with whom Mr. Justice White and Mr. Justice Marshall join, dissenting.
This is a suit by petitioner against the United States to recover on a contract between petitioner and the Atomic Energy Commission. The contract included a “disputes clause,” which provided that the Commission would decide any factual disputes that arose under the contract and that its decision would “be final and conclusive unless determined by a court of competent jurisdiction to have been fraudulent, or capricious, or arbitrary, or so grossly erroneous as necessarily to imply bad faith, or not supported by substantial evidence.” *24The disputes clause also provided that while it did “not preclude consideration of law questions in connection with [disputes] decisions,” it was not to “be construed as making final the [Commission’s] decision ... on a question of law.” Disputes arose during performance of the contract, and the Commission decided them in petitioner’s favor. The General Accounting Office, however, when rendering an advisory opinion requested on behalf of the Commission as to one of the disputed items, disagreed with the Commission’s decision, and for that reason the Commission refused to pay. In petitioner’s subsequent suit in the Court of Claims, petitioner relied upon the Commission’s decision as a “final and conclusive” resolution of the disputes, entitling petitioner to summary judgment. The Department of Justice defended the suit on the grounds that the Commission’s decision was not supported by substantial evidence and was erroneous on questions of law. The issue before us is whether the Government, through the Department of Justice, may assert those defenses.
It may be helpful at the outset to put this case in perspective by reviewing briefly the law developed over the past century to regulate the enforcement of disputes clauses in Government procurement contracts. Until 1954, with the passage of the Wunderlich Act, disputes clauses provided that the decision of a designated Government official upon a matter in dispute under the contract would be final and binding upon both parties. Although in terms the disputes clauses precluded judicial review of disputes decisions, this Court beginning in 1878 consistently held that the finality of a disputes decision could be challenged in court by either party on the ground of fraud or bad faith by the deciding Government official. Thus the “fraud” exception to the finality of disputes decisions was not written into disputes clauses but was judicially fashioned.
*25Under this system, then, a contractor dissatisfied with an adverse disputes decision could contest the finality of that decision only by proving in court that it was fraudulent. The Government, of course, bore an identical burden when it contested the finality of a disputes decision in favor of the contractor. That situation arose when GAO, congressional watchdog of Government expenditures, refused to sanction payment to a contractor of the amount found due under a disputes decision in his favor and thereby forced him to bring suit. GAO’s view of the disputes decision, however, was of no consequence in court; indeed, whether or not the Government defended the contractor’s suit was a matter solely for the judgment of the Government’s lawyer, the Department of Justice. Once in court, the contractor relied upon the finality of the disputes decision and recovered on that basis unless the Government proved that the decision was fraudulent.
Over the years, the Court of Claims gradually broadened the fraud exception to the finality of disputes decisions. In 1951, however, this Court stopped the trend by holding that a disputes decision, rendered pursuant to a disputes clause purporting to make that decision final, was conclusive upon both parties unless the challenger proved in court that the deciding Government official was guilty of “conscious wrongdoing, an intention to cheat or be dishonest.” United States v. Wunderlich, 342 U. S. 98, 100 (1951). Wunderlich’s narrow definition of the fraud exception alarmed the Government as well as contractors, for, in practical effect, it meant that disputes decisions were virtually invulnerable to challenge.
The result of this concern was the so-called Wunder-lich Act, drafted by GAO and supported by GAO, Government procurement agencies, and contractors. The Act overruled Wunderlich by directing that no disputes clause, purporting to make disputes decisions final, “shall *26be pleaded in any suit ... as limiting judicial review of any [disputes] decision to cases where fraud by [the Government] official ... is alleged.” The Act did more than simply overrule Wunderlich, however, for it also explicitly stated the grounds upon which courts could set aside disputes decisions: “any [disputes] decision shall be final and conclusive unless the same is fraudulent or capricious or arbitrary or so grossly erroneous as necessarily to imply bad faith, or is not supported by substantial evidence.” Finally, the Act provided that “[n]o Government contract shall contain a provision making final on a question of law the decision of any [Government] official . . . .”
