(B). The court then held that, as a matter of law, the arbitrator improperly concluded that the Trustees’ actuarial assumptions were “unreasonable in the aggregate.” Id. at 18. The court further stated that “one assumption, no matter how unreasonable, may not be deemed to render the assumptions unreasonable in the aggregate,” reasoning that it was Classic’s burden to produce evidence of the unreasonableness of the other assumptions as well. Id. at 19.

The court also determined that the arbitrator should not have used post-1980 evidence to find the 1980 assumptions unreasonable. The court noted that forcing the Trustees to use evidence generated subsequent to the withdrawal would discourage the Trustees from making actuarial updates. This, the court concluded, “cannot be what Congress had in mind when it formulated the ‘best estimate’ language [in § 1393(a)(1)].” Id. at 18.

The district court, therefore, granted summary judgment in favor of the Trustees and reinstated the Trustees’ original determination of withdrawal liability. This appeal followed.

II. Discussion

In reviewing a grant of summary judgment, a legal rather than a factual determination, we consider de novo the same questions originally presented to the district court. See Beatty v. Washington Metropolitan Area Transit Authority, 860 F.2d 1117, 1119-20 (D.C.Cir.1988). In the present case we reach the same conclusion as the district court, although through somewhat different reasoning.

We review, as did the district court, the arbitrator’s determination of whether the Trustees’ “actuarial assumptions and methods [were], in the aggregate, ... reasonable.” Under the statute,

the determination is presumed correct unless a party contesting the determination shows by a preponderance of evidence that — (i) the actuarial assumptions and methods used in the determination were, in the aggregate, unreasonable ....

29 U.S.C. § 1401(a)(3)(B)(i). 1 “In meeting this burden, the test is not which withdrawal determination is the most reasonable but rather whether the challenged determination is unreasonable or clearly erroneous.” Board of Trustees, Michigan United Food and Commercial Workers Unions v. Eberhard Foods, Inc., 831 F.2d 1258, 1261 (6th Cir.1987) (emphasis added).

Actuarial valuations are based upon and reflect the experience of the plan, the professional judgment of the actuary, and the theories and expectations to which the actuary ascribes. Great differences of opinion exist as to actuarial methods. Congress, therefore, created the statutory presumption in favor of withdrawal determina tions expressly to forestall endless disputes “over technical actuarial matters with respect to which there are often several equally ‘correct approaches.’ ” S. 1076, The Multiemployer Pension Plan Amendments Act of 1980: Summary and Analysis of Consideration, 98th Cong., 2d Sess. 21 (1980). In the absence of this presumption, “employers could effectively nullify their obligation by refusing to pay and forcing the plan sponsor to prove every element involved in making an actuarial determination.” H.R.Rep. No. 869, pt. I, 96th Cong., 2d Sess. 1, 86, reprinted in 1980 U.S.Code Cong. & AdminNews 2918, 2954. Moreover, because of congressional recognition that an actuary’s calculation of withdrawal liability is not an “exact science,”

[MPPAA] does not anticipate that the actuary will always choose the figure the court would choose as the most reasonable from among this range [of reasonableness]. Rather, the only requirement is that in every case the actuarial determination will fall within the range of reasonableness.

Eberhard Foods, 831 F.2d at 1261. This “range of reasonableness” permits the actuary wide latitude in determining withdrawal liability.

Consequently, under the MPPAA’s withdrawal liability procedures,

[c]omplex technical determinations of the amount of liability are allocated to those with the greatest expertise. Plans are protected under the Act from inconsistent judicial or arbitrator holdings by a range of acceptable actuarial methods, and a concrete formula by which liability may be calculated.

