Halliburton Co. v. Commissioner

100 T.C. No. 15, 100 T.C. 216, 1993 U.S. Tax Ct. LEXIS 14, 16 Employee Benefits Cas. (BNA) 1929

• Court: United States Tax Court
• Decided: March 24, 1993
• Docket: Docket Nos. 26290-90R, 2397-91R
• Status: Published

Opinion

OPINION

Wells, Judge:

The instant case is before us on a petition for declaratory judgment under section 7476. The issue presented for decision is whether a profit sharing plan sponsored by Halliburton Co. (Halliburton) experienced a partial termination under section 411(d)(3) for the plan year ending on December 31, 1986.

Unless otherwise indicated, all section references are to the Internal Revenue Code, and all Rule references are to the Tax Court Rules of Practice and Procedure.

Background

The instant case was submitted for declaratory judgment on a stipulated administrative record pursuant to Rule 217.

Halliburton is a corporation with its principal place of business in Dallas, Texas. Ken Nash (Mr. Nash) resided in Houston, Texas, at the time he filed his petition. The plans whose qualified status is in issue are the Halliburton Profit Sharing and Savings Plan (the Halliburton plan or plan) and the imco Profit Sharing and Savings Plan (the imco plan).

The Halliburton plan was established in 1944 and is a defined contribution profit sharing plan. The plan received favorable determination letters with respect to all amendments through January 16, 1986. From 1944 through 1985, Halliburton contributed approximately 10 percent of annual profits to the plan, totaling $575 million, and the plan has paid out over $600 million in benefits. No contribution was made in 1986 because Halliburton had no profits in such year.

Eligible plan participants employed by Halliburton on the last day of the year, or who terminated participation during the year due to retirement, death, or disability, were entitled to a share of Halliburton’s annual contribution to the plan. Such contribution was allocated to each participant’s account based on the ratio of such participant’s basic earnings to the total basic earnings for all participants. Employer contributions were subject to a 5-to-15-year graded vesting schedule, as permitted by section 411(a) for 1986;¹ account balances vested upon death, normal retirement at age 65, early retirement, disability, or attaining age 50.

On April 27, 1987, Halliburton submitted an application to respondent’s Dallas, Texas, Key District for determination that the plan continued to qualify under section 401(a). On the same date, Halliburton submitted an application to such office for an initial determination for the IMCO plan, a spinoff² of the Halliburton plan. On September 7, 1988, respondent issued a proposed adverse determination letter which concluded that the Halliburton plan had experienced a partial termination in 1986 due to the large workforce decline which caused a drop in participation in the plan during such year, and that, because affected employees had not been vested in their accrued benefits as required by section 411(d)(3), the plan no longer qualified under section 401(a). Respondent also issued a proposed adverse initial determination letter with respect to the IMCO plan, holding that it failed to qualify because the Halliburton plan, of which the IMCO plan was a spinoff, did not qualify.

After additional correspondence and administrative conferences, Halliburton filed the instant declaratory judgment action disputing and seeking review of respondent’s proposed determination. Mr. Nash, a former employee of Halliburton who had been laid off in 1986, filed a petition in support of respondent’s proposed determination. Respondent filed a motion to dismiss both petitions for lack of jurisdiction, claiming that each petitioner had failed to exhaust administrative remedies. We denied such motion. Halliburton Co. v. Commissioner, 98 T.C. 88 (1992). Subsequently, respondent filed a second motion to dismiss, claiming that Halliburton had failed to notify affected parties of the request for determination, which motion we also denied. Halliburton Co. v. Commissioner, T.C. Memo. 1992-534. Additionally, in Halliburton Co. v. Commissioner, T.C. Memo. 1992-533, we denied respondent’s motion to compel discovery to supplement the administrative record.

Accordingly, we now reach the merits of the partial termination issue. If we decide that a partial termination of the Halliburton plan occurred in 1986, the plan may be disqualified because participants affected by such partial termination have not been vested in their accrued benefits. In respondent’s response to the amicus brief filed by the Profit Sharing Committee of the Halliburton and IMCO plans, respondent stated that, if such affected participants were immediately vested in accrued benefits, respondent would issue a favorable determination letter retroactive to the date of the partial termination. Furthermore, if we decide that a partial termination occurred, both the Profit Sharing Committee and respondent ask that we identify the affected participants entitled to immediate vesting of accrued benefits. A decision as to the status of the Halliburton plan will resolve the IMCO plan’s qualified status.

Halliburton and its controlled group constitute one of the world’s largest and most diversified oil field services and engineering and construction organizations. Halliburton’s business is divided into three principal categories: Oil field services and products, industrial engineering and construction services, and marine engineering and construction services. During 1986, the oil field services and products business was a major provider of products and services, both onshore and offshore, to the exploration and production segments of the petroleum industry. The engineering and construction categories of Halliburton’s business provided services to the petroleum industry, as well as to other types of customers. The marine engineering group performed services predominantly for the petroleum industry, including the design, fabrication, and installation of offshore structures. Approximately three-quarters of the consolidated net revenues of the Halliburton controlled group is derived from services performed for the petroleum industry.

During 1986, however, a sharp decline in the oil business occurred, which severely disrupted Halliburton’s business. During December 1985, the price of oil began to collapse due to overproduction in the Middle East and slack demand in oil-consuming countries. During the period beginning December 1985 to August of 1986, the price of oil fell from $27 per barrel to $10 per barrel. The U.S. oil production rate also dropped to its lowest level since 1977. As a consequence of tumbling prices, oil exploration and drilling fell drastically. Reinvestment in exploration and production fell to their lowest levels since World War II. The steepest decline occurred during the first 7 months of 1986. The rotary rig count fell from 1,950 during December 1985 to 686 during July 1986, which was the fewest in the 46 years such data had been compiled. Halliburton’s profit is closely tied to the domestic rotary rig count. As drilling operations shut down, demand for oil field services virtually disappeared. A report by the National Petroleum Council, a Federal advisory committee to the Secretary of Energy, stated:

While the recent oil price decline has affected all segments of the industry, it has been particularly onerous for the oil field service industry. Eighty percent of the recent increase in unemployment in the oil and gas extraction industry has occurred in this area. [“Factors Affecting U.S. Oil & Gas Outlook”, Report of the National Petroleum Council 23 (February 1987).]

The report went on to state that employment in the oil service industry dropped from 315,000 during December 1985 to 206,000 during December 1986. Id. at 135.

Halliburton was faced with two options: Cut costs or go out of business. Halliburton responded to the crisis with a broad array of cost-cutting measures and did not merely attempt to reduce its staffing levels. The capital budget was slashed, and the dividend was cut by 44 percent. In December 1985, before taking other measures to reduce personnel, Halliburton announced that, during the period beginning January through March 1986, the age at which an employee was eligible to take early retirement would be lowered from 55 to 50. The letter announcing such program stated:

This temporary change in the retirement policy is being made due to our existing work load and business forecasts for the immediate future. * * * It is for the purpose of encouraging early retirements at this time in our attempts to reduce our workforce in line with our business. While we had hoped that normal attrition would accomplish our objectives, additional reductions may be necessary. For those of you who have been contemplating early retirement, we urge your serious consideration during this period which might possibly save the job of another employee.

Two additional 3-month “window periods” for early retirement were subsequently announced for April through June and October through December 1986. Early retirees could take advantage of certain benefit distribution options not available to vested terminees and could participate in a retiree health plan not available to vested terminees. Employees who took early retirement did so voluntarily and were not forced out of their jobs. Halliburton also implemented a wage freeze during March 1986 and a furlough program during June 1986 under which many employees, including Halliburton’s executives, were asked to take every eighth week off without pay. The furlough program permitted Halliburton to retain more of its employees and so reduce job losses.

Halliburton also resorted to involuntary layoffs in order to reduce its work force. The number of layoffs during each month of 1986 generally tracked the size of the drop in oil prices for such month: in months where the price drop was sharp, the number of layoffs was also large, but, as the price of oil leveled off toward the end of the year, the number of layoffs also stabilized at a lower level. Halliburton’s 37-per-cent overall decline in service personnel closely approximated the industry’s 35-percent overall decline in employment. Even with its cost-cutting measures, Halliburton posted record operating losses during 1986.

