At issue in this appeal is whether the layoffs at Sea Ray Boats, Inc. between 1989 and 1991 constituted a partial termination of the company’s stock ownership and profit sharing plan (“Plan”), governed by the Employee Retirement Income Security Act of 1974 (“ERISA”), 29 U.S.C. §§ 1001-1461. Sea Ray’s administrative committee (“Committee”) concluded that the company had not terminated the Plan. Upon review, the district court granted summary judgment in favor of the Committee and a class of Sea Ray employees (“Class II”) who are current or fully vested participants in the Plan and their beneficiaries. Another class of former Sea Ray employees (“Class I”), non-vested former participants in the Plan and their beneficiaries, contends that the district court erred in applying an arbitrary and capricious rather than a de novo standard of review of the Committee’s decision. Class I also contends that the Committee erred in determining that a partial termination of the Plan had not occurred. For the reasons that follow, we affirm the judgment of the district court.
I.
Sea Ray manufactures sport boats, sport cruisers, and yachts — a particular line of boats known as pleasure crafts. Founded in the late 1950s by C.N. Ray in Oxford, Michigan, the company enjoyed steady growth from the late 1960s through the 1970s. During this time, it systematically opened new plants and renovated existing ones. During the 1980s, the company experienced unprecedented growth, enjoying an average annual growth of 35 percent from 1983 to 1986 and record sales from 1986 to 1988. In 1986, Brunswick Corporation purchased Sea Ray, but allowed it to operate as a separate entity.
*984As part of its employee benefits, Sea Ray offers employees membership in the Plan, a defined contribution stock bonus and profit-sharing plan established under the provisions of ERISA. The Committee, as a fiduciary under ERISA, acts as the administrator for the Plan. Sea Ray funds the Plan in its entirety; participating employees do not contribute to the Plan. The Plan’s assets are allocated to the individual accounts of its participants, with the benefits of each participant based solely upon the amount contributed to the participant’s account and any income, expenses, gains and losses, or forfeitures of accounts of other participants which may be allocated to such participant’s account. Although membership in the Plan is voluntary, a majority of Sea Ray’s employees participate.
Sea Ray begins contributing to the Plan on behalf of an employee after he or she has completed one full year of service. As permitted under ERISA, 29 U.S.C. § 1053(a)(1)(B), the Plan adopts a gradual vesting formula:
Vested (or Non-Forfeitable) Credited Service Percentage
fewer than 3 years 0%
3 years 20%
4 years 40%
5 years 60%
6 years 80%
more than 7 years 100%
Employees become fully vested in then-account balances after they have worked at Sea Ray for seven full years. Under normal circumstances, if an employee leaves Sea Ray before completing seven years of service, the non-vested (or forfeitable) portion of his or her account balance is reallocated among the employees who remain at Sea Ray and are currently participating in the Plan.
The one exception to reallocation for non-fully vested employees, stated in § 19 of the Plan and codified in ERISA, 26 U.S.C. § 411(d)(3), occurs if Sea Ray engages in a partial termination of the Plan. In this event, the terminated employees become fully vested and are entitled to retain their full account balances irrespective of the years they have actually served.
During Sea Ray’s period of unprecedented growth, membership in the Plan grew from 1,501 employees in July 1985 to 3,832 employees in July 1989. Beginning in March 1989, however, Sea Ray began to experience serious declines in sales, affecting all boats except yachts. During many of the months during 1989 through 1991, Sea Ray laid off part of its work force; some layoffs began as temporary, but almost all became permanent. By June 1990, as the company reduced its number of employees, the number of participants in the Plan declined to 3,060 employees. In 1990, Congress passed a federal luxury tax targeting five high-priced consumer goods — private boats, private planes, cars, jewelry, and furs — as part of its annual budget agreement: effective January 1991, any boat was subject to a 10 percent tax for the amount above $100,000. In the aftermath of the luxury tax, sales of all boats continued to decline as did the number of participants in the Plan. By June 1991, membership in the Plan had shrunk to 1,968 participants and twelve of the thirteen plants had either reduced their workforce or closed altogether.