Dorn’s Transportation, Inc. v. I.A.M. Nat’l Pension Fund Benefit Plan A, 578 F.Supp. 1222, 1238 (D.D.C.1984) (emphasis added), aff'd without opinion, 753 F.2d 166 (D.C.Cir.1985). The existence of a statutory presumption, however, does not preclude all review of the Trustees’ determinations. In Washington Star Co. v. International Typographical Union Negotiated Pension Plan, 729 F.2d 1502 (D.C.Cir.1984), this Court stated that it “interpreted] the presumptions under the MPPAA, in favor of a plan’s withdrawal liability assessments, not to insulate the trustees’ determinations from effective review, but rather to require the trustees affirmatively to provide some support for the reasonableness of their determinations in the event that the employer introduces evidence to the contrary.” Id. at 1511. Pursuant to this standard, the arbitrator must uphold a plan’s determinations if the plan is able to support the reasonableness of its assumptions in the face of an employer’s challenge. See Eberhard Foods, 831 F.2d at 1260. Although the arbitrator here noted the presumption of reasonableness due the Trustees’ determinations, and the wide range of “reasonableness” available, we agree with the district court that he did not in fact afford the Trustees’ determination the benefit of the presumption.

In its attack on the Trustees’ assumptions, Classic focused solely on the interest rate assumption. The Trustees responded that the interest rate assumption was reasonable and that, in the alternative, even if unreasonable, it was reasonable in the aggregate because the lower interest rate accounted for other intangibles. The arbitrator, basing his decision on the evidence tendered by Classic, rejected the Trustees’ first contention, and found the interest rate assumption, standing alone, to be unreasonable. However, neither the arbitrator’s decision nor Classic’s evidence confronts the Trustees’ second argument that the assumptions were reasonable in the aggregate. Indeed, the arbitrator himself noted that “Classic’s evidence related entirely to the interest assumption issue.”

The arbitrator carefully explored the fact that the Trustees’ actuary, in selecting the interest rate assumption, did not account for the higher rate of return actually experienced by the Plans, projected future investment returns over a period of too many years, and incorrectly estimated the period over which withdrawal liability would be paid. However, the arbitrator did not consider whether the Trustees used the lower interest rate assumption to offset the possible understatement of other assumptions. Accordingly, he made no determination as to whether the assumptions were, in the aggregate, unreasonable.

The fundamental premise that an erroneously low interest rate assumption alone is sufficient to destroy the validity of the entire calculation is not flawed. The application of that premise, however, must take into account the question of whether an apparently unreasonable assumption is offset by other assumptions, so as to produce an assumption package reasonable in the aggregate. The arbitrator failed to apply the premise correctly. The Trustees produced evidence to show that their lack of participant data resulted in actuarial assumptions, in the aggregate, that offset or reduced the effect of the conservative interest rate assumption. The Trustees further identified that the Plans’ weak funding position and their resulting inability to participate to any meaningful extent in then-current high interest rates, coupled with the generally unstable condition of the coal industry, affected the overall package of actuarial assumptions. According to the Trustees’ actuary, liberal assumptions or uncertainty as to. these elements offset the conservative assumption in the interest rate.

In support of the aggregate reasonableness of their actuarial assumptions, the Trustees offered evidence of a standard actuarial test, the “gain-and-loss analysis,” which measures the accuracy of a plan’s aggregate assumptions by comparing the plan’s actual liability as of a given date with the liability that would have existed had all the assumptions used in the previous year’s valuation been realized. The difference between the actual liability and the liability computed from the assumptions represents the actuarial gain or loss. A high gain means the assumptions were too conservative; a high loss means they were too liberal. The Plans’ actuary conducted such a gain-and-loss analysis for each year from 1974 through 1980 using the same assumptions he employed in computing Classic’s withdrawal liability. His computations revealed only slight actuarial variation. Thus, the Trustees’ evidence supports the proposition that the assumptions in the aggregate were not unreasonable.

Classic did not rebut the Trustees’ evidence of the aggregate reasonableness of their assumptions. Nor did the arbitrator make any findings rejecting the appellees’ evidence. Although the arbitrator may be correct that the interest rate assumption is the most important variable in determining withdrawal liability, reasonableness in the aggregate is still possible where the total actuarial package contains other assumptions offsetting any undue conservatism in the selection of a particular assumption. The appellees’ evidence in the present case supports a conclusion that this reasonableness was not only possible, but achieved.

