Pages 5-6 and footnote 28: The IRS revised and updated the procedures governing automatic and discretionary approval of funding method changes in Revenue Procedure 2000-40, 2000-2 C.B. 357 (automatic approval), and 2000-41, 2000-2 C.B. 371 (discretionary approval). The automatically approved changes in the funding method of a terminated plan are the same as under the prior procedure. Revenue Procedure 2000-40, §4.02. As before, all other method changes in the year of termination require IRS discretionary approval, even if they would be approved automatically in an ongoing plan. Id. at §6.01(5).

Page 8: The decline in the number of terminations has continued. In the year ended September 30, 2002, the PBGC received 1,214 standard termination notices, compared to 1,565 in the year ended September 30, 2001, reaching a new all-time low. Distress terminations numbered 93, the same as the (revised) figure for the preceding year. Pension Benefit Guaranty Corporation, Pension Insurance Data Book 2002 at 37. PBGC-insured plans now number about 33,000, covering about 35 million participants. The former number continues a slight downward trend, while the latter, despite much talk about corporate America’s abandonment of defined benefit pension plans, a new high. Id. at 5, 43. More recent information on plan terminations and participation in PBGC-insured plans, showing no significant upward or downward movement since the late 1990’s is included in the PBGC’s annual Pension Insurance Data Book.

In Shedco, Inc., 76 T.C.M. 267, 22 EB Cases 1723 (1998), a Tax Court memorandum opinion held that an imprudent investment was not, in and of itself, grounds for plan disqualification. The taxpayer’s pension plan, in which its sole shareholder (the plan trustee) and his wife were the only remaining participants, made an unsecured loan of about 90 percent of its assets to a company with which the shareholder had extensive past, but no significant current, financial relationships. The borrower eventually went into bankruptcy, and the IRS revoked the plan’s qualification on the ground that the loan violated the principles of prudence and diversification, thus demonstrating that the plan was not operated for the exclusive benefit of participants as required by section 401(a)(2).

. . . an isolated violation of the prudent investor rule, although a factor to be considered, does not, of itself, require a finding that the plan was not operated for the exclusive benefit of the employees or their beneficiaries. We must look at the entire picture in assessing whether the plan violated the exclusive benefit rule. See Feroleto Steel Co. v. Commissioner, 69 T.C. at 107. Mr. Shedd made an error in judgment in having the plan lend money to Estes Co. without security. We are persuaded, however, that Mr. Shedd intended the plan, and thereby the participants, to benefit from the loan. Indeed, on the basis of the record, we believe that the plan would have profited from the loan if the depression in the real estate market had not occurred in Arizona during the late 1980’s. The loan proceeds did not flow back to petitioner, nor were they diverted for the personal benefit of the Shedds or Mr. Estes. Interest was stated on the note at market rate, and payments were being made until Estes Co. and Estes Homes began to experience financial difficulties. Viewing the total picture, we conclude that the loan was an isolated violation of the prudent investor rule, but that violation was not so serious as to constitute a violation of the exclusive benefit requirement of section 401(a). [22 EB Cases at 1737]

The facts in another case decided by memorandum opinion, Westchester Plastic Surgical Associates, P.C., 78 T.C.M. 756, 23 EB Cases 2127 (1999), more closely resembled those in the cases noted in the main text (p. 20, fn. 98). The sole participant in a one-man pension plan took out loans from the plan, secured only by his accrued benefit, and made no repayments. Presumably he was past the plan’s normal retirement age, since the IRS did not hold that the plan had become disqualified as a result of prohibited in-service distributions. Cf. Revenue Ruling 71-437, 1971-2 C.B. 185. Instead, the plan was disqualified for violation of the exclusive benefit rule. The Tax Court agreed, finding the same “indifference toward the continued well-being of the plan that we found in Winger's Depart. Store, Inc. v. Commissioner . . . and Ada Orthopedic, Inc. v. Commissioner”, in contrast to Shedco, where investment decisions, foolish as they may have been in retrospect, had the goal of benefitting participants rather than the plan sponsor or trustee.

We believe that these cases support the view expressed in the text, that disqualification can be imposed as a sanction for fiduciary violations only in extreme cases that are not likely to arise in the course of plan termination. Nothing supports the position, occasionally taken by IRS officials in the past, that section 401(a)(2) can be used as a pretext for fashioning quasi-fiduciary tax rules independent of the fiduciary standards imposed by ERISA.

Pages 25-31: In Brothers v. Miller Oral Surgery Inc. Retirement Plan, 230 F.3d 1348, 25 EB Cases 1369 (3d Cir., 2000 [marked “Unpublished - Not Precedential”]), the court held that failure to inform participants of a plan amendment that converted the plan from a money purchase pension plan into a discretionary profit sharing plan required the continuation of the prior money purchase contribution formula. That conclusion is entirely unsurprising.

1. A plan sponsor that fails to give the required notice of significant reductions in future benefit accruals is subject to an excise tax of $100 day per uninformed “applicable individual”. I.R.C., §4980F(b). “Applicable individuals” are participants whose rate of future benefit accrual “may reasonably be expected to be significantly reduced” by the amendment, plus any alternate payees affected by the reduction, I.R.C., §4980F(f)(1) [a]. Where the sponsor exercises “reasonable diligence” in giving the notice, it is not liable for the tax with respect to failures whose existence was not known and that are corrected within 30 days of discovery, and its total section 4980F tax liability is capped at $500,000 a year. I.R.C., §4980F(c). The tax can be waived by the IRS for reasonable cause “to the extent that the payment of such tax would be excessive or otherwise inequitable relative to the failure involved”. I.R.C., §4980F(c)(4).

2. Failure to give notice no longer prevents an amendment from taking effect, so long as the failure is not “egregious”. In the case of egregious failures, “applicable individuals” must in the future accrue benefits at the higher of the rates in effect before and after the plan amendment. The regulations specify that all “applicable individuals”, not just those who did not receive proper notices, benefit from the continued accrual requirement. Accruals under the unamended plan cease after the notice failure has been corrected and the waiting period for a valid notice (generally 15 or 45 days, depending upon the size of the plan) has elapsed. Treas. regs., §54.4980F-1, Q&A-14(a)(3) (but accrual need not continue in event beyond the date of plan termination if the plan terminates in accordance with Title IV of ERISA, Treas. regs., §54.4980F-1, Q&A-17(b)). On the other hand, it is clear from the statutory language that accruals by participants who were not required to receive notice (e. g., those who first began accruing benefits after the amendment was adopted) are governed by the amended plan. ERISA, §204(h)(6)(A). An “egregious failure” is defined as one that (i) is intentional, (ii) results in failure “to provide most of the individuals with most of the information they are entitled to receive” or (iii) is defined as “egregious” under IRS regulations. ERISA, §204(h)(6)(B). The regulations offer this definition:

For purposes of section 204(h) of ERISA and this Q&A-14, there is an egregious failure to meet the notice requirements if a failure to provide required notice is within the control of the plan sponsor and is either an intentional failure or a failure, whether or not intentional, to provide most of the individuals with most of the information they are entitled to receive. For this purpose, an intentional failure includes any failure to promptly provide the required notice or information after the plan administrator discovers an unintentional failure to meet the requirements. A failure to give section 204(h) notice is deemed not to be egregious if the plan administrator reasonably determines, taking into account section 4980F, section 204(h) [of ERISA], these regulations, other administrative pronouncements, and relevant facts and circumstances, that the reduction in the rate of future benefit accrual resulting from an amendment is not significant (as described in Q&A-8 of this section), or that an amendment does not significantly reduce an early retirement benefit or retirement-type subsidy. [Treas. regs., §54.4980F-1, Q&A-14(a)(2)]

3. The old law required that the section 204(h) notice be given at least 15 days before the effective date of the amendment. EGTRRA changed 15 days to “a reasonable period”, ERISA, §204(h)(3), I.R.C., §4980F(e)(3), which the regulations interpret as 45 days for single employer plans that have, as of the effective date of the amendment,100 or more participants with accrued benefits. Treas. regs., §54.4980F-1, Q&A-9(a). The period is shortened to 15 days for smaller plans, Treas. regs., §54.4980F-1, Q&A-9(b), multiemployer plans, Treas. regs., §54.4980F-1, Q&A-9(c), and for amendments “adopted in connection with the acquisition or disposition”, Treas. regs., §54.4980F-1, Q&A-9(d)(1). Notices of reduction or elimination of future early retirement benefits or retirement-type subsidies that occur in connection with an acquisition or disposition may be given up to 30 days after their effective date. Treas. regs., §54.4980F-1, Q&A-9(d)(2). The meaning of “acquisition or disposition” is the same as in Treas. regs., §1.410(b)-2(f), i. e., “an asset or stock acquisition, merger, or other similar transaction involving a change of employer of the employees of a trade or business”. Nothing is said about what it means for an amendment to be “in connection with” such a transaction.

Since participants cannot make any effective use of the information provided by the notice and, except in the rarest of circumstances, are unaffected by whether they learn about changes in a plan’s benefit formula before or after its effective date, there can be no non-arbitrary standard for deciding how much notice is “reasonable”. The IRS’s 45 days could just as easily have been five or five hundred. Section 204(h) remains a trap for unwary plan sponsors rather than a rational protection for participants.

5. The regulations devote considerable space to fairly obvious statements about what sort of amendments result in a reduction in the rate of future benefit accrual. Treas. regs., §54.4980F-1, Q&A-8. One potentially troublesome IRS position is that, where a plan has different benefit formulas for different groups of participants (e. g., for hourly and salaried or for employees of different divisions), a reduction in the rate of future accruals for one group may necessitate giving notice to members of other groups if a significant number of participants may in the future change job categories and accrue lower benefits as a result. Treas. regs., §54.4980F, Q&A-10(f), Examples 5, 6 and 7. Because it may not be clear that only a “small percentage” of participants in one group can reasonably be expected to move to the other, employers may feel compelled to distribute superfluous notices out of an abundance of caution.

The regulations inform us that the determination of whether a reduction is “significant” is made by comparing “reasonable expectations taking into account the relevant facts and circumstances” of future benefit accruals, with and without the amendment. Treas. regs., §54.4980F-1, Q&A-8(a). An example makes it clear that the ultimate impact on normal retirement benefits is considered, not possible year-to-year fluctuations. Thus an amendment that cannot result in a lower normal retirement benefit for any participant than he would have received under the unamended plan does not require notice, even if accruals are lower in some future years than in the current year. Treas. regs., §54.4980F-1, Q&A-8(d). What is not discussed at all is how large a reduction must be in order to be “significant”. Cautious sponsors will issue section 204(h) notices for all plan amendments that could conceivably cause a future reduction in any participant’s benefits to even the smallest extent.

6. Reversing the position taken in the prior IRS regulations, EGTRRA requires a section 204(h) notice for significant reductions in early retirement benefits or retirement type subsidies. ERISA, §204(h)(9); I.R.C., §4980F(f)(3). For example, notice would have to be given by a defined benefit plan that eliminated a lump sum option with respect to future benefit accruals. The anti-cutback rule prohibits taking away that option for benefits already accrued. ERISA, §204(g)(2); I.R.C., §411(d)(6)(B). Benefit provisions that are not protected by the anti-cutback rule (for example, ancillary death or disability benefits) can be eliminated without the necessity of complying with section 204(h). Treas. regs., §54.4980F-1, Q&A-7(b).

7. EGTRAA adds to the law a few generalities about the contents of the notice, which the proposed regulations try to expand into useful guidance. The statute says that it “shall be written in a manner calculated to be understood by the average plan participant and shall provide sufficient information . . . to allow applicable individuals to understand the effect of the amendment”. ERISA, §204(h)(2); I.R.C., §4980F(e)(2). The proposed regulations deal at length with the need to describe the effects of the amendment in detail and to include examples illustrating of its impact. Treas. regs., §54.4980F-1, Q&A-11. Little of this material is pertinent to a terminating plan.

8. The proposed regulations confirm that EGTRRA did not alter the relationship between section 204(h) and Title IV of ERISA. For a plan that terminates in accordance with Title IV, benefit accrual ceases on the date of plan termination, regardless of whether a section 204(h) notice was ever issued. Treas. regs., §54.4980F-1, Q&A-17(b). As before, a section 204(h) notice is advisable if the plan sponsor wishes to be certain that accruals will stop on the proposed date of plan termination even if the actual DOPT for some reason is postponed. Treas. regs., §54.4980F-1, Q&A-17(c).

A section 204(h) notice is essential to the termination of a non-Title IV plan that is subject to the notice requirement. A consequence of the new section 4980F is that plans exempt from Title I of ERISA (and hence from section 204(h)) may still face excise tax exposure if they terminate (or otherwise reduce future benefit accruals) without providing a section 204(h) notice. Governmental plans and non-electing church plans are exempted from the excise tax by statute, I.R.C., §4980F(f)(2), and the regulations exempt those that cover only self-employed individuals and have fewer than 100 participants, Treas. regs., §54.4980F-1, Q&A-3(b). The sponsor of a plan solely for partners (or LLC owners taxed as partners) under which 100 or more had accrued benefits would be liable for the excise tax if it did not distribute a required section 204(h) notice. The failure to give the notice would not, however, prevent the amendment from going into effect (meaning that the notice requirement in this instance benefits only the federal fisc).

[a] The definition of “applicable individual” includes “beneficiaries”, but the reference to them is meaningless, as the accrual rate of the decedent from who a beneficiary derives his benefit entitlement is necessarily zero for reasons that have nothing to do with plan amendments. Notice must also be given to labor unions that represent applicable individuals. Since altering benefits for represented employees is rarely, if ever, possible without collective bargaining, unions will almost always have adequate notice of amendments in any event, making a formal section 204(h) notice a mere formality.