The Wunderlich Act, then, rendered the old forms of disputes clauses unserviceable, for no longer could the parties bind themselves to the finality of a disputes decision, judicially reviewable only if the challenger proved that it was fraudulent. Consequently, the disputes clause in the contract before us did not even attempt to provide for the finality of the Commission’s disputes decisions, but instead expressly tracked the language of the Act. Under this disputes clause and the Act, the party dissatisfied with a disputes decision is no longer limited to challenging the finality of that decision only on the ground that it was “fraudulent,” for the dissatisfied party is now entitled also to prove in court that the decision was “capricious,” “arbitrary,” “so grossly erroneous as necessarily to imply bad faith,” “not supported by substantial evidence,” or incorrect “on a question of law.” In this case, the Government relied upon the last two grounds to challenge the finality of the Commission’s disputes decision in favor of petitioner.1
*27As noted above, under pre-Wunderlich Act disputes clauses, which purported to make disputes decisions final, the Government, like the contractor, could avail itself of the judicially created fraud exception to the finality of disputes decisions. The Government obtained judicial review when GAO refused to sanction payment after a disputes decision in favor of the contractor, thus forcing him to bring a suit in which the Department of Justice represented the Government. That was precisely the path followed in this case, for GAO, in response to a request for an advisory opinion, informed the Commission that payment would be improper because the disputes decision did not meet the standards of the Wunderlich Act, and, in petitioner’s subsequent suit, the Department of Justice represented the Government. Had this case arisen under earlier forms of disputes clauses, which purported to make disputes decisions final, and before the Wunderlich Act, the Government could have defended the suit only on the judicially created ground that the disputes decision was fraudulent. Under the current clause and the Act, *28however, the Government is not limited to that narrow ground. Like the contractor, the Government may now also rely upon any or all of the other grounds enumerated in the clause and the Act. The Commission’s disputes decision is not “final and conclusive,” under the clause and the Act, if the Court of Claims determines, as the Government asserted here, that the decision was “not supported by substantial evidence” or was incorrect “on a question of law.” 2
Yet the Court today holds that the Government has no right to defend petitioner’s suit. Had the Commission’s disputes decision been adverse to petitioner, of course, petitioner would have been free to challenge its finality in court, under the disputes clause and the Act, on the grounds that it was “not supported by substantial evidence” and was incorrect “on a question of law.” The Court holds, however, that the Government may not challenge the finality of the disputes decision in favor of petitioner because the Government, under the disputes clause and the Act, has no right to judicial review of disputes decisions.3 The Court reaches this *29conclusion on the strength of its assertions that GAO had no business exercising its statutory authority and advising the Commission that the disputes decision was erroneous, that the Department of Justice had no business exercising its statutory authority and appearing in the Court of Claims to defend petitioner’s suit, and that the Government is always entitled to relief if the contractor perpetrates a fraud. Noticeably absent from the Court’s opinion is any justification for interpreting the disputes clause and the Act to apply only when a disputes decision is adverse to the contractor. Somehow the Court construes a contract and a statute that bar finality for all disputes decisions to require finality for disputes decisions in favor of contractors.
Today’s decision is demonstrably wrong. The Court holds that Congress enacted the Wunderlich Act for the benefit of contractors, to arm them with grounds in addition to fraud to challenge in court the finality of disputes decisions unfavorable to them. Yet, without an iota of support in the language of the Act, which expressly governs “any” disputes decision in “any suit,” or in the Act’s legislative history, which confirms that the expanded grounds of judicial review were to be available to both the Government and contractors, the Court holds that the Government, unlike contractors, may not rely upon the Act to challenge in court the finality of disputes decisions. Indeed, the Court goes further, for, as noted, the disputes clause before us did not purport to make the Commission’s disputes decisions final. The Court thus holds that the Act denies the Government the privilege of entering into a contract that affords it as well as the contractor the right to judicial review of disputes decisions. Hence, while the *30Act ensures that contractors are entitled to judicial review even when the disputes clause provides for finality, the Act also, according to the Court, ensures that the Government is denied judicial review even when the disputes clause does not provide for finality. Today’s decision produces the absurd result that when the Government agreed to a disputes clause with no provision for judicial review, it could nevertheless challenge the finality of a disputes decision at least for fraud, but now that the Government has agreed to a disputes clause specifying five grounds of judicial review, including fraud, it is entitled, holds the Court, to none at all.4 The Government’s position is thus worse than it was before the Act, for it is deprived of even the limited review for fraud to which it was entitled under Wunder-lich. Finally, the Act flatly prohibits disputes clauses that make disputes decisions final on questions of law. The clause before us, following the Act, expressly pro*31vided that the Commission’s disputes decisions could not be final on questions of law. Yet, in the face of the Act and the disputes clause, the Court holds that the Commission’s decision is final on questions of law.