On January 1, 1986, the Halliburton plan had 19,017 participants, and during such year, 581 new participants joined the plan, making a total of 19,598 participants. Participants leaving the plan during 1986 fell into the following categories: 1,141 plan participants elected to take early retirement either at the normal early retirement age or at the lower age permitted during the window periods; 38 participants retired at normal retirement age; 22 participants died; 264 were terminated for cause; and 797 participants voluntarily left their jobs with Halliburton. Nearly all of the voluntarily resigning participants made statements of their reasons for leaving their jobs, but none said that they were leaving because of anticipated discharge. Most of such employees cited personal reasons for leaving; others were dissatisfied with their jobs. Additionally, 5,015 participants in the plan were involuntarily terminated, of which 224 were fully vested, 4,560 were partially vested, and 231 had not been participants long enough to have received an allocation of employer contributions to the plan, and so such 231 participants had zero balances in the accounts subject to the plan vesting schedule.

A number of participants in the plan transferred to employment with employers related to Halliburton, but not covered by the plan, ceasing participation in the plan for such reason. During 1986, Halliburton combined its IMCO Services Division with a division of Dresser Industries to form a joint venture called M-I Drilling Fluids Co. (M-I), which would permit Halliburton to compete more effectively in the drilling fluids business. As a consequence of the M-I venture, 437 participants transferred to M-I and consequently left the Halliburton plan. Four hundred twenty-three IMCO employees who terminated employment and participation in the plan in 1986 for other reasons, such as retirements, layoffs, and voluntary resignations, are included in other categories of terminated employees.³ All participants who transferred to M-I were fully vested in their account balances at transfer. Additionally, Halliburton allowed the participants who transferred to M-I and were employed on November 30, 1986, to receive an allocation of employer contributions and forfeitures regardless of their employment status on December 31, 1986, the date on which the plan provided such allocations be made. During 1987, Halliburton spun off the account balances of the 437 IMCO transferees from the Halliburton plan to form the IMCO plan for the benefit of the IMCO transferees. Effective March 1, 1987, M-I adopted its own profit sharing plan, in which the IMCO transferees were immediately eligible to participate. Additionally, six participants transferred to affiliates of Halliburton other than M-I. Such transferees received vesting service credit under the plan for work performed for the affiliate, and so may continue to vest in their unvested accrued benefits under the plan. Such transferees became participants in the affiliates’ plans, and are credited with vesting service for the entire period of service with members of the Halliburton controlled group.

Toward the end of 1986, the price of oil began to recover somewhat, because oil producers agreed to return to a fixed price system. The number of operating rigs also began to rise with the oil price recovery. The administrative record reveals that during the period beginning 1987 through 1989, Halliburton rehired a total of 1,279 participants who had left the plan during 1986. Eight hundred sixteen of such persons were rehired during 1987, 371 were rehired during 1988, and 92 were rehired during 1989. Of the rehired group, 1,124 persons had been involuntarily laid off during 1986, of whom 45 were fully vested, 1,039 were partially vested, and 40 had zero balance accounts. Of such rehired participants, 742 were reemployed during 1987, 311 were rehired during 1988, and 71 were recalled during 1989. Halliburton contacted many more former participants, but it was unsuccessful in its attempts to rehire them.

Under the plan, if a participant terminated employment prior to full vesting, and received a cash distribution of his vested account balance, the unvested portion of such account balance would be forfeited upon the later of a 1-year break in service or 1 year from the date of the distribution. If a terminated participant did not receive a cash-out distribution, the nonvested portion of such participant’s account balance would be forfeited after a 5-year break in service.

Under the plan, a rehired employee automatically reenters as a participant if such employee is rehired within a year of termination, such employee’s period of severance amounts to less than a 5-year break in service, or such employee was vested at the time of severance. If such an employee did not receive a cash distribution from the plan on severance, the forfeited portion of such employee’s account balance was restored on reentering the plan. If such a rehired employee had received a cash-out distribution on severance, such employee could “buy back” any forfeited amounts by recontributing such distribution within 5 years of rehire.

The changes in plan participation may be summarized in the following table:

1986 Changes in Plan Participation

Deaths 22

Retirees

Normal retirements 38

Early retirements 1,141

264 Terminations for cause

Transferees

To M-I 437

Other affiliates 6

797 Voluntary separations

Involuntary separations

Fully vested 224

Partially vested 4,560

Zero balance accounts 231

5,015

Total 1986 decrease in participation 7,720

Participants at beginning of 1986 19,017

Participants added in 1986 581

Total 19,598

Under the plan, forfeitures are allocated to participants employed as of the end of the year the forfeiture occurs, or to participants who have terminated employment in such year by reason of death, disability, or retirement. Forfeitures are apportioned among participants according to the ratio which the participant’s basic earnings bear to the total basic earnings of all eligible participants. In accordance with section 411(d)(3), the plan provides that, upon its partial termination, the rights of each participant and beneficiary affected by such partial termination will be fully vested and non-forfeitable.

The amount of potential forfeitures resulting from the 1986 reduction in participation in the plan totaled $55.7 million. Restorations of account balances to rehired participants reduced the amount of such potential forfeitures to $52.7 million. Since April 30, 1987, such funds have been held in a segregated fund, together with the earnings thereon, and will be distributed to either affected former participants or the accounts of remaining participants when a final decision of the partial termination issue determines which group is entitled to them. During 1986, the plan held approximately $1.1 billion in assets, and potential forfeitures represented approximately 4.8 percent of its asset value.

Under the Halliburton plan, no possibility exists for plan assets to revert to Halliburton upon termination or partial termination of the plan. No possibility exists for any amount of potential forfeitures resulting from terminations to reduce Halliburton’s contributions to the plan or otherwise benefit Halliburton. Neither the reduction in participation nor the allocation of forfeitures among remaining participants resulted in prohibited discrimination in favor of highly compensated employees.

Discussion

Before turning to the merits of the parties’ contentions, we note preliminarily that respondent bears the burden of proof herein because respondent failed to issue a final determination prior to the filing of the instant declaratory judgment action. Rule 217(c)(1)(B). Such burden, however, plays only a limited role where, as here, the decision is based on a fully stipulated administrative record. Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. 154, 157 (1984). Such burden has no bearing on the weight to be given a party’s legal arguments. BBS Associates, Inc. v. Commissioner, 74 T.C. 1118, 1121-1122 (1980), affd. without published opinion 661 F.2d 913 (3d Cir. 1981). With the foregoing in mind, we consider whether the Halliburton plan sustained a partial termination for tax purposes in 1986.

General Discussion of the Partial Termination Rule

Section 411(d)(3) provides that, in the event of a partial termination of a qualified plan, the rights of affected employees to benefits accrued to the date of such partial termination shall be nonforfeitable, to the extent the plan is funded, as of the date of such partial termination. Such statutory provision, enacted as part of the Employee Retirement Income Security Act of 1974 (erisa), codified a rule in the pre-ERISA regulations amplifying the prior statute, which had simply required vesting of benefits upon “termination” of a plan. Employee Retirement Income Security Act of 1974, Pub. L. 93-406, sec. 1012(a), 88 Stat. 912; H. Conf. Rept. 93-1280, at 277 (1974), 1974-3 C.B. 415, 438; S. Rept. 93-383, at 50 (1973), 1974-3 C.B. (Supp.) 80, 129; sec. 1.401-6(b)(2), Income Tax Regs., T.D. 6675, 1963-2 C.B. 151; see also Anderson v. Emergency Medicine Associates, 860 F.2d 987, 991 (10th Cir. 1988). The regulations promulgated under section 411(d)(3), which essentially restate the pre-ERISA regulations, provide that the question of whether a partial termination has occurred is to be resolved on the basis of all the facts and circumstances of the particular case. Sec. 1.411(d)-2(b)(1), Income Tax Regs. Such circumstances include “the exclusion by reason of a * * * severance by the employer, of a group of employees who have previously been covered by the plan”. Sec. 1.411(d)-2(b)(1), Income Tax Regs. Such an exclusion has come to be known as a “vertical partial termination”. In re Gulf Pension Litigation, 764 F. Supp. 1149, 1163 (S.D. Tex. 1991). It is such a situation that we consider in the instant case.

Neither the Code nor the regulations explicitly define what constitutes a partial termination. Consequently, courts have relied on legislative history, as well as prior judicial and administrative interpretations of the rule. One long-standing concern of Congress, the courts, and respondent is that a reduction in plan participation may result from abuse, bad faith, or misconduct by the employer. Congress was concerned that an employer could establish a forfeitable plan, enjoying the tax benefits flowing therefrom, and then terminate it, thus causing the assets to revert to the employer with favorable tax consequences. H. Rept. 87-378, at 16 (1961), 1962-3 C.B. 261, 275-276. An abuse also arises when an employer excludes plan participants in a predatory effort to profit from subsequent forfeitures or diminished contribution requirements. Kreis v. Charles O. Townley, M.D. & Associates, P.C., 833 F.2d 74, 80 (6th Cir. 1987). Additionally, a bad motive may be indicated if the financial health of a plan suffers because of reduced contributions arising out of a decline in participation. Id. Furthermore, the size of the forfeitures falling to continuing participants may also indicate an improper motive for the reduction in plan participation. Id.

