Title: Employer Trustees of The American Federation of
Musicians and Employers’ Pension Fund and Union Trustees of The American
Federation of Musicians and Employers’ Pension Fund
the Union Trustees
the Employer Trustees
February 10, 2003
On October 18, 2000, the Employer and Union Trustees of the American Federation of Musicians and Employers’ Pension Plan (“Plan”) deadlocked on four motions. In accordance with Article 6.6 of the Agreement and Declaration of Trust Establishing the American Federation of Musicians and Employers’ Pension Fund (“J-2), the Trustees chose to “submit the question for decision” to me.
I held hearings in New York City, at the offices of the Plan, on July 9, 10, and 11, 2001. At those hearings, both groups of Trustees were independently represented by counsel. They entered a Joint Stipulation of Facts describing how the deadlock occurred. Both groups of Trustees were given a full opportunity to present evidence, documents, and arguments in support or opposition to the motions. At the close of the hearings, the parties agreed to have closing arguments at their next scheduled Trustees’ meeting. The events of September 11, 2001, intervened and made that impossible. I heard closing arguments on November 30, 2001, after which I declared the hearings closed.
The parties agreed that technically, and under federal law, my task is simply to vote yes or no on each of the four motions. Since these motions have caused atypical disharmony among the Trustees, however, they asked for an Opinion to accompany this Award to provide some guidance for future actions. Finally, the parties agreed that if I were to vote yes on any of the deadlocked motions, I would make the change effective as soon as administratively practicable, with the understandings that the Trustees would confer and attempt to agree on a date for implementation. If they were unable to agree, I would be asked to decide that issue in a telephone conference call.
1. MOTION to modify the current pre-retirement death benefit offered by the Fund as follows:
2. MOTION to modify the current post-retirement death benefit offered by the Fund as follows:
3. MOTION to modify the current covered earnings threshold requirements solely for purposes of earning credit for vesting and participation under the Fund as follows:
4. MOTION to modify the current covered earnings threshold requirements for purposes of avoiding breaks-in-service under the Fund as follows:
The arbitrator’s role in a trustee deadlock is to act as a tiebreaker by voting “yes” or “no” on each of the deadlocked motions. The arbitrator cannot change a single word of a deadlocked motion. While acting as a tiebreaker, the arbitrator must observe the same standards as a Trustee, whose fiduciary duties are found in ERISA Section 404(a)(1). Thus, in deciding how to vote on each deadlocked motion, I must act solely in the interest of the participants and beneficiaries.
The trustees have a good faith difference of opinion as to what is in the best interest of the participants and beneficiaries, as evidenced by the deadlock. The Employer Trustees made four motions to change the way death benefits, vesting, and breaks-in service are calculated. As the proponents of change, they have the obligation of showing that each of the changes is necessary and appropriate to further the best interest of the participants and beneficiaries. If they fail to make that showing on any motion, I must vote against it.
The Employer Trustees have propounded three questions that embody the general principles by which the propriety of their motions should be judged. First, is the rule in its current form appropriate for pension funds in general? Second, is the rule appropriate in light of this Plan’s design? Third, how do the existing and proposed rules affect the administrative efficiency of the Plan? The Union Trustees generally agree that these are appropriate principles for testing the Employer Trustees’ motions. Their Union Trustees’ argument for the status quo is based on the assertion that the Employer Trustees cannot demonstrate a need for any of the proposed changes.
In this Opinion, I evaluate each of the motions in accordance with these standards and principles.
Motion 1: Modify pre-retirement death benefit to provide 100 times the monthly benefit that actually accrued, or would have accrued, if the participant died at 55.
Vote: Yes. The motion is adopted.
The Employer Trustees argue that the current benefit is inappropriately large for a pension plan death benefit, inconsistent with the Plan design, and costly to the Plan.
