Opinion for the court filed by Circuit Judge NEWMAN, in which Circuit Judges CLEVENGER, SCHALL, BRYSON, and GAJARSA join. Opinion concurring in result filed by Circuit Judge RADER, in which Chief Judge MAYER and Circuit Judge LOURIE join. Opinion dissenting-in-part and eoncurring-in-part filed by Circuit Judge PLAGER.
NEWMAN, Circuit Judge.
We took this appeal and cross-appeal en banc to reconsider the questions of law presented, upon certification for interlocutory appeal, concerning the applicability of § 8118 of the Defense Appropriations Act of 1987 to a contract between the Department of the Navy and the American Telephone and Telegraph Company. The Court of Federal Claims ruled that in view of the failure of the Department of Defense to comply with § 8118, the contract, which had been performed, was void ab initio. We now hold that the contract was not void, and remand to the Court of Federal Claims for further proceedings in accordance with this premise.
The Reduced Diameter Array Contract
This contract arose in Cold War response to the new ultra-quiet Soviet submarines, which were difficult to monitor using available technology and equipment. Effective antisubmarine response requires that hostile submarines be reliably detected, classified, located, and tracked. The Navy, among its programs for this purpose, employed an integrated undersea acoustic sonar system called the Surveillance Towed-Array Sensor System (SUR-TASS). In SURTASS a suitably equipped surface vessel tows .an array of undersea detection equipment through the ocean, while the equipment collects and transmits appropriate data for processing on shipboard and for transmission to shore-based facilities. The President’s Annual Report [1370]*1370to the Congress for fiscal 1987, on the topic of Antisubmarine Warfare Forces, referred to SURTASS as “[o]ne of our most important ongoing programs in this area.” Id. at 188.
On December 31, 1987, after competitive bidding, the Navy awarded AT & T a fixed price incentive fee contract for a subsystem of SURTASS, referred to as the Reduced Diameter Array. The contract was a “Total Package Procurement,” requiring design of shipboard and shore-based electronics, ship-winch interface and tow cable, and an acoustic and electronic array some 8,000 feet long, to meet the new Soviet submarine capabilities. The contract required research, development, and the delivery and testing of an engineering development model, at a fixed ceding price of $19,221,630, and included an option to the Navy to acquire a second engineering development model at a fixed ceiling price of $3,510,253, and an additional option to acquire three production-level models at a fixed ceiling price of $8,475,466.
The contract was successfully performed by AT & T over a period of five years. With the price adjustments to which the Navy agreed during performance, the final fixed -price was approximately $34.5 million. AT & T states that technical problems and unknowns arose throughout performance, and that its total cost was at least $91 million. The Navy rejected AT & T’s requests for restructuring the contract and other relief, although AT & T directed attention to § 8118 of the Defense Appropriations Act and relevant Department of Defense policy directives concerning procurement of research and development for new technologies.
AT & T duly brought suit in the Court of Federal Claims under the Contract Disputes Act. On cross motions for summary judgment the issues arising from the enactment of § 8118 were presented and argued. The Court of Federal Claims ruled that § 8118 applied to this contract, that it had not been complied with by the Department of Defense, and that the contract consequently was void ab initio. Responding to AT & T’s proposal that the appropriate remedy was to reform the contract into the cost-reimbursement form favored by § 8118, the Court of Federal Claims held that since there had never been a valid contract it could not be reformed. The court held, however, that AT & T was entitled to compensation for its work on the basis of quantum meruit, on a theory of implied-in-fact contract. Before proceeding to determine quantum, the court certified for interlocutory appeal, in accordance with 28 U.S.C. § 1292(d)(2), the following questions:
(i) whether a contract executed in violation of statutory restrictions on the obligation and expenditure of appropriated funds may be declared void from the start at the instance of the performing contractor, and, if so,
(ii) whether compensation for benefits conferred upon the Government (pursuant to the voided contract) can be predicated on an implied-in-fact contract with the amount of recovery to be determined pursuant to unjust enrichment principles.
