Charles R. Mcdaniel v. The Chevron CorporationCharles R. Mcdaniel v. The Chevron Corporation
COUNSEL: Edward S. Zusman, Zelle & Larson, San Francisco, California, for the plaintiff-appellant.
Robert A. Gordon (on the briefs) and Craig E. Stewart (argued), Pillsbury Madison & Sutro, San Francisco, California, for the defendants-appellees.
Edgar Pauk, Legal Services for the Elderly, New York, New York for the amicus. Mary Ellen Signorille, American Association of Retired Persons, Washington, District of Columbia, for the amicus. Paula Brantner, National Employment Lawyers Association, San Francisco, California, for the amicus.
OPINION
ALARCON, Circuit Judge:
Charles R. McDaniel, a participant of the Chevron Corporation Retirement Plan, and a class of similarly situated plaintiffs (collectively, the “class“) appeal from the district court‘s final order granting summary judgment in favor of the Chevron Corporation (“Chevron“), the Chevron Corporation Retirement Plan (the “Chevron Plan“), and the Chevron Corporation Retirement Plan Administrator (the “Plan Administrator“).1 The class primarily contends that the Plan Administrator violated the unambiguous requirements of the Chevron Plan in calculating the amount of retirement benefits that must be paid to participants who elect to receive actuarially equivalent benefits, including a single lump sum payment of cash. The restatement of the plan provided that those benefits would be calculated either “based on” or “in accordance with” the mortality assumptions set forth in the “UP-1984 Mortality Table.” The Plan Administrator interpreted the Chevron Plan‘s reference to the “UP-1984 Mortality Table” as authorizing a nine month set forward to avoid understating the correct mortality assumptions due to the difference in life expectancy of males and females.
We have jurisdiction pursuant to
I
The Chevron Plan is a defined benefit pension plan that is governed by ERISA and tax qualified under S 401(a) of the Internal Revenue Code (“IRC“). See
The Chevron Plan calculates the amount of contributions to be made, and determines the amount of benefits that the plan must pay, by using an interest rate assumption and a mortality assumption. An interest rate assumption is used to estimate the present value of benefits to be paid in the future. A mortality assumption, which is derived from a standard mortality table, is used to estimate the probability that a person will survive, and therefore receive benefits, to a given age. At issue in this case is whether Chevron used the correct mortality assumptions when it calculated the class members’ benefits.
Actuaries prepare standard mortality tables based upon the mortality experience of a large segment of the population. See William W. Fellers & Paul H. Jackson Noninsured Pensioner Mortality: The UP-1984 Mortality Table, 25 The Proceedings 456, 456 (1975). In the late 1970s, Chevron began to use the UP-1984 Mortality Table for calculating the amount of contributions that it makes to the Chevron Plan. The UP-1984 Mortality Table is a “unisex” mortality table that is based on the mortality experience of a blended population that is roughly eighty percent male and twenty percent female.2 See id. at 483, 485. If used without adjustment, the table is appropriate for calculating mortality assumptions for populations that have a ten percent to thirty percent female content. See id. For populations whose female content is either greater than thirty percent or less than ten percent, the authors of the UP 1984 Mortality Table recommend adjusting the table backward or forward a number of years to avoid overstating or understating the correct mortality assumptions.3 See id.
For funding purposes, the Chevron Plan uses an all male version of the table to calculate its male benefit liabilities, and it used an all female version of the table to calculate its female benefit liabilities. Following the author‘s instructions, the plan‘s actuaries created the all male version by setting the UP1984 Mortality Table forward one year, and they created the all female version by setting the table backward four years. See Fellers & Jackson, at 460, 483, 485.
McDaniel is a former Chevron employee and plan participant who elected to have his benefits paid in a single lump sum distribution of cash. The Plan Administrator calculated McDaniel‘s benefits using the UP-1984 Mortality Table with a nine month set forward and paid him roughly $49,900. Had the Plan Administrator used the UP-1984 Mortality Table without a set forward, McDaniel would have been entitled to an additional $405.