According to the Segal Report of January 18, 2000 (J-7), the Plan’s current pre-retirement benefit is unique among entertainment industry funds. With a single exception, the funds provide the qualified pre-retirement survivor annuity (“QPSA”) required by ERISA, or a benefit based upon the participant’s accrued benefit payable upon her death (or the earliest retirement date, if that is later). A single industry fund determines the pre-retirement death benefit by uniformly multiplying the accrued age 65 benefit by 48, regardless of age at death. Thus, only the Plan calculates the pre-retirement death benefit by using both the accrued age 65 benefits and a multiplier that ranges from 65 to 100. It is a uniquely generous calculation.
The change to the pre-retirement death benefit proposed by the motion would still provide a more generous benefit than the QPSA required by ERISA, or any pre-retirement death benefit in the industry. Only one industry plan provides a pre-retirement death benefit that multiplies the accrued benefit at death (or earliest possible retirement date) by more than 100, but that plan does not provide for a lump sum benefit. Thus, the change effected by the Employer Trustees’ motion would result in a pre-retirement death benefit that is more generous than any in the industry.
The change is also consistent with plan design. In an October 1966 actuarial review, the Segal Company responded to a suggestion that the Trustees increase the death benefit so that a vested participant would be guaranteed 100% of the amount contributed on his behalf. Segal explained how the pre-retirement death benefit could be made to return 100% of the participant’s contributions, based upon the then-existing benefit formula that paid $1.25 per $100 of contributions at age 65. (J-16, p.45) In 1969, in connection with a proposed increase to $1.40 per $100 contribution, the Segal Company noted that the death benefit would have to be reviewed if the pension benefit were increased. It suggested reducing the multiplier from 100 to 90, which would still guarantee a return of 126% of contributions. (J-17, p.36) Over the years, the regular monthly pension has increased to $4.65 per $100 of contribution, so that the value of the pre-retirement death benefit for anyone under 55 is more than three times the contribution. Under the proposed change, it is 2.3 times the contribution, which retains a full return of contributions, plus some interest.
Finally, while the Employer Trustees do not argue that the cost of the benefit is prohibitive, they do assert that it is inappropriate. The purpose of the Fund is to provide pensions, not life insurance. To the extent that the pre-retirement death benefit is greater than the actuarial value of the accrued benefit, the Fund is subsidizing the benefit. If it were not subsidizing this benefit, the money that represents a subsidy could be used to increase pension benefits. Since pensions are the purpose of the Fund, the benefit is far more generous than the pre-retirement death benefit provided by other industry funds, and the proposed change keeps the benefit among the most generous in the industry, the Employer Trustees argue they have shown the necessity and propriety of this change.
The Union Trustees concede, as they must, that the benefit is generous by industry standards. They argue, however, that it is popular with participants, who have come to rely upon it instead of buying life insurance. The benefit’s popularity is incontrovertible. On February 21, 1991, the Trustees unanimously adopted a change in the pre-retirement death benefit that is identical to the current motion. That change was rescinded on September 25, 1991. The minutes read: “It was noted by several Trustees that the modifications had prompted adverse reactions by numerous plan participants.”(J-12, p.10) Union Trustee Schwarz testified to circulating a petition and getting 1150 names in opposition to the change before the vote to rescind. (Tr. 549:7-22)
It is more difficult to argue, however, that participants have detrimentally relied upon the benefit so that it cannot be changed. Mr. Rosen, a 35-year member of Professional Musicians Local 47, testified that he considered the death benefit when deciding how much life insurance he needed when he re-married ten years ago. He subsequently developed certain medical conditions that made him a “substandard risk” for life insurance purposes. Consequently, life insurance to cover the reduction in the pre-retirement death benefit is now significantly more expensive for him than it would have been at the time of his decision. (Tr.521-523) He testified that others depended upon the benefit to forego buying additional life insurance and then had illnesses that made buying additional insurance extremely expensive.
