ALTHOUGH PENSION PLANS are initiated as permanent programs, they are subject to discontinuance, as is the existence of the sponsoring company. The recent ending of a few large plans, coupled with a widespread assumption that terminations lead to a considerable loss of earned pensions, has kindled interest in data on terminations. With the cooperation of the Internal Revenue Service, the Bureau of Labor Statistics has studied the causes and effects of termination and the characteristics of plans closed out between 1955 and 1965.

Of 8,100 qualified retirement plans terminated during the years 1955-65, over half were pension plans,' as the following tabulation shows:

The remainder were profit-sharing, stock bonus, and thrift plans, and are not involved in the present inquiry.

Data abstracted from the termination record files of the Internal Revenue Service (IRS) show that terminating plans tended to be young and small in size. More than half were no more than 6 years old, and two-thirds covered fewer than 25 employees. Although a variety of circumstances may lead to termination, the most frequent reasons given were company and plan mergers, financial difficulties, and business dissolution.

The number of terminations included in this study differs significantly from that reported in quarterly IRS releases. This study relied on IRS termination records (Form 517T); the quarterly releases report the number of determination letters Issued.

This estimate does not account for employees who lost their Jobs and, unless vested, their pension rights in a business decline preceding termination.

11 Years' Experience

These 4,300 plans covered approximately 225,000 employees at the time of termination. Thus, on the average, about 20,000 workers a year were affected-about one-tenth of 1 percent of total pension plan coverage. All accrued pension rights were not lost, however. Some rights were undoubtedly preserved by fund accumulations or, in many instances, by the transfer of accrued pension credits to other plans. Only a continuance of corerage in another plan, however, assures participants of benefits for future employment.

Frequency of Terminations

A marked upward trend in the frequency of pension plan terminations is evident from the data shown in the table. The increase is a reflection more of the spread of private pension plans than of any significant change in the rate of plan termination, which has consistently remained around 1 percent of active plans.

The rise in the number of terminations did not follow a smooth path-experience fluctuated in a manner that, in part, reflects the influence of changing business conditions. For example, the greatest rise in the number of pension terminations occurred in 1961, a year of relatively low economic activity; the number fell during the following year of general recovery. The relationship, however, is by no means perfect. In 1963, a relatively good business year, not only did terminations rise, but much of the increase was attributed to financial difficulties.

Although changes in general economic conditions and in the economic characteristics of firms with pension plans influence the rate of plan termination, there is little reason to expect a radical change in the rate in the coming decade. The termination rate, as previously mentioned, was largely unaffected by the moderate economic downturns of the 1955-65 period. Similar experience is

Reasons for Termination

Financial difficulties were frequently cited as the primary reason for terminating a plan. Other reasons, such as company and plan mergers or the sale of individual plants, were also prevalent. Financial difficulty was given as the reason for 1 out of 4 terminations, and business dissolution for 1 out of 5. Mergers and sales, which are difficult to separate, accounted for 3 out of 10. In at least a third of these terminations by merger or sale, pension coverage of employees was not continued.

The frequency of these several reasons for terminating a plan tends to vary from year to year (chart 1). Some reflect adverse business conditions, others may not. Financial difficulty and business dissolution accounted for a larger part of the annual totals when the level of general economic activity was relatively low and business conditions less favorable. It is interesting to note that the role played by business dissolution, the final step, conforms quite closely with that for financial difficulty if adjustment for a 1-year lag is made. Changes in the role of mergers and sales, on the other hand,

were not closely related to general business activity. Rather, experience followed a relatively consistent upward trend. In 1961, for example, merger or sale was cited more often than in the preceding year, but represented a smaller part of the total. The remaining reasons, which were far less prevalent, did not show either a definite upward trend or fluctuations related to general economic conditions.

The reasons for plan termination tend to differ significantly with plan size (chart 2), and to a lesser extent with age. Financial difficulty is most frequently cited in terminations involving small plans; organization change-business dissolution, sale or merger-are more prominent among the larger plans. Financial difficulty accounted for a third of the terminations affecting less than 50 employees, a fifth of those with 50 to 100 employees, and a lesser proportion among larger plans. Mergers and sales tended to increase with plan size, accounting for a fourth of the terminations involving fewer than 10 workers, and about a third among large plans. Business dissolution followed a similar pattern. Experience among the less prevalent reasons revealed either no pattern or one that is easily explained. Lack of interest and few eligible employees, for instance, were reasons offered mainly by small plans.

Extent of Benefit Losses

Most pension plans do not, at any one point in time, have sufficient resources to fully discharge all of their liabilities. While benefits earned after the inception of a plan are funded as they accrue, a substantial unfunded liability usually is created when a plan is established by giving participants either full or partial credit for earlier service. Additional unfunded liability may, and usually does, arise because of subsequent, liberalizing plan amendments. Under both circumstances, immediate full funding of these liabilities is not practicable, not only because IRS rules discourage funding at the rapid rate that would be required, but also for reasons of prudent management of a company's financial resources. Although some em

On IRS forma, a single reason is cited for each termination. Since the various reasons are not necessarily mutually exclusive, some distortion of their relative importance is unavoidable. Financial dificulties, for example, may be a contributory factor lending to the sale, merger, or dissolution of a company.