Analysis of the judicial history of disputes clauses, both in this Court and in the Court of Claims, will unfortunately unduly extend the length of this opinion. But the devastation today’s decision wreaks upon Government procurement practices is sufficient justification, and Congress should be alert to the urgent need for immediate remedial legislation. Congress alone can restore the former balance between Government and contractor, for today’s decision not only holds that the Act’s expanded scope of judicial review is available solely for contractors, but also holds that the Act, in some unspecified way, prohibits the contracting parties from agreeing to a disputes clause that affords the Government that same scope of review. Congress must therefore make more explicit what is already explicit in the Wunderlich Act, but this time in terms so plain that even this Court will be unable to thwart the congressional will.
I
A
The contract in Kihlberg v. United States, 97 U. S. 398 (1878), as the Court construed it, provided that the decision of a designated Government official would be “conclusive.” The official rendered a decision adverse to the contractor, and the contractor brought suit. Because there was “neither allegation nor proof of fraud or bad faith” by the official, the Court held that his decision could not “be subjected to the revisory power of the courts without doing violence to the plain words of the contract.” Id., at 401. The Court then enunciated the standard of judicial review that has been the *32basis for the decision of every subsequent disputes clause case, both in this Court and in the Court of Claims: “in the absence of fraud or such gross mistake as would necessarily imply bad faith, or a failure to exercise an honest judgment, his action in the premises is conclusive upon the [contractor] as well as upon the government.” Id., at 402 (emphasis added).
The very first case in this Court, then, laid down the rule that a decision rendered pursuant to a disputes clause was equally binding upon both parties; the contractor and the Government could impeach a disputes decision that the contract purported to make final, but only by proving that the decision was fraudulent. Until today, this Court never departed from the Kihlberg view that the same standard of judicial review is available to both parties.
Sweeney v. United States, 109 U. S. 618 (1883), reiterated the Kihlberg rule in another suit by a contractor dissatisfied with a disputes decision rendered by a Government official. Because “there was neither fraud, nor such gross mistake as would necessarily imply bad faith, nor any failure to exercise an honest judgment on the part of the officer,” the Court held, “on the authority of Kihlberg v. United States,” that the official’s decision was conclusive. Id., at 620.
The Court next decided three cases involving contracts between private parties. In Martinsburg & Potomac R. Co. v. March, 114 U. S. 549 (1885), a contractor agreed to do certain work for a railroad company, and the contract provided that disputes would be decided by a company official whose decision would be “final and conclusive.” Id., at 553. The official’s decision was in favor of the company, and the contractor brought suit. The Court, stating that the “case is within the principles announced in Kihlberg v. United States and *33Sweeney v. United States,” id., at 550 (here, and in subsequent similar quotations, citations not repeated), held that the officiars decision was conclusive because there was no proof that he “had been guilty of fraud, or had made such gross mistake in his estimates as necessarily implied bad faith, or had failed to exercise an honest judgment in discharging the duty imposed upon him,” id., at 553.
The contract in Chicago, S. F. & C. R. Co. v. Price, 138 U. S. 185 (1891), was essentially the same as the contract in March. In Price, however, the official’s disputes decision was in favor of the contractor. The company refused to pay in accordance with the decision, and the contractor brought suit. The Court first reviewed March and stressed that March had applied “the principles announced in Kihlberg v. United States and Sweeney v. United States.” Id., at 193. The Court then pointed out that “[t]he only difference between that case [March] and the present one is that the alleged mistakes of the engineer in the former were favorable to the railroad company, while in this case they are favorable to the contractors.” Id., at 194. “[T]hat difference,” said the Court, “cannot affect the interpretation of the contract.” Ibid. Because there was no proof of “fraud upon the part of the company’s engineers, or such gross mistakes by them as imply bad faith,” the Court held that the disputes decision was binding upon the company. Id., at 195.