Page 37: EGTRRA’s liberalization of the section 404(a)(3) deduction limits for profit sharing plans made money purchase pension plans obsolete, so many are now being converted into, or merged with, profit sharing plans. Revenue Ruling 2002-42, 2002-28 I.R.B. 76, answers two (not very difficult) questions concerning the cessation of benefit accruals under the plan. First, neither a conversion nor a merger results in termination or partial termination of the money purchase plan. Hence, participants may continue vesting under the plan’s existing schedule, with no requirement of immediate full vesting. Since the transaction does not deprive any participant of the possibility of accruing future vesting service and cannot create a prospective reversion to the employer, there is clearly no basis for requiring immediate vesting.

Second, the merger or conversion is a cessation of future benefit accruals. Profit sharing allocations are not pension accruals, so the post-conversion accrual rate has to be zero. Hence, the employer must give advance notice under ERISA, §204(h) and I.R.C., §4980F. Note: In the course of its analysis, the ruling refers to Treas. regs., §1.401(a)-2(b), which states that, upon the termination of a defined contribution plan, the employer may receive a reversion of unallocated assets in any section 415 suspense account. As discussed in the main text, pp. 232-3, ERISA, §403 forbids all reversions from individual account plans. The regulation is thus pertinent only to non-ERISA individual account plans.

Pages 50-1: An error in the Notice of Plan Benefits does not bind the plan to pay a higher benefit than the participant is entitled to receive under the terms of the plan. The only circumstance under which a participant could reasonably make such a claim is where he acted in detrimental reliance on the misstatement - not a likely situation when the plan is terminating. Cf. Hofsas v. Montgomery Hospital Medical Center, 2000 U.S. Dist. LEXIS 14602, 25 EB Cases 1410 (E.D.Pa., 2000) and Crosby v. Rohm & Haas Company, 2007 U.S. App. LEXIS 6084; 2007 FED App. 0104P (6th Cir., March 16, 2007).

Pages 69-71: EGTRRA greatly liberalized the rules governing the deduction of contributions made to enable a plan to close out in a standard termination, obviating the problems discussed in the text. New section 404(a)(1)(D)(iv) provides that, when a plan subject to ERISA, §4041 terminates, the deduction limit for the tax year related to the plan year of termination will never be less than “the amount required to make the plan sufficient for benefit liabilities (within the meaning of §4041(d) of [ERISA])”. Except in the rare case in which the close-out contribution is so large that the combined defined benefit-defined contribution limit of section 404(a)(7) comes into play, all contributions made on account of plan termination will henceforth be deductible immediately.

Plans that are not subject to Title IV and therefore cannot take advantage of section 404(a)(1)(D)(iv) may be helped by the extension of the general section 404(a)(1)(D) deduction limit (unfunded current liability as determined under section 412(l)) to plans with 100 or fewer participants. These plans now have the same deduction limit as larger ones, except that benefit liabilities attributable to accruals by highly compensated employees that result from amendments adopted or effective during the last two years are disregarded. I.R.C., §404(a)(1)(D)(ii). Note: The legislative history of section 404(a)(1)(D)(ii) states that it is to apply only to plans subject to Title IV. H. Rep. No. 107-84 at 166. The statute omits that limitation. It appears that either the conferees changed their minds at the last minute or the report is in error.

The EGTRRA amendments are effective for plan years beginning after December 31, 2001, and hence are available for any plan that terminates during a plan year beginning after that date. The effective date does not depend on when the sponsor’s tax year begins. If the sponsor and the plan are on different years, it is possible for a tax year beginning before 2002 to base its deduction limit on a plan year beginning after 2001.

The Pension Protection Act of 2006 increased the deduction limits for the 2006 and 2007 tax years to 150 percent of current liability minus the value of plan assets. I.R.C., §404(a)(1)(D)(i), as amended by P.L. 109-280, §801(d)(1). For 2008 and later years, the limitations have been thoroughly revamped to reflect the revised minimum funding standards that then come into effect. The new limitations are sufficiently high that they should eliminate any vestigial constraints on fully deducting contributions made to close out a terminating plan.

Funding deficiencies are subject to a recurring 10 percent excise tax, I.R.C., §4971(a), which increases to 100 percent if the deficiency is not corrected, I.R.C., §4971(b). Revenue Ruling 79-237, 1979-2 C.B. 190, held that the sponsor of a terminating plan was subject to the 100 percent tax unless the funding deficiency was corrected in the year of termination. The contribution needed to eliminate the funding deficiency may be larger than the amount needed to fund all benefit liabilities, thus creating residual plan assets where none would otherwise exist. If the employer receives the residual assets as a reversion, it must pay both regular income tax and the section 4980 reversion excise tax. Worse yet, before the EGTRRA liberalization, all or part of the corresponding deduction might be spread over ten years.

Page 86, end of 1st partial paragraph: The calculation of lump sums on plan termination is subject, of course, to the same rules as in an ongoing plan. The actuarial assumptions used to convert participants’ accrued benefits into single-sum values must be based on the actual date of distribution, not on the date of plan termination. Pension Benefit Guaranty Corporation v. Wilson N Jones Memorial Hospital, 374 F.3d 362, 32 EB Cases 2921 (5th Cir., 2004).

(II) the interest rate and mortality table used to determine the amount of any benefit payable under the plan in the form of an annuity payable at normal retirement age shall be the rate and table specified under the plan for such purpose as of the termination date, except that if such interest rate is a variable rate, the interest rate shall be determined under the rules of subclause (I). [ERISA, §204(b)(5)(B)(vi); I.R.C., §411(b)(5)(B)(vi)]

First, the participant’s hypothetical account balance must be credited with interest after plan termination at a frozen rate equal to the average rate credited over the past five years (or to the fixed rate at termination, for those rare plans that have fixed rates). If a participant elects a lump sum distribution, he will be entitled to his account balance, as would be the case for any other post-PPA’06 cash balance plan. ERISA, §203(f)(1); I.R.C., §411(a)(13).

The effect of these provisions is, on the whole, beneficial. Purchasing annuity contracts to close out cash balance plans could be difficult if the interest crediting rates were subject to unpredictable future fluctuations. Complications may arise from the effective date: “periods beginning on or after June 29, 2005”, Pub. L. 109-280, §701(e)(1). The statute was not enacted until August 17, 2006, leaving a period of over a year during which the new requirements apply retroactively. Plans that terminated during that interval face the prospect of renegotiating their annuity contracts.

Pages 141-5: In an opinion marked by patent judicial illiteracy, the Third Circuit characterized a shutdown-type benefit as an “early retirement subsidy” protected from cutback by ERISA, §204(g) and I.R.C., §411(d)(6). Bellas v. CBS Inc., 221 F.3d 517, 25 EB Cases 1206 (3d Circuit, 2000), cert. den., 531 U.S. 1104, 25 EB Cases 2280 (2001). Until 1994, the pension plan in which the plaintiff participated included an unreduced “job separation benefit” for any participant with 30 years of service whose employment terminated “through no fault of his own through lack of work for reasons associated with the business . . . [and the employer] determines there is no reasonable expectation of recall”. In 1994 this benefit was eliminated. The plaintiff was laid off in 1997. Having 30 years of service, he contended that he was entitled to the job separation benefit, on the ground that its elimination with respect to benefits accrued as of the date of amendment violated the anti-cutback rule.

The court agreed with the employer's contention that the disputed provision was in the nature of a shutdown benefit. (One judge disagreed, taking the view - hard to justify on the basis of the plan language - that it was a garden variety 30-and-out clause.) The court did not, however, agree that shutdown benefits are outside the purview of the anti-cutback rule. Instead, it twisted the legislative history (“a plant shutdown benefit (that does not continue after retirement age) will not be considered a retirement-type subsidy”) to mean the opposite of what it says.

The job separation benefit provided nothing after retirement age. From that point on, a participant who met its conditions received exactly the same normal retirement benefit as if the provision did not exist. The subsidy was the benefit paid between separation from service and normal retirement. The IRS expounded this point in GCM 39869, but the judges were unimpressed. In their view, the receipt after normal retirement age of the plan's normal retirement benefit, rather than of an actuarially reduced benefit, is a subsidy that continues after retirement age. If that were true, the only shutdown benefit not within the purview of the anti-cutback rule would be a temporary benefit that ceased at or before normal retirement age, which, except for Social Security supplements paid before age 62, is illegal. (A participant’s normal retirement benefit is “the greater of the early retirement benefit under the plan, or the benefit commencing at normal retirement age”, ERISA, §3(22), I.R.C., §411(a)(9). A supposedly temporary benefit in excess of the ostensible normal retirement benefit would therefore be substituted for the normal retirement benefit at normal retirement age, unless it qualified as a Social Security supplement.) It is clear that Congress did not intend that shutdown benefits would be exempt from the anti-cutback rule only if they qualified as Social Security supplements, since the latter are discussed separately in the same passage of the legislative history.

In Notice 2003-10, 2003-5 I.R.B. 1 (Jan. 10, 2003), the IRS announced that it intended “to propose regulations that would provide guidance on benefits that are treated as early retirement benefits and retirement-type subsidies for purposes of [section] 411(d)(6)(B) of the Internal Revenue Code. It is expected that the regulations would include guidance resolving conflicting court decisions that address the extent to which payments that are contingent on the occurrence of an unpredictable event, such as a plant shutdown, are protected under [section] 411(d)(6)(B).”

Astonishingly, the upshot was proposed regulations that resolved the conflict in favor of Bellas without any analysis or rationale, 69 Fed. Reg. 13769 (March 24, 2004); adopted without material changes, Treas. Regs., §1.411(d)-3(b), 70 Fed. Reg. 47109, 47116 (August 12, 2005). An example in the proposal discusses a plan that provides that participants who are severed on account of a plant shutdown and have completed at least ten years of service may commence receiving their pension immediately with a three percent per year reduction for commencement before normal retirement age (less than a full actuarial reduction). We are told, in conclusory fashion, that this benefit is a retirement-type subsidy, with no attempt to explain in what sense it “continue[s] after normal retirement age”. Starting at normal retirement age, participants entitled to the shutdown benefit receive nothing more than they would have gotten if they had accrued no further benefits after the shutdown and had begun drawing their pensions at normal retirement. It is therefore evident that nothing that they got as a result of the shutdown has extended beyond the pre-normal retirement period.

The regulations state that a shutdown benefit consisting solely of a Social Security supplement is not protected by section 411(d)(6) and thus may be eliminated, but its elimination is permissible because it is an ancillary benefit, not because it is a shutdown benefit that does not continue after normal retirement age. In short, the proposal reads the shutdown benefit exception out of the statute’s legislative history, without explanation or even, so far as one can tell, awareness of what it is doing. Treas. Regs., §1.411(d)-3(b)(4), Example (2).

Looking to the future, it is advisable to fashion shutdown benefits as “ancillary benefits”, which can be freely reduced or eliminated, rather than as “retirement-type subsidies”, which cannot. The key distinction is that the participant’s benefit starting at the plan’s normal retirement age must equal his accrued benefit at the time of the shutdown; whatever payments he receives before NRA are then “ancillary”. Treas. Regs., §1.411(d)-3(g)(2)(vi). An example is a plan that pays 75 percent of the accrued benefit between shutdown and age 65 and 100 percent thereafter. A shutdown benefit with no reduction for early commencement would also qualify. By contrast, merely lowering the plan’s normal early retirement reduction factor to a level greater than zero results, in the IRS’s odd view, in a subsidy that “continue[s] after retirement age” (the benchmark that the legislative history employs to identify a “retirement-type subsidy”).

The definition of “ancillary benefit” cautions that a shutdown benefit falls into that category only if it “is permitted in a qualified pension plan”. Treas. Regs., §1.401-1(b)(1)(i) prohibits pension plans from providing inter alia “layoff benefits”, a term that is not defined in the statute or regulations. Its meaning is explicated to some extent in G.C.M. 39869 (April 6, 1992):

Although layoff benefits are not defined in the Code or regulations, generally they are perceived to be event-based benefits that are provided for a limited period of time and are intended to be a short-term replacement of the employee’s income. The event that triggers this short-term benefit is usually a permanent or temporary layoff or a reduction in the employer’s workforce. These benefits are similar to severance benefits as defined under Title I of ERISA. See 29 C.F.R. § 2510.3-2(b). [footnote omitted] For example, a layoff benefit may be offered for a 6-month or 1-year period following a reduction in workforce. The form of this benefit may be either a lump-sum payment or monthly or quarterly payments. . . .

Some shutdown benefits are provided in a form similar to severance benefits. They are short-term in nature and are a supplement or replacement of the participant's income during the period of unemployment. These benefits are provided for a limited period of time after the termination of the individual's employment and the amount of the benefit is generally determined either as a flat amount or as a multiple or fraction of the individual's annual compensation. For example, these benefits may be paid for a period equal to the number of weeks that an individual is out of work (e.g., $500 a week for each week unemployed) up to a maximum number of weeks, or for a definite period of time regardless of the length of the period of unemployment. . . .

Notice 2007-14, 2007-7 I.R.B. 501 (Feb. 12, 2007) announced a project to issue proposed regulations on permissible benefits, with a focus on “nontraditional benefits that are not subject to the protections of §411 and other qualification rules of §401(a)”, including some forms of shutdown benefit. Whether this guidance will be narrower than G.C.M. 39869 is at present unknowable.