A panel of the Federal Circuit, by split decision, affirmed the ruling that the contract was void ab initio. The court also held that no relief was available to AT & T on any theory, except perhaps to replevin the goods that had been delivered to the Navy. Upon the petitions of both sides we have reheard the matter en banc.1
Section 8118 of the Defense Appropriations Act of 1987
Concern about the use of fixed price contracts for research and development [1371]*1371phases pervades defense procurement. In 1971 Department of Defense Directive (DODD) 5000.1 stated that “[i]t is not possible to determine the precise production cost of a new complex defense system before it is developed,” and established the policy of using cost-reimbursement price terms for procurement of research and development. The Directive 'stated: “Fixed price contracts are normally not appropriate for research and development phases.” DODD 5000.1 & D.9.g (as amended, Sept. 1, 1987). The Federal Acquisitions Regulations governing R & D contracts also embodied this policy. See, e.g., 48 C.F.R. § 35.006(c) (1984-1998) (“Because the absence of precise specifications and difficulties in estimating costs with accuracy (resulting in a lack of confidence in cost estimates) normally precludes using fixed-price contracting for R & D, the use of cost-reimbursement contracts is usually appropriate.”)
The record states that in the 1980s, despite these policy directives, the Navy returned to fixed price contracting for R & D as part of the Total Package Procurement concept. This in turn led to congressional investigations and hearings. An investigation conducted by the House Appropriations Committee concluded that for the development phases of new technologies, the Navy’s use of fixed price contracting resulted in program delays, cost overruns, contractor claims, non-participation, and litigation. See Surveys & Investigations Staff, Report to the Comm, on Appropriations, U.S. House of Representatives: Navy Fixed Price Contracting in the Research, Development, Test and Evaluation (RDT & E) Account, 100th Cong., 1st Sess. (1987). The Report stated that: “Although Navy officials at the headquarters level have predicted immense success for the acquisition policy, the opinions expressed by Navy and other Service field procurement officials and technical experts indicated that [fixed price contracting] generally [has] proved unsuitable in an R & D environment.” Id. at ii. The Report concluded that the nature of the work in research and exploratory development contracting “most frequently necessitates” use of the cost-re: imbursement type contract. Id. at 11.
At ensuing hearings on the 1988 Defense budget, concern was expressed about the continuing use of fixed price contracts for high-cost, high-risk development projects, as well as concern for meeting congressional oversight and allocation obligations under this form of procurement. Department of Defense Appropriations for 1988: Hearings Before the Defense Sub-comm. of the Comm, on Appropriations, 100th Cong., 454-55 (1987). Legislatively implementing these concerns, the House included in the Defense Appropriations Act of 1987 the provision that became § 8118:
§ 8118. None of the funds provided for the Department of Defense in this Act may be obligated or expended for fixed price-type contracts in excess of $10,-000,000 for the development of a major system or subsystem unless the Under Secretary of Defense for Acquisition determines, in writing, that program risk has been reduced to the extent that realistic pricing can occur, and that the contract type permits an equitable and sensible allocation of program risk between the contracting parties: Provided, That the Under Secretary may not delegate this authority to any persons who hold a position in the Office of the Secretary of Defense below the level of Assistant Under Secretary of Defense: , Provided further, That the Under Secretary report to the Committees on Appropriations of the Senate and House of Representatives in writing, on a quarterly ba[1372]*1372sis, the contracts which have obligated funds under such a fixed price-type developmental contract.
Pub.L. No. 100-202, § 8118, 101 Stat. 1329, 1329-84 (Dec. 22, 1987). The accompanying Conference Report reiterated congressional concern that the risks of failure and of cost uncertainties be allocated equitably between government and contractor, and stressed the desire to “maintain the government’s credibility as a reliable business partner.” H.R. Conf. Rep. No. 100-498 at 623 (Dec. 22, 1987). Congress referred to the burden of a fixed price contract on the contractor when the miscalculation of development cost may have been that of the government agency as well as the contractor, and to the reluctance of some highly qualified firms to enter into such contracts. The Conference Report was unambiguous: “Fixed price contracts are normally not appropriate for research and development phases.” Id. at 624. Thus Congress acted to adjust the risks of developing the advanced technologies needed in the service of national defense.