In June of 1995, McDaniel filed an administrative claim for additional benefits, contending that the Plan Administrator was required to calculate plan participants’ actuarially equivalent benefits using the UP-1984 Mortality Table without a set forward. Chevron denied that claim. McDaniel appealed to Chevron‘s review panel. The review panel consisted of three randomly selected senior level Chevron employees, who were also plan participants. The review panel concluded that the correct interpretation of the disputed mortality assumption was the UP-1984 Mortality Table with a nine month set forward and, accordingly, denied McDaniel‘s claim.
McDaniel‘s claim prompted Chevron to conduct an internal review of the plan‘s mortality assumption. Concerned that the description of the mortality assumption was not sufficiently specific to qualify the plan for special tax treatment, Chevron voluntarily approached the Internal Revenue Service (the “IRS“) through its anonymous Walk-in Closing Agreement Program (“CAP“). See
On August 13, 1996, McDaniel filed a complaint in the district court alleging that Chevron underpaid McDaniel‘s benefits in violation of S 502(a) of ERISA. See
All persons who between January 1, 1984 and October 1, 1996 commenced receiving benefits from the Chevron Retirement Plan that were calculated using 85% of the UP-1984 Mortality Table with a nine month age set forward as the mortality assumption, or the spouses, beneficiaries, executors, successors, representatives, heirs or assigns of said persons to the extent that the application of the nine-month age set forward resulted in a reduction of benefits when compared to the benefit calculated using 85% of the UP-1985 Mortality Table without a nine-month age set forward as the mortality assumption.
After the class was certified, both sides filed cross motions for summary judgment. The class argued that it was entitled to judgment on the basis that ERISA required Chevron to pay benefits using eighty-five percent of the mortality factors set forth in the UP-1984 Mortality Table without a set forward and that Chevron‘s use of the UP-1984 Mortality Table with a nine-month set forward violated the anticutback rule in S 204(g) of ERISA. See
II
As a threshold matter, we must decide the proper standard for reviewing the Plan Administrator‘s decision to deny the class members the benefits they claim are due to them under the Chevron Plan. In so doing, we are mindful that the de novo standard of review normally applies when a court reviews a claim that a plan administrator improperly denied benefits under S 502(a)(1)(b) of ERISA. See
The Chevron Plan provides that the Plan Administrator has the “sole discretion to interpret the terms of the Plan” and that those interpretations “shall be conclusive and binding.” We conclude that the reservation of discretion to the Plan Administrator is sufficiently clear to overcome the presumption in favor of de novo review. See Bendixen v. Standard Ins. Co., 185 F.3d 939, 943 & n.1 (9th Cir. 1999) (applying the abuse of discretion standard where the plan stated “we have full and exclusive authority to interpret the Group Policy . . . . [and] any decision we make . . . is conclusive and binding“); see also, e.g., Jones v. Laborers Health & Welfare Trust Fund, 906 F.2d 480, 481 (9th Cir. 1990) (applying the abuse of discretion standard where a plan specified that the plan administrator “shall have power . . . to construe the provisions of this Trust Agreement and the Plan, and any such construction adopted . . . shall be binding“). Nevertheless, the class contends that the district court should have applied a de novo standard of review because the Plan Administrator “interpreted” a federal statute or, alternatively, that it should have applied a “less deferential” standard of review because the Plan Administrator acted under a conflict of interest. We review de novo a contention that the district court did not apply the proper standard of review. See Lang v. Long-Term Disability Plan, 125 F.3d 794, 797 (9th Cir. 1997).
We reject the class‘s argument that the de novo standard of review applies to the Plan Administrator‘s interpretation of the disputed language. That argument is based on a mistaken assumption that the Plan Administrator interpreted parts of ERISA or the IRC when it denied McDaniel‘s claim. While it is true that we cannot defer to a plan administrator‘s construction of a federal statute, the class has failed to demonstrate that Chevron has invaded the “province of statutory interpretation.” Arnold v. Arrow Transp. Co., 926 F.2d 782 (9th Cir. 1990). The record shows that the Plan Administrator merely interpreted the meaning of the plan‘s mortality assumption when it denied McDaniel‘s claim. It did not interpret federal law.