Page 151: In Sheet Metal Workers’ National Pension Fund, 117 T.C. 220, 27 EB Cases 1001 (2001), affirmed, 318 F.3d 599, 29 EB Cases 2377 (4th Cir., 2003), the Tax Court, reversing an unfavorable IRS determination letter, held that a multiemployer pension plan did not cut back accrued benefits in violation of ERISA, §204(g) and I.R.C., §411(d)(6) when it was amended to eliminate automatic cost-of-living increases for participants who had retired before the COLA was instituted. Until 1991, retirees had received COLA’s from a separate “COLA Fund”, a supplemental plan treated, under Department of Labor regulations, as a welfare rather than a pension plan. The COLA Fund’s target was annual three percent pension increases, but actual supplements were limited to available funds, which proved inadequate from 1985 onward. The main plan began granting ad hoc increases to make up the difference. In 1991 the trustees adopted an automatic two percent annual increase. In 1995, they concluded that the COLA was not a good use of the plan’s increasingly hard-pressed resources and eliminated it for pre-1991 retirees. In deference to the case law discussed in the main text, holding that an automatic COLA accrues along with the underlying benefit, they retained it for participants who retired after its institution. Though the facts are not entirely clear, it appears that pensions were reduced to their levels at the time of retirement, so that retirees lost past, as well as future, COLA increments.

Viewing this change as a prohibited cutback of accrued benefits, the IRS disqualified the plan, which brought an action for a declaration preserving its qualified status. The Tax Court thus had to decide whether a benefit increase granted to someone who is no longer working falls within the definition of “accrued benefit” (ERISA, §3(23); I.R.C. §411(a)(7)). (It quickly disposed of the frivolous alternative argument that a COLA is a “retirement-type subsidy” and thus protected from cutback even if it is not part of the accrued benefit.) The court’s analysis cannot be called razor sharp - in particular, it places far too much emphasis on the statute’s use of the phrase “employee’s accrued benefit” as excluding retired employees - but it stumbles to the correct result. The anti-cutback rule was enacted to prevent employers from taking away benefits earned during employment. The reasons underlying this protection do not extend to benefit increases granted after a participant has stopped performing services. The 1991 retirees temporarily received more than they had earned, but it is not one of the purposes of ERISA to protect that windfall.

(c) Irrevocability. The Company and any Adopting Employer shall have no right, title, or interest in the contributions made by it to the Trustee or Trustees, and, except as permitted by Section 15.4, no part of the Trust Fund shall revert to the Company or any Adopting Employer, except that after satisfaction of all liabilities of the Plan, any funds remaining in the Trust Fund as a result of overpayment may revert to the Company or Adopting Employer. In the event of a Change in Control, such reversion shall not take place and the provisions of Section 11 shall apply.

“Resolution” of this ambiguity [created by not entirely consistent plan provisions regarding reversions], the district court reasoned, “turns on the word ‘overpayment’ in Section 9.2(c). The district court defined the word as “payment in excess of what is due” . . . and found “that a return on an investment does not constitute an ‘overpayment.’” Because “the surplus assets in the Plan resulted from a high return on monies invested in the Plan, not from excess money contributed to the fund,” the district court concluded that the surplus did not result from “overpayment,” and therefore that Coleman was not entitled to the surplus. Applying the principle of contra proferentem, the district court construed any ambiguity it could not otherwise resolve against Coleman and declared that the participants are entitled to the Plan’s surplus.

Page 200: The wave of conversions of mutual insurance companies into stock companies has resulted in the receipt of demutualization dividends by many owners of annuity contracts purchased to close out terminated pension plans. (For a general discussion of demutualization, focused on welfare rather than pension plans, see Jeffrey W. Knapp and Judy C. Bauserman, “Demutualization: Coming to a Welfare Plan Near You”, 12 Benefits Law Journal 7 (Winter 1999).) Technically, these “dividends” represent distributions of the policy holders’ equity interests in the pre-conversion mutual company. Since, however, this new form of equity, unlike its predecessor, is a security separate from the insurance policy and can be sold for cash, the practical effect is to create what may be post-termination plan assets.

Where the terminated plan has distributed annuity contracts to participants in satisfaction of benefit liabilities, the demutualizing insurer will issue the dividends (normally in the form of company stock, though there may be an option to receive cash) directly to the individual policy holders. Since the former plan sponsor will have no control over - and perhaps no knowledge of - these distributions, they should be treated as involving only the insurer and the policy holders (who will recognize no income upon the receipt of stock and capital gain equal to the sales proceeds when they dispose of it). The fact that the policies were purchased in connection with a plan termination is a matter of mere historical interest.

The former situation has been the subject of two IRS rulings. PLR 200214031 treats demutualization dividends received after plan termination as residual plan assets. Although the facts are not set forth with perfect clarity, it appears that the plan was funded prior to termination with two group annuity contracts, whose issuer converted from a mutual into a stock company shortly before the plan terminated and distributed its assets. The insurer subsequently paid the dividend, in the form of stock, to the employer, which sold the shares, held the proceeds of sale in a segregated account on behalf of the plan and decided that it wished to use the new-found funds to increase benefits.

The IRS ruled that the dividend could be treated as plan assets for purposes of section 4980, even thought the amount involved had not been knowable at the date of plan termination, and could be used to fund additional benefit accruals. The employer would thus receive no reversion and would owe no excise tax, notwithstanding the fact that it had held the dividend for a time in an account subject to the claims of its creditors. PLR 200317049 reached the same result on closely similar facts.

The PLR’s do not address the practical aspects of a post-termination amendment to utilize the dividend for a benefit increase. For instance, was the plan required to file Form 5500 for the year in which the dividend was received, and did it have to be amended to comply with statutory and regulatory changes adopted since its original termination date? The most sensible course of action would be to treat the amendment increasing benefits as if it had been adopted and effective on the date of plan termination, particularly if a considerable period has passed since then.

An employer that wishes to reduce its reversion excise tax by contributing 25 percent of the dividend to a replacement plan should be careful to transfer the stock that it received from the insurer rather than sell the stock and transfer cash. Transferring the stock should not have any tax consequences, since it is clear from I.R.C., §4980(d)(2)(B)(iii) that the transfer is not treated as a contribution. A sale, by contrast, would produce taxable capital gain, not offset by any deduction for the transfer of the sales proceeds to the plan.

The second situation, a dividend on a close-out group annuity contract, is more controversial. The Department of Labor addressed it in ERISA Advisory Opinion 2003-05A (April 10, 2003), its only significant statement concerning demutualization dividends received by pension, rather than welfare, plans. The opinion responded to an inquiry from a plan sponsor that held a group annuity contract purchased by a plan that had terminated in 1990. In 2001 the insurer demutualized. The employer, upon receiving the demutualization dividend, deposited it into a separate account pending the DoL’s advice on whether it belonged to the employer or to the plan. The Department decided not to decide. In its view, the answer lay outside the scope of ERISA. The key passage in the opinion states:

If, as you represent, the Plan was properly terminated [footnote omitted] and all obligations and claims under the Plan were satisfied prior to the termination annuity contract provider’s demutualization, there is no obligation under Title I of ERISA to treat demutualization proceeds as plan assets. Therefore, no violation of Title I of ERISA would occur if Church takes possession of the proceeds. The question of whether the employer or the beneficiaries of the termination annuity contract are the actual owners of the demutualization proceeds received by the employer as the named policyholder of the annuity is not within the jurisdiction of the Department of Labor under Title I of ERISA. Rather, this issue is governed by the terms of the contract and applicable state law.

The absence of an “obligation under Title I of ERISA to treat demutualization proceeds as plan assets” means only that ERISA does not establish property rights. In other words, the ownership of the dividend depends upon “the terms of the contract and applicable state law”, a position consistent with what the Department has said elsewhere about rights to demutualization dividends received by welfare plans. See ERISA Advisory Opinions 2001-02A (Feb. 15, 2001) and 2005-08A (May 11, 2005), and Information Letter to Theodore R. Groom (February 15, 2001).

Being under no obligation to examine “the terms of the contract and applicable state law”, the DoL didn’t. That task was undertaken by the courts in Bank of New York v. Janowick, 470 F.3d 264, 39 EB Cases 1631 (6th Cir., Nov. 22, 2006). The National-Southwire Aluminum Company Pension Plan terminated in 1986 and was closed out by the purchase of two Prudential group annuity contracts. Sixteen years later, Prudential demutualized and issued a demutualization dividend of over 35,000 shares, worth $1.3 million, with respect to the contracts. In an interpleader action brought by the nominal policyholder, claims to the funds were advanced by the former plan participants and by two corporate successors to the former employer. A district court decided in favor of one of the corporations, but the court of appeals reversed, holding, in a split decision, that the money belonged to the participants (or, more precisely, to those participants who were entitled to benefits under the annuity contracts).

The majority’s starting point was ERISA Advisory Opinion 2003-5A. Under “the terms of the contract”, i. e., the Prudential group annuities, the dividend was to be paid to the contract holder, which nominally was the former trustee of the defunct pension plan. Since the plan no longer existed, and in fact had ceased to exist the moment that the contracts were purchased, the contractual terms could not be applied literally. There were three claimants to the position of successor contract holder: the plan sponsor’s former parent, the purchaser of the sponsor’s operations, and the plan participants entitled to pensions under the contracts. The court chose the last, essentially on the ground that the demutualization dividend had no connection with the pre-termination plan. The annuity contracts were purchased as a means of distributing plan assets, not as an investment. Hence, the dividend could not reasonably be considered a plan asset. The claims of both successor employers depended upon the succession of the property from the plan to its sponsor via a reversion on plan termination. With the first link in that chain eliminated, both were complete strangers to the res. The only claimants left were the individual participants.

The dissenting judge argued that the majority’s decision was tantamount to granting participants a right to surplus plan assets in contravention of Hughes Aircraft Co. v Jacobson, 525 U.S. 432, 22 EB Cases 2265 (1999). That objection, however, assumes the matter to be proved, namely, whether the dividend was a plan asset. If it wasn’t – and the absence of any connection between the dividend and the pre-termination plan makes that a plausible conclusion – it cannot have been a surplus asset. It was more like gains from the investment of funds distributed to participants upon plan termination and hence most naturally viewed as their property.

Page 202, fn. 127: A National Office memorandum to Assistant Regional Commissioners (Examination), dated February 18, 1986, but not publicly disclosed until 1999 (Tax Notes Today, 1999 TNT 85-27) stated that the IRS was “considering the effect, if any, upon plan qualification of transferring assets from defined benefit plans, including terminating plans, to defined contribution plans” and placed a hold on the issuance of determination letters concerning such transfers.

Pages 211-2: In 1999 Congress reversed course yet again, restoring the pre-1995 rule that the cost of retiree health benefits (rather than the benefits themselves) may not be cut back during the five-year period following a section 420 transfer. I.R.C., §420(c)(3). The new/old rule applies to transfers made on or after December 18, 1999. Ticket to Work and Work Incentives Improvement Act of 1999, Pub. L. 106-170, §535(b)(1). It requires that the employer’s average per-retiree expenditure on health benefits for each of the five tax years following a transfer remain at least equal to that in the higher of the two years immediately preceding the transfer. The averages may be calculated separately for Medicare-eligible and other employees.

To prevent employers from circumventing the goals of the maintenance-of-cost requirement, the IRS is authorized to “prescribe such regulations as may be necessary to prevent an employer who significantly reduces retiree health coverage during the cost maintenance period from being treated as satisfying the minimum cost requirement”, I.R.C., §420(c)(3)(E). Final regulations under this section were published on June 15, 2001, effective as of December 18, 1999. Treas. regs., §1.420-1. Their basic principle is that a 10 percent one-year decline, or a 20 percent cumulative decline over five years, in the number of retirees with employer-provided health coverage is “significant”. Only retirees, spouses and dependents who lose coverage due to employer action (both explicit amendments and less direct actions, such as the sale of assets to a buyer that doesn’t continue the plan) are included in the calculation. Not counted is loss of coverage due to expiration under the original terms of the plan (e. g., where benefits are provided for a fixed period following a plant closing).

The consequences of losing the protection of section 420 through violation of the cost maintenance requirement are not well defined but are bound to be severe. At the mildest, the employer will be taxed on the funds siphoned off from the pension plan. At the most drastic, it may also have to make repayment to the plan with interest, since the transfer will have retroactively turned into a prohibited transaction.

A 2004 amendment, included in the American Jobs Creation Act, allows employers whose retiree health expenses (on a cash basis) equal five percent or more of their gross receipts to make de minimis cost reductions. The maximum reduction allowed by this rule is the amount that would have been saved if the company had made the maximum coverage reduction permitted under the regulations. I.R.C., §420(c)(3)(E)(ii).

The Pension Protection Act of 2006 made a technical change to the definition of “excess pension assets”, revising it to reflect the new funding standards the become effective in 2008, I.R.C., §420(e)(2), and, more significantly, added an alternative to year-by-year transfers. Under the new provision, which became effective upon enactment (August 17, 2006), transfers under section 420 may equal up to ten years’ estimated benefit payments to retirees. The following rules apply:

Assets eligible for transfer. Only assets in excess of 120 percent of the plan’s current liability (sum of funding target [the text, in an obvious typo, reads “funding shortfall”] and target normal cost for years beginning in 2008 or later) may be transferred. I.R.C., §420(f)(2)(B)(i). Note that this is more liberal than the 125 percent floor for annual transfers. On the other hand, assets must be kept above the 120 percent floor throughout the period over which the transfer prefunds future benefits. If their value falls below the floor as of any valuation date, the employer must correct the deficiency by either making additional contributions or transferring funds back from the 401(h) account to the plan’s retirement assets (entailing a 50 percent excise tax under section 4980). I.R.C., §420(f)(2)(B)(ii).