Between 1989 and 1991, decisions regarding hiring, firing, and dismissal of employees were not vested in any one centralized figure or manager but left to the discretion of each plant. Since each plant at Sea Ray was operated separately as a profit center, each plant’s manager determined the plant’s criteria for laying off employees at his or her plant.1 During the downsizing, however, Sea Ray’s corporate headquarters issued written directions to affected plant managers on how to handle the layoffs.
In June 1992, prior to any litigation leading to the present ease, the Committee convened to discuss the partial termination question and the governing legal standards, and made the following determinations:
1. The reductions in employees was motivated by unfavorable changes in the *985economy and not by targeted employment numbers.
2. Partial determinations were defined by the Internal Revenue Code and determined on the basis of all facts and circumstances.
3. The historic application of the partial termination rule focused primarily on whether the termination resulted in funds reverted to the employer or discrimination in favor of highly compensated employees; recently, however, the emphasis had shifted to percentage decline in employment.
4. Case law was unsettled on what constitutes a significant decline and the level of deference, if any, owed to employer or IRS determinations.
5. Lastly, the courts’ determinations regarding partial terminations were binding on the IRS only if it joined the suit or the suit was an appeal of a final administrative determination by the IRS.
Later that month, the Committee reconvened and concluded that for the period between 1989 and 1991, “there were no facts or circumstances other than the layoff due to the cyclical business which would dictate a finding of a Partial Termination.” In making that determination, the Committee decided to calculate the percentage of layoffs for each year separately. The Committee also decided that the company would retain legal counsel to commence a lawsuit on behalf of the plan to seek a judicial ruling upholding the Committee’s determination that a partial termination did not occur. The Committee filed an action before the district court in August. Following pre-trial matters, Class I filed a motion for summary judgment on December 1995. The Committee and Class II each filed a cross-motion for summary judgment in March 1996. The district court referred the motion and cross-motions to a magistrate judge, who recommended that the district court grant the cross-motions by the Committee and Class II and deny the motion by Class I. Specifically, it recommended holding that 1) the Committee’s determination that a partial termination had not occurred was subject to the arbitrary and capricious standard of review; 2) the layoffs, calculated as caused by two separate corporate events, were not at a level generally considered determinative of a partial termination; and 3) the Committee’s determinations regarding partial termination were neither arbitrary nor capricious. Although Class I filed an objection to the magistrate judge’s report and recommendation, the district court accepted it in its entirety. Class I filed a timely notice of appeal.
II.
We review a district court’s order granting summary judgment de novo. See Roush v. Weastec, Inc., 96 F.3d 840, 843 (6th Cir.1996). Summaiy judgment is appropriate “if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.” Fed.R.Civ.P. 56(c). This court draws all reasonable inferences in favor of the non-moving party. See Terry Barr Sales Agency, Inc. v. All-Lock Co., 96 F.3d 174, 178 (6th Cir.1996) (citing Matsushita Elec. Indus. Co., Ltd. v. Zenith Radio Corp., 475 U.S. 574, 587-88, 106 S.Ct. 1348, 89 L.Ed.2d 538 (1986)). This court recognizes, however, that “the mere existence of a scintilla of evidence in support of [a party’s] position will be insufficient; there must be evidence on which the jury could reasonably find for the [party].” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 252, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986).
III.
We first address whether the Plan vests discretionary authority in the Committee, the administrator of the Plan. If so, we then apply the arbitrary and capricious standard of review to the Committee’s decision that a partial determination of the Plan did not occur at Sea Ray between 1989 and 1991. See Perez v. Aetna Life Ins. Co., 150 F.3d 550, 555 (6th Cir.1998) (en banc) (citing Firestone Tire and Rubber Co. v, Bruch, 489 U.S. 101, 111, 109 S.Ct. 948, 103 L.Ed.2d 80 (1989)).