Under the “less deference” test, a plan participant must come forward with “material, probative evidence, beyond the mere fact of the apparent conflict, tending to show that the fiduciary‘s self interest caused a breach of the administrator‘s fiduciary obligations.” Atwood v. Newmont Gold Co., 45 F.3d 1317, 1322-23 (9th Cir. 1995) (setting forth the standard); see also Tremain, 196 F.3d at 976; Bendixen, 185 F.3d at 189. Only if the plan participant satisfies that burden does the burden of proof shift to the plan administrator to produce evidence that the apparent conflict of interest did not affect the decision to deny benefits. See Atwood, 45 F.3d at 1323 (stating that the plan administrator‘s decision is “presumptively void“). If the burden shifts and the plan administrator does not produce evidence that the apparent conflict of interest did not affect the decision to deny benefits, we will “review the decision de novo, without deference to the administrators tainted decision.” Atwood, 45 F.3d at 1323; see also Lang, 125 F.3d at 797-98; Tremain, 196 F.3d at 977. Otherwise, we will review the plan administrator‘s decision under the traditional abuse of discretion standard. See Bendixen, 185 F.3d at 943-44; see also Atwood, 45 F.3d at 1322 (stating that the less deference standard is two-tiered).
In this case, the class contends that Chevron was operating under a “serious conflict” of interest when it interpreted the disputed mortality assumptions as providing for a nine-month set forward. It bases that contention on the following allegations:
- Chevron was the plan sponsor and the plan administrator, and Chevron employees denied its claim;
- the class‘s claim, if successful, will have a significant monetary impact on the plan;
- the individuals who were responsible for implementing the plan‘s mortality assumptions influenced the review panel;
- the attorney who drafted the plan‘s mortality assumptions are representing Chevron in connection with this lawsuit;
- Chevron did not disclose to the IRS the existence of either McDaniel‘s claim or the class action; and
- Chevron misled McDaniel in order to delay resolution of his claim pending the outcome of the CAP process.
The class has failed to demonstrate that the Plan Administrator‘s interpretation of the disputed mortality assumptions was “born out of the apparent conflict of interest” that it faced while serving as both the plan sponsor and the plan administrator.
The class has also failed to present any material, probative evidence that the individuals responsible for drafting and implementing the plan‘s mortality assumptions, including Chevron‘s attorneys, influenced the review panel out of “selfinterest.” Those who were responsible for creating the plan‘s mortality assumptions did not participate in the decision making process. While they provided information to the review panel regarding their methodology, that is not sufficient to create a serious conflict of interest. The information they provided was consistent with the prior operation of the plan. See Tremain, 196 F.3d at 977-78 (finding that the use of two different definitions of “disability” was probative of a serious conflict); Lang, 125 F.3d at 798-99 (finding that inconsistent positions are evidence that a plan‘s decision was tainted by self interest) (citing Brown v. Blue Cross & Blue Shield of Ala., Inc., 898 F.2d 1556, 1569 (11th Cir. 1990)). The fact that some of the attorneys responsible for drafting the amendment may be representing Chevron in the present lawsuit is irrelevant. This action did not arise until after the review panel issued its decision in McDaniel‘s administrative proceeding.
The class‘s contention that Chevron purportedly misled the IRS and delayed resolution of McDaniel‘s claim pending the outcome of the CAP proceeding are also unsupported by the record. Chevron complied with the deadlines for handling claims. It informed McDaniel of its participation in the CAP program and disclosed to the IRS the nature of the plan‘s defect and that a participant had filed an administrative claim. The fact that Chevron failed to disclose to the IRS the existence of this lawsuit, which arose five months after the review panel denied McDaniel‘s claim, is irrelevant to the inquiry of whether the Plan Administrator was operating under a conflict of interest when it denied McDaniel‘s claim. Accordingly, the district court did not err in reviewing Chevron‘s interpretation of the plan‘s mortality assumption under the traditional abuse of discretion standard.