Amount that may be transferred. The maximum transfer into the plan’s section 401(h) account is the estimated total reimbursements that will be paid of two to ten years, as elected by the employer, beginning with the year of the transfer. I.R.C., §§420(f)(2)(C) and 402(f)(5). Regulations will presumably specify that the future reimbursements must be reduced to present value.

Cost maintenance requirement. The cost maintenance period continues through the end of the fourth year following the end of the transfer period. The average per capita cost of retiree health benefits must be maintained for each year in that period, applying the same principles as for year-by-year transfers. I.R.C., §420(f)(2)(D)(i)(I). The employer may, however, elect to follow the repealed benefit maintenance rules (described in the main text at 211–12) instead of maintaining costs. I.R.C., §420(f)(2)(D)(ii). It is unlikely that the choice, once made, can be changed, though the statute is not explicit on that point.

Collectively bargained transfers. Special, generally more liberal, provisions govern multi-year transfers made pursuant to collective bargaining agreements, but they are available only on a very limited basis (to employers whose cash basis retiree health expenses in 2005 were at least five percent of gross receipts). Because of this narrow applicability, they are not summarized here.

Page 220: State courts are split on whether a reversion is business or non-business income for state corporate income tax purposes. Business income is typically apportioned among all states in which a corporation conducts business, while non-business income is taxed only by the state of domicile. The North Carolina Supreme Court held, in Union Carbide Corporation v. Offerman, 351 N.C. 310, 526 S.E.2d 167, 23 EB Cases 2998 (2000), that reversions are non-business income, because they arise from the investment of trust assets, not from business operations. The fact that the trust is established to fund a pension plan that is related to the corporation’s business does not alter the result, because the plan’s residual assets are not, by definition, needed for purposes of the plan.

The California Supreme Court reached the opposite conclusion in Hoechst Celanese Corporation v. Franchise Tax Board, 25 Cal. 4th 508; 22 P.3d 324 (2001), arguing that the investment of pension trust assets was “integral” to the operation of the plan. It did not explain in any clear fashion what that fact has to do with the accumulation of surplus assets that the plan does not use to provide benefits. A particularly questionable element of its reasoning was that it treated what is, from an economic point of view, investment income as taking on a “business” character simply because, while in the plan, it was earmarked for a particular business activity. By that standard, all investment income is business income, because it all is or can be used in the corporation’s business. The mere placing of investments in a special-purpose trust ought not to make them different from all other investments.

PLR 8812006, dealing with the tax consequences of a reversion to a social club (exempt under section 501(c)(7)) notes in passing the club’s concession that the entire amount received from the plan will be subject to excise tax, even though only a de minimis portion constituted unrelated business taxable income. (The investment income of section 501(c)(7) organizations is UBTI. The ruling held that the reversion was investment income to the extent that it exceeded the club’s contributions to the plan, which it did by a very modest amount.) The taxpayer had little reason to dispute that conclusion at a time when the excise tax rate was only 10 percent, though the extent to which it had “derived a tax benefit from the qualified plan” was questionable.

PLR 9622037, issued under section 4972 (the excise tax on nondeductible contributions, which incorporates section 4980’s exclusions by reference and has legislative history almost identical to that quoted above, S. Rep. No. 100-445 at 188 (1988)), takes a broad view of the effect of the receipt of UBTI on the exclusion from excise tax, ignoring completely the question of whether the employer has derived any tax benefit from its contributions. There the employer paid taxes on unrelated business taxable income but had apparently not deducted any plan contributions in computing UBTI. The IRS held that any contributions that it made in excess of what would be deductible by a taxable employer would be subject to excise tax under section 4972, basing that conclusion solely on the legislative history: “[T]he above-mentioned committee report is very specific in denying the exception in section 4972(d)(1)(B) to an otherwise tax-exempt organization that has been subject to unrelated business income taxes.” The words “to the extent that” and “otherwise derived a tax benefit” are given no consideration.

PLR 9304033 did pay attention to “to the extent that”. There a tax-exempt plan sponsor made quarterly pension contributions that it believed were required under section 412(m). The plan’s actuary subsequently determined that the maximum deductible contribution (applying the rules applicable to taxable employers) was zero. The sponsor had a small amount of UBTI ($27,809; total revenues for the year are not disclosed, but compensation paid to plan participants was approximately $150,000,000, so the UBTI was obviously trivial in relation to the size of the entity). In calculating UBTI, it did not deduct any contributions to the plan. It asked for a ruling that either (i) it was exempt from section 4972, because it had derived no tax benefit from its contributions or (ii) it could withdraw the “nondeductible” contributions in accordance with Revenue Procedure 89-35, 1989-1 C.B. 917.

The IRS agreed that the plan contributions had not resulted in any tax benefit to the employer. Nonetheless, it was subject to section 4972, because it “actually had unrelated business income”. The ruling then computed the extent of the benefit by dividing “the salaries and wages (attributable to employees involved in the generation of unrelated business income) which were taken as a deduction on the 1989 Form 990-T ($5317)” by the total compensation of all participants (about $150,000,000). The quotient was then multiplied by the amount of contributions in excess of what a taxable employer could have deducted ($227,730). The product ($8.07) was deemed to be de minimis. The IRS did not let the employer take its eight bucks back from the plan and waived the 81 cent excise tax.

The latest IRS ruling on the subject, PLR 200222035 (dealing with a potential reversion rather than nondeductible contributions), perhaps retreats from the positions taken in the two preceding rulings. There the employer had unrelated business income, though not enough to incur any tax, and had never deducted any pension contributions in computing UBTI. The ruling states:

Organizations described in 501(c)(3) of the Code generally are exempt from taxation and would derive no tax benefit from contributions to a qualified pension plan. However, an organization described in 501(c)(3) could receive a tax benefit from a contribution to a qualified plan if it deducted the amount of the contribution from its unrelated business taxable income.

The information submitted with the ruling request indicates that Companies A, B, and C have at all times been exempt from tax under 501(c)(3) of the Code, have never had unrelated business taxable income, and have at no time deducted any amounts contributed to the Plan from unrelated business taxable income. Based on this information, the proposed asset reversion received by Company B upon termination of the Plan will not be includible in income as unrelated business taxable income under 511 of the Code, and there will be no excise tax under 4980(a) of the Code.

The IRS’s analysis in these rulings suffers from two defects. First, it treats the legislative history of section 4980(c)(1)(A) as if it were incorporated into the statute. Reliance on legislative history is, of course, commonplace, but is highly dubious in this particular instance. The statutory language does not limit the exclusion from excise tax to tax-exempt employers that have not incurred UBTI, and there is no discernible rationale for doing so. An exempt organization whose investment income is tax-exempt gains no tax benefit from putting money into a qualified plan instead of retaining it in general assets. A strong argument can be made for the position that the statute means what it says: that reversions are tax-free for employers that have “at all times” been tax-exempt. Irrational legislative history is no basis for ignoring the plain meaning of the law.

Second, the IRS’s reading of the legislative history, at least in PLR’s 9304033 and 9622037, ignores the natural meaning of the phrase “has been subject to unrelated business income tax or has otherwise derived a tax benefit from the qualified plan” [emphasis added]. The writer apparently had in mind situations in which an employer “subject to unrelated business income tax . . . derived a tax benefit from the qualified plan”. Any other interpretation makes the word “otherwise” meaningless. Cf. PLR 8951066, which is too generalized to make the IRS’s reasoning clear but does conclude that an employer is not subject to the reversion excise tax where “it has not derived any tax benefit from its contributions”, with no discussion of whether it had ever been subject to the tax on unrelated business income.

The other interpretive question is the meaning of “at all times . . . exempt from tax under Subtitle A”. Rulings under both section 4980 and the parallel provisions of section 4972 indicate that the IRS regards this standard as met only if no taxable employer’s employees have ever accrued benefits under the plan. The most recent ruling to take that position is PLR 9837036. There an exempt organization’s plan covered employees of both taxable and tax-exempt affiliates. It requested a ruling that, upon plan termination, only those residual assets that reverted to the taxable employers (apparently a very small portion) would be subject to the section 4980 excise tax. The IRS held to the contrary:

Because employees of Corporations B and C participated in and received benefits under Plan X respectively from the period beginning June 1, 1979, and ending August 1, 1997, and from the period beginning July 1, 1987, and ending June 30, 1995, Corporations B and C are considered employers which maintained the plan. Because Corporations B and C are non-exempt organizations, Plan X has not, at all times, been maintained by an employer that has, at all times, been exempt from tax under Subtitle A.

Accordingly, with respect to the first ruling request, we conclude that Plan X is a “qualified plan” for purposes of section 4980(a) of the Code. Therefore, the Surplus attributable not only to contributions made by Corporations B and C but also the contributions made by Corporation A and its other affiliates is subject to the tax on reversion of qualified assets to the employer under section 4980.

In the instant case, the controlled group contains both tax-exempt and nonexempt employers, and the Plan maintained by members of the controlled group has reached the full funding limitation of Code section 412(c)(7). Under Code section 414(b), all employers in the controlled group are regarded as a single employer for purposes of Code section 404. In order for the controlled group to be considered a tax-exempt organization, each and every member of the controlled group must be a tax-exempt employer. Thus, because the controlled group contains at least one nonexempt employer, the entire controlled group is considered to be a nonexempt organization for purposes of section 404 and 4972. Because the controlled group is a nonexempt organization, the Plan to which members of the controlled group contribute is not entitled to the exemption from the tax under section 4972 for a "qualified employer plan" described in section 4972(d)(1)(B). These qualified employer plans are further described in section 4980(c)(1) as plans exempt from tax under Subtitle A (of the Internal Revenue Code) or governmental plans. Therefore, although there is no specific prohibition against a tax-exempt employer (which is a member of such a controlled group) making contributions to the Plan, any such contributions would be nondeductible and subject to the excise tax imposed under Code section 4972.

Both of these arguments take it for granted that all members of a controlled group are a single employer for purposes of sections 4980 and 4972. That is the only basis on which one can attribute the taxable employers’ status to the tax-exempt employers that also maintain the plan. If the employers are not combined into one, a plan that covers employees of both tax-exempt and taxable organizations is, literally, “maintained by an employer” that “has, at all times, been exempt from tax under Subtitle A”. It has also been maintained by other employers, but the text does not say that all of the employers must have been tax-exempt.

Sections 414(b), 414(c) and 414(t) specify the provisions for purposes of which all controlled group members are treated as a single employer; sections 4972 and 4980 are absent from that list (as is section 404 - the ruling quoted above misunderstands section 414(b), under which employers are aggregated only for the specific purpose of calculating the deduction limitations imposed by section 414). Hence, there is no basis for the IRS’s conclusion that all members of the controlled group must be tax-exempt in order for any of them to receive the benefit of the excise tax exemption.

It would be overreaching to turn the IRS’s stance on its head and assert that participation by any exempt employer to the most limited extent frees from excise tax reversions received by taxable entities, although that is a possible ultra-literal interpretation. Rather, the most reasonable approach is to treat the plan, for excise tax purposes, as if the portion attributable to each employer’s contributions were separately maintained. Hence, only the portion of the reversion attributable to taxable employers would be subject to excise tax.

Page 226: In PLR 199911058, the IRS held that, where a transfer of residual assets is made to a section 401(k) plan, all employees who are eligible to make elective deferrals under the replacement plan are considered “active participants” in that plan, whether or not any contributions are ever allocated to their accounts. Hence, they count toward satisfying the requirement that at least 95 percent of the employees who participated in the terminated defined benefit plan be active participants in the replacement plan. This ruling obviates the concern, expressed in the main text, that it might be necessary to make nonelective allocations in order to ensure that the replacement plan covered a sufficient proportion of prior plan participants.

Pages 230-1: A series of private letter rulings issued in 1998 held that, contrary to the statement in the main text, transfers to a replacement plan that exceed 25 percent of the terminated plan’s residual assets are exempt from income tax and may be allocated over a period of up to seven years, in the same manner as the qualifying 25 percent transfer. The portion in excess of 25 percent was, however, held to be subject to the 20 percent excise tax on reversions. PLR’s 9823051 (excise tax issue only), 9837036 (allocation issue only), 9839030 and 9839031. The letters presented no rationale for their holdings, which are difficult to reconcile with the language of the statute. After their issuance, about a dozen companies sought similar rulings, but the IRS National Office delayed responding to their applications, and there were reports that field offices had taken audit positions contrary to the rulings, with at least one case expected to go to litigation. It was thus not surprising that PLR’s 200227040 and 200227041 revoked PLR 9839030 and held that the portion of a transfer in excess of 25 percent of the terminated plan’s surplus fell outside the scope of section 4980(d). The rulings’ analysis follows that in our text: The excess transfer is treated as a taxable reversion to the employer followed by a contribution to the replacement plan, and the contribution is deductible only to the extent permitted by section 404(a). The rulings do not discuss the allocation of the transferred amount, but the only reasonable inference is that the excess must be allocated in the year of the transfer, like any ordinary contribution to an individual account plan, not spread over up to seven years under section 4980(d)(2)(C)(i)(II). See also PLR 200252094, which follows PLR’s 200227040 and 200227041.