Price thus established that the party whose employee was delegated authority to make the disputes decision could also challenge the finality of that decision, although, like the contractor, only under the Kihlberg test of fraud. The Court reaffirmed this application of the Kihlberg rule in Sheffield & Birmingham Coal, Iron & R. Co. v. Cordon, 151 U. S. 285 (1894), holding that *34“in the absence of fraud or mistake” by the company official, his decision in favor of the contractor “was conclusive upon the company.” Id., at 292.
United States v. Gleason, 175 U. S. 588 (1900), involved a Government official’s disputes decision adverse to the contractor. The Court again affirmed the rule of Kihlberg and the intervening cases
“that it is competent for parties to a contract, of the nature of the present one, to make it a term of the contract that the decision of an engineer, or other officer, of all or specified matters of dispute that may arise during the execution of the work shall be final and conclusive, and that, in the absence of fraud or of mistake so gross as to necessarily imply bad faith, such decision will not be subjected to the revisory power of the courts. Martinsburg & Potomac Railroad v. March; Chicago, Sante Fe &c. Railroad v. Price.” Id., at 602.
The Court also followed the Kihlberg rule in Ripley v. United States, 223 U. S. 695, 701-702, 704 (1912), and Merrill-Ruckgaber Co. v. United States, 241 U. S. 387 (1916).
In United States v. Mason & Hanger Co., 260 U. S. 323 (1922), the contractor was paid in accordance with a disputes decision in his favor, but the Comptroller of the Treasury disagreed with the decision and subsequently deducted the amount paid from other sums due the contractor. Id., at 325. The contractor brought suit, relying upon the finality of the disputes decision. The Court’s holding was direct and simple:
“We have decided that the parties to the contract can so provide and that the decision of the officer is conclusive upon the parties. Kihlberg v. United States; Martinsburg & Potomac R. R. Co. v. March; *35United States v. Gleason; Ripley v. United States. This is extending the rule between private parties to the Government.” Id., at 326.
Mason & Hanger, then, applied the Kihlberg rule when the contractor in a Government contract relied upon the disputes decision by a Government official and the Government challenged it. Hence, both parties to a Government contract, like both parties to a private contract, as in Price and Gordon, were free to challenge the finality of a disputes decision, although only upon the limited grounds permissible under Kihlberg.
Mason & Hanger also held that “the Comptroller of the Treasury has no power” over a disputes decision, 260 U. S., at 326, meaning that his disagreement with the decision was irrelevant and had no effect in court, where the parties’ rights under the contract were determined. The Government, like the contractor, could prevail only by proving that the disputes decision was fraudulent. The Comptroller’s authority was limited to his power to refuse to sanction payment to the contractor, thereby forcing the contractor to bring suit for a judicial determination of his right to payment in accordance with the disputes decision in his favor.5
In sum, the rule first announced in Kihlberg in 1878 had, with Mason & Hanger in 1922, been held to apply to any disputes decision, whether in a Government or in a private contract, and to apply no matter which party relied upon the finality of the decision. If the Government (or, in a private contract, the party whose official decided the dispute) relied upon the finality of the decision, the contractor had to prove that it was fraudulent. Kihlberg; Sweeney; March; Gleason. If *36the contractor relied upon the finality of the decision, the Government (or, in a private contract, the party whose official decided the dispute) had to prove that it was fraudulent. Price; Gordon; Mason & Hanger.6
In United States v. Moorman, 338 U. S. 457 (1950), the Court once again gave extended consideration to the proper judicial interpretation of disputes clauses. The Court pointed out that “ [contractual provisions such as these have long been used by the Government. No congressional enactment condemns their creation or enforcement.” Id., at 460. The Court then reviewed Kihlberg, Sweeney, and March, and said that “[t]he holdings of the foregoing cases have never been departed from by this Court. They stand for the principle that parties competent to make contracts are also competent to make such agreements.” Id., at 461. The Court added that “[i]f parties competent to decide for themselves are to be deprived of the privilege of making such anticipatory provisions for settlement of disputes, this deprivation should come from the legislative branch of government.” Id., at 462.