While it is entirely understandable that Mr. Rosen and others in his circumstances would properly feel disadvantaged by a change in the pre-retirement death benefit, that does not create a detrimental reliance that the Trustees must recognize. It would have been preferable for the Trustees to develop a phased implementation of the change, so that participants could plan for it. But it is beyond my authority to change the resolution so that it phases in the change. It is impossible to determine the number of people who are similarly situated to Mr. Rosen. The number of vested participants is known. They will all be better served by having the pre-retirement death benefit made more consonant with industry standards and the savings made available for Fund stabilization or benefit improvement. I am persuaded that the pre-retirement death benefit should be reduced to the level provided in the Employer Trustees’ motion.
Motion 2: Modify the post-retirement death benefit to provide benefits equal to 100 times the benefit actually accrued at benefit commencement or, if greater, the amount the participant was entitled to immediately before the effective date of this change.
Vote: Yes. The motion is adopted.
The Employer Trustees argue that the current benefit is inappropriately large, inconsistent with the plan design, and costly to the Fund. It is all of these.
The post-retirement death benefit appears to be more generous than any of the other industry funds. There is considerable variation within the industry, with guarantees ranging from 24 to 120 monthly payments, reduced by the number of payments already received. (E-10, J-7, pp.6-8) The most common benefit is a guarantee of 60 months, with some guarantees payable in a lump sum. None of the plans with a 120-month guarantee permits a lump sum payment. This Fund is unusual in having a 100 month guarantee, combined with a lump sum payment. Consequently, a retiree’s beneficiary may collect a lump sum amount that exceeds the actuarial value of the pension to which the retiree was entitled. This is a unique subsidization of the death benefit.
The post-retirement death benefit grew to its current level in the same way as the pre-retirement death benefit. As the Fund grew, it increased the amount paid per $100 in contributions, without changing the formula for the post-retirement death benefit. It began as a return of contributions—and perhaps something more. It currently returns 4.65 times contributions. For early retirees, it may pay more than the amount reserved by the Fund for the individual’s retirement payments. Since the benefit is payable in a lump sum, it can be invested to produce the same retirement payment as promised by the Fund, without invading principal. The Union Trustees have not argued that the plan was originally designed to provide greater death benefits than retirement benefits to retirees. Nor does there appear to be any evidence from which such an argument can be made. Thus, at some ages the post-retirement death benefit is a subsidized, disproportionate benefit. To the extent that the plan design originally intended to return contributions with some interest, the current post-retirement death benefit does far more than that.
The post-retirement death benefit is also inappropriately costly. The Employer Trustees demonstrated the benefit has grown disproportionately generous by industry standards, and is subsidized because it returns more than has been reserved for retirement. The Union Trustees have failed to offer a cogent rationale for providing more in post-retirement death benefits than a participant would have been entitled to in retirement benefits. As with the pre-retirement death benefit, the money saved by reducing this benefit to a level that is consistent with the best plans in the industry can be used to stabilize the Fund or increase retirement benefits.
Motion 3: Triple the covered earnings threshold for receiving a quarter of credit for vesting and participation.
Vote: No; the motion is defeated.
The Employer Trustees’ motion triples the amount of covered earnings required to receive a quarter of vesting or service credit (from $375 to $1125), thereby tripling the covered earnings required for a year of vesting or service credit (from $1500 to $4500). They make three arguments to support trebling the “covered earnings” threshold. First, they argue that the current threshold of $375 in covered earnings for a quarter of vesting credit is unusually low in comparison to other plans in the entertainment industry. In the Employer Trustees’ view, it permits people with no real attachment to the industry, or people who have demonstrated no real commitment to covered work, to receive pensions, albeit very small ones. This, in the Employer Trustees’ view, is out of keeping with industry standards. Second, the Employer Trustees argue that the plan design is to have the dollar threshold act as a proxy for an amount of work that demonstrates attachment to the industry. The threshold was raised once after the inception of the plan, but not since then. Because of inflation related increases in salaries, the amount of covered work needed to earn a quarter of vesting credit has decreased. Thus, in the Employer Trustees’ view, the earnings threshold must be increased in order to continue to reflect the amount of work originally required. Third, they argue that the cost of administering the small pensions obtainable under the current covered earnings threshold is excessive in comparison to the benefit received. Thus, increasing the covered earnings threshold will save administrative expenses that are currently being inappropriately expended. I examine each of these arguments below.