III
The class contends that the district court erred in concluding that the mortality assumptions were ambiguous. It argues that the reference to “UP-1984 Mortality Table ” in the disputed mortality assumption plainly required Chevron to calculate benefits using the UP-1984 Mortality Table without a set forward. The class also contends that the district court failed to consider the “unique position” of class members who elected to have their actuarially equivalent benefits paid “in accordance with” eighty-five percent of the mortality assumptions set forth in the UP-1984 Mortality Table. Chevron asserts that the disputed mortality assumptions were ambiguous and that the Plan Administrator‘s interpretation was reasonable. It argues that the words “based on” and “in accordance with,” which qualified the references to the UP1984 Mortality Table in the disputed mortality assumptions, reasonably support an alternative interpretation that the assumptions could be calculated using the UP-1984 Mortality Table either set forward or set backward a number of months.
Whether the terms of a plan are plain or ambiguous is a question of law that we must review de novo. See Patterson v. Hughes Aircraft Co., 11 F.3d 948, 950 (9th Cir. 1993). We have held that terms in a pension plan should be interpreted “in an ordinary and popular sense as would a[person] of average intelligence and experience.”
The Chevron Plan calculates actuarially equivalent benefits using mortality assumptions set forth in a standard mortality table. Actuaries prepare standard mortality tables based upon the mortality experience of a large segment of the population. See Fellers & Jackson, at 456. The most important consideration in preparing and selecting a mortality table to be used in calculating pension benefits is whether the population from whom the mortality experience is developed is sufficiently broad and has characteristics that are typical of the plan‘s participants. See id.
The proportion of males and females in a population is one such characteristic that is particularly important. See id. at 468. Because females, on average, enjoy a greater life expectancy than males, a standard mortality table that is prepared from the mortality experience of a population with a greater percentage of females than is typical of the plan will tend to understate the mortality assumption for each plan participant. For the same reason, a mortality table that is based on the mortality experience of a population with a relatively lower percentage of females will tend to overstate the mortality assumption for each plan participant. The practical effect of either understating or overstating a mortality assumption is to treat each plan participant as either surviving longer or dying earlier than should actuarially be expected. Changing the life expectancy of a plan participant, of course, can increase or decrease the amount of actuarially equivalent benefits to which he or she is entitled. It is not surprising, then, that standard mortality tables typically recommend that the composition of males and females in the study population be adjusted in order to reflect the true gender profile of a particular plan. See id. at 459.
The UP-1984 Mortality Table is a standard mortality table. It was prepared using the mortality experience of a “blended” population that was roughly twenty percent female and eighty percent male and is valid for calculating the mortality assumptions for a population that is between ten percent and thirty percent female. See Fellers & Jackson, at 483, 485. The authors of the table instruct plan actuaries to set the table forward whenever the percentage of females participants falls below ten percent and to set the table backward whenever the percentage of female participants exceeds thirty percent. See id. In fact, it is common for plan actuaries to adjust the UP1984 Mortality Table to reflect the true percentages of males and females participating in a given plan. See id. at 459. Some of those adjustments are reflected in regulations that have been prescribed by the Pension Benefit Guarantee Corporation.
Given that the UP-1984 Mortality Table can be, and in fact is typically, adjusted forward or backward, the exact meaning of a plan provision that incorporates the table depends on the language that describes the method in which the table is to be used. A plan provision whose language contemplates the UP-1984 Mortality Table as a starting point or as a foundation for determining mortality factors does not reasonably preclude an interpretation that sets the table forward or backward. Nor does a provision whose language provides that the plan administrator will calculate mortality factors in conformity with the methods provided for in the table itself. In each of those cases, it would be reasonable to interpret the mortality assumption either as providing for a set forward or a set back or as not providing for a set forward or a set back.