The story did not end there, however. In Revenue Ruling 2003-85, 2003-32 I.R.B. 1, the I.R.S. reversed course again, reinstating the income tax positions taken in the 1998 PLR’s and, in an act of remarkable generosity, going on to hold that no portion of the amount transferred to the replacement plan is subject to the reversion excise tax. In the ruling's example, the reversion that the employer would have received $60x in a reversion if it had simply terminated its defined benefit plan without making any transfer to a replacement plan. Instead, it took $40x as a reversion and arranged for $20x to be transferred to the replacement plan. The IRS ruled that only $40x was taxable as a reversion or includible in the employer's taxable income. The ruling does not address the allocation of the transferred amount, but presumably the whole sum may be allocated over up to seven years, as in PLR 9837036.

The ruling opened up planning opportunities. Consider a company whose pension plan had a $10 million surplus and that also maintained a section 401(k) plan, to which it annually contributed $2 million to match participants’ elective deferrals. It could terminate the pension plan, transfer the full $10 million surplus to the 401(k) plan and be relieved for five years of the need to put up corporate cash for matching contributions. For many companies, that deal was almost as good as outright repeal of section 4980. Like many good deals, unfortunately, it did not last.

The revised regulations under section 401(k), effective for years beginning after December 31, 2005, prohibit funding matching contributions with amounts contributed to a plan before the related services were performed. Treas. Regs., §1.401(m)-1(a)(2)(iii). There are limited exceptions to allow the use of forfeitures and shares released from ESOP suspense accounts, but nothing that would let assets transferred under section 4980 offset future matching contributions. It remains possible to allocate the transferred amounts as nonelective, nonmatching contributions. Hence, the strategy discussed in the last paragraph would still be viable if the employer regularly made profit sharing contributions.

Page 250: In Hughes Aircraft Co. v. Jacobson, 525 U.S. 432, 22 EB Cases 2265 (1999), the Supreme Court, unanimously reversing the Ninth Circuit decision criticized in the main text, held that ERISA’s fiduciary standards rule do not confer on particular participants any interest in surplus assets of an ongoing plan. The most significant element of the Court’s opinion is its clear statement of the principle that ERISA’s prohibition against the inurement of plan assets for the benefit of the employer is not violated when assets are used for the benefit of employees, even if the plan design is motivated by the employer’s self-interest.

. . . respondents argue that Hughes - by effectively increasing certain employees’ wages through either providing increased retirement incentives or including those employees in the Plan's noncontributory structure - reduced its labor costs by spending down the Plan's surplus to cover its own obligations. In Spink [Lockheed Corporation v. Spink, 517 U.S. 882, 20 EB Cases 1257 (1996)], we noted in passing that if a transaction constituted “merely a sham transaction, meant to disguise an otherwise unlawful transfer of assets to a party in interest, . . . that might present a different question.” Id. at 895, n. 8. Even assuming that a sham transaction may implicate a fiduciary duty, the incidental benefits conferred upon Hughes when it amended the Plan are not impermissible under the statute. It is irrelevant whether Hughes received lower labor costs or other such incidental benefits from implementing the noncontributory structure under the 1991 amendment. As we noted in Spink:

“Among the ‘incidental’ and thus legitimate benefits that a plan sponsor may receive from the operation of a pension plan are attracting and retaining employees, paying deferred compensation, settling or avoiding strikes, providing increased compensation without increasing wages, increasing employee turnover, and reducing the likelihood of lawsuits by encouraging employees who would otherwise have been laid off to depart voluntarily.” Id. at 893-894 (citation omitted).

Receipt of these types of benefits no more constitutes a breach of fiduciary duties than they would constitute improper inurement or otherwise violate ERISA. To find that such benefits somehow violated the statute would forestall employers’ efforts to implement a pension plan. ERISA, by and large, is concerned with “ensuring that employees will not be left empty-handed once employers have guaranteed them certain benefits,” id., at 887, not with depriving employers of benefits incidental thereto. [525 U.S. at 445-6, 22 EB Cases at 2271-2]

Page 257: Two appellate court decisions have dealt serious blows to the effectiveness of the PBGC’s claims in bankruptcy. In re CF&I Fabricators of Utah, Inc., 150 F.3d 1293, 22 EB Cases 1481 (10th Cir., 1998), cert. den., 526 U.S. 1145, 23 EB Cases 1112 (1999); In re CSC Industries, Inc., 232 F.3d 505, 25 EB Cases 1550 (6th Cir., 2000), rehearing den., 2001 U.S. App. LEXIS 2086, cert. den., 122 S.Ct. 50 (2001). Both cases, reaching the substantially the same conclusions as the withdrawn Chateaugay opinion (discussed in the main text at 255ff.), rejected legal theories that the corporation had relied on to calculate - critics said inflate - the value of its claims and to assert their priority over claims of unsecured creditors.

The facts in the cases were broadly similar. A defined benefit plan sponsor declared bankruptcy and failed to make minimum funding contributions required by ERISA, §302, and I.R.C., §412. The plan was subsequently terminated, and the PBGC filed claims against the bankruptcy estate that included the value of the delinquent minimum contributions. The bankruptcy trustee disputed the PBGC’s positions regarding both the priority to which the minimum contribution claim was entitled and the valuation of claims arising from the sponsor’s liability for unfunded benefits.

The PBGC’s main argument for priority rested on ERISA, §302(f) and I.R.C., §412(n), which provide that a lien in favor of the plan arises if minimum required contributions are not made by their due dates and certain other conditions are satisfied. The amount of this lien is to be “treated as taxes due and owing the United States”. ERISA, §302(f)(4); I.R.C., §412(n)(4). Under the statute as in effect at the time of the litigation, the lien arose on the 60th day following the deadline for making the contributions. (The law has since been amended, so that the lien now arises immediately on the contribution due date.)

The PBGC reasoned that, because ERISA characterizes the contributions secured by the lien as taxes, the lien is entitled to the same status in bankruptcy as a federal tax lien, rendering it superior to unsecured claims. The bankruptcy trustees presented two counter-arguments: first, that the lien never arose, because the plan sponsors filed for bankruptcy less than 60 days after (indeed before) the contribution deadline; second, and more fundamentally, that delinquent contributions do not meet the Bankruptcy Code’s standards for characterization as a “tax” and hence receive no priority regardless of what ERISA says.

The CSC Industries court found the first argument sufficient and did not consider the District Court’s examination of “the interplay between the ERISA and Bankruptcy Code provisions involved in order to determine whether Congress intended the ‘tax’ in section 412(n)(4) to be given tax treatment in the bankruptcy context”. 232 F.3d at 510, 25 EB Cases at 1553-4. The contribution deadline fell after the date of the bankruptcy filing. The filing “trigger[ed] an automatic stay which prevented ‘any act to create, perfect, or enforce any lien against property of the estate.’ 11 U.S.C., §362(a)(4)”. Id. Hence, the lien on which the PBGC based its asserted priority did not exist.

The same issue could have been resolved in the same way in CF&I Fabricators, but the court relegated the existence of the lien to a footnote and instead addressed the substantive question of whether the “tax” treatment conferred by ERISA should be read into the Bankruptcy Code. Its conclusion was that the ERISA provision was ineffective, because (i) it had not been incorporated into bankruptcy law and (ii) minimum funding standards, unlike what the Bankruptcy Code means by “taxes”, are designed to serve the private interest of participants rather than public purposes. The analysis followed the lines enunciated by the Supreme Court in an earlier decision involving the same company, United States v. Reorganized CF&I Fabricators, Inc., 518 U.S. 213, 20 EB Cases 1289 (1996), which had held that the tax on minimum funding deficiencies imposed by I.R.C., §4971 was not a “tax” for bankruptcy purposes.

The PBGC also argued in CF&I that the delinquent contributions were entitled to administrative priority as a post-petition expense of preserving the bankruptcy estate. The court agreed with this position to the extent that contributions were attributable to post-petition services but rejected it for costs properly attributable to the pre-bankruptcy period. By the time of the CSC case, the PBGC had apparently abandoned this theory.

The most significant portion of the two courts’ decisions was their determination, based on slightly different rationales, that the present value of the PBGC’s claims should be computed using the anticipated rate of return that a prudent, long-term investor would anticipate rather than typically much lower rates prescribed by PBGC regulations. CSC illustrates the dramatic impact of rate differences. On PBGC assumptions (6.2 percent annual return for the first 20 years, 5.75 percent thereafter), the present value of the plan’s unfunded benefit liabilities was just under $50 million. Recalculating with the ten percent “prudent investor rate” reduced that figure by 96 percent, to about $1.8 million.

Page 264: PBGC Technical Update 00-3 describes that criteria that the corporation uses to identify transactions that, in its view, pose a threat to the pension insurance program. The PBGC’s standard tactic in such instances is to threaten to initiate involuntary termination proceedings under ERISA, §4042, unless the sponsor takes steps to reduce the risk. The credibility of this threat is not clear, but the corporation’s demands are usually modest, and the parties to the transaction almost always acquiesce rather than risk delay.

According to the update, the PBGC’s primary area of interest is sponsors whose PBGC-insured plans have total liabilities valued in excess of $25 million and where either (i) the sponsor has a less than investment grade bond rating or (ii) the plan has unfunded current liability of more than $5 million. The PBGC follows these companies, using publicly available sources, and takes action when it learns that one of them is contemplating a transaction that may “substantially weaken the financial support” for its plans. Examples include major divestitures of assets, spinoffs of subsidiaries, leveraged buyouts, extraordinary dividend payments and the substitution of secured for unsecured debt.

If it feels threatened, the PBGC will ask for further information. Anecdotal evidence suggests that it is rarely satisfied by what it hears. Its next step generally is to “seek to negotiate with the company to obtain protections for the pension insurance program in lieu of terminating the plan”. Settlements typically include additional contributions to the plan, the posting of letters of credit or pledging of corporate assets to ensure that future contributions will be made, or guarantees of future contributions by financially strong units that are leaving the plan sponsor's controlled group.

Page 265: In re UAL Corporation (Pilots’ Pension Plan Termination), 468 F.3d 444, 39 EB Cases 1129 (7th Cir., Oct. 25, 2006), rehearing den., 2006 U.S. App. LEXIS 28830 (Nov. 13, 2006) is the first Appeals Court decision to consider in depth the extent of judicial review of the PBGC’s decision to terminate a plan involuntarily under ERISA, §4042. District Courts had generally applied the highly deferential “arbitrary and capricious” standard to the Corporation’s actions in this area. In re Pan American World Airways, Inc. Cooperative Retirement Income Plan, 777 F.Supp. 1179, 14 EB Cases 1785 (S.D.N.Y., 1991), aff’d without opinion, 970 F.2d 896 (2d Cir., 1992); Pension Benefit Guaranty Corporation v. Fel Corporation, 798 F.Supp. 239, 15 EB Cases 2401 (D.N.J., 1992); Pension Benefit Guaranty Corporation v. Haberbush, 2000 U.S. Dist. LEXIS 22818, 25 EB Cases1481 (C.D.Cal., 2000); Pension Benefit Guaranty Corporation v. WHX Corporation, 2003 U.S. Dist. LEXIS 7526, 30 EB Cases 2567 (S.D.N.Y., 2003); Association of Flight Attendants – CWA v. Pension Benefit Guaranty Corporation, 2006 U.S. Dist. LEXIS 1318, 36 EB Cases 2233 (D.D.C., 2006); but see Pension Benefit Guaranty Corporation v. Rouge Steel Company, 2006 U.S. Dist. LEXIS 2685, 36 EB Cases 2882 (E.D.Mich., 2006).

Superficially, UAL was a PBGC victory. The court upheld the involuntary termination of the United Airlines pilots’ pension plan against objections by both the company and the pilots’ union. That result saved the pension insurance fund nearly $100 million. Unfortunately, from the PBGC’s point of view, the way in which the decision was reached draws many teeth from the “Early Warning Program”, which has long been the agency’s prime instrument for extracting concessions from sponsors of ongoing plans.

The PBGC brought an action to terminate the pilots’ plan involuntarily, because UAL had agreed with the union not to initiate a distress termination before June 2005. The PBGC preferred termination at the end of 2004, to preclude additional benefit accrual, a scheduled cost-of-living adjustment and the annual increase in the maximum PBGC-guaranteed benefit. The district court overseeing the UAL bankruptcy agreed that, if the plan did not terminate in 2004, “the possible long-run loss of the corporation with respect to the plan may reasonably be expected to increase unreasonably”, meeting the standard for a §4042 termination. 436 F.Supp. 2d 909, 38 EB Cases 1321 (N.D.Ill., 2006). The PBGC liked that outcome but did not like the fact that the court conducted a trial on the merits, instead of considering only whether the agency’s action was “arbitrary or capricious”. The upshot was a highly unusual appeal by a successful litigant. As the appeals court described the action,

Of the four appeals, the PBGC's has the distinction of not seeking any relief – for the PBGC prevailed in the district court. . . . The PBGC’s nose is out of joint because the court held a trial and made its own judgment about how much extra it would have cost to keep the plan in force until the end of June 2005, and whether that amount (which the court fixed at $84.2 million, about $17 million less than the PBGC’s calculation) would be an “unreasonable increase” in federal liability. The PBGC does not want to go through such a procedure again and asks us to hold that, instead of conducting an independent inquiry, the court should have limited review to the administrative record and deferred to the PBGC’s evaluation. [468 F.3d at 449]

The only authority that the PBGC has under [§4042] is to ask a court for relief. That implies an independent judicial role [citations omitted]. All the PBGC had done is commence litigation, and its position is no more entitled to control than is the view of the Antitrust Division when the Department of Justice files suit under the Sherman Act [citations omitted]. As the plaintiff, a federal agency bears the same burden of persuasion as any other litigant. [Id. at 449–50]

The PBGC’s counterargument was that the Supreme Court, in PBGC v. LTV Corporation, 496 U.S. 633, 12 EB Cases 1593 (1990), held that a decision to restore a plan under ERISA, §4047 was entitled to a high degree of deference under the “Chevron doctrine”. Section 4042, it argued, should be treated the same way. The difference, as the 7th Circuit saw it, was that section 4047 –

enables the PBGC to make self-executing orders, which is what leads to deferential review under the APA. Section [4042], by contrast, requires the PBGC to initiate litigation. Review under the APA differs substantially from the sort of position that an agency must assume when, like any other litigant, it must demonstrate a preponderance of the evidence in order to prevail. [Id. at 450].

small in relation to the total obligations of the plan (several billion dollars), the assets in the PBGC’s insurance fund, the cost of an aircraft carrier, or the national government’s annual budget. . . . [it] is substantial by any normal calculation. The district court concluded that the right question is whether the federal Treasury receives value for money; if not, the marginal outlay is “unreasonable.” This $84 million would not buy the insurance fund anything of value. Nor was it necessary to provide the pilots with a minimally satisfactory retirement: they are well compensated even after the termination, since many will receive benefits at the statutory maximum.