First, the Employer Trustees argue that the plan’s current threshold for earning vesting credit is low by general and industry standards. They note that in general the law requires 1,000 hours of service as a threshold for earning pension credit. (Tr.702:10-17) In other industry plans cited by the Employer Trustees, the vesting requirements varied from “any service” to “100 days of service” and from $4,000 to $25,000 in covered earnings. (E-11) Of the ten plans cited by the Employer Trustees, four calculate the benefit based solely on years of service, three provide a benefit tied to years of service plus a formula that accounts for contributions made, one has a formula based upon the average daily contributions by the employer, one has a formula based on salary over and under a certain level, and one has a formula based solely on earnings. (U-49) The AFTRA plan seems most like the AFM plan in that it is based solely on covered earnings. Unlike the AFM plan, however, the AFTRA plan does not appear to limit contributions to scale. The AFTRA plan formula covers annual earnings between $50,000 and the “IRS maximum.” The Employer Trustees have not shown that the employment circumstances of AFTRA members are so similar to those of AFM members that the AFTRA $5000 threshold for a year’s vesting credit provides an appropriate standard.
The general standard for pension vesting of more than 1,000 hours of work in a year is not helpful in determining the appropriate level of work that should make musicians eligible for a year of vesting credit. Musicians are not compensated for the enormous amount of time they must practice in order to keep their skills at a level that makes them salable. The three-hour studio recording session (for which a musician receives scale of around $240) is preceded by unpaid six-hour daily practice sessions in the days and weeks before the job. Similarly, industry comparisons are not helpful since they show so much variability. The measure of annual work needed for a vesting credit, the basis for calculating the regular pension, and the role of individual earnings differs from plan to plan. The only statement that can be made with assurance is that each plan reflects the realities of work in the craft or profession it covers. Consequently, the evidence does not support the Employer Trustees’ argument for increasing the earnings threshold based on comparisons with general standards or other plans in the entertainment industry.
Second, the Employer Trustees make three “plan design” arguments to support tripling the amount of “covered earnings” required for a vesting credit. First, they generally argue that “covered earnings” are a measure of “attachment to the industry,” which should be the basis for determining whether a musician gets a pension from the Fund. That is, the Fund does not exist to provide tiny pensions to unsuccessful musicians who have since gone on to other work but still play enough weekends to avoid a break-in-service. AFM pensions are for musicians who make their living in the industry. Second, the Employer Trustees note that the plan has had a threshold for earning vesting credits since its inception. They assert that this threshold – although expressed in dollars –previously measured the amount of covered work required to demonstrate “attachment to the industry.” As the result of increases in wages over the years, the amount of work a musician has to do to earn a vesting credit has decreased. Thus, in the Employer Trustees’ view, people who do not have an “attachment to the industry” now qualify for pensions. Third, the Employer Trustees argue that the purpose of pensions is income replacement in retirement and the tiny pensions that can be earned under the current vesting rules do not serve that purpose.
The Employer Trustees’ first plan design argument fails because it is unclear that “covered earnings” is an adequate surrogate for “attachment to the industry.” The Employer Trustees use “covered earnings” to mean only earnings under a collective bargaining (or participation) agreement that require pension contributions. As the testimony shows, however, not all AFM collective bargaining agreements require pension payments. According to one witness: “There is a great deal of union work that is not covered by pension.” (Tr.445:13-14) Indeed, the pension records of some extremely successful and well-known professional musicians suggest that for much of their work no pension contributions were made. (See, U-31 through 34) There is no question that these are professional musicians with a full-time attachment to the industry. Their work under collective bargaining agreements, however, did not always result in a pension contribution. Thus, the evidence does not support the Employer Trustees’ assertion that the amount of a musician’s “covered earnings” generally demonstrates that musician’s “attachment to the industry”.
