Plaintiff seeks review of a decision of the Armed Services Board of Contract Appeals1 which held that the Government was entitled to disallow a portion of plaintiff's claims for reimbursement of general and administrative (G&A) expense, which plaintiff had allocated to its cost-plus-fixed-fee contracts performed during its fiscal years 1959 through 1965. The case is presently before the court on defendant's request for review of an opinion and recommended conclusion of law by Commissioner George Willi.

The two issues decided by our commissioner are whether, consonant with the finality standards of the Wunderlich Act, 41 U.S.C. §§ 321-322 (1970), the Board correctly decided, first, that plaintiff's long-standing method of allocating G&A expense between its fixed price contracts and its cost reimbursement contracts was inequitable and therefore subject to mandatory change and, second, that a new allocation method should be retroactively applied to contracts performed during the years 1959 through 1965. The result of the Board's decision to retroactively apply a new allocation method is to disallow approximately $1,900,000 in G&A expense which plaintiff had allocated to its cost reimbursement contracts, and to reallocate this amount to fixed price contracts which plaintiff had negotiated on the basis of using the cost of sales method in allocating such costs.2

Our commissioner agreed with the Board, although for different reasons, that plaintiff's allocation method was inequitable and did not comply with Section XV, Part 2 of the Armed Services Procurement Regulations. Since we agree with his opinion and conclusions on this issue, we have adopted and incorporated them in this opinion. We disagree, however, with his recommendation that the enforced change in plaintiff's allocation method should be effective as of the date of the Board's decision. We have decided that the change should be effective as of December 3, 1962, as discussed in Part II of this opinion.

* Plaintiff has been a wholly-owned subsidiary of Litton Industries, Inc., since that conglomerate's beginning in 1953.

This lawsuit stems from the contracting activities of the two largest of plaintiff's seven operating divisions. Those two units have always been devoted exclusively to the development, production, and sale of advanced electronic equipment, principally inertial navigation systems, used by the United States military and that of certain NATO countries.

During the years 1959 through 1965, the period covered by the administrative decision that plaintiff seeks to reverse in this proceeding, it had more than a thousand contracts, both fixed price and CPFF, by which it sold various types of military electronic equipment to governmental customers and their suppliers. Which of these two types of contracts was utilized in any given transaction was determined by the nature of the particular articles or services involved in the sale. In terms of the economic substance of the sales transaction, the two types of contracts were regarded as the same.

The operations required to produce the electronic equipment in which plaintiff dealt generated various overhead-type indirect expenses that, while incontestably attributable to the aggregate of its fixed price and CPFF contract work,3 were not directly traceable to any particular contract or group of contracts. Because these general and administrative expenses were specifically reimbursable to the extent that they related to CPFF contract work, but not so as to the fixed price contracts, it was necessary for the plaintiff to adopt a method by which each contract's fair share of the total G&A burden could be constructively determined.

In addition to a standard record-keeping provision4 expressly sanctioning the contractor's accounting practices provided that they conform to generally accepted accounting principles, plaintiff's CPFF contracts typically included a provision5 defining allowable cost in terms of the standards and criteria set forth in the applicable portion of the Armed Services Procurement Regulations. As to the G&A items here involved, and the means of allocating them, the governing principles are found in Part 15 of the Regulations, 32 C.F.R. § 15.203 (a), (b), (c), and (d) (Rev. as of January 1, 1961):6

Sec. 15.203 Indirect costs.

(a) An indirect cost is one which, because of its incurrence for common or joint objectives, is not readily subject to treatment as a direct cost. Minor direct cost items may be considered to be indirect costs for reasons of practicality. After direct costs have been determined and charged directly to the contract or other work as appropriate, indirect costs are those remaining to be allocated to the several classes of work.

(b) Indirect costs shall be accumulated by logical cost groupings with due consideration of the reasons for incurring the costs. Each grouping should be determined so as to permit distribution of the grouping on the basis of the benefits accruing to the several cost objectives. Commonly, manufacturing overhead, selling expenses, and general and administrative expenses are separately grouped. * * *

(c) Each cost grouping shall be distributed to the appropriate cost objectives. This necessitates the selection of a distribution base common to all cost objectives to which the grouping is to be allocated. The base should be selected so as to permit allocation of the grouping on the basis of the benefits accruing to the several cost objectives. * * *

(d) The method of allocation of indirect costs must be based on the particular circumstances involved. The method shall be in accord with those generally accepted accounting principles which are applicable in the circumstances. The contractor's established practices, if in accord with such accounting principles, shall generally be acceptable. However, the methods used by the contractor may require re-examination when:

(1) Any substantial difference occurs between the cost patterns of work under the contract and other work of the contractor; or

(2) Any significant change occurs in the nature of the business, the extent of subcontracting, fixed asset improvement programs, the inventories, the volume of sales and production, manufacturing processes, the contractor's products, or other relevant circumstances. [Emphasis added.]