The record in the instant case, however, shows that Halliburton had no improper or bad motive for terminating plan participants in 1986. Halliburton did not stand to gain anything by the 1986 reductions in plan participation. The parties agree that Halliburton’s contributions to the plan will not be reduced because of the potential forfeitures resulting from the 1986 reductions in participation. Furthermore, the plan does not allow Halliburton to create a reversion of plan assets to itself by excluding participants. The events of 1986 had no adverse effect on the financial health of the plan. No discrimination in favor of highly compensated participants resulted from the reduction in participation. Additionally, the relative size of the forfeitures to the total assets of the plan⁴ and the fact that they would be divided among over 11,000 continuing participants does not indicate that Halliburton had any intent to create an improper windfall for the remaining participants.

The record clearly establishes that the reduction in plan participation resulted from emergency measures designed to reduce Halliburton’s costs, including personnel costs, in the face of a collapse in the demand for its services. Halliburton’s business is providing services to oil drillers. Consequently, its financial health is heavily dependent upon the price of oil and level of drilling activity. The price collapse which occurred during 1986 removed economic incentives for the exploration and drilling activity on which Halliburton’s businesses depended. Halliburton’s actions appear to us to be prompted by a crisis in its business, which created a need to pare expenses, including personnel costs, in order to survive. Although certain interested party comments suggest that Halliburton possessed an improper motive, we find such comments outweighed by other facts in the record. Rather, we think it is clear that the staff reductions taken in response to Halliburton’s business emergency were not the result of bad faith and were not intended to take improper advantage of the rules related to tax-qualified retirement plans.

Halliburton argues that our inquiry into the facts and circumstances surrounding the 1986 drop in plan participation should end at the point of finding a lack of abuse, contending that the sole purpose of the partial termination rule is to prevent employer abuse. We disagree. While the foregoing discussion demonstrates that one purpose of the rule is to prevent abuse, the rule has the additional purpose of protecting employees’ legitimate expectations of benefits. Gluck v. Unisys Corp., 960 F.2d 1168, 1182-1183 (3d Cir. 1992); Amato v. Western Union Intl., Inc., 773 F.2d 1402, 1409 (2d Cir. 1985); United Steelworkers of America v. Harris & Sons Steel Co., 706 F.2d 1289, 1298 (3d Cir. 1983). Tax laws governing qualified plans contain a variety of rules which penalize employers who seek to abuse such rules for tax advantages, such as the permanence, exclusive benefit, and discriminatory vesting rules. Sec. 411(d)(1)(A); sec. 1.401-1(b)(2) and (3), Income Tax Regs. Such rules operate by stripping away the tax benefits associated with qualified plans, while the partial termination rule operates by requiring that plan assets be applied to their ostensible purpose of benefiting participants. Consequently, the partial termination rule does not serve exclusively as a sanction for employer abuse, but also serves to protect employees “who would otherwise be left out in the cold after a drastic and sudden change in the plan.” Chait v. Bernstein, 835 F.2d 1017, 1021 (3d Cir. 1987). To say that the reason for awarding plan assets to participants is simply to punish employer abuse ignores one of the fundamental purposes of the rules governing qualified plans, the protection of employee retirement benefits. See S. Rept. 93-383, at 1 (1973), 1974-3 C.B. (Supp.) 80.

The 1974 legislative history of section 411(d)(3) does not indicate that Congress intended to limit the application of the rule to abusive situations, for it states simply that “Examples of a partial termination might include, under certain circumstances, a large reduction in the work force”. S. Rept. 93-383, at 50 (1973), 1974-3 C.B. (Supp.) 80, 129. The foregoing statement is not limited to partial terminations that occur only when employer abuse is present. We note that Congress did not indicate any disagreement with the developing administrative interpretation of the partial termination rule, which held that a layoff due to business circumstances could cause a partial termination. See, e.g., Rev. Rul. 72-510, 1972-2 C.B. 223, superseded by Rev. Rul 81-27, 1981-1 C.B. 228.

Instead, Congress stated in the 1974 legislative history that section 411(d)(3) continued the partial termination rule as in effect prior to the enactment of such paragraph. H. Conf. Rept. 93-1280, at 277 (1974), 1974-3 C.B. 415, 438; S. Rept. 93-383, at 50 (1973), 1974-3 C.B. (Supp.) 80, 129; see also Anderson v. Emergency Medicine Associates, 860 F.2d 987, 991 (10th Cir. 1988). At such time, the partial termination rule was stated in the regulations issued under the predecessor to section 411(d)(3), which regulations provided that

Whether or not a partial termination of a qualified plan occurs when a group of employees who have been covered by the plan are subsequently excluded from such coverage * * * by reason of being discharged by the employer, will be determined on the basis of all the facts and circumstances. [Sec. 1.401-6(b)(2), Income Tax Regs., T.D. 6675, 1963-2 C.B. 151.]

Congress thus indicated approval of the facts and circumstances approach to the partial termination issue developed in such regulations, an approach which did not limit the operation of the rule to situations involving employer abuse. The facts and circumstances approach of the pre-ERISA regulations was incorporated in the regulations issued under section 411(d)(3). Sec. 1.411(d)-2(b)(1), Income Tax Regs., T.D. 7501, 1977-2 C.B. 133. The foregoing history of the partial termination rule refutes Halliburton’s argument that such rule should apply only in cases of employer abuse.

Consistent with the foregoing, this Court has expressly held that layoffs caused by economic circumstances can cause partial termination of a plan. Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. at 160-161; Peter M. Boruta, M.D., P.C. v. Commissioner, T.C. Memo. 1988-172. In Tipton, we stated as follows:

We infer * * * a broader ameliorative purpose than simply preventing the reversion of contributions to an owner-employee. We think that Congress in enacting * * * [the predecessor of section 411(d)(3)] sought to protect employees from forfeiting their retirement benefits upon termination of a plan. This broader congressional purpose would be ill-served by holding the presence or absence of an intent or purpose to deprive employees of benefits determinative of whether a termination had occurred. Where a significant percentage of plan participants are discharged, the effect is the same regardless of the employer’s intent; unless the discharge is treated as a partial termination, a significant percentage of plan participants are compelled to forfeit accrued but unvested retirement benefits.

Petitioner’s argument boils down to the simple assertion that it was not at fault. Although this may be true, neither were the discharges the fault of the discharged employees, who nevertheless were compelled to forfeit their retirement benefits.

[83 T.C. at 160-161; fn. ref. and citations omitted.]

Accordingly, the fact that there is no evidence of abuse or bad faith on the part of Halliburton in connection with the 1986 reduction in force will not per se insulate Halliburton from a finding that a partial termination of the plan occurred in such year.⁵

Halliburton also contends that the partial termination rule was not designed to protect employees from periodic layoffs which occur in cyclical industries. Whatever the merits of such contention in another context, we do not find it applicable in the instant case. The collapse of the oil market was not a mere recurrent downturn in the business cycle. Rather, it appears to have been an extraordinary event resulting from a pronounced imbalance of market forces. The unusual nature of the events of 1986 is illustrated by the drastic adjustments which Halliburton had to make in all areas of its business, from its capital spending and dividend to its workforce. Halliburton’s losses in 1986 prevented it from making a contribution to its plan for the first time since the plan’s establishment in 1944. Consequently, we do not view the events of 1986 as a mere cyclical downturn.

An additional circumstance which courts attempt to identify in deciding whether a partial termination has occurred is the occurrence of a major corporate event in connection with the reduction in plan participation. Generally, the closing of a plant or division, or a restructuring incident to a merger, has been held to constitute a major corporate event. Weil v. Retirement Plan Admin. Comm., 750 F.2d 10, 12 (2d Cir. 1984) (Weil I); In re Gulf Pension Litigation, 764 F. Supp. at 1169-1170. A “major corporate event” is not necessarily a prerequisite to finding that a partial termination has occurred, as layoffs due to nothing more than a downturn in business can trigger a partial termination. Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. at 160-162. Whether or not the reason for a reduction in participation may be termed a “major corporate event”, however, such cause needs to be identified because the drop in participation attributable to such cause which is considered in deciding whether a partial termination has occurred. Bruch v. Firestone Tire & Rubber Co., 828 F.2d 134, 151 (3d Cir. 1987), affd. in part and revd. in part on another issue 489 U.S. 101 (1989); Weil I, supra. Reductions in participation unrelated to such cause are not relevant to the question of whether a partial termination occurred, for they would have occurred even in its absence. In re Gulf Pension Litigation, 764 F. Supp. at 1168.