The incremental $84 million was a bargaining chip between United and the Union: to obtain ALPA’s assent to the termination, United offered the pilots an extra six months of benefits to be underwritten by a third party, the PBGC. No wonder it objected. The deal between United and the unions exemplifies the moral hazard to which insurance gives rise. Insured parties alter their behavior to take advantage of the third-party payor; insurers must respond by making such maneuvers more difficult. That’s why the PBGC proposed early termination. The district court did not abuse its discretion in concluding that any extra outlay attributable to the moral hazard created by insurance would be an “unreasonable increase” in the federal commitment, given the substantial amount that the PBGC already was committed to pay toward the pilots’ benefits. [Id. at 452]

The negative implications of this decision for the Early Warning Program are clear. The PBGC typically acts under EWP when it perceives that a corporate transaction will lead to an underfunded plan’s transfer to a financially weaker sponsor. Its bludgeon for persuading the parties to improve the plan’s funding or otherwise protect the insurance fund is the threat of a §4042 termination. Now that bludgeon is less wieldy. If the targets of EWP don’t cave in, the PBGC will face the task, daunting in many cases, of proving to a judge that the expectation of unreasonable losses is in fact reasonable. In the past, most practitioners have cautiously assumed that the PBGC possessed the great advantage of Chevron deference, and no plan sponsor has been bold enough to test that position in court. In light of UAL, those dynamics may well change.

Page 265: In October 1999, the Pension and Welfare Benefits Administration initiated the “Orphan Plans Project”, with the goal of identifying plans that have been “abandoned” by their sponsors and ensuring that participants receive their accrued benefits. As of February 2002 the project had reportedly opened 440 cases, including eight that led to criminal prosecution. Michael W. Wyand, “Strategic Enforcement Leverages Resources for PWBA, Enforcement Director Smith Says”, 29 BNA Pension & Benefits Reporter at 394 (Feb. 5, 2002). See also Michael W. Wyand, “PWBA Developing ‘White Knight’ Guidelines to Assist Fiduciaries; Asks for Information”, 29 BNA Pension & Benefits Reporter at 1996 (July 23, 2002).

Page 293: In an aberrant decision on which no prudent practitioner will place much reliance, even within its home territory, the Sixth Circuit Court of Appeals has ruled that participants are not entitled to full vesting on plan termination where, owing to the dissolution of the plan sponsor, there is no prospect that they might someday be rehired and meet the plan’s vesting requirements. Borda v. Hardy, Lewis, Pollard & Page, P.C., 130 F.3d 1062, 21 EB Cases 2842 (6th Cir., 1998).

The court’s theory was that I.R.C., §411(d)(3), which requires full vesting on the termination or partial termination of a plan, applies only to “affected participants” and that a nonvested participant is not “affected” by plan termination unless there is a possibility of returning to employment and gaining additional service credit toward vesting. A review of the history of the section shows that this superficially plausible notion is wrong.

A trust shall not constitute a qualified trust under this section unless the plan of which such trust is a part provides, that, upon its termination or upon complete discontinuance of contributions under such plan, the rights of all employees to benefits accrued to the date of such termination or discontinuance, to the extent then funded, or the amounts credited to the employees’ accounts are nonforfeitable.

The regulations extended the nonforfeitability requirement to employees affected by the “partial termination of a plan”, that is, those who lost their jobs in connection with the partial termination. Treas. regs., §1.401-6(b)(2) (1963), superseded by Treas. regs., §1.411(d)-2(b). (Partial terminations are discussed in Chapter 14 of the main text.) Section 411(d)(3) elevated partial terminations to the statute and used the phrase “affected participants” to make it clear that those who remain employed following an event that gives rise to a partial termination do not get the benefit of full vesting. Neither common sense nor the legislative history suggests that Congress meant to alter the prior rule that all participants must become fully vested when a plan actually terminates.

As the court’s opinion observes, section 411(d)(3) affects only the manner in which funds will be allocated among plan participants; there is no possibility that any money will revert to the employer. The choice is between letting each participant, vested or not, have the account balance that has been accumulated for his benefit or redistributing balances from nonvested to vested participants. Either alternative would be rational, but the former is simpler and precludes windfalls to the vested group. If the Sixth Circuit’s view takes root, one can imagine fascinating cases turning on a careful weighing of the plan sponsor’s future business prospects, but courts have quite enough to do already.

Pages 297-301: Revenue Ruling 2000-27, 2000-1 C.B. 1016, significantly broadened the circumstances under which distributions could be made to section 401(k) plan participants who moved to employment with a company that purchased assets of the plan sponsor. EGTRRA, §646 then took liberalization further by repealing the application of the “same desk” rule to section 401(k) plan distributions and allowing distributions upon any “severance from employment”, a term that the IRS defines quite differently from “separation from service”. The concept is explained in GCM 39824 (Aug. 15, 1990), dealing with the prohibition against distributions by pension plans to participants who have not had a severance from employment. The GCM states that participants employed by a business unit whose assets are sold to a third party are considered to have severed from employment with the plan sponsor, regardless of whether they have a separation from service under the same-desk rule. Notice 2002-4, 2002-2 I.R.B. 298 (Jan. 14, 2002) elaborates slightly on how the IRS will apply that principle in the section 401(k) context:

Under Code § 401(k)(2)(B)(i)(I), as amended by § 646 of EGTRRA, amounts attributable to elective contributions may be distributed upon the employee's severance from employment with the employer maintaining the plan. For this purpose, the employer includes all corporations and other entities treated as the same employer under Code §414(b), (c), (m), or (o). An employee does not have a severance from employment if, in connection with a change of employment, the employee's new employer maintains the section 401(k) plan with respect to the employee (for example, by assuming sponsorship of the plan or by accepting a transfer of plan assets and liabilities (within the meaning of Code § 414(l)) with respect to the employee). Thus, for example, if all employees of a controlled group of corporations (within the meaning of § 414(b)) are covered by a section 401(k) plan and a transaction occurs such that one subsidiary corporation in the group is no longer aggregated with other members in the group under § 414(b), (c), (m), or (o), and in connection with the transaction no assets are transferred from the section 401(k) plan to a plan maintained by the former subsidiary corporation, then, participants in the section 401(k) plan who continue employment with the subsidiary corporation will have a severance from employment with the employer maintaining the section 401(k) plan and may receive a distribution of amounts attributable to elective contributions from that plan. However, if the subsidiary corporation maintained a section 401(k) plan for its employees before the transaction and continues to maintain the section 401(k) plan following the transaction, the employees who continue employment with the subsidiary do not have a severance from employment with the employer maintaining the plan.

Page 304: PLR 200034039 reverses the IRS’s position on the section 415 treatment of allocations of residual funds that remain in an ESOP suspense account after the plan’s loan has been paid off. The ruling, whose release had long been predicted in the trade press, holds that ESOP participants have no section 415 annual additions when the plan repays its loans by selling stock held in its suspense account, allocates to participants’ accounts the shares that were not needed to repay the loan, and then terminates. There is no reason to think that selling all of the stock and allocating the surplus cash proceeds would change the result. Private letter rulings are not, of course, precedents for other taxpayers, and similarly situated plan sponsors are probably best advised to obtain their own rulings, lest the IRS change its mind once again.

Page 308, first paragraph: The Pension Protection Act of 2006, P.L. 109-280, §410(a) added a new section 4050(d) to ERISA, under which individual account plans and defined benefit plans of professional service employers with no more than 25 active participants will be permitted, though not required, to transfer the account balances or accrued benefits of missing participants to the PBGC at the time of plan termination. If the PBGC later finds the participants, it will pay their benefits in a lump sum or in whatever other form its regulations specify. The PBGC is also authorized to require individual account plans and small professional service employer defined benefit plans to furnish information on how missing participants’ benefits will be provided.

The expanded PBGC missing participant will not become effective until the corporation adopts implementing regulations. P.L. 109-280, §410(c). In the meantime, the Department of Labor’s Field Assistance Bulletin 2004-02 (Sept. 30, 2004) gives some guidance on what kind of search plan administrators must conduct in order to meet their fiduciary obligations and what action they should take if the search fails.

The FAB quite sensibly states that the effort put into searching for a participant should depend on the size of his account balance. Relatively easy and inexpensive steps – checking plan records, asking the participant’s beneficiary for an address, sending a notice by certified mail and using either the IRS or SSA letter forwarding program (described in the main text) – should be taken in all cases. More costly and effective methods should be utilized if the amount involved justifies them. Specifically mentioned are “Internet search tools, commercial locator services, and credit reporting agencies”. The inclusion of the first in the “advanced” category tells us something about the agency’s technological sophistication. (Hasn’t anybody there heard of Google?) If finding someone costs more than a nominal sum, the plan is permitted to charge it back to his account.

Once a reasonably diligent search has failed, the plan administrator must figure out what to do with the missing distributee’s balance. The FAB suggests rollover to an IRA set up on the participant’s behalf, transfer to a taxable bank account or transfer to a state unclaimed property fund. If the first option is chosen, the plan administrator can discharge its fiduciary duties by following the safe harbor regulations for automatic rollovers, 29 C.F.R., §2550.404a-2. The other two choices are permissible but seem clearly inferior in almost all cases. The FAB specifically disapproves an idea that some practitioners have advocated: remitting the full account balance to the IRS as income tax withholding. The flaw in that strategy is that the participant will get no benefit unless he files an income tax return before the statute of limitations expires, and many missing participants are dead or otherwise no longer part of the U.S. tax system.

Page 310: Revenue Ruling 2000-36, 2000-2 C.B. 140, approved a plan provision under which lump sum distributions of $5,000 or less would be transferred to an individual retirement account via direct rollover if the participant failed to give other directions. The sponsor assumed responsibility for selecting the trustee and taking other actions necessary to set up the account. The ruling took no notice of the 1987 proposed regulation (referred to in footnote 74 of the main text) that allowed rollovers only at the direction of the participant.

EGTRRA, §657, amending I.R.C., §401(a)(31), makes direct rollover to an IRA mandatory for all distributions in excess of $1,000, unless the participant elects otherwise. This procedure eventually will have to be used for all missing participants who are entitled to more than $1,000 from an individual account plan but will not become compulsory until the Department of Labor issues regulations on how ERISA’s fiduciary standards apply to the employer’s role in establishing the IRA.

Page 312, addition: A surprising number of qualified and section 403(b) plans have no active plan sponsor. The employer that used to maintain them has vanished for one reason or another (bankruptcy, dissolution, death, disability or imprisonment of the company’s principals, etc.), and the plan’s assets are left in the unsupervised custody of a trustee or insurance company. In 1999, the Department of Labor initiated a National Enforcement Project on Orphan Plans, which tries to identify abandoned plans and oversee their termination. As of early 2006, more than 1,000 plans, with a total of 50,000 participants, had been closed out via NEPOP, with another 400 cases in progress.

In April 2006, after years of dealing with abandoned plans on an ad hoc basis, the Department of Labor issued regulations and a prohibited transaction class exemption, under which the holders of the plans’ assets could make distributions and otherwise wind up plan affairs through a simplified termination procedure that would minimize out-of-pocket costs and exposure to fiduciary liability. 29 C.F.R., §2520.103-13, 71 Fed. Reg. 20819, 20853; 29 C.F.R., §2550.404a-3, 71 Fed. Reg. 20819, 20850; 29 C.F.R., §2578.1, 71 Fed. Reg. 20837; PTCE 2006-06, 71 Fed. Reg. 20855 (Apr. 21, 2006). The procedure is limited to individual account plans; the PBGC disposes of abandoned Title IV plans through involuntary termination under section 4042, and non-Title I plans do not fall within the Department of Labor’s jurisdiction. NEPOP remains open for Title I, non-Title IV plans for individual account plans whose circumstances are extraordinarily complex. The preamble to the regulations mentions plans with significant illiquid assets as an example of the latter.

After ascertaining that the employer responsible for the plan cannot be located or is unable to carry out its responsibilities, an entity may become a QTA by filing a Notification of Plan Abandonment and Intent to Serve as Qualified Termination Administrator with the DoL. A plan may have only a single QTA. If its assets are held by several trustees or insurers, they will have to decide among themselves which takes the lead. Service in this capacity will be a burden, which the DoL has tried to mitigate by providing that all reasonable expenses of winding up the plan may be paid from plan assets and that the QTA will be protected from fiduciary liability so long as it follows the regulations and monitors the activities of any service providers that it hires. The class exemption allows it to hire itself or its affiliates.