The Employer Trustees’ second plan design argument, that tripling the dollar threshold for vesting merely accounts for growth in musician’s wages and continues the plan’s intended use of the vesting threshold as a surrogate for work in the industry, suffers from two infirmities. First, there is no convincing evidence that musician’s scale wages (upon which pension contributions are made) have tripled since 1977. The evidence, such as it is, is to the contrary. Wages reported to the fund upon which pension contributions were made increased by 69% from 1979 to 1997 and on a per capita basis 105%. (E-9) Second, the Employer Trustees have not established that there was any intention to use the dollar threshold as a surrogate for work in the industry. The evidence does not reveal why the original threshold was set at $1200 or why it was raised to $1500 in 1977.
The Employer Trustees also suggest that the original threshold was actually designed to create a certain level of “breakage.” At the plan’s inception in 1965, half the musicians for whom contributions were made each year did not qualify for any vesting credit. Currently, one-third of the musicians for whom contributions are made each year do not qualify for any vesting credit. If the threshold is tripled, half the musicians for whom contributions are made each year will not qualify for any vesting credit. (Tr.119 ff) That will return the “breakage” to what it was at plan inception. The Employer Trustees provide no good reason for returning a mature plan, with sufficient earnings to pay benefits, to a design that pays benefits to the few through forfeitures by the many. None of the plans for which the Employer Trustees expert Dr. Kra currently does actuarial work has a breakage rate as high as one-third. (Tr.127:21-128:8) There is no justification for significantly increasing the number of participants whose employer contributions will be used to finance the pensions of the highest earning musicians.
There is no support for the Employer Trustees’ third plan design argument that the historical purpose of this Plan was to provide income substitution for musicians who retire, and that this purpose is not served by the small pensions earnable by musicians under the current vesting threshold. The testimony shows that many professional musicians work a multiplicity of jobs in order to piece together a living. One of the witnesses described the life of a professional musician in his area as “driving for dollars.” (Tr. 467:19) They drive between regional orchestras to play or substitute in order to get enough “services” to make a living. For many, the pension earned by their “covered earnings” is only a portion of what they are trying to put together for retirement. The pensions awarded by the Fund support this testimony about the life of professional musicians. In 2000, the average monthly pension awarded (adjusted for optional forms of payment) was $533. Over the last ten years that average has risen from $391 per month. (J-14, p.2-3) As of March 31, 2000, of the 8,000 persons receiving pension benefits, a quarter receive under $100 a month and more than half receive under $400 per month. (J-14, p. 2-4) It does not appear that this plan was designed to provide musicians, who may play for hundreds of employers in a single year, with a single source of income substitution.
The Employer Trustees’ final argument is that the administrative cost of the small pensions that can be earned under the current vesting scheme is excessive. If the vesting threshold were tripled, it would eliminate the administrative cost associated with servicing the small pensions-- by eliminating the small pensions. The Employer Trustees have not presented any direct data about the number of extremely small pensions or the cost of servicing them. The per person administrative cost to the Fund is $200, which is the result of dividing the administrative costs by the total number of people on whose behalf contributions are made. If the vesting threshold were raised, much of the administrative cost would remain the same, since all of the contributions have to be recorded and tracked over the years. The only direct difference would be that half the musicians for whom contributions were made would get nothing. Thus, the Employer Trustees have failed to show that savings in administrative costs justifies trebling the vesting threshold.
One additional factor argues against the Employer Trustees proposal. The predictable effect of trebling the threshold will be to provide a disincentive for negotiating pension payments with unionized employers who do not currently make pension contributions, and for negotiating increases in the percentage contribution of those employers contributing less than 4%. As those with collective bargaining experience noted, the employer generally has a finite amount of money for a wage package. It can be put into wages or benefits, but it is finite. If musicians believe it is unlikely they will ever benefit from the employer’s pension contribution, there is no point in negotiating a pension contribution in lieu of wages. If the earning requirement is trebled, increasing the breakage rate to fifty percent, there will be excellent reason for non-vested participants to believe they will never receive any benefits. The excellent record of the Fund in increasing benefits is irrelevant if you do not believe you will ever qualify for a benefit.