In response to the ASPR's guideline criteria for apportionment of indirect costs, plaintiff used the familiar accounting concept termed "cost of sales" as the distribution base common7 to all of its contracts and the means of measuring the extent to which each benefited8 from the G&A services (i. e., overall management and supervision) for which the expenses to be apportioned were incurred.

The assumptions implicit in this approach to allocation are: first, that a contract benefits from general management and supervisory services in direct proportion to the extent that it is performed during the period in which such services are rendered; second, that the total cost bill for articles produced and sold under a supply contract is a direct reflection of the extent to which the contract has been performed and that, accordingly, an individual contract's performance progress during a specified period, relative to that of other concurrently active contracts, is directly proportionate to the relationship that its cost-of-sales total bears to the combined cost-of-sales figure for the group of contracts as a whole.9

As the term implies, cost of sales represents the cost of production of items sold. Kohler, A Dictionary for Accountants (2d ed. 1957), p. 145.

Since the concept includes only those costs pertaining to articles "sold" during the relevant accounting period, the question of what should be deemed a cognizable "sale" is fundamentally important and must be resolved at the outset. Common sense dictates that the definition selected be based on criteria that fit the purpose for which the defined standard is to be used. Here, a comparative anaylsis of cost-of-sales figures is intended to reflect the relative extent to which each one of many supply contracts for various articles of military electronic gear has been completed during a particular accounting period. Thus, while the hallmark of a "sale" for purposes of determining a seller's income tax liability may be little more than his bare receipt of funds from the buyer,10 such a criterion would be wholly inapposite where, as here, the focus is not on the amount or character of the proceeds that the seller receives but on how far he has progressed towards completion of the work required of him by the contract from which those proceeds flow. If performance progress is to be measured by an equation of work done with costs incurred, but only to the extent that those costs have become a part of an item "sold" under the contract, a "sale" should mark the point at which the contractor has finished his work on the article involved and has therefore incurred substantially all of the costs entailed in doing that much of the job.

Since the purpose of the cost-of-sales calculation in this case was to depict the relative progress towards completion attained by each one of a group of concurrently active fixed price and cost-plus contracts during a specified accounting period, its acceptability as an allocation method demanded that it operate uniformly on contracts of both types in order to serve as the common denominator required by the ASPR.11 Thus, while a logically preferred definition of a "sale" would follow the functional lines previously suggested, the attribute indispensable to a useful application of the cost-of-sales formula is that, whatever be deemed a "sale" for cost computation purposes under one type of contract be similarly regarded under the other. If the selected definition brands a given set of economic circumstances a "sale" under one type of contract, it must accord substantially similar circumstances the same classification under the other.

As will be seen, failure to achieve a substantive parity in the definition of a sale under the two types of plaintiff's contracts was the central deficiency in its cost-of-sales method of allocation and was therefore the root cause of such distortion12 as occurred in the apportionment of G&A expense.

The litigants freely acknowledge, as did the Board,13 that in all substantive respects relevant to the present controversy, plaintiff's fixed price and its cost-plus contracts were alike. This actual similarity was not altered by the fact that each type of contract typically contained differently labeled interim payment provisions. As the Board earlier decided in Daystrom Instrument Division of Daystrom, Inc., 58-1 BCA ¶ 1588, at 5758-59, there is no real difference between a progress payment application under a fixed price contract and a reimbursement billing under a cost-plus contract. The parallel function of each is solely to provide the contractor with production financing. Despite their essential equivalence, plaintiff ascribed materially differing significance to these two credit devices.

On its fixed price contracts plaintiff, in the Board's words,14 "* * * followed the common accounting practice of treating the sale of an article as having occurred when the article was delivered to the customer." As to its cost-plus contracts, however, plaintiff used what the Board described15 as "* * * a rather artificial concept of what constitutes a sale, * * *." It treated each cost reimbursement billing as a sale irrespective of whether any part of the contract's fixed fee had been earned and even though it did not similarly classify its applications for progress payments under fixed price contracts.