Application of Section 8118
Section 8118 prohibited the award of certain fixed price-type contracts unless the program risk was evaluated at a high level within the Defense Department, and required quarterly reports of such awards to the House and Senate Appropriations Committees. The government argues first that ’8118 did not apply to the Reduced Diameter Array contract, thus eliminating any need for the Navy to have complied with the statute. The Court of Federal Claims correctly held otherwise.
Section 8118 by its terms applies to “fixed price-type contracts in excess of $10,000,000 for the development of a major system or subsystem.” The government argues that the Reduced Diameter Array is not a “major system,” referring to a memorandum issued six weeks after enactment of § 8118 wherein the Under Secretary of Defense defined “major system” for the purposes of § 8118 as a system having a contract cost of over $75,000,000. In a Memorandum for Service Acquisition Executives, Directors of the Defense Agencies issued February 11, 1988, Under Secretary of Defense for Acquisition Costello instructed that “[t]he definition of major system at 10 U.S.C. § 2302(5) is the definition of that term for the purpose of [§ 8118].” This content was incorporated into SECNAV Instruction 4210.6A (April 13,1988).
Section 2302(5) is a provision of chapter 137 of Subtitle A — General Military Law, which as then written defined “major system” as a system costing more than $75,000,000 for research, development, test, and evaluation:
10 U.S.C. § 2302(5). The term “major system” means a combination of elements that will function together to produce the capabilities required to fulfill a mission need. The elements may include hardware, equipment, software or any combination thereof, but excludes construction or other improvements to real property. A system shall be considered a major system if (A) the Department of Defense is responsible for the system and the total expenditures for research, development, test, and evaluation for the system are estimated to be more than $75,000,000 (based on fiscal year 1980 constant dollars) or the eventual total expenditure for procurement of more than $300,000,000 (based on fiscal year 1980 constant dollars)....
The government argues that the agency had discretion to define the § 8118 “major system” in accordance with ’2302(5), and thereby to place a $75,000,000 floor on the systems to which § 8118 would apply. However, it was not within the agency’s discretion to rewrite § 8118 to replace the statutory threshold of $10,000,000 with that of $75,000,000. Although an agency’s [1373]*1373interpretation of a statute it administers is indeed entitled to deference, agency discretion does not extend to changing a clearly stated dollar figure. See Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 842-43, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984) (“if the intent of Congress is clear, that is the end of the matter”).
In addition, the AT & T contract itself, and the Space and Naval Warfare Systems Command’s guide to the SURTASS, described the Reduced Diameter Array as a “subsystem.” Subsystems were not defined in § 2305(5) and were not mentioned in the Memorandum of the Under Secretary. However, subsystems costing more than $10,000,000 were explicitly included in § 8118. Although the government now argues that the Under Secretary’s Memorandum and SECNAV Instr. 4210.6A really covered a major system or a subsystem of a major system, this interpretation is contrary to the plain text of these documents. It is apparent that the Memorandum was contrary to the statute, and in all events that it did not include subsystems such as the Reduced Diameter Array.
The government also argues that not all of the funds expended under the Reduced Diameter Array contract were appropriated in the corresponding Appropriations Act, and thus that the § 8118 prohibition on obligating or expending funds does not apply. Indeed, the contract was structured for multi-year incremental funding. However, it is undisputed that the starting research and development effort drew on several millions of dollars of appropriated funds. The multi-year funding does not excuse the Defense Department from compliance with § 8118.