The downsizing caused by the 1986 collapse in the oil market was not the type of major corporate event generally found in other cases involving partial terminations. Although such crisis prompted a major restructuring of Halliburton, much like a plant closing or postmerger consolidation, there are, nonetheless, certain features of Halliburton’s response to the crash in oil prices which distinguish it from a typical major corporate event. Halliburton did not permanently restructure its operations, as occurred in other cases involving the closure of a division or plant. Instead, Halliburton engaged in a temporary economizing measure, affecting all its operations, which was aimed at meeting a business emergency. The 1986 downturn in business seems, given the facts in the record, to have been a transitory economic phenomenon, and the staff reductions which accompanied it seem to have been a response which was temporary, rather than permanent, in nature. The fact that Halliburton was able to rehire a significant number of laid-off employees in the immediately succeeding years demonstrates the transitory nature of the staff reductions. Consequently, the circumstances surrounding the reduction in plan participation are less suggestive of a partial termination than would be the case had a permanent restructuring of the firm occurred, e.g., the closure of a plant or division.

One of the most important facts and circumstances bearing on the vertical partial termination issue is the size of the drop in plan participation resulting from the event causing such reduction. Kreis v. Charles O. Townley, M.D. & Associates, P.C., 833 F.2d 74, 79 (6th Cir. 1987), and cases cited therein. In order to assess whether such a drop is sufficiently large to warrant the conclusion that a plan has partially terminated, respondent has developed two tests: The significant number test and the significant percentage test.

The Significant Number Test

Respondent urges that we should apply the significant number test to determine whether a partial termination of the Halliburton plan occurred, while Halliburton argues that such test should not be applied. In Tipton & Kalmbach, we specifically reserved for another day the decision on whether the significant number test should be used to determine whether a plan has experienced a partial termination. Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. at 160 n.5.

The significant number test first appeared in respondent’s revenue rulings on the partial termination issue. The test was first announced in a 1972 revenue ruling, which was superseded by a 1981 ruling reaffirming the test. Rev. Rul. 81-27, 1981-1 C.B. 228, superseding Rev. Rul. 72-510, 1972-2 C.B. 223. In that ruling, respondent stated that a partial termination occurred where 95 out of 165 participants of a plan were terminated due to the closing down of one of the employer’s divisions. The ruling’s stated reason for such conclusion was that a “significant number” of participants were terminated.

Because the significant number test was announced in a revenue ruling, it is entitled to no special deference. Higgins v. Commissioner, 312 U.S. 212, 215 (1941); Helvering v. New York Trust, 292 U.S. 455 (1934); Stark v. Commissioner, 86 T.C. 243, 250-251 (1986). As has been stated by the Court of Appeals for the Fifth Circuit, the court to which an appeal in the instant case would lie:

a [revenue] ruling is merely the opinion of a lawyer in the agency and must be accepted as such. It may be helpful in interpreting a statute, but it is not binding on the Secretary [of the Treasury] or the courts. It does not have the effect of a regulation or a Treasury Decision. [Stubbs, Overbeck & Associates v. United States, 445 F.2d 1142, 1146-1147 (5th Cir. 1971).]

Accordingly, a ruling or other interpretation by the Commissioner is only as persuasive as her reasoning and the precedents upon which she relies. Gordon v. Commissioner, 88 T.C. 630, 635 (1987); Reinhardt v. Commissioner, 85 T.C. 511, 520 (1985); Wolfers v. Commissioner, 69 T.C. 975, 989 (1978); Hartman v. Commissioner, 34 T.C. 1085 (1960), affd. per curiam 296 F.2d 726 (2d Cir. 1961).

The rationale for the significant number test, however, is not readily apparent. The fact situation presented in Rev. Rul. 81-27, supra, shows that the plan experienced a 57.57-percent reduction in participation, which would have supported a finding that a partial termination occurred under the significant percentage test, discussed infra. Consequently, it does not appear that, at least in the situation presented in the ruling, resort to the significant number test was required to remedy a deficiency in the significant percentage test. Moreover, in Rev. Rul. 72-439, 1972-2 C.B. 223, the Commissioner did not rely upon the significant number test in concluding that a partial termination occurred when 120 of 170 plan participants were terminated, but based her holding exclusively on the significant percentage test.

The courts have been hesitant to accept the significant number test. One recent decision noted that

Although courts have given lip service to the significant number test, they have been reluctant to use this test as a basis for holding that a partial termination has or has not occurred. * * * No court has found a partial termination under the significant number test. [In re Gulf Pension Litigation, 764 F. Supp. at 1163-1164; citations omitted.]

In one case, a District Court refused to find a partial termination had occurred under the significant number test even though 415 participants were terminated where such terminees constituted only 2.5 percent of plan participants. Ehm v. Phillips Petroleum Co., 583 F. Supp. 1113, 1115-1116 (D. Kan. 1984); see also Morales v. Pan Am. Life Ins. Co., 718 F. Supp. 1297, 1302-1303 (E.D. La. 1989), affd. 914 F.2d 83 (5th Cir. 1990) (court considered only percentage test in holding no partial termination occurred where 128 participants, constituting 15.3 percent of participants, were laid off).

Only in Weil I, 750 F.2d at 12, has a court suggested that the absolute number of terminations should be determinative of the partial termination issue. Such suggestion, however, appears to be dicta, as the issue before the Court was whether the record in the case was sufficient to enable the District Court to grant a motion for summary judgment on the partial termination issue. Weil I, supra. The District Courts hearing such cases, moreover, have relied only on the percentage test in deciding the partial termination issue. In re Gulf Pension Litigation, 764 F. Supp. at 1164 n.6. Furthermore, the Court of Appeals for the Third Circuit has held that, in view of the language of section 411(d)(3) and the purpose of the partial termination rule, the appropriate focus of the inquiry was the “percentage of employees in the plan who were affected by the transaction said to constitute a partial termination.” Bruch v. Firestone Tire & Rubber Co., 828 F.2d 134, 151 (3d Cir. 1987), affd. in part and revd. in part on other grounds 489 U.S. 101 (1989). The Court of Appeals for the Second Circuit has apparently adopted the view of Bruch, for, in a subsequent ruling on the case which gave rise to Weil I, it held that the percentage decrease in plan participation was to be considered in deciding the partial termination issue, and made no mention of the significant number test. Weil v. Retirement Plan Admin. Comm., 913 F.2d 1045 (2d Cir. 1990) (Weil II), vacated in part on other grounds 933 F.2d 106 (2d Cir. 1991) (Weil III).

The Significant Percentage Test

The other test used to measure the drop in plan participation for purposes of considering whether a partial termination has occurred is the significant percentage test. Such test was also first announced in one of the Commissioner’s revenue rulings, Rev. Rul. 72-439, 1972-2 C.B. 223. Such ruling held that a 70.59-percent reduction in plan participation resulted in a partial termination of the plan. The significant percentage test has generally been accepted by courts ruling on the partial termination issue. See, e.g., Bruch v. Firestone Tire & Rubber Co., supra at 151; Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C at 159-160; In re Gulf Pension Litigation, 764 F. Supp. at 1163-1164. Indeed, one court has noted that the percentage of participants excluded from the plan is the most persuasive evidence of whether a partial termination has occurred. Kreis v. Charles O. Townley M.D. & Associates, P.C., 833 F.2d at 79.

The Test To Be Applied

After due consideration, we apply the significant percentage test in the instant case, rather than the significant number test. Doing so appears to us to be more consistent with the overall nature of the partial termination inquiry, an inquiry which focuses on the effect that a reduction in participation has on the plan as a whole. The legislative history of ERISA also suggests that emphasis on the relative size of the reduction is appropriate, giving as an example of a partial termination “a large reduction in the work force”. S. Rept. 93-383, at 50 (1973), 1974-3 C.B. (Supp.) 80, 129; H. Rept. 93-807, at 65 (1974), 1974-3 C.B. (Supp.) 236, 300. Accordingly, we think that Congress intended that the magnitude of a reduction in plan participation generally should be judged in relation to the size of the plan, and not according to the absolute number of persons terminated from participation. In other words, a workforce reduction of 1 percent should not be significant even if it involves 100 or 200 employees.