With only those costs associated with a "sale" entering into the cost-of-sales figures that determined each active contract's share of total G&A expense, it is apparent that the effect of plaintiff's varying and incongruous definitions of a sale under each type of contract was to postpone the impact of costs incurred under fixed price contracts while simultaneously accelerating that of similar costs distinguishable only by the fact that they were incurred under cost-plus contracts.

By a lucid and informative hypothetical example at the outset16 of its opinion, the Board demonstrated that the conceptual inconsistency that occurred here in the definition of a sale would, by itself, result in loading all G&A expense for a given period on an unfinished cost-plus contract even though in the same period the contractor was concurrently performing a second contract, that was, except for its fixed price format, identical in all respects, including the type and quantity of costs incurred and the degree of actual performance achieved. Such distortion would not occur, the Board explained, only in the most unlikely event that the contractor's volume of fixed price work remained absolutely constant so that the balance of uncompleted work under such contracts was the same at both the beginning and end of each period for which G&A costs were to be allocated. Any variation between opening and closing balance would automatically create distortion — a net increase for the period resulting in too much G&A expense being attributed to the cost-plus contracts and a decrease resulting in too little. So it is that here, the imbalance favored the fixed price work to the detriment of the cost-plus contracts for the period 1959 through 1964 because as the record shows, plaintiff's fixed price business increased during each of those years. That trend was reversed in 1965 and thereby advantaged the cost-plus contracts at the expense of the fixed price work for that year.

In sum, it is not questioned that, as a proposition considered in the abstract, cost of sales is a method for allocating indirect costs that, in the words of the ASPR,17 is fully "* * * in accord with generally accepted accounting principles * * *." It is equally clear, however, that the divergent definitions of a sale applied to each of the two types of contract on which the method operated in this case constituted an inherent source of distortion that effectively prevented it from meeting the ASPR's altogether sensible and absolute requirement18 that such a method function as "* * * a distribution base common to all cost objectives [i. e., contracts] * * *."

Accordingly, the unsuitability of the cost-of-sales method, as it was applied in this instance, cannot be responsibly challenged. That conclusion does not, however, put an end to the controversy. It simply leads to the further question of whether the Board properly approved the Government's insistence that a new G&A allocation method19 be applied to contracts performed during plaintiff's fiscal years beginning August 1, 1958.

Plaintiff has used the cost of sales method as a basis for allocating G&A expense since at least as early as 1954. Since that time, plaintiff has employed the allocation base not only to compute cost reimbursement billings under its CPFF contracts, but also to develop provisional G&A rates which it used in bidding on fixed price contracts. These provisional rates, which plaintiff submitted to the Government for approval during the period involved here, were formulated by adjusting prior actual G&A rates to reflect projected increases or decreases in company activity. By applying the provisional rate to the anticipated total production costs of a particular proposed fixed price contract, plaintiff could estimate the amount of G&A expense which would be generated by that proposed contract and accordingly include the estimated amount in its bid price as a G&A allowance. In this manner, plaintiff would recover the amount of G&A expense which it determined under its cost of sales method to be allocable to its fixed price work.

Plaintiff argues here, as it did before the Board, that it would be unfairly prejudiced if the Government were permitted to employ a new allocation method as of the beginning of plaintiff's 1959 fiscal year. According to plaintiff, such a retroactive change would be unfair because it would cause approximately $1,900,000 of G&A expense which plaintiff had originally allocated to its CPFF contracts to be reallocated to its fixed price contracts, and that plaintiff would not be able to recover the reallocated amounts because its fixed price contracts have already been finally priced. Had it known that its fixed price contracts would be required to absorb these extra costs, plaintiff asserts that it would have increased the bid prices of its fixed price contracts accordingly.

The Board rejected plaintiff's argument of unfair prejudice by concluding that plaintiff received larger amounts for its fixed price work under its cost of sales method than it would have received had it bid its fixed price contracts on a cost input basis, the method which the Government is seeking to impose. Briefly stated, the Board's theory was that, in a period of rising inventories, plaintiff's provisional bidding rates for its fixed price contracts would be higher under its cost of sales method than under the cost input method, be cause no G&A expense would be allocated to fixed price closing inventory if plainiff's cost of sales base were employed; under the cost input method, fixed price closing inventory would be assessed a charge for G&A expense so that less G&A expense would be allocated to CPFF contracts and to completed fixed price contracts. According to the Board, this reduction in the amount of G&A expense allocable to plaintiff's fixed price contracts under the cost input base would produce lower G&A bidding rates than those actually used by plaintiff in pricing those contracts and, as a result, plaintiff received higher G&A allowances for its fixed price contracts than it would have received had it used the Government's cost input method.