The Committee contends that the Plan grants it discretion to determine partial determinations, pointing to the language in § 16(c) of the Plan:
*986The Committee shall have all powers necessary to enable it to administer the Plan and Trust Agreement in accordance with their provisions, including without limitation the following:
(2) determining the appropriate allocations to Participants’ Accounts pursuant to Section 6;
(3) determining the amount of benefits payable to a Participant (or Beneficiary), and the time and manner in which such benefits are to be paid;
(7) construing and interpreting the Plan and the Trust Agreement and adopting rules for administration of the Plan that are consistent with the terms of the Plan documents and of ERISA and the Code;
The Committee shall perform its duties under the Plan and the Trust Agreement solely in the interests of the Participants (and their Beneficiaries). Any discretion granted to the Committee under any of the provisions of the Plan or the Trust Agreement shall be exercised only in accordance with rules and policies established by the Committee which shall be applicable on a nondiscriminatory basis.
(emphasis added).
Class I contends that the Committee lacked discretion in this matter because any such determination turned on a matter of law. Although the Plan allows the Committee to determine whether the Plan had been partially terminated, it does not define “partial termination.” Class I contends that, as a matter of law, this determination should be provided not by the Committee, but by the court. Therefore, this court should apply a de novo standard of review, following other circuits which have held that eligibility under ERISA and its regulations are a question of law. See e.g., Weil v. Retirement Plan Admin. Comm. of Tersan Co., Inc., 913 F.2d 1045, 1048-49 (2nd Cir.1990); Gauer v. Connors, 953 F.2d 97, 99-100 (4th Cir.1991); Pratt v. Petroleum Production Management. Inc. Employee Sav. Plan and Trust, 920 F.2d 651, 658 (10th Cir.1990).
Because the Plan is governed by ERISA, we apply federal common law rules of contract interpretation in making our determination. See Perez, 150 F.3d at 556. In doing so, we “take direction” from state law and general contract law principles and interpret the Plan’s provisions “according to then-plain meaning, in an ordinary and popular sense.” Id. (internal citations omitted). Plain meaning requires that we “must give effect to the unambiguous terms of an ERISA plan.” Id. (internal citations omitted).
We find that a plain reading of the Plan reveals that it clearly grants discretion to the Committee. We have earlier held that discretionary authority does not “hinge[] on incantation of the word ‘discretion’ or any other ‘magic word,’ ” Johnson v. Eaton Corp., 970 F.2d 1569, 1572, n. 2 (6th Cir.1992) (quoting Block v. Pitney Bowes, Inc., 952 F.2d 1450, 1453 (D.C.Cir.1992)), but whether the ERISA plan grants the administrator “authority to determine eligibility for benefits or to construe the terms of the plan.” Firestone, 489 U.S. at 115, 109 S.Ct. 948. While terms in § 16 of the Plan such as “construe” and “interpret” by themselves would likely be enough to establish discretion, we need not decide that issue because § 16 also expressly gives the Committee “discretion.”
We reject Class I’s contention that because the Plan does not define “partial termination,” the Committee’s discretion does not extend to partial terminations. The Committee’s discretion to interpret and construe the Plan logically extends to terms of the Plan, since any interpretation of a plan turns on the meaning of terms contained therein. To limit an administrator’s discretion to only those terms explicitly defined would undermine the administrator’s discretionary power or require companies to write interminably long plans to account for every term.
We also reject Class I’s argument that eligibility determinations are questions of law subject to de novo review. Although other circuits have adopted this view, see Weil, 913 F.2d at 1048-49; Gauer, 953 F.2d at 99-100; Pratt, 920 F.2d at 658, we have declined to do so. The Committee, by virtue of its grant of discretion, may interpret the *987language of terms in the plan. See Davis v. Kentucky Finance Cos. Retirement Plan, 887 F.2d 689, 694 (6th Cir.1989).
Accordingly, we review the Committee’s determinations regarding partial terminations under the arbitrary and capricious standard.
IV.
Having established the appropriate standard of review, we now evaluate whether the Committee’s determination that a partial termination did not occur between 1989 and 1991 was arbitrary and capricious. The Committee contends that the percentage of terminations at Sea Ray during these years, in the absence of improper motive or bias on the part of Sea Ray, did not constitute a partial termination. Class I counters that the percentage of terminations was higher than the Committee calculated, and the real percentage was high enough in itself to be determinative of a partial termination. The differences in calculations stem from the Committee’s decision to calculate the layoffs from each year separately and to exclude certain classes of former employees in its tally of terminated employees.