The assets of a terminating plan may also prove to be inadequate to satisfy all of its obligations because of investment losses, failure of the employer to make contributions, or adverse actuarial errors resulting from an overly optimistic projection of a plan's income or an understatement of its Liabilities.

Pension plans, with few exceptions, limit an employer's financial obligations to the amount of his contributious, i.e., any deficit in the plan's finances is not chargeable against company assets in case of default. The participants, consequently, must assume the burden of any asset deficiency upon termination, unless their pension credits are transferred to another plan. Available resources are allocated among the participants as specified in the plan or as agreed to at the time of termination. Priority orders are frequently set up to favor the older and long-service members, although pro rata allocations are sometimes made to all participants.

A case in point is the Studebaker plan termination. When the Studebaker plant in South Bend, Ind., was closed and the pension plan terminated, workers with at least 10 years of service and age 60 or over-i.e., those retired or eligible to retirelost no benefits. Workers with 10 years of service or more and between ages 40 and 59-those with vested rights to benefits-received 15 percent of the value of their accrued benefits. The rest of the participants—those without rested_rights—received nothing.

Aside from such well-publicized cases, only the most fragmentary data are available on the extent of participant losses of expected benefits through plan terminations. IRS termination records do not contain the information needed to determine the frequency and magnitude of accrued benefit

Liabilities as reported under the Disclosure Act are generally greater than the value of accrued benefits and still greater than the value of vested benefits. Although reported Habilities are determined by a variety of actuarial methods. nearly all methods level out the sharply rising cost of promised benefits by estimating liabilities during a plan's early years at a level well above the average cost of providing those benefits. The difference is usually so great that it is only partly offset by the accrued lability owing to credits given for service before the introduction of the plan. See Frank L. Griffin, Jr., "Pension Se curity and Funding Regulation," Proccedings, Conference of Actuaries in Public Practice, 1964–65, p. 135.

losses. In an effort to obtain some information, reports filed under the Welfare and Pension Plans Disclosure Act were examined for a group of 99 terminated plans, each with 100 participants or more. Due to one deficiency or another, rough estimates of the extent of potential participant losses could be made for only 26 of the plans.

By comparing assets to liabilities, it was evident that in 10 of the 26 plans some participant losses were incurred. These 10 plans had slightly over 10,000 members, including 8,500 reported by Studebaker. The assets of these plans, as a group, averaged about one-half of their reported liabilities, but benefit losses probably averaged less than this fraction suggests. Six other plans, with 2,400 members, also reported insufficient assets to fund their accrued liabilities; however, there were no apparent losses since the participants were transferred to other plans. The remaining 10 plans, with 2,300 members, appeared to be fully or almost fully funded; if any losses occurred in these in

47-752 O 7033

stances they were probably nominal. These 26 plans may have been more thoroughly funded than the typical terminating plan because they were older and, consequently, had a longer opportunity to improve their funding positions.

The general lack of pertinent financial information frustrates any effort to determine the value of benefits lost through the plan terminations included in this study. Some reasonable inferences as to the magnitude of loss in typical cases can be drawn, however, by assuming hypothetical, but more or less traditional, funding patterns. Funding practices largely determine the relation between plan resources and accrued benefit obligations, especially during the early years in the life of a plan when asset appreciation usually is minor and liberalizing amendments are least likely.

Employers generally adopt one of several actuarial methods that eliminate abrupt fluctuations and sharp increases in the amount of yearly contributions. The more customary methods will, on the average, fund between 20 and 40 percent of a plan's accrued benefit obligations by the end of its fifth year of operation." Between 45 and 65 percent of the accrued benefits will be funded by the end of the 10th year, if there have been no major plan amendments or changes in asset values during the intervening years. Even if these occur, funding experience is likely to fall within this range because amendments and changes in asset

values tend to offset each other. Their net influence becomes less predictable with advancing age; hence, it is not practicable to suggest a meaningful range of ratios for older plans.

Translated into benefit losses, these funding patterns suggest that, unless coverage is continued through the transfer of credits to another plan, workers stand to lose between 60 and 80 percent of their total accrued benefits if their plan terminates in its fifth year. Most of this loss, however, is attributable to service prior to the inception of the plan. At 10 years, the total loss will range between 35 and 55 percent, assuming the net effect of plan amendments and changes in asset values is rather modest. (In all cases, a system of priority may allocate the loss among participants from none to 100 percent.)

Actuarial conjectures such as these may be of limited value for many purposes. Reasonably accurate estimates of the magnitude of benefit losses cannot be obtained from any government reporting system now in operation. Unless such reporting systems are changed, only a special survey program can produce more reliable data.

The system an employer selects may provide for substantially lower contributions once the liabilities for service before inception of the plan have been funded, e.g., 20 or 30 years. However, such reductions are seldom realized because of periodic plan amendments.

Op. cit., footnote 4.