Classic was free to argue that the evidence of unreasonableness regarding the interest rate assumption was not sufficiently rebutted by the Trustees’ evidence demonstrating the aggregate reasonableness. Absent the statutorily mandated presumption in favor of the Trustees’ determination, the arbitrator’s finding in Classic’s favor might well be supportable on the record. The problem for Classic is that we do not operate in the absence of that presumption.

Classic attacks the district court’s statement that “one assumption, no matter how unreasonable, may not be deemed to render the assumptions unreasonable in the aggregate.” Memorandum and Order, supra, at 19. The court noted that “[t]he burden was on Classic to produce evidence of the unreasonableness of the other assumptions utilized so as to render a complete picture; its failure to do so deprives the arbitrator’s decision of necessary support.” Id. Under the MPPAA, the reasonableness of an actuarial assumption must not be judged in isolation; the statute directs us to consider the assumptions “in the aggregate.” 29 U.S.C. § 1401(a)(3)(B)(i). However, this does not mean the employer cannot show that a single assumption is so unreasonable that it affects the assumptions overall. Indeed, “[a] small adjustment in the interest rate assumption can lead to a major change in the withdrawal liability calculation.” Eber- hard Foods, 831 F.2d at 1260. It is not, therefore, necessary that the employer offer evidence that each assumption is unreasonable in order to show that the assumptions are unreasonable in the aggregate. It is, however, necessary that the employer, in order to prevail, show by some means that the assumptions are unreasonable in the aggregate. Id. Given the totality of the evidence in the present record, the employer has failed to demonstrate this. Therefore, although the district court’s language may appear to overstate its rationale, its conclusion is correct.

Classic invokes a further statutory presumption in support of its position — that the arbitrator’s findings of fact are correct unless rebutted by a clear preponderance of the evidence. 29 U.S.C. § 1401(c). Classic argues that the present record will not support a reversal of the arbitrator when that presumption is applied. The difficulty with Classic’s position is that the district court also had the duty of determining “whether applicable statutory law has correctly been applied and whether the findings comport with the evidence.” I.A.M. Nat’l Pension Fund Benefit Plan C v. Stockton TRI Industries, 727 F.2d 1204, 1207 n. 7 (D.C.Cir.1984). In this case we agree with the district court that the arbitrator did not correctly apply the statutory law, specifically the statutory presumption of correctness due the Trustees’ determination in § 1401(a)(3)(B).

Classic also argues that the district court erred in its conclusion that the arbitrator wrongly considered evidence gathered after Classic’s withdrawal to find the assumptions used to calculate Classic’s withdrawal liability unreasonable. We conclude, however, that the district court was correct in its determination. As the court noted, “this calculation is like a snapshot, in that it represents the actuary’s ‘best estimate’ given the evidence then available.” Memorandum and Order, supra, at 18 n. 10 (emphasis added). Although it was true that the Trustees considered reevaluating their interest rate assumptions, they had made no firm decisions to reevaluate at the time Classic withdrew from the Plans. The Trustees relied upon evidence “then available,” i.e., available at the time of Classic’s withdrawal, to calculate Classic’s liability. To require the Trustees to base their assumptions on information gathered after the fiscal year-end of the Plans would discourage actuarial updating. Once a withdrawn employer’s liability is fixed, changes in the UVB are irrelevant to the inquiry regarding withdrawal liability. Just as an employer’s liability is not increased if the plan suffers losses in the withdrawal year, the employer is not entitled to benefit from actuarial changes subsequent to its withdrawal.

III. Conclusion

For the reasons set forth above, we conclude that the district court did not err in its application of the law of withdrawal liability to the uncontested facts before it. Even given the presumption afforded the arbitrator’s findings, the district court properly concluded that the arbitrator did not comply with the statutory standards and appropriately set aside the arbitration award. Accordingly, we affirm.

1 . The only other statutory basis for upsetting the Trustees’ determination, that "the plan’s actuary made a significant error in applying the actuarial assumptions or methods,” 29 U.S.C. § 1401 (a)(3)(B)(ii), is not relevant to the present case.