Moreover, contrary to the government’s argument, which is made but not strongly pressed, this case does not involve a funding deficiency or implicate the Anti-Deficiency Act, 31 U.S.C. § 1341. See Hercules Inc. v. United States, 516 U.S. 417, 427, 116 S.Ct. 981, 134 L.Ed.2d 47 (1996) (“The Anti-Deficiency Act bars a federal employee or agency from entering into a contract for future payment of money in advance of, or in excess of, an existing appropriation.”); see generally Ferris v. United States, 27 Ct.Cl. 542, 546 (1892) (“An appropriation per se merely imposes limitations upon the Government’s own agents ... its insufficiency does not pay the Government’s debts, nor cancel its obligations, nor defeat the rights of other parties.”) There is no issue in this case of lack of appropriated funds.
We affirm the determination of the Court of Federal Claims that § 8118 applies to this contract. The government does not dispute that the requirements of § 8118 were not met by the Department of Defense. There is no assertion that the Under Secretary of Defense for Acquisitions made or had made the program risk and pricing determinations required by § 8118, and no report of this contract is stated to have been made to the Senate and House Appropriations Committees. Although the government stresses that the contract was awarded only nine days after the enactment of § 8118, this does not excuse the failure of all compliance.
Consequences of Agency Noncompliance With § 8118
We turn to the certified question of the consequences of this failure of compliance by the Department of Defense. AT & T states that § 8118 was enacted at least in part for its protection, and that the agency, by failing to obey the law, can not deprive AT & T of the protection of the law. AT & T argues that § 8118 is a “mandatory statute” restricting the agency’s authority to obligate and expend funds, and that the Navy’s direct contravention of § 8118 rendered the Reduced Diameter Array contract void ab initio.
The government responds that Congress chose and intended to enforce § 8118 [1374]*1374through its oversight powers, and that AT & T can not benefit from whatever lapses may have occurred within the Department of Defense in its compliance with congressional oversight legislation. The government stresses that § 8118 did not provide that these fixed price contracts were prohibited, but only that the Defense Department must review the risk and its allocation at a specified executive level, and must report to Congress on a quarterly basis.
Legislative intent and precedent both lead to the conclusion that the AT & T contract was not void ab initio as a consequence of the agency’s noncompliance. Invalidation of the contract is not a necessary consequence when a statute or regulation has been contravened, but must be considered in light of the statutory or regulatory purpose, with recognition of the strong policy of supporting the integrity of contracts made by and with the United States. In United States v. Mississippi Valley Generating Co., 364 U.S. 520, 81 S.Ct. 294, 5 L.Ed.2d 268 (1961) the Court explained that when a statute “does not specifically provide for the invalidation of contracts which are made in violation of [its provisions]” the court shall inquire “whether the sanction of nonenforcement is consistent with and essential to effectuating the public policy embodied in [the statute].” Id. at 563, 81 S.Ct. 294. Thus the policy underlying the enactment must be considered in determining the remedy for its violation, when the statute itself does not announce the sanction of contract invalidity.
The policy embodied in § 8118 is elucidated in the congressional response when § 8118 did not receive full compliance from the Department of Defense. See Alabama Rural Fire Ins. Co. v. United States, 215 Ct.Cl. 442, 572 F.2d 727, 733 (1978) (“illegality may be proved with reference to legislative history”). Congress simply tightened the reporting provision, by moving from after-the-fact quarterly reports to before-award reports. Indeed, the House version of § 8118 had initially required before-award reports, but this was dropped in Conference in favor of the Senate version “to reduce the appearance of congressional micromanagement.” H.R. Conf. Rep. No. 100-498 at 623 (Dec. 22, 1987).
The Conference Report stated that if Defense Department policy did not become more uniform, “more severe restrictions” would be imposed. Id. This remark carries no hint of, and indeed belies, an interpretation that § 8118 was intended, upon enactment, to invalidate any contract made without meeting its internal review and reporting requirements, for such a “restriction” would already be extremely “severe.” The statutory shift to before-award reports in succeeding years would be a trivial discipline indeed, if meanwhile all of the fixed price contracts within the statutory scope, although in the process of performance, or as in this case fully performed, were void ab initio.