Finally, came United States v. Wunderlich, 342 U. S. 98 (1951). The contract contained the usual disputes clause providing that the disputes decision was “final and conclusive.” Id., at 99. After noting that the *37same disputes clause had been upheld in Moorman, the Court stated:
“Contracts, both governmental and private, have been before this Court in several cases in which provisions equivalent to [this disputes clause] have been approved and enforced ‘in the absence of fraud or such gross mistake as would necessarily imply bad faith, or a failure to exercise an honest judgment . . . .’ Kihlberg v. United States; Sweeney v. United States; Martimsburg & P. R. Co. v. March; Chicago, S. F. C. R. Co. v. Price.” Id., at 99-100.
We thus have an unbroken line of cases in this Court, from 1878 to 1951, applying a simple, straightforward rule of judicial review. A contractual disputes clause making final a decision by an agent of one of the parties was given full effect in court, subject to the judicially created exception that allowed relief to the party challenging the decision if he was able to prove that it was fraudulent. This rule applied whether the contract was Government or private and no matter which party challenged the finality of the decision. In short, a disputes clause was equally binding upon both parties.
B
Most disputes clause cases, of course, have been decided not by this Court but by the Court of Claims. That court followed the Kihlberg rule when a contractor challenged a disputes decision against him, see, e. g., Kennedy v. United States, 24 Ct. Cl. 122 (1889); P. H. McLaughlin & Co. v. United States, 37 Ct. Cl. 150 (1902); Pacific Hardware Co. v. United States, 49 Ct. Cl. 327 (1914); Brinck v. United States, 53 Ct. Cl. 170 (1918); Southern Shipyard Corp. v. United States, 76 Ct. Cl. 468 (1932), as well as when the Government challenged a disputes decision in the contractor’s favor.
*38In Pacific Hardware, supra, the contract provided that a Government official would deduct specified amounts from the contract price if the contractor delayed in performing the contract. Deductions were made, and the contractor brought suit. The court applied the Kihlberg rule and upheld the deductions. 49 Ct. Cl., at 336. The contract also provided that the official could waive deductions under certain circumstances. The contractor argued that this power violated public policy and therefore vitiated the contract. The court rejected the argument, but added that the power to decide in favor of the contractor by waiving deductions, like the power to decide against the contractor by making deductions, was subject to the Kihlberg rule:
“Of course, if there were fraud or such gross error as implies bad faith or a failure to exercise an honest judgment in deciding that the deductions be not made, the Government would not be bound and the contractor would remain liable.” Id., at 337.
In Yale & Towne Mfg. Co. v. United States, 58 Ct. Cl. 633 (1923), the disputes decision was in favor of the contractor, but the Government refused to pay because the Comptroller of the Treasury disagreed with the decision. The contractor argued “that the contract reposed in the contracting officer . . . the right to determine whether or not and the extent to which the contractor was entitled to extension of time, and that the finding of that officer was conclusive upon the parties in the absence of fraud or mistakes so gross as to imply bad faith.” Id., at 637.
The court, noting “that a long line of decisions not only by this court but by the Supreme Court requires the sustaining of the .[contractor’s] contention,” stated:
“Provisions in Government contracts reposing in some designated official the right to determine cer*39tain questions and making his determination thereof conclusive are of frequent occurrence. Such provisions are inserted largely for the protection of the Government, and the cases in which such a determination by the designated official has been upheld by the courts have been largely cases in which the rule has been invoked in favor of the United States and against the [contractor], but the rule is none the less effective if perchance it occasionally may operate the other way.” Id., at 638 (emphasis added).
In Penn Bridge Co. v. United States, 59 Ct. Cl. 892 (1924), the disputes decision was in favor of the contractor, but the Comptroller General disagreed with the decision and deducted the amount from other sums due the contractor. The Court, referring to the Comptroller’s attempt to “substitute his judgment for that of the contracting officer and thereby eliminate from the case the finding of the contracting officer when the rights of the parties are in this court for adjudication,” id., at 898, stated that “action by the comptroller could [not] in any way conclude this court in the determination of the rights of the parties under the contract,” id., at 896. The court then applied the Kihlberg rule. Id., at 897.
Penn Bridge, then, aside from reaffirming that the same rule of judicial review applied whether the Government or the contractor challenged the finality of a disputes decision, also demonstrates that GAO’s view of the correctness of a disputes decision was of no effect in court. GAO’s only power — the power of the purse — was to force the contractor to bring suit and thus to obtain judicial review for the Government. But once the case reached court, review was the same for both parties.