In the context of statutory interpretation, courts have held that the plain meaning of “based on” is synonymous with “arising from” and ordinarily refers to a “starting point” or a “foundation.” See, e.g., Gould, Inc. v. Mitsui Min. & Smelting Co., 947 F.2d 218, 221 (6th Cir. 1991) (stating that “based upon” as used in the Foreign Sovereign Immunities Act should be read plainly as “arising from“); Werre v. David, 913 P.2d 625, 631-32 (Mont. 1996) (concluding that, under the plain meaning of “based on,” a claim is “based on intentional conduct” if the intentional conduct is the starting point or the foundation for the claim); State ex rel. Snidow v. State Bd. of Equalization, 17 P.2d 68, 69-73 (Mont. 1932) (holding that “gross value . . . shall be determined . . . based upon the average quotations” plainly meant that the gross value would be calculated using the average quotations as a “starting point“); Nevada v. Hutchings, 795 P.2d 497, 497, 501 (Nev. 1990) (holding that salaries would be “set based upon ” prevailing rates did not mean that salaries would be set “at ” prevailing rates); C.J.C. v. Corporation of the Catholic Bishop, 985 P.2d 262, 267 (Wash. 1999) (concluding that the plain meaning of “based on” refers to the “starting point or foundation“). We have also used the phrase “based on” in a similar manner when describing the law of ERISA. Compare Rodrigues v. Herman, 121 F.3d 1352, 1355 (9th Cir. 1997) (stating that S 404 of ERISA is “based on the common law of trusts“) with Varity Corp. v. Howe, 516 U.S. 489, 496 (1996) (stating that fiduciaries under ERISA “draw much of their content from the common law of trusts“). That construction not only is consistent with the definition that most dictionaries attribute to the phrase, but it also reasonably supports an interpretation that the Chevron Plan allows its lump sum benefit to be calculated using a set forward or a set back so long as the standard version of the table serves as the starting point or as the foundation. See, e.g., Webster‘s 3d New Int‘l Dictionary 180 (Philip Babcock Gove ed. 1979) (stating that “base” means “found” and is typically used with “on” or “upon“).
In further support of our conclusion, we note that the IRS also interprets mortality assumptions that are “based on” a standard mortality table either as providing for a set forward or a set back or as not providing for a set forward or a set back. Section 417(e)(3)(A)(i)(I) of the IRC provides that the applicable mortality table for calculating certain forms of mandatory distributions shall be “based on the prevailing commissioner‘s standard table.” Id. Section 412(1)(7)(C) (ii)(I) of the IRC similarly provides that the mortality table for calculating certain current liabilities shall be “based on the prevailing commissioner‘s standard table.” Id. The “prevailing commissioner‘s table” is the 1983 Group Annuity Mortality Table (“GAM 83“). See
The class points to evidence in the record indicating instances in which the IRS and expert actuaries do not interpret mortality assumptions that are “based on ” a standard mortality table as providing for an adjustment. The fact that the IRS or expert actuaries sometimes interpret such language as not providing for a set forward does not render the disputed mortality assumptions unambiguous. Indeed, those views, while relevant, provide only one view as to how the mortality assumptions should be interpreted. To find the mortality assumptions at issue in this case unambiguous, we must exclude all other alternative constructions of the mortality assumptions as unreasonable. See Kearney, 175 F.3d at 1090 (finding a reservation of discretion clause ambiguous in a disability benefits plan). That we cannot do.
The class relies heavily on two district court cases, Korn v. Levine Bros. Iron Works. Corp., 574 F. Supp. 836 (S.D.N.Y. 1983), and Kiefer v. Ceridian Corp., 976 F. Supp. 829 (D. Minn. 1997), in maintaining that the phrase “based on” does not imply a right to adjust calculations under the UP-1984 Mortality Table forward or backward. That reliance is misplaced. In Korn, the court merely recognized in a footnote that the plan administrator had replaced one factor with another factor. 574 F. Supp. at 843 n.8. The court did not indicate that the plan administrator had set a specific mortality table forward or backward. Nor did it indicate whether the plan administrator was interpreting the plan when it substituted factors. Id. In Kiefer, the court simply held that an interest rate assumption was unambiguous. 976 F. Supp. at 848 (holding that “the Pension Benefit Guarantee Corporation [interest] rate” did not mean 120 percent of that rate). Unlike a standard mortality table, however, an interest rate is a specific number that cannot reasonably be interpreted as adjustable.