The evidence shows that an area in which contributions to the Fund have doubled in the past six years (from under $6 million to over $12 million) is “Symphony, Opera, and Ballet.” (U-15a) In recent years, many of the regional orchestras have become contributors to the Fund, while others have not yet agreed to pension contributions. For musicians playing in some of these smaller orchestras, the current vesting threshold would provide a year’s vesting credit for a year’s service in that orchestra. (U-18) Tripling the threshold will make that impossible. Thus, if the Employer Trustees proposed tripling were adopted, there would be no incentive to negotiate pension contributions from these regional orchestras. And those who have negotiated contributions under 4% would have no incentive for increasing those contributions. In general, it is healthier for a fund to broaden its contribution base. Tripling the threshold for vesting credit will work against broadening the contribution base.
Motion 4: Triple the covered earnings required for avoiding breaks-in-service.
Vote: No, the motion is defeated.
The Employer Trustees’ arguments for tripling the covered earnings required to avoid a break in service are closely related to their arguments for tripling the earnings required for vesting and participation. They offered a separate resolution on breaks-in-service so the arbitrator would have the option of giving current vested participants an opportunity to avoid breaks-in-service while trying to meet a trebled threshold for earning vesting credit. Since I rejected the resolution to raise the covered earnings required for vesting, it would be illogical to support this resolution, since it would have similar adverse effects.
By reason of the foregoing, I make the following:
On deadlocked resolutions 1 and 2, adjusting the pre-retirement death
benefit and the post-retirement death benefits, I vote with the Employer
Trustees. The resolutions are
On deadlocked resolutions 3 and 4, tripling the covered earnings
threshold for earning vesting and participation credit and tripling the covered
earnings required to avoid breaks-in-service, I vote with the Union Trustees.
The resolutions are defeated.
The Trustees will determine when the changes in death benefits can
be made, taking into account administrative practicability.
In accordance with the stipulation of the parties, if the Trustees are
unable to agree on the implementation date of the changes, I retain jurisdiction
to issue a supplemental Award, after a telephone conference call with counsel.
San Francisco, CA _ ____ _________________
February 10, 2003
 The current pre-retirement death benefit is the monthly benefit the participant would have accrued at age 65 times 100 if 60 or older at death, times 90 if between 55 and 59, and times 65 if the participant is under 55 at death.
 Local One Stagehands (E-10).
 I recognize that this change was previously implemented and later rescinded. That might not have left the Trustees with the option of attempting a phased implementation.
 The formula is: age 65 accrued benefit times 100, minus monthly benefits paid prior to death, regardless of age at death. The Employer Trustees motion on the post-retirement death benefit differs from the motion on the pre-retirement death benefit. It provides a guarantee that no participant who retired in reliance on the existing post-retirement death benefit will be affected by the change. Thus, no participant who has retired at the time the benefit is changed will be affected by the change.
 Five out of ten (E-7); four out of seven (J-10).
 The illustration used by Dr. Kra ignored the effect of taxes, but nonetheless demonstrated that the benefit is inappropriately large.
 Local One Stagehands, Treasurers and ticket sellers, Porters and Cleaners, and Wardrobe.
 Scenic Artists, Equity League, and Society League.
 The Union Trustees assert that the diminution of work opportunities in the industry makes it more difficult to find covered work. (See, e.g. Tr.389:1-5)
 While the Employer Trustees recognize the gulf between earnings under a collective bargaining or participation agreement and earnings in the industry, their view is that if the bulk of a musicians work is non-covered, that musician should not use the AFM as a vehicle for retirement.
 The plan originally had a $300 threshold in 1965, which was raised to $375 in 1974, but implementation was delayed until 1977.
 The parties agreed that current law permits the fund to make a single lump sum payment to participants who would be awarded pensions with a net present value of less than $5,000.
 The evidence shows that collective bargaining agreements containing pension contributions of less than 4% require special approval by the Trustees. They have usually approved those agreements. (Tr. 354:19-355:8) The Union Trustees view this as a “foot in the door” that will allow them to negotiate increases in pension contributions in future years.