GAO’s opinion of a disputes decision was irrelevant in court even when GAO favored the contractor. In *40Eaton, Brown & Simpson, Inc. v. United States, 62 Ct. Cl. 668 (1926), the disputes decision was in favor of the Government, but the Comptroller General disagreed and paid the contractor. In the contractor’s suit to recover on other claims, the court held that the disputes decision controlled and deducted the amount GAO had paid from other sums due the contractor. “The action of the comptroller is not conclusive upon this court in determining the rights of the parties. See Penn Bridge Co. v. United States.” Id., at 685.
In Carroll v. United States, 76 Ct. Cl. 103 (1932), the Comptroller General disagreed with a disputes decision in favor of the contractor and assessed damages in a sum greater than the amount due under the contract. The contractor brought suit, and the Government argued that it was entitled to the excess. The court replied:
“The issue is not a new or novel one insofar as judicial precedents are concerned. At least beginning with the case of Kihlberg v. United States to the present time, the Supreme Court has uniformly-held that in Government contracts containing provisions similar to the one in suit, the parties are competent to bind themselves to the conclusiveness and finality of the action and findings of the department with which the contract is made, and that such action is not open to the supervisory power of the courts unless overturned by proof of fraud or such gross error as to warrant the implication of fraud.” Id., at 124-125.
In Albina Marine Iron Works v. United States, 79 Ct. Cl. 714 (1934), the disputes decision was in the contractor’s favor, but the Comptroller General disagreed and assessed damages. The court held that the disputes decision
“was a final disposition of the matter. Neither *41fraud nor bad faith is alleged or proven. The court cannot go behind the decision of the contracting officer where the contract makes him the final arbiter of the facts of the case unless there has been fraud or such gross error which, in effect, would imply bad faith. The cases in this court and the Supreme Court so holding are numerous.” Id., at 720.
After repeating that it could not review the disputes decision “without the establishment of fraud or such gross error which would imply bad faith,” the court concluded:
“It is seldom that a case arises like the instant case, where the contractor is upholding the decision of the contracting officer and the Government is attempting to overthrow the decision of the officer appointed and designated by it to contract and carry out the terms of the undertaking. Unless proven to the contrary, full faith and credit should be accorded an officer of the Government in arriving at a decision which requires fair and impartial action on his part.” Id., at 721.
In McShain Co. v. United States, 83 Ct. Cl. 405 (1936), the designated Government official decided that the contractor’s delay in completing the contract was unavoidable. The Comptroller General later decided that part of the delay was the contractor’s fault and deducted damages from the amount due under the contract. The contractor brought suit, relying upon the finality of the disputes decision. The court said:
“Neither fraud nor bad faith is alleged or proven. This court and the Supreme Court by numerous decisions have held there is no going behind the decision of the contracting officer when the contract provides that ‘his finding of facts therein shall be final and conclusive on the parties thereto.’ The action of the Comptroller General was without *42legal authority. Kihlberg v. United States; United States v. Gleason.” Id., at 409.7
In B-W Construction Co. v. United States, 97 Ct. Cl. 92 (1942), the Comptroller General deducted damages for delay after a disputes decision in the contractor’s favor. The court held that because of the disputes clause “[i]t is . . . the action of the head of the department that is before us for review. On the question now before us that action is binding on us unless we find that it was arbitrary or grossly erroneous. In no event are we bound under this contract by the action of the Comptroller General.” Id., at 123.
In Mitchell Canneries v. United States, 111 Ct. Cl. 228, 77 F. Supp. 498 (1948), the Comptroller General disagreed with a disputes decision in favor of the contractor and set off that amount against other sums due the contractor on other contracts. The court applied “{t]he established principle of law that the findings of fact of a contracting officer are binding upon both the Government and the contractor if there is no fraud, gross error or arbitrariness by the contracting officer amounting to bad faith.” Id., at 247, 77 F. Supp., at 502.
These Court of Claims cases are further cogent authority that the Government was, until today, entitled to exactly the same judicial review as contractors. A disputes clause providing for a final decision by a Govern*43ment official was equally binding upon both parties. GAO’s opinion of that decision was irrelevant in court. GAO’s only power was to refuse to sanction payment under a disputes decision favorable to a contractor and thereby compel the contractor to bring suit. Once in court, the standard of review applicable to contractor challenges likewise controlled the Government’s challenge.