An additional reason for using the significant percentage test over the significant number test is that use of the significant number test would have a disproportionate impact upon larger plans and would unnecessarily complicate their administration. If, as Rev. Rul. 81-27, 1981-1 C.B. 228, superseding Rev. Rul. 72-510, 1972-2 C.B. 223, holds, exclusion of 95 participants can cause a partial termination, it would seem that even minor adjustments to a large workforce, or even normal turnover (if no major corporate event were required), could cause partial terminations to occur with some frequency.⁶ Consequently, the partial termination vesting rule could be turned into a mechanism for providing severance benefits instead of retirement benefits, a purpose which Congress did not intend the vesting rules to serve. 119 Cong. Rec. 30384-30385 (1973) (statement of Sen. Javits). Widespread application of the significant number test might thus discourage large employers from maintaining pension plans, contrary to the congressional policy of encouraging employers to provide such retirement benefits. S. Rept. 93-383, at 1-2 (1973), 1974-3 (Supp.) 80, 80-81.

Respondent argues that the regulations promulgated under section 411(d)(3) seem to permit use of the significant number test, as they provide that one of the facts and circumstances to be considered in deciding whether a partial termination has occurred is “the exclusion, by reason of a * * * severance by the employer, of a group of employees who have previously been covered by the plan”. Sec. 1.411(d)-2(b)(1), Income Tax Regs. The regulation, however, does not define the phrase “group of employees”. Accordingly, the regulation provides limited guidance in considering whether the significant number test is a valid test for deciding when a partial termination has occurred. Furthermore, the regulation does not suggest that the severance of a group of employees is to be the only circumstance considered; rather, the decision is to be made on the basis of all the facts, of which the number of affected employees is only one.

Given the foregoing considerations, we do not find that use of the significant number test would be necessary here to remedy inadequacies in the significant percentage test. In the more than 20 years since the latter test was first applied, it has proven useful in a wide array of situations. Consequently, we decline to find that a partial termination has occurred in the instant case solely on the basis of the significant number test. Accordingly, even though the number of employees affected by Halliburton’s reduction in force may be substantial or significant, such circumstance standing alone does not warrant a finding that the Halliburton plan experienced a partial termination in 1986.

The Measuring Percentage To Be Utilized in the Test

The significance of the percentage drop in establishing whether a partial termination has occurred generally depends on its size. The exact level at which it becomes determinative of the issue, however, is still somewhat ill defined. As one court noted, there is no “magical figure at which a partial termination occurs”. Bruch v. Firestone Tire & Rubber Co., 640 F. Supp. 519, 530 (E.D. Pa. 1986), affd. on this issue 828 F.2d 134 (3d Cir. 1987). Another court, however, has noted that often the percentage standing alone may be sufficiently large or small to decide the partial termination question.⁷ Kreis v. Charles O. Townley, M.D. & Associates, P.C., 833 F.2d at 80. For instance, based on the decided cases, a drop in participation exceeding 50 percent was sufficient standing alone to establish that a partial termination had occurred. See Peter M. Boruta, M.D., P.C. v. Commissioner, T.C. Memo. 1988-172, and cases cited therein. On the other hand, a drop of 2 to 3 percent has been thought to be sufficiently small to preclude a finding that a partial termination has occurred. Bruch v. Firestone Tire & Rubber Co., 828 F.2d at 151.

Between the two extremes, where the percentage drop has not been sufficient in and of itself to establish that a partial termination has occurred, the surrounding facts and circumstances have been considered in conjunction with the percentage drop in order to decide whether the percentage drop is significant. Kreis v. Charles O. Townley, M.D. & Associates, P.C., 833 F.2d at 79-80. A rule of thumb has developed that a percentage drop of at least 20 percent is sufficient, if coupled with other circumstances, e.g., the closing of a plant or division, to suggest that a partial termination has occurred.⁸ In re Gulf Pension Litigation, 764 F. Supp. at 1168; Morales v. Pan Am. Life Ins. Co., 718 F. Supp. at 1302-1303, and cases cited therein. A drop of less than 20 percent has been considered significant only if accompanied by egregious abuse on the part of the employer, such as discrimination in favor of the highly compensated, manipulation of the pension rules to obtain tax benefits, creation of a reversion of plan assets to the employer, or an attempt to prevent employees from becoming vested in accrued benefits. In re Gulf Pension Litigation, 764 F. Supp. at 1164.

The regulations direct that all facts and circumstances are to be considered in deciding whether a partial termination has occurred. Sec. 1.411(d)-2(b)(1), Income Tax Regs. Accordingly, we decline to accord talismanic significance to the 20-percent rule of thumb, but will regard the percentage drop in the light of the other facts and circumstances. As one court has said:

The significant percentage which serves to determine whether there has been a partial termination has not been reduced to any specific number. What constitutes a significant percentage is preeminently a matter of fact. * * * [Babb v. Olney Paint Co., 764 F.2d 240, 242 (4th Cir. 1985).]

As noted above, however, no evidence of abuse exists in the instant case that would justify lowering the level at which the percentage drop in plan participation would be deemed significant.

Defining the Class of Affected Employees

Our next step is to decide the size of the drop to be utilized in applying the percentage test. The basic area of disagreement between the parties and the principal question in the instant case is how to define the class of employees to be included in calculating the percentage reduction. Halliburton contends that the class should be defined restrictively, thus minimizing the number to be considered in reaching the percentage, while respondent contends that the class should be defined broadly, maximizing such number.

A body of case law has developed on the participants who should be counted in the class. It is clear that not every employee separated from service in 1986 should be included in the affected employee class. Certain employees would have terminated employment regardless of Halliburton’s circumstances, and so should not be counted. Others who left may not have been within the class of plan participants which Congress intended to protect under the partial termination rules, and are likewise to be excluded. Consequently, each category of employee separations must be examined to decide whether it should be included for purposes of the partial termination inquiry.

Classes the Parties Agreed To Exclude

The parties agree that certain classes of separated employees should not be counted in the partial termination percentage, and we accordingly exclude them. Such classes include: Deaths, 22; normal retirements, 38; terminations for cause, 264. The discontinuance of such employees’ participation in the plan is deemed not to be attributable to Halliburton’s reductions in force, and so such employees are deemed not to have been affected by such action. See In re Gulf Pension Litigation, 764 F. Supp. at 1168; Morales v. Pan Am. Life Ins. Co., 718 F. Supp. at 1302-1303. Moreover, the account balances of employees who died or retired at normal retirement age were fully vested upon occurrence of such events.

Halliburton also contends that 437 employees of the IMCO Division who were transferred to the M-I joint venture and ceased participating in the Halliburton plan should not be counted. Halliburton bases its contention on the fact that they were fully vested in benefits accrued through the end of 1986, which were spun off into the IMCO plan. Halliburton maintains that the IMCO plan should be treated as a continuation of the IMCO employees’ portion of the Halliburton plan and that a spinoff does not cause the discontinuance of the employees’ participation in the original plan. See sec. 414(1). Furthermore, Halliburton points out that the IMCO participants are covered under a successor pension plan, the M-I plan, maintained by a successor employer, the M-I Drilling Fluids joint venture. The decided cases support the view that the transferred employees should not be counted for purposes of deciding whether a partial termination has occurred. In re Gulf Pension Litigation, 764 F. Supp. at 1165-1166; Morales v. Pan Am. Life Ins. Co., 718 F. Supp. at 1302-1303. Moreover, on brief, respondent concedes that the transferred employees should be excluded.

Accordingly, the foregoing categories of former participants will not be considered for purposes of deciding whether a partial termination of the plan occurred in 1986.

Classes the Parties Agreed To Include

All parties agree that 4,560 partially vested participants who were involuntarily terminated in 1986 should be counted for purposes of deciding whether a partial termination of the Halliburton plan occurred in such year.

Classes oh Which the Parties Disagree

1. Voluntarily Separated Employees

Halliburton and respondent disagree over the question of whether employees who voluntarily relinquished employment during 1986, either through early retirement or voluntary resignation, should be counted in deciding whether a partial termination of the Halliburton plan occurred in that year. Respondent contends that such separations occurred as part of Halliburton’s efforts to reduce the size of its work force in such year, and so should be counted. Halliburton contends that such separations should not be counted because Congress intended the partial termination rules to protect only plan participants involuntarily separated from service. After due consideration, we agree with Halliburton.