Because we conclude that the plan‘s mortality assumptions for actuarially equivalent benefits are ambiguous, our next task is to determine whether the Plan Administrator abused its discretion by interpreting those terms as providing for a nine month set forward to the UP-1984 Mortality Table. A plan administrator‘s decision to deny benefits must be upheld under the abuse of discretion standard if it is based upon a reasonable interpretation of the plan‘s terms and if it was made in good faith. See Bendixen, 185 F.3d at 944 (citing Estate of Shockley v. Aleyeska Pipeline Serv. Co., 130 F.3d 403, 405 (9th Cir. 1997)). The question we must ask is not “whose interpretation of the plan documents is most persuasive, but whether the . . . interpretation is unreasonable.” Canseco v. Construction Laborers Pension Trust, 93 F.3d 600, 606 (9th Cir. 1996) (citations and internal quotations omitted). A consistent pattern of interpretation is “significant evidence” that the plan administrator acted reasonably in interpreting ambiguous plan language. See Hansen v. Western Greyhound Retirement Plan, 859 F.2d 779, 781 (9th Cir. 1988) (upholding a consistent interpretation of an early retirement provision under the arbitrary and capricious standard); see also James E. Jordan et al., Handbook on ERISA Litigation, S 4-61 (1999 Supp.) (stating that consistency of application and interpretation are important factors in determining whether a plan administrator abused its discretion).
Chevron‘s interpretation of the plan‘s mortality assumptions was also consistent with the plan‘s historical funding practices. Chevron has used the mortality factors from an all male version of the UP-1984 Mortality Table when making contributions for ninety-five percent of the plan‘s projected benefits, and it has used the mortality factors from an all female version of the table when making contributions for the remaining five percent of the plan‘s projected benefits. The same result may be achieved by making contributions for all of the plan‘s projected benefits using the standard version of the UP-1984 Mortality Table weighted ninetyfive percent for males and five percent for females. In fact, based on the recommendations of the actuaries that developed the table, that weighting roughly corresponds to a nine month set forward. See Jackson & Fellers, at 483 (stating that a one year set forward is appropriate for plans with an all male population). The district court, therefore, properly concluded that Chevron did not abuse its discretion when it interpreted mortality assumptions as providing for a nine-month set forward to the UP-1984 Mortality Table.
IV
Having concluded that the Plan Administrator did not abuse its discretion in interpreting the disputed mortality assumptions, we next address the class‘s assertion that Chevron violated certain provisions of ERISA by interpreting the Chevron Plan to permit a set forward. The class contends that we must reverse the judgment pursuant to S 204(c)(3) and S 204(g) of ERISA, if we find the plan‘s mortality assumptions ambiguous. The class asserts that rules prescribed under S 411 of the IRC, which are incorporated into S 204(c)(3) and S 204(g) of ERISA pursuant to S 3002(c) of ERISA,11 pursuant to S 101 and S 104 of Reorganization Plan No. 4 of 1978,12 and pursuant to Labor Regulation 2530.200a-2,13 preclude a pension plan from specifying actuarial assumptions in a manner that allows an employer to exercise discretion. The class maintains that Chevron violated S 204(c)(3) of ERISA by calculating benefits using a nine-month set forward. It also argues that Chevron violated S 204(g) of ERISA and S 18(b) of the Chevron Plan by “illegally” reducing the class members’ benefits through a 1996 amendment that specified the nine-month set forward to the UP-1984 Mortality Table. We address each argument separately.