On these grounds, the Board rejected plaintiff's argument that if it had been on a cost input basis from the beginning, it would have negotiated prices on contracts calling for deliveries in a particular year so as to recover, not only the G&A expense allocable to the products delivered during that year, but also all the G&A expense allocable to the closing inventory for the same year. In applying its decision retroactively to plaintiff's fiscal year 1959, we think the Board erred in basing its result, almost entirely, upon its conclusion as to the amounts plaintiff would have received on its fixed price contracts under the accounting theories relied upon by the Board. A proper construction of the contract clauses involved, and the ASPR's cost principles incorporated by reference into plaintiff's CPFF contracts, requires that all of the relevant circumstances be examined in determining whether, and to what extent, a change in a contractor's accounting practices should be retroactive.

The ASPR's cost principles do not prescribe rigid accounting procedures for contractors who perform Government contracts. Instead, the regulations indicate that a high degree of flexibility is permitted in determining costs chargeable to the Government. Thus, a contractor's own accounting system will be acceptable if it conforms to "generally accepted accounting principles," and if it meets the other general requirements of the regulations. This theme of flexibility, recurring throughout the cost principles, is especially apparent in Part 2 of those regulations, around which the dispute here is centered. Such terms as "reasonableness," "ordinary and necessary," "relative benefits," and "equitable" preclude the application of any general rule or litmus paper test. Lockheed Aircraft Corp. v. United States, supra. Rather, we think the cost principles require us to search for a result which, as we said in Lockheed, is "fair and reasonable." Id., 375 F.2d at 796, 179 Ct. Cl. at 563. Such a standard seems especially appropriate here, where the Government seeks to retroactively impose upon plaintiff a new allocation method.

The facts relating to the Government's decision to disapprove plaintiff's cost of sales method are set out in Part VI of the Board's decision. 66-1 BCA at 26, 162-63. Briefly summarized, the Board found that the earliest event pertinent to the present controversy was an audit meeting held on June 27, 1960, at which plaintiff submitted proposed G&A bidding rates. At this meeting, one of the Government auditors mentioned that plaintiff's fixed price deliveries from inventory were creating a distortion in its allocation base. The record does not indicate that any proposals to change plaintiff's practice were made at that time. The subject apparently did not re-emerge until September 7, 1961, at a conference held to discuss plaintiff's proposed 1962 G&A bidding rates for its Data Systems Division. During this conference, the resident Navy auditor suggested changing from a cost of sales base to a cost input base because of plaintiff's rising inventories. Subsequently, on November 2, 1961, the resident auditor submitted a memorandum to plaintiff in which he recommended that the proposed bidding rate for Data Systems Division not be approved, because he considered plaintiff's cost of sales method to be inequitable. On December 27, 1961, plaintiff by memorandum defended its cost of sales base and protested the auditor's recommendation that it change to a cost input base.

After further discussion, a conference was held on May 8, 1962, between plaintiff's comptroller and the Officer in Charge of the Navy Area Audit Office. Prior discussions were concerned principally with plaintiff's use of its cost of sales method in formulating proposed bidding rates for fixed price contracts, but during this particular conference, the parties also discussed the feasibility of shifting to a cost input basis for determining amounts due plaintiff under its cost reimbursement contracts. Plaintiff expressed concern that such a change in allocation base would cause a prejudicial reallocation of costs from its CPFF contracts to its fixed price contracts, which had been bid and negotiated on a cost of sales basis. Because of the effect that the proposed shift would have on plaintiff's outstanding contracts, the Officer in Charge recommended that the change be prospective only, effective at the beginning of plaintiff's 1963 fiscal year (August 1, 1962). Plaintiff rejected this proposal on May 25, 1962, asking instead that the Government continue to approve its cost of sales base. The Officer in Charge declined to do so. In his reply of June 7, 1962, he said that his position was that plaintiff's allocation base was inequitable and that only a cost input base would be accepted.

After further correspondence, when it became apparent that the controversy could not be authoritatively resolved at the local level, plaintiff requested formal action under the Disputes provisions of its contracts. On December 3, 1962, the Government auditor issued a DD Form 396 (Notice of Costs Suspended and/or Disapproved), disallowing claimed reimbursements under plaintiff's CPFF contracts in excess of the amount allocable to a cost input basis for plaintiff's 1959 fiscal year.