The case law overwhelmingly holds that employees who voluntarily separate from their employer should not be counted for purposes of deciding whether a partial termination has occurred. Anderson v. Emergency Medicine Associates, 860 F.2d 987, 990 (10th Cir. 1988); Sage v. Automation, Inc. Pension Plan & Trust, 845 F.2d 885, 891 (10th Cir. 1988); Kreis v. Charles O. Townley, M.D. & Associates, P.C., 833 F.2d 74, 79 (6th Cir. 1987); Weil v. Retirement Plan Admin. Comm., 750 F.2d 10, 11-13 (2d Cir. 1984); In re Gulf Pension Litigation, 764 F. Supp. 1149, 1166 (S.D. Tex. 1991); Morales v. Pan Am. Life Ins. Co., 718 F. Supp. 1297, 1302 (E.D. La. 1989), affd. 914 F.2d 83 (5th Cir. 1990).⁹ Furthermore, the regulations indicate that only employer-initiated terminations suggest that a vertical partial termination has occurred by defining such event in terms of a “severance by the employer” of plan participants. Sec. 1.411(d)-2(b)(1), Income Tax Regs.

Even if an employee resigns, however, the termination may still be counted if it is shown that the participant was constructively discharged, meaning that the employee resigned due to intolerable working conditions created by the employer. Kreis v. Charles O. Townley, M.D. & Associates, P.C., supra at 81-82; Young v. Southwestern Savings & Loan Association, 509 F.2d 140, 144 (5th Cir. 1975). Moreover, employees who leave of their own accord after it is clear that there is no prospect of continuing to work for the employer may also be considered as having been involuntarily terminated. Collignon v. Reporting Services Co., 796 F. Supp. 1136, 1141-1142 (C.D. Ill. 1992); Morales v. Pan Am. Life Ins. Co., 718 F. Supp. at 1303.

In Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. 154 (1984), we did not directly address the issue of whether voluntary separations were to be included for purposes of the partial termination calculation; however, the employees were referred to as having been “discharged” or “terminated”, suggesting that employer-initiated separations are the relevant ones for such inquiry. Id.¹⁰ In the instant case, however, we hold that voluntary, non-employer-initiated terminations are not to be considered in deciding whether a partial termination has occurred.

Turning to the question of whether the employees who took Halliburton’s offer of early retirement or who resigned their positions were constructively discharged, resulting in involuntary separation from employment, we find that such was not the case. As Halliburton’s downsizing was company-wide, we do not accept, as respondent argues, that employees had no choice but to leave. The situation of the instant case is therefore distinguishable from one in which the company’s assets are sold or a discrete portion of the company, such as a division or plant, shuts down. Accordingly, the instant case is distinguishable from Collignon and Morales, involving a sale of the employer’s assets and termination of an entire division which led the courts to conclude that employees who resigned from their jobs should be included in the partial termination calculation. See Collignon v. Reporting Services Co., supra at 1142; Morales v. Pan Am. Life Ins. Co., supra at 1303. Furthermore, some employees might leave their jobs of their own accord, even when economic conditions are bad, because of dissatisfaction with their situations or for personal reasons. The partial termination rule was not intended to protect such persons — they do not have the same expectation with respect to the vesting of their accrued benefits that involuntarily separated employees do. Moreover, vesting their benefits would not necessarily deter abuse by employers.

Additionally, the circumstances of such separations do not indicate that such persons were constructively discharged from their positions. Almost all of the employees counted as voluntarily resigning filled out a form giving the reasons for such action. None of such employees stated that he or she was leaving because of anticipated discharge or layoff, nor did any claim that he or she had been forced to resign.¹¹ Accordingly, employees who are identified in the record as having voluntarily resigned their positions will not be counted in deciding whether a partial termination occurred.

Employees who chose early retirement similarly should not be counted for purposes of deciding whether a partial termination occurred. The circumstances indicate that such employees voluntarily relinquished employment. In re Gulf Pension Litigation, 764 F. Supp. at 1168. The record shows that Halliburton simply made a general offer of early retirement benefits to all eligible employees, and each individual was free to accept it or not. Halliburton did not force its employees to choose between taking early retirement or being laid off.¹² Although it is possible that the condition of Halliburton’s business was a factor weighed by some retirees in deciding whether to accept Halliburton’s offer of early retirement, such consideration does not make the resulting choice to retire involuntary.

Moreover, an early retirement offer does not amount to a constructive discharge where the employer gives the employee a choice between taking previously unavailable benefits or of continued work under the same conditions. Mitchell v. Mobil Oil Corp., 896 F.2d 463, 467 (10th Cir. 1990). The fact that an employee declining the early retirement offer may face the same risk of layoff as every other employee does not make acceptance of the offer coerced. Id.; Bodnar v. Synpol, Inc., 843 F.2d 190, 194 (5th Cir. 1988). Only where the offer sufficiently alters the status quo so that the choices available to the employee leave the employee in a worse position than before can acceptance of the offer be considered involuntary. Bodnar v. Synpol, Inc., supra at 193. In the instant case, the employees were given a choice between taking early retirement, thus obtaining the benefit of distribution options not available to terminees and a retiree medical benefit plan, and continuing to work for Halliburton, taking the same chance of termination faced by other employees. Furthermore, because the Halliburton plan provided that accrued benefits vested at age 50, no employees who were eligible for early retirement stood to lose any accrued benefits if such employees chose not to retire.

Based upon the record in the instant case, we find that the decision to take early retirement was the result of a rational weighing by the employees of the benefits offered early retirees by Halliburton versus the risks and rewards of continuing to work for Halliburton. Henn v. National Geographic Soc., 819 F.2d 824, 826 (7th Cir. 1987). Consequently, although Halliburton’s liberalized early retirement program was prompted by its need to reduce its workforce due to adverse economic conditions, we find that the purposes of the partial termination rule would not be advanced by taking the early retirees into account, because the employees’ decision to take advantage of the company’s offer was voluntary.

Accordingly, we will not take into account participants who either voluntarily resigned or took early retirement in deciding whether a partial termination occurred.

2. Fully Vested, Involuntarily Separated

An additional point on which the parties disagree is whether fully vested employees laid off by Halliburton should be counted toward the partial termination number and percentage.

We have not previously expressly ruled on such question; in Tipton & Kalmbach, we counted all participants leaving the plans, without stating whether they were fully or partially vested, in ruling that a partial termination had occurred. Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. 154 (1984). In the instant case, 224 fully vested participants were involuntarily terminated. Halliburton argues that such participants should be excluded because they do not stand to gain or lose any benefits based on whether the plan experienced a partial termination. Halliburton argues that the purposes of the partial termination rule, namely to protect participants from forfeiting rights to accrued, but unvested, benefits and to prevent employers from reaping a windfall due to reversions of amounts on which they have paid no tax, would not be served by considering participants who forfeit nothing on leaving the plan. Respondent argues that it is the significant contraction of a plan, and not the consequences of such event, which is determinative of whether a partial termination has occurred, and urges that all involuntarily terminated participants be counted, regardless of vesting status. We, however, need not decide the question because, even if we include such participants, as we set forth more fully below, the plan did not experience a partial termination. We accordingly include such employees in the calculation purely for illustrative purposes.

3. Participants With No Accrued Benefits at Separation

Two hundred thirty-one participants who terminated participation in the plan during 1986 had no accrued benefits at the time they separated from service because they had not become eligible to receive an allocation by such time. Under the Halliburton plan, an employee must be a participant on December 31, 1986, or have terminated participation due to death, disability, or retirement in the year in order to receive an allocation of employer profit-sharing contributions or forfeitures for such year. Consequently, employees in their first year of participation separating prior to December 31, 1986 for reasons other than death, disability, or retirement, accrued no employer-funded benefits and so forfeited nothing when they left the plan.

Halliburton contends that, because such participants forfeited nothing, they should not be treated as “affected employees”, as they stand to neither gain nor lose from a decision that a partial termination of the plan occurred. Halliburton argues that only participants whose vesting rights to accrued benefits are put at risk by the partial termination should be considered affected employees. Halliburton accordingly uses the same rationale advanced to exclude fully vested employees from the relevant class. Respondent, on the other hand, argues that such terminees should be counted because resolution of the partial termination issue should turn on an examination of the reduction in plan participation as a whole, regardless of the vesting status of the separated participants. As with the fully vested involuntary terminees, our decision in the instant case would not be affected whether we included them or not, so we decline to decide the issue. We include them in the calculation, however, purely for illustrative purposes.