The QTA’s key duties are to assemble plan records, calculate participants’ account balances, notify participants and beneficiaries of the termination, distribute plan assets, inform the DoL when the asset distribution has been completed, and file a simplified final Form 5500. It does not have to amend plan documents (not even to conform to changes in qualification requirements since the plan’s last update), investigate whether fiduciary violations occurred in the past, collect delinquent contributions, let participants elect any form of benefit other than a lump sum (unless the plan is unluckily subject to qualified joint-and-survivor annuity requirements), apply for an IRS determination letter on plan termination, or file any Form 5500’s other than the final report. Also, the IRS has promised not to disqualify plans terminated in accordance with the regulations.

Distributions to missing participants may be made through rollover to a QTA-established IRA, regardless of the account balance, subject to substantially the same safe-harbor rules as for employer-established IRA’s. A 2007 amendment to 29 C.F.R., §2550.404a-3, 72 Fed. Reg. 7516 (Feb. 15, 2007) requires that distributions to nonspouse beneficiaries of deceased participants be rolled into individual retirement accounts, as permitted by new I.R.C., §402(c)(11). See also Proposed Amendment to PTCE 2006-06, 72 Fed. Reg. 7461 (Feb. 15, 2007). Distributions of less than $1,000 may be made to a taxable bank account or a state unclaimed property fund if setting up an IRA is not practicable. Unfortunately, the QJSA rules must be followed for money purchase pension plans and plans that hold balances transferred from money purchase plans. In those cases, participants must be given the choice of receiving an insurance company annuity rather than a lump sum. Those who cannot be located will be unable to waive the annuity, which means that the QTA will have to go to the trouble of arranging an annuity purchase (probably at a highly uneconomical price) on their behalf.

The procedure that the DoL has devised is about as simple and painless as it could be, which is not to say that the QTA role will be attractive. The chief motive for undertaking it will be to remove the risk that continuing to hold an abandoned plan’s assets without operating the plan could lead to negative publicity or conceivably fiduciary liability. Basically, though, QTA’s will have to think of themselves as Good Samaritans and anticipate little material reward for their endeavors.

Pages 334-5: In re Unisys Savings Plan Litigation, 173 F.3d 145, 22 EB Cases 2945 (3d Cir.), cert. den. sub nom. Meinhardt v. Unisys Corp., 528 U.S. 950, 23 EB Cases 1952 (1999), affirmed a district court decision in favor of the fiduciaries in 12 consolidated lawsuits arising from the purchase of three Executive Life GIC’s by Unisys Corporation’s individual account plans. The lower court had originally granted summary judgment in favor of the defendants but was reversed on appeal on the ground that issues of material fact existed, 74 F.3d 420, 19 EB Cases 2393 (3d Cir.), cert. den., 519 U.S. 810, 20 EB Cases 1872 (1996). After a trial on the merits, the district court found both that the defendants had followed prudent procedures when they chose to invest plan assets in Executive Life products and that a hypothetical prudent fiduciary would have made the same investments. The majority opinion of the Court of Appeals (one judge dissented on an issue involving the admissibility of expert testimony) consists essentially of a summary of how the plan’s investment decisions were reached, from which their prudence is not left in much doubt. To have found for the plaintiffs would have been tantamount to turning fiduciaries into insurers of plan investments.

Bussian v. RJR Nabisco Incorporated, 223 F.3d 286, 25 EB Cases 1120 (5th Cir., 2000) refused to adopt the Department of Labor’s “safest annuity” standard but did state that ERISA requires “a thorough, impartial investigation of which provider or providers best served the interests of participants” and held that there was a material issue of fact as whether plan fiduciaries had fulfilled that duty.

In 1988 RJR Nabisco terminated the overfunded pension plan of a subsidiary that it had sold in 1984. The plan purchased close-out annuity contracts from Executive Life. After that company’s collapse, participants brought suit, alleging the that selection of an annuity provider violated ERISA’s fiduciary standards. The District Court examined RJR’s selection process in detail and concluded that it had been prudent. 1998 U.S. Dist. LEXIS 14338. According to the trial judge, ruling on the defendant’s motion for summary judgment,

After seven months of work, Buck [RJR Nabisco’s consultant on the plan termination] concluded that Executive was a sensible candidate for the annuity. In its investment strategy, Executive was not reasonably discernable as imprudent. Nabisco had no knowledge that the particular portfolio of investments held by Executive would be too risky - to volatile, more precisely. Although some people in Nabisco management admit that they knew about Executive’s bad publicity, giving full credit to the beneficiaries’ conspiratorial view requires Nabisco to have known that Executive was suicidal in its investment strategy.

The judge rejected the “safest annuity” standard and criticized the DoL for trying to impose it after the fact. “It is one thing to scrutinize conduct with the benefit of hindsight; it is altogether another to impose retroactively standards that arose in direct response to that very conduct.” Overall, he summarized the plaintiffs’ arguments as, “Nabisco could not reasonably rely on anything from anybody” and “If a lot a money was lost, somebody must be at fault”.

The Appeals Court, while expressing itself less colorfully, agreed that the DoL’s standard was the wrong one. After determining that it was not required to defer to Interpretive Bulletin 95-1 under the Chevron doctrine (Chevron U.S.A., Inc. v. Natural Resources Defense Council, 467 U.S. 837 (1984)), it concluded that the Department had offered no sound basis for devising a special fiduciary rule for annuity purchases.

. . . we are not persuaded that [ERISA, §404(a)] imposes on fiduciaries the obligation to purchase the “safest available annuity” in order to fulfill their fiduciary duties. We hold that the proper standard to be applied to this case is the standard applicable in other situations that involve the potential for conflicting interests: fiduciaries act consistently with ERISA’s obligations if “their decisions [are] made with an eye single to the interests of the participants and beneficiaries”. [Donovan v.] Bierwirth, 680 F.2d [263,] 271[, 3 EB Cases 1417, 1425 (2d Cir.), cert. den. 459 U.S. 1069, 3 EB Cases 2490 (1982)] . . . . That standard does not require that a fiduciary under the circumstances of this case purchase the “safest available annuity”. [223 F.3d at 298, 25 EB Cases at 1129]

The court did not elaborate on the details of its preferred standard but did hold that there were material issues of fact as to whether RJR had met it, precluding summary judgment. The evidence that might lead a reasonable fact finder to conclude that the plan fiduciaries had not acted with “an eye single to the interests of the participants and beneficiaries” consisted of instances in which persons involved in the selection process arguably gave priority to obtaining the lowest possible annuity bid or ignored data unfavorable to Executive Life. For example, Executive Life was not on the initial list of companies to receive the request for proposals. It was added in the expressed hope that its “inclusion would facilitate bringing other bidders down in price because it would come in with a lower quote”. The lesson appears to be that making a final selection based on price is permissible, so long as all reasonable and necessary steps have been taken to ensure that all of the finalists will be able to meet their obligations to participants.

Page 362, after the first full paragraph: In defined contribution plans, the sole effect of a partial termination is that affected participants must be fully vested in their account balances. In Bennett v. Conrail Matched Savings Plan Administrative Committee, 168 F.3d 671, 22 EB Cases 2717 (3d Cir.), cert. den., 521 U.S. 871, 23 EB Cases 1856 (1999) and Rummel v. Consolidated Freightways, Inc., 1992 U.S. Dist. LEXIS 15144 (N.D. Cal., 1992), courts rejected arguments that a partial termination of an ESOP requires the allocation of all stock in the plan’s unallocated suspense account. Since the unallocated shares are not part of anyone’s accrued benefit, section 411(d)(3) does not have any effect on their allocation. The Rummel court did suggest in dictum that a suspense account established to hold assets transferred from a terminated defined benefit pension plan to a replacement plan under section 4980(d) would have to be allocated following a partial termination of the replacement plan in order to comply with I.R.C., §4980(d)(2)(C)(iv). It was wrong on that point, however: The cited provision requires full allocation is required only on the termination of the replacement plan, and a partial termination, despite the name, is not a termination.

In Burstein v. Retirement Account Plan for Employees of Allegheny Health Education and Research Foundation, 2002 U.S. Dist. LEXIS 22290, 28 EB Cases 2800 (E.D.Pa., 2002), the court rejected the plaintiffs’ attempt to force the PBGC to guarantee unvested benefits that were unfunded at the time of a partial termination. The partial termination was followed by a distress termination and the PBGC’s assumption of the plan’s assets and liabilities. (The court’s recitation of facts is slightly confused, suggesting that the termination preceded the partial termination, which of course is impossible.) The plaintiffs were participants who had not satisfied the plan’s five-year cliff vesting requirement as of either the partial termination or the date of plan termination. They based their claim on a minor inaccuracy in the summary plan description, which stated flatly that, “if the plan is terminated you will automatically become vested in your account, regardless of how many years of service you earned.” The plan document distinguished between termination and partial termination, stating, with regard to the latter, “Upon ... partial termination of the Plan, the rights of all affected participants accrued to the date of such ... partial termination shall become nonforfeitable ... to the extent funded as of such date.”

No one disputed the fact that the plaintiffs’ benefits were unfunded at the time of the partial termination. They claimed, nevertheless, that they were entitled to PBGC benefit guarantees, because the language of the SPD gave them full vesting upon partial termination regardless of their benefits’ funded status. Being vested, the benefits were guaranteeable by the PBGC when the plan terminated.

The court accepted sub silentio the premise that the SPD’s statement about what would happen “if the plan is terminated” embraced partial terminations, too. It could have refuted the plaintiffs’ case quite adequately by pointing out that a partial termination is not, pace the terminology, the same as, or even very much like, a termination. No ordinary reader (the ostensible audience of the SPD) is likely to think that information about what will happen if the plan goes out of existence applies to a significant reduction in the number of plan participants. Since the SPD was silent about the consequences of partial termination, the first step in the plaintiffs’ reasoning was fallacious.

Instead of taking that approach, the court assumed that the SPD promised full vesting of unfunded benefits upon the plan’s partial termination, then asked whether that promise bound the PBGC, concluding that it didn’t. The opinion follows those authorities that give SPD language preference over conflicting statements in the plan document only if participants can show that the plan administrator is equitably estopped from disregarding the SPD. It is unlikely in this case that equitable estoppel existed. The alleged misinformation in the SPD was not of a kind that would affect any participant’s conduct. In any event, the court held that the PBGC, as a government agency, is immune to claims based on equitable estoppel, particularly when the basis of the claim is the act or omission of an unrelated party. Whatever the price may be of careless SPD draftsmanship, the PBGC’s premium payers are not the right parties to bear it.

On appeal, the Third Circuit issued a confused and incompetent opinion holding that the SPD language did indeed result in full vesting of unfunded benefits upon the plan’s partial termination but agreeing with the lower court that those benefits were not eligible for PBGC guarantees. 334 F.3d 365, 30 EB Cases 2121 (3d Cir., 2003). Among other strange notions strewn along the way, the court asserted that ERISA’s minimum funding standards do not apply to nonvested benefits, that satisfying minimum funding requirements is a condition for plan qualification, and that a partial termination of a plan is a procedural step toward a complete termination. It also seemed unaware that its conclusion that the claimed benefits were not PBGC-guaranteed rendered its discussion of the SPD meaningless, as there is no source from which the supposedly vested benefits can be paid.

Page 378: Matz v. Household International Tax Reduction Investment Plan, 1999 U.S. Dist. LEXIS 14842, 23 EB Cases 2603 (N.D.Ill., 1999), aff’d, 227 F.3d 971, 24 EB Cases 2825 (7th Cir., 2000) rejected the plan’s assertion that several years cannot be aggregated for the purpose of determining whether there has been a significant reduction in participation. The District Court’s opinion concentrates on trying to divine whether the IRS, in its administrative practice, does or does not combine different years. Curiously, no one drew the court’s attention to the Great Atlantic & Pacific Tea Co. technical advice memorandum discussed in the main text (pp. 377-8), in which the IRS aggregated a period of seven years in order to reach the conclusion that a partial termination had occurred.

Page 382: In Administrative Committee of the Sea Ray Employees’ Stock Ownership and Profit Sharing Plan v. Robinson, 164 F.3d 981, 22 EB Cases 2513 (6th Cir., 1999), the employer, a boat manufacturer, suffered a downturn in business beginning in 1989. The slump was aggravated by the enactment of a “luxury tax” on pleasure boat purchases. In the course of two years, layoffs reduced participation in the company’s profit sharing plan by about one-third. The plan administrative committee, acting preemptively, conducted an administrative review of whether the plan had undergone a partial termination, determined that it had not and filed suit against a class of nonvested and partially vested terminated participants for a declaratory judgment to that effect. The District Court granted its motion for summary judgment, and the Sixth Circuit affirmed.

The two most important points of contention were (i) whether the decline should be considered a single event or as two distinct events (viz., the economic downturn in 1989 and the luxury tax imposed in 1990) and (ii) whether participants who voluntarily left the company in anticipation of layoffs should count as part of the reduction in participation. (All parties apparently agreed that purely voluntary terminations did not count.)

Regarding terminations in anticipation of layoffs, the court took the view that an employee who left voluntarily should be counted as affected by the hypothesized partial termination only if he had been “constructively discharged”, a stringent standard whose suitability in this context was left unexplained. Constructive discharge (conditions “so difficult or unpleasant that a reasonable person in the employee’s shoes would have felt compelled to resign”) is a concept derived from civil rights law. It is not obvious that the existence of partial termination should turn on the extent to which an employer accomplishes a significant reduction in force through forced attrition rather than overt layoffs. In Sea Ray, however, it does not appear that a different holding on this point would have altered the outcome of the case.