Only a few months after enactment of § 8118 the House Appropriations Committee reported that the “enforcement of existing policy in this area has not yet been demonstrated,” H.R.Rep. No. 100-681 at 147 (June 10, 1988), and recommended a pre-award reporting requirement (which was included in the enactment for the next fiscal year). The Senate Armed Services Committee, considering this renewal, stated explicitly that noncompliance was not intended to be “the basis for litigating the propriety of an otherwise valid contract”:
The committee recognizes that there are circumstances in which fixed-price development contracts are appropriate (e.g., when costs and foreseeable program risks can be reasonably anticipated), and the committee expects the Department to establish clear guidelines under this section for use of such contracts.
[1375]*1375It is the intent of the committee that this section be applied in a manner that best serves the government’s interests in the long term health of the defense industry, and that this section not be used as the basis for litigating the propriety of an otherwise valid contract. Nothing in this section shall be construed to affect the requirements of section 8118 of the Department of Defense Appropriations Act, 1988.
(Emphasis added.) S.Rep. No. 100-326, 100th Cong., 2d Sess. at 105 (May 4, 1988). This explicit statement of intent weighs heavily against judicial invalidation of “an otherwise valid contract,” for the clearly stated congressional purpose is contrary.
These congressional responses, made with knowledge of the agency’s imperfect compliance with § 8118, negate any reasonable inference that Congress intended simply to render void ab initio, even after full performance, any fixed price contract for which the Under Secretary’s review of risk allocation and the report to the Committees were omitted. Congress can not have intended to charge the contracting partner with adverse consequences depending on whether the Defense Department carried out the internal responsibilities and filed the reports that Congress required.
Nor is it the judicial role to discipline the agency’s noncompliance with the supervisory and reporting instructions of congressional oversight. See Longshore v. United States, 77 F.3d 440, 443 (Fed.Cir.1996) (“Congress has undoubted capacity to oversee the performance of Executive Branch agencies, consistent with its constitutional authority. It is not for this court to instruct Congress on how to oversee and manage its creations.”); E. Walters & Co. v. United States, 217 Ct.Cl. 254, 576 F.2d 362, 367 (1978) (“The fact that a procurement practice is prohibited does not necessarily mean that it is therefore actionable. The discipline to be administered in such cases is a responsibility of the cognizant procurement officials within the agency [and not] by this court”); cf. National Treasury Employees Union v. Campbell, 654 F.2d 784, 794 (D.C.Cir.1981) (by statutory requirement that the Comptroller General report on certain expenditures “Congress itself is in a position to monitor and enforce its spending limitations. It is not for us to question the effectiveness of existing remedies and infer additional remedies.”)
Both the DoD administration of § 8118, and the congressional response to this administration, make clear that Congress did not intend that this enactment would terminate fully performed contracts because of this flawed compliance.
Precedent reinforces our conclusion that the Reduced Diameter Array contract is not void ab initio. The invalidation of a contract after it has been fully performed is not favored. Precedent shows that those contracts that have been nullified, based on a failure to meet a statutory or regulatory requirement, are contracts that have not been substantially performed. E.g., Alabama Rural Fire Ins. Co. v. United States, 215 Ct.Cl. 442, 572 F.2d 727, 733-34 (1978). In Prestex, Inc. v. United States, 162 Ct.Cl. 620, 320 F.2d 367, 374-75 (1963), the court held a contract invalid, and refused to allow any recovery because no performance had occurred. It is not surprising that much of the litigation raising issues of violation of statute or regulation at the inception of government contracts has arisen in the bid protest context, where the asserted illegality has been explored before substantial performance has occurred. E.g., CACI, Inc. v. Stone, 990 F.2d 1233, 1235 (Fed.Cir.1993); Schoenbrod v. United States, 187 Ct.Cl. 627, 410 F.2d 400, 403-04 (1969). We take incidental note that the case at bar also involved a disappointed bidder raising post-award ob[1376]*1376jections, and that none of the objections were based on § 8118.