We disagree, however, with the class‘s contention that those rules compel the conclusion that a pension plan violates S 204(c)(3) of ERISA if the description of its mortality assumptions is ambiguous. Section 411(c)(3) of the IRC is identical to S 204(c)(3) of ERISA.14 Compare
The class contends that the Commissioner has determined that the “employer discretion” standard of Revenue Ruling 79-90 must be applied under S 204(c)(3) of ERISA when a court considers whether a nontraditional benefit is the actuarial equivalent of a benefit that would commence at normal retirement age. The standard set forth in Revenue Ruling 79-90 requires actuarial assumptions, including mortality assumptions, to be “specified within the plan in a manner which precludes employer discretion” when “the amount of a benefit in a defined benefit plan is to be determined by some procedure . . . which requires the use of actuarial assumptions.”16 Id. In effect, the class suggests that a nontraditional benefit cannot be the actuarial equivalent of a benefit that would commence at normal retirement age unless it is expressed in a manner that precludes employer discretion. To support that contention, the class points to IRS Announcement 95-33, which, as a collateral issue, establishes procedures for testing whether a plan complies with the requirements of S 411(c)(3) of the IRC and whether a plan complies with the definitely determinable requirements of Revenue Ruling 79-90.
The class‘s reliance on IRS Announcement 95-33 is misplaced. In citing that announcement, the class fails to distinguish between the meaning of “definitely determinable” and the meaning of “actuarially equivalent.” IRS Announcement 95-33 establishes guidelines for IRS specialists to follow in examining pension plans for “qualified joint and survivor issues and single sum distribution issues.” IRS Announcement 95-33, 1995-19 I.R.B. 14. Those issues include whether “plan benefits are definitely determinable” under S 401(a)(25) of the IRC and Revenue Ruling 79-90 and whether a plan‘s nontraditional payment options provide an actuarially equivalent amount of benefits. See id. at SS 361, 362. Contrary to the class‘s assertion, however, the announcement examines each of those issues separately. Compare id. at S 361.2 (examining plans for employer discretion under the definitely determinable requirement) with id. at S 362.2 (examining plans for actuarial equivalence).
The class‘s contention that S 204(c)(3) of ERISA incorporates the employer discretion standard of Revenue Ruling 79-90 also ignores the legislative history of ERISA. “ERISA is a comprehensive and reticulated statute which Congress has adopted after careful study of private retirement plans.” See Alessi v. Raybestos-Manhattan, Inc., 451 U.S. 504, 510 (1981); Hughes Salaried Retirees Action Comm. v. Administrator of Hughes Non-Bargaining Retirement Plan, 72 F.3d 686, 696 (9th Cir. 1995). Congress codified Revenue Ruling 79-90 in S 401(a)(25) of the IRC when it enacted the Retirement Equity Act of 1984.
Despite the class‘s suggestion, Treasury Regulation 1.411(d)-4 does not apply to simple ambiguities in the text of a covered plan. The legislative history of S 411(d)(6) of the IRC and S 204(g) of ERISA, under which the regulation was prescribed, indicate that neither Congress nor the Treasury Department intended the regulation to apply to ambiguous descriptions of actuarial assumptions. Congress enacted ERISA in 1974 “to promote the interests of employees and their beneficiaries in employee plans and to protect contractually defined benefits.” Williams, 30 F.3d at 1430-31 (citing Firestone Tire & Rubber Co. v. Bruch, 489 U.S. 101, 113 (1989)) (internal quotations omitted). To further that goal, Congress amended S 411(d)(6) of the IRC and S 204(g) of ERISA in 1984 in order to prevent employers from “pulling the rug out from under” plan participants by eliminating or reducing certain forms of benefits through a plan amendment. Counts, 986 F.2d at 1324; see also
Interpreting that change, the Treasury Department prescribed Treasury Regulation 1.411(d)-4. See Williams, 30 F.3d at 1431. In its notice of proposed rulemaking, the Treasury Department indicated that it was concerned that employers would circumvent the prohibitions in S 411(d)(6) of the IRC and S 204(g) of ERISA by eliminating or reducing benefits without formally amending the terms of a plan. See 51 Fed. Reg. 3798, 3798. That notice explained that, without the regulation, “consent or discretion provisions . . . . would effectively enable an employer to eliminate or reduce the availability of alternative forms of benefits without regard to the protection given plan participants under section 411(d)(6).” Id. (emphasis added). Given that rulemaking history, it is apparent that Treasury Department prescribed Treasury Regulation 1.411(d)-4 only to prevent plans from including specific provisions that purport to authorize an employer to decide which employees may receive covered benefits and in what amounts. The Treasury Department did not intend to create a system of strict liability for mere ambiguities in the text of a covered plan.