Applying the preceding principles, the court found that plan participation had declined by 15.9 percent in 1989/90 and by 27.9 percent in 1990/91, neither of which, it declared, was dispositive on its own. It therefore proceeded, following the lead of its decision in Kreis v. Townley [discussed in the main text at 375-6] to consider “two additional factors to determine whether a partial termination had occurred: one, the effects on the plan of the exclusion of employees from participation; two, the employer’s motives”. Since the reduction in participation had not “financially impaired the Plan” (a meaningless notion in an individual account plan) and the employer had not “initiated the layoffs to generate profit from the forfeitures from non-vested employees, either for itself or a third party”, neither of the “additional factors” weighed in favor of a partial termination. It is difficult to see how they could ever be usefully employed to demonstrate the presence or absence of a partial termination.

As discussed in the main text (pp. 371-3), the Second Circuit briefly adopted, then repudiated, the view that vested participants should be disregarded in determining whether there has been a significant reduction in plan participation. Weil v. Retirement Plan Admin. Comm. of the Terson Co., 933 F.2d 106, 13 EB Cases 1945 (2d Cir., 1991). That was also briefly the position of the Seventh Circuit, which changed its mind twice, first agreeing with Weil’s ultimate conclusion, then deciding that only nonvested participants are pertinent to partial termination calculations, and finally reverting to its initial holding. Matz v. Household International Tax Reduction Investment Plan, 265 F.3d 572, 26 EB Cases 2121 (7th Cir., 2001), cert. den., 122 S.Ct. 1357, 27 EB Cases 2088 (2002).

The plaintiff contended that the plan underwent a partial termination between August 1994 and May 1996, a period during which the sponsor divested itself of a large number of non-core businesses and plan participation declined to an undisclosed extent. The parties disagreed on how to evaluate whether the decline was significant, with the plan arguing that the number of affected participants should be computed without regard to those who were fully vested when they left the plan. Thus the numerator of the reduction percentage would equal nonvested participants involuntarily separated in connection with the divestitures, and the denominator would equal the total number of participants before the alleged partial termination began. The quotient presumably was a small number.

The first District Court decision in the case held that, in the absence of a clear statutory rule, courts were required to defer to the IRS on the method of determining whether a substantial reduction in the number of participants had taken place. Matz v. Household International Tax Reduction Investment Plan, 1998 U.S. Dist. LEXIS 19160 (N.D.Ill., 1998). On an interlocutory appeal, the Seventh Circuit affirmed but expressed dissatisfaction with the arguments presented by the IRS in the Weil litigation (which were indeed weak; cf. the main text, pp. 372-3). “Purely from a policy standpoint, we believe that the method adopted in Gulf Pension Litigation [excluding vested participants from both the numerator and the denominator; see main text, p. 374] best furthers the purposes of the partial termination statute. However, we are constrained in our analysis of the statue and must decide only whether the IRS’ construction is reasonable.” 227 F.3d 971, 976, 24 EB Cases 2825, 2829 (7th Cir., 2000).

The Supreme Court lessened that constraint when, in United States v. Mead Corporation, 533 U.S. 218 (2001), it reduced the degree of deference that courts are required to accord to pronouncements of administrative agencies. It granted certiorari and remanded Matz for reconsideration in light of Mead. 533 U.S. 925 (2001). Given that invitation, the Circuit Court had no trouble deciding that its own theory was more rational than the one espoused by the IRS. Because vested participants have nothing to gain or lose from the determination that a plan has undergone a partial termination, the court held that they should not be counted at all for that purpose. 265 F.3d 572, 576, 26 EB Cases 2121, 2123 (7th Cir., 2001), cert. den., 535 U.S. 954, 27 EB Cases 2088 (2002). It remanded the case to the District Court to apply the new theory to the still-to-be-found facts.

Although the court’s position was plausible on the surface, its analysis was not at all complete. The concept of partial termination has a long judicial and regulatory history, none of which draws any distinction between vested and nonvested participants. The IRS’s rejected interpretation is “not a post hoc rationalization” and can reasonably be said to reflect “the agency’s fair and considered judgment” as developed over many years of guidance in an area that unquestionably lies within the scope of its Congressionally delegated jurisdiction. Cf. Id. at 575, 26 EB Cases at 2122. The parallel to the tariff classification guide that was denied deference in Mead is not especially compelling.

On a practical level, omitting vested participants from both the numerator and the denominator would lead in many instances to eccentric results. Thanks to the drastic shortening of vesting schedules since the Tax Reform Act of 1986, nonvested participants are a much smaller proportion of plan populations than they used to be may be affected disproportionately by events that have negligible impact on the plan as a whole. In an extreme, but probably not uncommon, case, a small company’s layoff of one of its two or three nonvested employees could be characterized as a partial termination, requiring full vesting for the discharged worker and rendering the plan’s vesting schedule more or less useless.

On remand, the district judge voiced some of the same doubts but was of course bound to carry out fact finding in accordance with the higher court's decree. After sorting through the data, she concluded that, at most, 17.6 percent of the plan’s nonvested participants terminated employment in connection with the corporate event, which was not sufficient to constitute a partial termination, although the judge gave the plaintiffs the opportunity to present evidence of “egregious conduct or bad faith on the employer’s part”. No figures were given for the overall percentage reduction in plan participants. 2003 U.S. Dist. LEXIS 16906 (N.D.Ill., 2003).

Back at the Appeals Court for a third time, on another interlocutory appeal, 388 F.3d 570, 33 EB Cases 2569 (7th Cir., 2004), the case was heard by a new set of judges, who, presumably out of politeness to their colleagues, pretended that the court below had misunderstood the previous appellate guidance. The new opinion strangely lacks an explicit holding, though its thrust is clear enough.

The court rejected the methodologies proposed by both the plaintiff (count only nonvesteds) and the defendant (include only nonvesteds in the numerator but everyone in the denominator), because both can produce ridiculous outcomes. Moreover, they confuse the purpose of the partial termination rule – construed by the court as the prevention of the tax abuse that might arise if employers could reclaim previously deducted contributions by discharging nonvested recipients – with the method chosen by Congress to carry out that purpose, viz., applying the requirement of full vesting on plan termination to events that approximate complete termination. The unavoidable implication is that the closer the decrease in the total number of participants is to 10 percent, the stronger the case for the existence of a partial termination.

In search of a usable test, the court turned to a rule of thumb that is often cited informally but had not previously rested on any real authority, namely, that a 20 percent or greater one-year reduction in the number of plan participants probably rises to the level of a partial termination. The Matz opinion“generalize[d] from the cases and the rulings a rebuttable presumption that 20 percent or greater reduction in plan participants is a partial termination and that a smaller reduction is not”, id. at 577–8. The court went on to suggest that the presumption should be conclusive where the reduction is 40% or more, on the one hand, or 10% or less, on the other. Within that range, “given that the statutory purpose is to prevent tax windfalls, it seems to us that the only relevant facts and circumstances should be the tax motives and tax consequences involved in the reduction in plan coverage”, id. at 578.

While it is pleasant to have “bright line” rules, the basis for this one is shaky. The opinion presents a table (adapted from our main text, Table 14-1 at 367) showing percentage reductions in past cases and rulings, and whether the court or the IRS found a partial termination in each instance. Unfortunately for attempts to draw empirical conclusions, there is a gap in the data. Except for one case (a 27.9% reduction with a finding of no partial termination, which the Matz opinion dismisses, on reasonable grounds, as the product of special circumstances), there are no data points between 19.9% (not a partial termination) and 33.8% (partial termination). How can one conclude that 20 percent is a breakpoint? Why not 25 percent or 30 percent?

The limitation of pertinent facts and circumstances to “tax motives and tax consequences” is fully justified by the statute’s history and original purpose, but it raises the question of whether any such motives or consequences will ever be present in the current environment. The partial termination concept originated in a period when vesting schedules could be, and often were, very long and “nondiscrimination” standards were rudimentary. The scenario that concerned the IRS was one in which an employer discharged most of its nonvested rank-and-file employees and reallocated the resulting forfeitures to the owner and a few other favored participants. Now that vesting is required after very short periods of service, that tactic has no allure. In all but the most extraordinary situations, a business decision that results in a decline in participation is not motivated to even the slightest extent by the kind of abuse that led to the formulation of the partial termination concept and its codification in I.R.C. §411(d)(3). A reasonable conclusion is that, except where the participant reduction reaches quite a high level (e. g., the Matz opinion’s 40% threshold for a conclusive presumption), findings of partial termination should be exceedingly rare.

The controversy is now back at the District Court, which, in the course of certifying it as a class action, observed that “Despite the lengthy and complex procedural history of this case, . . . many issues that may well be dispositive have never been decided.” 232 F.R.D. 593, 597, 36 EB Cases 2525 (N.D. Ill., 2005). Some might suggest that the nearly decade-long history of this case offers a good example of why judges ought to be cautious about certifying issues for appeal before fact-finding has been conducted.

(e) Treatment of Substantial Cessation of Operations. – If an employer ceases operations at a facility at any location and, as a result of such cessation of operations, more than 20 percent of the total number of his employees who are participants under a plan established and maintained by him are separated from employment, the employer shall be treated with respect to that plan as if he were a substantial employer under a plan under which more than one employer makes contributions and the provisions of §§4063, 4064, 4065 shall apply.

Section 4063 deals with liability to the Pension Benefit Guaranty Corporation when a “substantial employer” (defined in ERISA, §4001(a)(2)) withdraws from a multiple employer plan (i. e., a plan the includes employers that do not belong to the same controlled group and that is not a multiemployer plan, as defined in ERISA, §4001(a)(3)). It requires the withdrawing employer to place in escrow an amount equal to its “share” of the plan’s unfunded vested benefits. If the plan terminates within five years, the money is applied to provide unfunded benefits (with any excess going back to the employer). After five years without a termination, it is returned to the employer, without interest. Alternatively, the employer may post a bond in an amount acceptable to the PBGC, up to 150 percent of its liability. ERISA, §4063(c).

The theory behind section 4062(e) is that plan closings with substantial layoffs may be a sign of current or impending financial trouble and that a bond or escrow in favor of the plan gives the PBGC insurance fund a degree of protection against a potential distress termination. For the first 20 or so years of ERISA’s existence, the provision was pretty much ignored. In recent years, as the PBGC’s balance sheet has grown more precarious, efforts have been made to enforce it on a case-by-case basis.

Employers are required to report “section 4062(e) events” to the PBGC within 60 days after the cessation of operations occurs. ERISA, §4063(a). The PBGC has not issued any regulations concerning this reporting obligation but has stated that it is distinct from the reportable event described in ERISA, §4043(c)(3) and 29 C.F.R., §4043.21 (reduction in the number of active participants) and is not satisfied simply by filing a notice of reportable event. 71 Fed. Reg. 34819, fn. 1 (June 16, 2006); Enrolled Actuaries Meeting, Summary of Questions to the PBGC and Their Responses, Q&A-21 (March 2006).

Section 4062(e) applies only to reductions resulting from cessations of discrete operations, not to a generalized decline in the employer’s work force. A controlled group’s plans are not aggregated for this purpose. Whether a 20 percent reduction has occurred is determined on a plan-by-plan basis. PBGC Opinion Letter 77-123 (Jan. 14, 1977). Because the statute does not define such terms as “facility”, “location” and “cessation of operations”, there is a great deal of room for uncertainty as to the circumstances under which it can be invoked.

The rule for allocating liabilities under section 4043, based on the percentage of total plan contributions made by the withdrawing employer during the preceding five years, obviously cannot be applied to a plan with only one contributor. The PBGC’s policy, now embodied in regulations, 29 C.F.R., §4062.8 (June 16, 2006), is to multiply the value of unfunded vested benefits by the ratio of active participants terminated in connection with the cessation to the total number of employees in the plan sponsor’s controlled group. The regulations provide a simple example:

Company X sponsors a pension plan with 50,000 participants of which 20,000 are current employees and 30,000 are retirees or deferred vested participants. On a PBGC termination basis, the plan is underfunded by $80million. Company X ceases operations at a facility resulting in the separation from employment of 5,000 employees, all of whom are participants in the pension plan. A section 4062(e) event has occurred, and the PBGC will determine the amount of employer liability under section 4063(b) of ERISA. The numerator described in paragraph(a)(1) of this section is 5,000 and the denominator described in paragraph(a)(2) of this section is 20,000.Therefore, the amount of liability under section 4063(b) of ERISA pursuant to section 4062(e) is $20 million (5,000/20,000 × $80 million). [29 C.F.R., §4062.8(b)]

This formula’s product bears only a chance relationship to the liability actuarially attributable to the separated participants. In the preamble to the regulations, the PBGC defended that result on the ground that the standard section 4043 formula for multiple employer plans, based on relative contribution levels over a five-year period, also allocates liabilities imprecisely. 71 Fed. Reg. 34819, 34820–21 (June 16, 2006). That is not a very convincing rationale, but the calculated figure is likely to be, in practice, only a starting point for discussions with the PBGC and thus not any great cause for complaint.

The regulations address only the narrow issue of computing the employer’s liability. The PBGC has promised guidance on other questions, but for the moment the operation of section 4062(e) is very much a matter of individual negotiations. For an analysis of the regulations and discussion of related issues, see Harold J. Ashner, “PBGC’S Final Rule on Liability for Facility Shutdowns Affects Downsizing Employers”, 33 BNA Pension & Benefits Reporter 1546 (June 27, 2006).