In determining whether a partial termination has occurred, this court has held that Treasury Regulation § 1.411(d)-2(b)(l)2 “directs us to consider all the facts and' circumstances, of which the percentage of excluded plan participants is but one, albeit generally the most persuasive one.” Kreis v. Charles O. Townley, M.D. and Associates, P.C., 833 F.2d 74, 79 (6th Cir.1987). In some cases, however, we have held that “the percentage may be so high or so low as to be determinative standing alone.” Id. at 80.
Neither this court nor other circuits have established clear benchmarks for these high or low percentages. This court noted that one court has held 34 and 51 percent terminations of the work force constituted partial terminations, while other courts have held that termination rates of 16.7 percent, 13.0 percent, and 12.4 percent were insufficient to trigger a partial termination. See id. at 83 (citing See Tipton and Kalmbach, Inc. v. Commissioner, 83 T.C. 154, 162, 1984 WL 15598 (1984) (34 and 51 percent); Wishner v. St. Luke’s Hosp., 550 F.Supp. 1016, 1019 (S.D.N.Y.1982) (16.7 percent); Taylor v. Food Giant, Inc. Salaried Employees Pension Plan, No. C 84-253A, 1984 WL 8144, at *3 (N.D.Ga. Nov.30, 1984) (13.0 percent); Babb v. Olney Paint Co., 764 F.2d 240, 243 (4th Cir.1985) (12.4 percent)). The United States Tax Court in Halliburton Co. v. C.I.R., 100 T.C. 216, 1993 WL 83409 (1993) stated that a percentage exceeding 50 percent may be sufficient standing alone to establish a partial termination; conversely, a drop of less than 20 percent is significant only if accompanied by egregious abuse on the part of the employer. See id. at 236-37. A drop of at least than 20 percent is sufficient, if coupled with other circumstances, to suggest a partial termination. See id. at 237. We decline to establish rigid percentage parameters for partial terminations, and repeat our holding in Kreis that while the percentage of excluded plan participants is the most persuasive factor, we must generally consider all the facts and circumstances, “unless and until Congress or the Treasury Department provides otherwise.” Kreis, 833 F.2d at 80.
Before evaluating the Committee’s determination, we must first ascertain the percentage of participants involuntarily excluded from the Plan between 1989 and 1991 through actual or constructive discharge. Whether an actual or constructive discharge has occurred is a question of fact, see id. at 81, and we must affirm the district court’s findings on this question unless they are clearly erroneous. See Fed.R.Civ.P. 52(a). A finding of fact is clearly erroneous only when, although there may be some evidence to support the finding, “the reviewing court on the entire evidence is left with the definite and firm conviction that a mistake has been committed.” Anderson v. City of Bessemer, 470 U.S. 564, 573, 105 S.Ct. 1504, 84 L.Ed.2d 518 (1985) (citation omitted).
*988The district court first concluded that the layoffs at Sea Ray between 1989 and 1991 were caused by two separate events: a two-year downturn in the small boat industry and a downturn in the large boat business caused by the federal luxury tax. It therefore decided to analyze the percentage of actual or constructive discharges as separate years. In deciding which terminated employees to count for the partial termination calculation, the court excluded classes of employees who left Sea Ray for the following reasons: death, normal retirements, terminations for cause, terminations due to disabilities, departure due to a transfer, and voluntary departures — -even if the departure were influenced by the economic downturn at the company. Class I disagrees with the district court’s findings, maintaining that the discharges should be considered as one event and that former employees who left in anticipation of an involuntary layoff should count as constructive discharges.
We see no occasion to disturb the factual findings of the district court. First, our review of the record does not firmly convince us that the district court erred in its conclusion that two events occurred. We agree that the economic downturn in 1989 and the federal luxury tax in 1990, while both leading to dire consequences at Sea Ray, stem from two independent factors. Second, while we do not dispute that some former Sea Ray employees felt compelled to leave in light of the economic downturn, a showing of constructive discharge requires more than dissatisfaction with the economic conditions at work. See e.g., Kreis, 833 F.2d at 82-83; see also Halliburton, 100 T.C. at 241 (stating same proposition). The record does not reflect that working conditions at Sea Ray were “so difficult or unpleasant that a reasonable person in the employee’s shoes would have felt compelled to resign.” Held v. Gulf Oil Co., 684 F.2d 427, 432 (6th Cir.1982) (internal citations omitted). To characterize the conditions at Sea Ray as intolerable is to impute liability to the company simply because of broad economic conditions which were beyond the company’s control. That imputation exceeds the scope of partial termination.