Plaintiff appealed the auditor's disallowance to the Director of the Navy's Contract Audit Division who, under the Disputes provisions in plaintiff's contracts, was the equivalent of the contracting officer for various questions relating to reimbursement of costs. Statements and documentation in support of the parties' positions were filed by plaintiff and the resident auditor, after which the Director rendered a decision against plaintiff on October 5, 1964. This decision was final under the Disputes clause, permitting plaintiff for the first time to appeal to the Board. The Board's decision was filed on May 18, 1966, shortly after which plaintiff brought suit in this court.

The history of the exchanges between the parties indicates that the controversy was, until the May 8, 1962, conference, limited to plaintiff's use of its cost of sales method as a basis for formulating bidding rates for proposed fixed price contracts. There is no indication in the record that prior to this conference the Government was seeking to impose its cost input method upon plaintiff's allocation of actual G&A expense in computing reimbursement billings under its CPFF contracts. Moreover, despite the resident auditor's prior criticisms of the plaintiff's method of formulating G&A bidding rates, there is nothing in the record indicating that the bidding rates were disapproved.

We think that the May 8, 1962, conference was also significant, because it was the first time, according to the record, that anything was said regarding the possibility of retroactively applying a new allocation method. Indeed, even this possibility was rejected at the conference by the Officer in Charge of the Navy Area Audit Office who, sympathetic to plaintiff's argument that a retroactive shift in allocation base would be prejudicial to plaintiff's fixed price contracts, proposed only that plaintiff change to a cost input method on a prospective basis.

It was not until December 3, 1962, when the DD Form 396 was issued, that plaintiff was authoritatively informed of the Government's previously ambivalent position. This document, we think, put plaintiff on notice that its cost of sales method would no longer be accepted as a method of computing its reimbursement billings under its CPFF contracts.

In view of plaintiff's long and consistent use of the cost of sales method with the Government's knowledge, approval and acquiescence, plaintiff was entitled to reasonably adequate notice that the Government would no longer approve the use of that method with respect to the CPFF contracts. United States v. Hanna Nickel Smelting Co., 253 F. Supp. 784 (D.Or.1966), aff'd on an alternate ground, 400 F.2d 944 (9th Cir. 1968). See also Elastic Stop Nut Corporation of America v. United States, 132 F. Supp. 466, 132 Ct. Cl. 631 (1955). Such notice was essential in order to enable plaintiff to recover the additional G&A expense that would be allocable to its fixed price contracts as a result of the shift in accounting methods. It is undisputed that the Government's requirement for the use of the cost input method results in a larger allocation of G&A expense to plaintiff's fixed price contracts than originally contemplated, and it seems inconceivable to us that plaintiff, had it known that its fixed price contracts would have to carry this additional burden, would not have sought to negotiate higher prices for them.

Subject to the Renegotiation Act of 1951, 50 U.S.C. App. §§ 1211-1233 (1970), and the Truth in Negotiations Act, Pub.L.No.87-653, 76 Stat. 528 (Sept. 10, 1962), there is no legal limit on plaintiff's right to negotiate the highest prices which it could obtain for its fixed price contracts. Plaintiff's opportunity to do so was, we think, improperly denied it by the Board's decision, which by ordering a retroactive change in plaintiff's accounting methods, would require its fixed price contracts to absorb more costs than was expected when the contracts were entered into. Consequently, we hold that, under the peculiar facts of this case, the Government was required to give plaintiff adequate notice that its cost of sales allocation base would no longer be accepted, so that plaintiff could thereafter negotiate its contracts accordingly. We also conclude that the Government's issuance of the DD Form 396 on December 3, 1962, constituted an adequate, authoritative notice, and that it was unreasonable for plaintiff to expect that the Government would permit plaintiff to allocate its G&A expenses to its CPFF contracts entered into after that date on the cost of sales basis.

Accordingly, plaintiff's G&A expenses should be allocated to the contracts in suit that were entered into prior to and including that date in accordance with the cost of sales method used by plaintiff, and plaintiff's motion for summary judgment is granted to that extent. As to all of such contracts that were entered into after December 3, 1962, plaintiff's G&A expenses should be allocated in accordance with the cost input method, and defendant's cross-motion for summary judgment is granted to that extent. The parties shall have 30 days from this date within which to file a stipulation showing the amount plaintiff is entitled to recover in accordance with this opinion. If they are unable to file such a stipulation within the time stated, proceedings in this case shall be suspended for a period of 90 days, commencing 30 days from the date of this decision, in order that plaintiff may obtain from the Armed Services Board of Contract Appeals a decision as to the amount plaintiff is entitled to recover pursuant to this decision.