4. Participants Transferred to Affiliates of Halliburton

Respondent contends that the six participants transferred to Halliburton affiliates besides M-I should be treated as affected employees because they were not accorded the same treatment as participants transferred to M-I, such as full vesting in their account balances under the Halliburton plan. It appears, however, that such participants were accorded sufficient protections to warrant their exclusion from the affected class. See In re Gulf Pension Litigation, 764 F. Supp. at 1165-1166. In the instant case, a transferred employee did not forfeit unvested benefits under the plan simply by reason of transfer, but such employee was allowed vesting credit under the plan for service to the affiliate, which enabled continued vesting of benefits accruing while he was a participant. Furthermore, such employee became a participant in the affiliate’s plan, and vesting of benefits accruing thereunder was based on all service with members of Halliburton’s controlled group. Moreover, there is no indication that the transfers were related to the events causing the decrease in plan participation. Accordingly, we will exclude such transferees from the affected class.

5. Separated Participants Subsequently Hired Back

Between 1987 and 1989, 1,124 participants who had been involuntarily separated from service in 1986 were hired back by Halliburton and recommenced participation in the plan.¹³ Halliburton argues that the rehired participants should not be considered “affected employees” because they were only temporarily laid off, and forfeited amounts were available for restoration to the accounts of partially vested participants upon reemployment. Respondent argues that such employees should be included in the affected class because only the events of 1986 are relevant to the issue presented here and permitting the number of affected employees to be adjusted by the number of rehires would prolong and complicate the process of ascertaining whether a partial termination had occurred.¹⁴ We do not agree that we should limit ourselves to considering the events of only 1 plan year in resolving the partial termination issue, as the events bearing on such issue may extend over more than 1 such year. In re Gulf Pension Litigation, 764 F. Supp. at 1167.

Halliburton further argues that the partial termination rule contemplates a permanent, rather than a temporary, reduction in plan participation, citing In re Gulf Pension Litigation, supra at 1163, and accordingly concludes that employees who are only temporarily absent from their jobs should not be counted in the partial termination calculation. Halliburton also notes that such rehires either had their forfeitures restored or were given the opportunity to obtain restoration upon rejoining the plan, and so were in the same position with respect to their accrued benefits as they were when they originally left the plan. Halliburton notes that various courts have suggested that temporary reductions in plan contributions and participation might not constitute terminations; however, no court has yet explored the issue in any detail. Indeed, in Tipton & Kalmbach, we specifically reserved ruling on the issue of whether a temporary work force reduction could cause a partial termination under the facts and circumstances test. Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. at 162 n.9.

After due consideration, we agree with Halliburton that the terminated participants who were rehired should not be considered in deciding whether a partial termination occurred. We cannot, as respondent argues, treat such circumstance as irrelevant. The regulations direct that all facts and circumstances must be considered, and, to the extent such information is in the record, we cannot simply ignore it. Moreover, a temporary workforce reduction does not appear to fit within the concept of the partial termination, and other courts have suggested as much. For instance, in Collignon v. Reporting Services Co., 796 F. Supp. at 1141, the court described partial termination of a plan as “An employer-initiated permanent reduction of a significant percentage of employees * * * as part of a major corporate event” (emphasis supplied). Moreover, where a suspension of contributions to a plan is undertaken as a temporary measure in response to unfavorable business and economic conditions, such suspension is not considered indicative of a termination of a plan. Babb v. Olney Paint Co., 764 F.2d 240, 243 (4th Cir. 1985).

Holding that a large drop in plan participation, no matter how short in duration, could cause a partial termination would greatly and unnecessarily complicate the administration of plans and would needlessly discourage their establishment. There are occasions in the ordinary course of business when an employer must temporarily lay off workers, where such layoffs do not portend any substantial change in the business’ structure or scope of activity. For instance, a layoff may occur while a factory is retooled to produce a new product, or a plant may be forced to shut down if a strike or natural disaster elsewhere temporarily interrupts the supply of parts or raw materials needed to maintain production. In such instances, the employees are usually rehired within a relatively short time, and treating the plan as partially terminated due to the temporary dip in participation caused by such work interruptions makes no sense. The oil price collapse and resulting reduction in demand for Halliburton’s services, which caused the drop in plan participation in the instant case, constituted a similar temporary business dislocation. We do not accept respondent’s argument that we should define a “temporary” layoff by reference to the standard employed under section 162 for deciding whether expenses incurred when a worker is temporarily away from home are deductible.

Nonetheless, respondent raises valid concerns that must be addressed in deciding the extent to which rehiring shall be taken into account. Respondent’s primary concern is that taking rehiring into consideration will complicate and delay the process of ruling on partial termination issues. Respondent notes that a worker may be rehired up to 5 years after termination and still be eligible for restoration of his forfeited account balance. Sec. 411(a)(6)(D). However, the longer the interval between the rehiring and the event allegedly causing a partial termination, the less probative such rehiring is with respect to such issue. If a drop in workforce is truly temporary, the bulk of the rehiring should occur within a relatively short time after the initial drop in participation. As time goes by, the likelihood of rehiring terminated participants declines significantly as they find new jobs or otherwise leave the labor pool available to the employer. In the instant case, the majority of the rehiring occurred in 1987, while much smaller numbers were hired back in 1988 and 1989, indicating that such workers were only temporarily discharged.

Employers who reduce their workforce due to adverse economic conditions must also distinguish their situations from that presented in Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. 154 (1984), by rehiring terminated participants in a relatively rapid manner. In Tipton, the workforce reduction was permanent, and participation levels did not recover in the years following such reduction. Id. at 162. In Tipton, a permanent downsizing or restructuring in response to adverse business conditions occurred, suggesting that a partial termination of a plan had taken place.

The instant case is distinguishable from Tipton because staffing levels began to recover almost immediately after the layoffs. Toward the end of 1986, the price of oil was increasing because oil producers decided to return to a fixed price system. The number of operating rigs also began to rise with the price of oil. Consequently, Halliburton put substantial numbers of the laid-off employees back on its payroll in the years immediately following 1986. Accordingly, in the instant case, the rehired participants should not be included in the class of affected employees for purposes of the percentage test.

We note further that, even if we were to include the rehired participants in the class of affected employees, we find it appropriate to consider the fact of such rehiring in deciding whether such reduction is significant, among all circumstances to be considered. Sec. 1.411(d)-2(b)(1), Income Tax Regs. Such rehiring is relevant to deciding whether such reduction was temporary or permanent in nature. The rehiring of a substantial number of terminated participants within a short time after the layoffs suggests that such layoffs were a temporary response to transient economic conditions, and a reduction in participation under such circumstances is not indicative of a partial termination of the plan, especially where forfeited amounts were restored to partially vested participants upon reemployment.

In summary, we hold that for purposes of deciding whether a partial termination has occurred, the only participants to be considered are those who are permanently excluded from the plan due to employer-initiated severances which are prompted by the circumstance or event causing the contraction of the plan.

Calculation of the Denominator of the Significant Percentage Fraction

Generally, the denominator of the significant percentage fraction is the number of plan participants at the beginning of the plan year in which the partial termination is alleged to have occurred plus the number of participants added during such year. Tipton & Kalmbach, Inc. v. Commissioner, supra at 156. In the instant case, there were 19,017 participants in the Halliburton plan on January 1, 1986, and 581 new participants were added in 1986, for a total of 19,598.

Respondent, however, argues that the transferred IMCO Division employees should be excluded from the denominator of the significant percentage fraction, arguing that fully vested participants who remained in employment, even with another employer, should be excluded from both numerator and denominator in order to avoid distortion of the significant percentage fraction. We do not agree. The transferred employees were participants in the plan at the beginning of the plan year in question, which has generally been accepted as a starting point for calculating the denominator. See id.; In re Gulf Pension Litigation, 764 F. Supp. at 1168.¹⁵ Consequently, the transferred employees who were participants in the plan at the beginning of the plan year would ordinarily be included in the denominator, regardless of how their circumstances changed during the year. At the beginning of 1986, the IMCO Division employees were indistinguishable from other plan participants. Some were fully vested and others partially vested in their accrued benefits. Consequently, in measuring the decline in workforce size for purposes of deciding whether a partial termination has occurred, including them is proper.

Furthermore, the cases that have dealt with transferred employees in the context of partial terminations have not excluded such employees from the denominator, even though they were not counted in the numerator. In re Gulf Pension Litigation, 764 F. Supp. 1149 (S.D. Tex. 1991); Morales v. Pan Am. Life Ins. Co., 718 F. Supp. at 1302. Respondent’s argument that “distortion” may occur if the IMCO transferees are included is vague, and we do not perceive any distortion by including such transferees. Moreover, respondent does not argue that all fully vested participants should be excluded from the denominator, as was done in In re Gulf Pension Litigation, supra at 1165 n.10, and so we do not consider whether the transferees should be excluded on such ground. Accordingly, we do not exclude the IMCO transferees from the denominator. Thus, the denominator of the significant percentage fraction is 19,598.