While adopting the findings of the Committee with the modifications noted by Class II, the district court generated three sets of calculations for the percentage of discharged employees at Sea Ray between 1989 and 1991, based on each party’s interpretation of the layoffs.3
Committee Class II Class I
1989-90: 1989-90: (a) 1989-91:
16.1% 15.9% 32.9%
1990-91: 1990-91: (b) 1989-91:
28.8% 27.9% 36.7%
Finding no clear error by the district court, we also adopt the percentages generated by Class II. Because we do not find the percentage for either 1989-90 or 1990-91 to be so high as “as to be determinative standing alone,” Kreis, 833 F.2d at 80, we consider additional facts and circumstances.
This court in Kreis considered two additional factors to determine whether a partial termination has occurred: one, the effects on the plan of the exclusion of employees from participation; two, the employ*989er’s motives. See id. at 79. We similarly apply these factors in the present case. We quickly dismiss the first factor, as none of the parties allege that the layoffs at Sea Ray between 1989 and 1991 financially impaired the Plan. For the second factor, we look to see whether the Committee initiated the layoffs to generate profit from the forfeitures from non-vested employees, either for itself or a third party. Evidence of this type of profit “supports but is not sufficient for a finding of bad faith.” Id. at 80. Nonetheless, we can similarly dismiss this factor, as Class I does not contend that Sea Ray was motivated by improper motives in terminating the employees.
We are left to decide whether the Committee’s determination that a partial termination did not occur at Sea Ray between 1989 and 1991 was arbitrary and capricious. The arbitrary and capricious standard is the least demanding form of judicial review and is met when it is possible to “offer a reasoned explanation, based on the evidence, for a particular outcome.” Davis 887 F.2d. at 693 (quoting Pokratz v. Jones Dairy Farm, 771 F.2d 206, 209 (7th Cir.1985)).
We find that the Committee’s determination regarding the partial termination issue was neither arbitrary nor capricious. The Committee made its decision only after' carefully deliberating both the governing legal standards of partial terminations and Sea Ray’s actions in light of these standards. Given its calculation of the percentage of terminations — both years were below 30 percent — and the absence of any damage to the Plan or improper motive for profit, the Committee was reasonable in concluding that the Plan had not been partially terminated.
Y.
For the reasons stated above, we affirm the district court’s grant of summary judgment in favor of the Committee and Class II.
. Each laid-off employee completed a "Change of Employee Status” ("COS") form, which provided the work history, and also served as an exit interview with the individual in charge of the plant’s personnel. Sea Ray systematically gathered COS forms from the plants, but did not use them in determining the number of participants in the Plan, instead relying on annual company census lists containing the name of all employees.
. Treasury Regulation § 1.41 l(d)-2(b)(l) states, Whether or not a partial termination of a qualified plan occurs (and the time of such event) shall be determined by the Commissioner with regard to all the facts and circumstances in a particular case. Such facts and circumstances include: the exclusion, by reason of a plan amendment or severance by the employer, of a group of employees who have previously been covered by the plan; and plan amendments which adversely affect the rights of employees to vest in benefits under the plan.
(emphasis added).
. The calculations for the Committee and Class II were generated by the following equation:
% of involuntary terminations = (total reduction of employees) - excludable employees) total number of participants
Class II’s calculations contain more up-to-date the following equations to generate slightly dif-figures on the total reduction of employees for ferent percentages: each year.
Class I, which argued that the layoffs between 1989 and 1991 should be analyzed together, used
(a) % of involuntary terminations = (terminees in Class I) + (fully vested participants) participants (6/30/89) +
(additional participants over period)
(b) % of involuntary terminations = (total reduction) - (excludable employees) total number of participants