In Harbor Gateway Commercial Property Owners’ Ass’n v. United States Environmental Protection Agency, 167 F.3d 602 (D.C.Cir.1999), a case stressed in the dissenting opinion hereto, the court voided an EPA action because the Governor had not signed the request as the statute required. However, there was no issue of performance, or reliance, or any other contractual element. It is not before us to decide whether either party to the Reduced Diameter Array contract could have voided the contract early in its life and without penalty; the contract was performed for over five years, with no record suggestion from either party that because of § 8118 there was no contract.
Judicial reluctance to annul performed contracts when the government did not comply with a statutory or regulatory requirement was explained by the Court of Claims in John Reiner & Co. v. United States, 163 Ct.Cl. 381, 325 F.2d 438, 440 (1963), stating that “the court should ordinarily impose the binding stamp of nullity only when the illegality is plain.” In Rein-er the court recognized the “dilemma” of a contractor who becomes aware, while deep in the performance of a contract, of a possible procurement illegality he did not cause: the contractor must either continue to perform a contract of uncertain validity, or discontinue performance and risk severe penalties if a court later disagrees with his assessment of the illegality.
When a contract or a provision thereof is in violation of law but has been fully performed, the courts have variously sustained the contract, reformed it to correct the illegal term, or allowed recovery under an implied contract theory; the courts have not, however, simply declared the contract void ab initio. For example, in LaBarge Products v. West, 46 F.3d 1547, 1552-53 (Fed.Cir.1995) there was an illegal disclosure by the government during bidding; this court noted that the contract had been substantially performed and held that a valid contract existed despite the violation. In Beta Systems, Inc. v. United States, 838 F.2d 1179, 1185-86 (Fed.Cir.1988) the court allowed reformation of the contract price term to correct a regulatory violation, stating that “[t]he risk of unintentional failure of a contract term to comply with a legal requirement does not fall solely on the contractor.” In Urban Data Systems, Inc. v. United States, 699 F.2d 1147, 1154 (Fed.Cir.1983) the court held that a contract price term that was contrary to law did not invalidate the fully performed contract. In Trilon Educational Corp. v. United States, 217 Ct.Cl. 266, 578 F.2d 1356, 1360 (1978) the court sustained a contract that was awarded after the contracting officer had negligently failed to meet a regulatory responsibility; the court held that the non-compliance with regulation was “a matter for internal resolution” and “did not render the resultant contract a nullity.” In Clark v. United States, 95 U.S. 539, 542, 24 L.Ed. 518 (1877) the Court held a parol contract void for violation of the statute of frauds, but allowed recovery on an implied contract theory.
The entirety of precedent strongly supports our conclusion that the Reduced Diameter Array contract is not void ab initio. Precedent does not favor the invalidation, based on governmental noncompliance with internal review and reporting procedures, of a contract that has been fully performed by either contracting party.2
[1377]*1377Although the parties discuss possible remedies, the issue of what relief may be available to AT & T is not before us, for the Court of Federal Claims did not consider AT & T’s claims on the premise that the underlying contract was not void. We have not considered this issue, and express no view thereon.
Answers to the Certified Questions
For the reasons we have discussed, we conclude that the agency’s failure to comply with the obligations of § 8118 did not render the Reduced Diameter Array contract void ab initio. Any failure by the Department of Defense in its internal compliance with § 8118 can not be invoked, particularly after full contract performance, either to strip the Navy of authority to have entered into the contract or to bar AT & T from presenting such claims, if any, that it may have.
The second certified question relates to remedy, but is based on the premise that the contract was void ab initio. Since that premise is incorrect, we do not reach the second certified question.
Costs
Each party shall bear its costs.
QUESTIONS ANSWERED; CASE REMANDED.