The class cites only one case, Korn v. Levine Bros. Iron Workers Corp., 574 F. Supp. 836 (S.D.N.Y. 1983), to support its contention that
B
Section 204(c)(3) of ERISA provides that, in the case of a defined benefit plan, an “accrued benefit” to be paid in an “amount other than an annual benefit commencing at normal retirement age” or in a form other than in a single life annuity commencing at normal retirement age must be the “actuarial equivalent” of the “amount” or “benefit ” that would commence at normal retirement age.19
Chevron paid the class members benefits in amounts that were the actuarial equivalent of their individual life pensions. The UP-1984 Mortality Table was prepared using the mortality experience of a population that was roughly eighty percent male and twenty percent female. See Jackson & Fellers, at 483, 485. It is appropriate and valid for use on plans with a ten percent to thirty percent female content. See id. at 460, 483, 485. That means that the standard version of the table produces actuarially correct mortality assumptions only when the proportion of females participating in a pension plan is between ten percent and thirty percent on a dollar basis. When used on a plan with a population that is less than ten percent female, however, the standard version of the table produces mortality assumptions that are understated. See id. at 483, 485. An understated mortality assumptions treats a person either as surviving longer than would actuarially be expected or as retiring earlier than is chronologically true.
In this case, the standard version of the UP-1984 Mortality Table is neither valid nor appropriate for use on the Chevron Plan. On a dollar basis, only five percent of the plan‘s benefits are paid to females. If used without a set forward, the UP-1984 Mortality Table would understate the true mortality assumption for each plan participant. That understatement would be equivalent to treating each plan participant as surviving nine months longer than should actuarially be expected or, alternatively, as retiring at an age that is nine months younger than the age at which the participant retires. According to the instructions of the authors of the UP-1984 Mortality Table, such an understatement can be corrected by setting the table forward a number of years. See Jackson & Fellers, at 459, 483, 485. The UP-1984 Mortality Table will produce the correct mortality assumption for each plan participant in an all male population when the table is set forward one year. See id. at 483, 485. Because Chevron‘s ninety-five percent male population represents three-fourthsof the range from eighty percent to one hundred percent, the correct set forward for the Chevron Plan is three-fourths of one year, or nine months. Chevron‘s actuarial assumptions, though set forward nine months, therefore satisfy the requirements of
C
The class contends that Chevron violated the anticutback provisions of
Paragraph 1 of
The class‘s argument that Chevron violated
In this case, Chevron denied the class members’ request for benefits based on its interpretation of an ambiguous description of the plan‘s mortality assumption. It did not rely on a formal plan amendment. In fact, the 1996 amendment, though retroactive to 1984, could not have provided the basis for calculating the class‘s benefits. The amendment was not executed until after Chevron interpreted the disputed mortality assumptions as providing for a nine-month set forward to the UP-1984 Mortality Table. The district court properly concluded that
V
We hold that the disputed mortality assumptions, which were qualified by the words “based on” or “in accordance with,” were ambiguous as used in the Chevron Plan and that the Plan Administrator did not abuse its discretion in applying a nine-month set forward to the UP-1984 Mortality Table. We affirm, because we conclude that none of the class‘s remaining challenges to the judgment is meritorious. The judgment of the district court is AFFIRMED.