Whether Percentage of Plan Participants Terminated Is Significant

Synthesizing the foregoing analysis, we hold that at most 3,891 participants who ceased participating in the Halliburton plan in 1986 should be counted for purposes of the significant percentage test. While we have included fully vested and zero balance participants in our calculation, we do so simply for illustrative purposes and do not decide whether such participants should be excluded or included, as such issue will not affect the outcome of the instant case. Such figure is calculated as follows:

Involuntarily separated participants

Fully vested 224

Partially vested 4,560

Zero balance accounts 231

5,015

Less:

Rehired involuntarily separated participants

45 Fully vested

1,039 Partially vested

40 Zero balance accounts

(1,124)

Net reduction in participation 3,891

For purposes of the significant percentage test, such figure serves as the numerator of the fraction, of which the denominator is 19,598, comprising all participants in the Halliburton plan on January 1, 1986, plus participants entering the plan in such year. The resulting percentage is 19.85.¹⁶

No egregious circumstances exist which would justify concluding that a percentage of less than the 20-percent rule of thumb is significant. Rather, it is clear that the layoffs were an emergency response to a temporary collapse in demand for Halliburton’s oil field services due to a drop in the price of oil. As drilling and production activity recovered, Halliburton rehired many laid-off participants, enabling them to regain amounts subject to forfeiture when they were laid off in 1986. Consequently, the 19.85-percent reduction is not sufficiently large to be significant.

We accordingly conclude, based on the entire record, that a partial termination of the Halliburton plan did not occur in 1986. Accordingly, the forfeitures arising out of the reduction in plan participation are to be allocated among continuing participants as provided under the plan. The parties appear to have agreed that if we find that no partial termination of the plan occurred, the plan and the IMCO plan are qualified under section 401(a). To reflect the foregoing,

Decision will be entered for petitioner in docket No. 26290-90R.

An appropriate decision will be entered in docket No. 2397-91R.

1 In the Tax Reform Act of 1986 (TRA), Congress changed the graded vesting schedule to require that a participant begin to vest after no more than 3 years of service and be completely vested after no more than 7 years of service. Pub. L. 99-514, sec. 1113(a), 100 Stat. 2446. Such amendment applies to plan years beginning after Dec. 31, 1988. TRA sec. 1113(e), 100 Stat. 2448.

2 A “spinoff’ is defined as “the splitting of a single plan into two or more plans.” Sec. 1.414(1)-1(b)(4), Income Tax Regs.

3 In a subsequent submission, Halliburton stated that 423 IMCO participants had transferred to the M-I joint venture; however, based on other information in the administrative record, we find that such statement is in error, and should have read 437. We have accordingly treated the number of IMCO transferees as 437.

4 The parties agree that the forfeitures constituted only 4.8 percent of plan assets in 1986.

5 We note that the Court of Appeals for the Third Circuit has indicated that a partial termination should be found only if the reductions in participation are so large that a plan appears to have been created as a mechanism to reduce taxes, rather than to provide benefits. Bruch v. Firestone Tire & Rubber Co., 828 F.2d 134, 151 (3d Cir. 1987), affd. in part and revd. in part on other issues 489 U.S. 101 (1989). However, we perceive no conflict between Bruch and our decision in Tipton & Kalmbach, Inc. v. Commissioner, 83 T.C. 154 (1984), because it seems that any reduction large enough to give rise to an inference of employer misconduct would also result in a finding that a partial termination had occurred under the facts and circumstances test as applied in Tipton. We also find significant that the Court of Appeals for the Third Circuit did not require direct evidence of abuse in order to sustain a finding of partial termination, but allowed the test of whether abuse was present to be inferred through the size of the reduction in plan participation. Consequently, the result under either approach would be the same.

6 For instance, Halliburton terminated 264 plan participants in 1986 for causes unrelated to its business downturn. Under the significant number test, such a number of layoffs could suggest that a partial termination occurred, a plainly absurd result.

7 We note, however, that the most recent revision of respondent’s Plan Termination Handbook states that there is no fixed turnover rate which determines whether a partial termination has occurred. Plan Termination Handbook, 4 Administration, Internal Revenue Manual (CCH), sec. 252(6), at 21,150.

8 There are a number of possible reasons such figure came into use. A drop of 20 percent or more must be reported by the plan administrator to the plan sponsor under the Employee Retirement Income Security Act of 1974 (ERISA), Pub. L. 93-406, sec. 4043(b)(3), 88 Stat. 1024. Additionally, respondent’s Internal Revenue Manual formerly instructed agents to question whether a partial termination had occurred when exclusion of 20 percent of plan participants occurred. Plan Termination Handbook, 4 Administration, Internal Revenue Manual (CCH), sec. 252, at 21,150. It has also been suggested that the 20-percent rule has its origin in a pre-ERISA annual report required to be filed with respondent by plan sponsors. The form asked whether a 20-percent drop in participation had occurred during the year. Laarman & Hildebrandt, Tax Management Portfolio 357, Plan Terminations and Mergers A-18(2) (1986).

9 Respondent as well has indicated acceptance of such position. In G.C.M. 39,344 (Oct. 16, 1984), it was concluded that a partial termination did not occur where the drop in participation resulted from the employees’ voluntarily relinquishing employment. The memorandum reasoned that employees do not need the protection from employer abuses afforded by the partial termination rule where they voluntarily leave their jobs.

10 We note that some commentators have suggested that the Tax Court has held that voluntary separations are to be counted toward a partial termination. Laarman & Hildebrandt, Tax Management Portfolio 357, Plan Terminations and Mergers, states at A-18(3) (1986):

In Tipton & Kalmback, [sic] Inc. v. Comr., for example, the Tax Court, agreeing with the IRS, held that a significant percentage of participants leaving in a single year is automatically grounds for a partial termination regardless of whether the employer initiated the departures. As a result, the opinion suggests that an employer can be involuntarily placed in a partial termination situation as a result of the departure of a significant percentage of employees, a result that can be particularly onerous for small employers because the departure of even a single participant may constitute a significant percentage.

A careful reading of Tipton & Kalmbach, however, reveals no such holding, nor can any portion of the opinion be read to imply such a view of the law. Rather, the opinion seems to assume that all participants excluded from the plan were discharged by the employer, and does not state that any left of their own accord.

11 Although some employee comment letters in the administrative record suggest otherwise, none of the commenting employees claim that they were compelled to falsely state that they were leaving voluntarily. Consequently, we do not find the statements in the comment letters to outweigh departing employees’ earlier statements in the forms. Respondent also challenges Halliburton’s characterization of the reasons offered by such participants for leaving their jobs, but, given that we must accept Halliburton’s representations in the administrative record as true for purposes of the instant case, Rule 217(b), we do not question their accuracy.

12 Certain comment letters claim that employees were forced to take early retirement, but such allegations seem based on secondhand information. No person filing a comment letter claimed to have been forced to retire. Accordingly, we give no weight to such comment letters.

13 A total of 1,279 employees were hired back during such period. Eighty had elected early retirement, 1 had retired at or over normal retirement age, 67 had voluntarily resigned, and 7 had been terminated for cause. Since employees in such categories have been excluded from the partial termination calculation, it is inappropriate to count them in arriving at the proper number of affected employees. Only employees who would otherwise have been counted should be deducted from the total as rehired employees in order to avoid excluding the same employees twice. Because fully vested, partially vested and zero balance participants are counted, hirebacks in such categories are taken into account.

14 Respondent’s argument is contradicted somewhat by the Plan Termination Handbook, which instructs respondent’s agents to consider “the extent to which terminated employees are replaced” in determining whether a partial termination has occurred. Plan Termination Handbook, 4 Administration, Internal Revenue Manual (CCH), sec. 252(6), at 21,161. Of course, respondent is not bound by statements in the Internal Revenue Manual. Capitol Fed. Sav. & Loan v. Commissioner, 96 T.C. 204, 216-217 (1991), and cases cited therein.

15 While a general counsel memorandum is not precedential, Biddle v. Commissioner, 302 U.S. 573, 582 (1938), we note that G.C.M. 39,344 (Oct. 16, 1984), reaches the same conclusion.

16 Counting only partially vested involuntary terminees produces a reduction percentage of 17.97 percent, calculated as follows: (4,560-1,039)/19,598.