The district courts reached the identical result. In James Graham Mfg. Co. v. United States, 91 F. Supp. 715 (ND Cal. 1950), the Comptroller General refused to accept a disputes decision in favor of the contractor. Although the agency adhered to the merits of its decision, it refused to pay because of the Comptroller’s contrary view. The court said;
“Another officer of the United States government, the Comptroller General, who has general control of the government’s purse strings, has refused to sanction payment of the account which the Navy Department has approved. The question . . .is: Has he power to determine that payment shall not be made?
“The powers of the Comptroller General are extensive and broad. But he does not, absent fraud or overreaching, have authority to determine the propriety of contract payments when the contracts themselves vest the final power of determination in the contracting executive department. United States v. Mason & Hanger Co.; United States v. Moorman.” Id., at 716.8
*44In Consolidated Vultee Aircraft Corp. v. United States, 97 F. Supp. 948 (Del. 1951), the contractor received an adverse disputes decision from the contracting officer but won reversal on appeal to the agency. GAO disagreed with the agency’s decision and refused to pay, forcing the contractor to bring suit. The court held for the contractor on the authority of Mason <fe Hanger, Penn Bridge, and James Graham. Id., at 951.
C
The law was thus crystal clear. The district courts, the Court of Claims, and this Court consistently applied the rule, originally announced almost a century ago in Kihlherg, that contractual clauses providing for the finality of disputes decisions rendered by an employee of one of the parties were enforceable in court, with the judicially created exception for fraudulent decisions. No court, nor even any contractor, ever questioned that GAO could obtain judicial review for the Government simply by refusing to approve payment on a disputes *45decision favorable to a contractor. It was accepted by all that the Government and the contractor both were entitled to judicial review.9 The problem that gave rise to the Wunderlich Act was not who was entitled to judicial review nor how judicial review was to be attained. The problem was the scope of judicial review.
As the Court noted in United States v. Bianchi & Co., 373 U. S. 709, 713 (1963), under the Kihlberg rule a court’s function “in matters governed by ‘disputes’ clauses was in effect to give an extremely limited review of the administrative decision”; the Court of Claims, however, had “somewhat expanded” the scope of judicial review “over the years.” See, e. g., Needles v. United States, 101 Ct. Cl. 535, 601-607 (1944). It was this expansion of the scope of judicial review that Wunder-lich addressed.
Certiorari was granted in Wunderlich “to clarify the rule of this Court which created an exception to the conclusiveness of such administrative decision[s].” 342 U. S., at 99. The Court gave a restrictive interpretation to this exception.
“Despite the fact that other words such as ‘negligence,’ ‘incompetence,’ ‘capriciousness,’ and ‘arbi*46trary’ have been used in the course of the opinions, this Court has consistently upheld the finality of the department head’s decision unless it was founded on fraud, alleged and proved. So fraud is in essence the exception. By fraud we mean conscious wrongdoing, an intention to cheat or be dishonest. The decision of the department head, absent fraudulent conduct, must stand under the plain meaning of the contract.” Id., at 100.
Within a month after Wunderlich was decided, its restrictive scope of judicial review was applied against the Government. In Leeds & Northrup Co. v. United States, 101 F. Supp. 999 (ED Pa. 1951), the contractor, after a favorable disputes decision, was reimbursed for certain costs. Several years later, GAO reviewed that decision, disagreed with it, and set off the amount already paid from sums due the contractor on another contract. The contractor was therefore compelled to bring suit. The court first pointed out that GAO’s power
“is subject to the rights of parties to a contract, including the Government, to provide for some designated person or persons, even if in the employ of one of the parties, to make a final determination of any question which may arise between them. This principle has been unequivocally declared by the courts, including the Supreme Court of the United States, in many cases.” Id., at 1002.
After quoting extensively from James Graham, the court stated the rule of judicial review as follows:
“The Bureau’s determinations of questions of fact under [the disputes clause] are final and conclusive in the absence of fraud. United States v. Wunder-lich. For a court to set aside such determinations under [the disputes clause], fraud, meaning conscious wrongdoing or an intention to cheat or be *47dishonest, must be alleged and proved. United States v. Wunderlich.” Id., at 1003.
See also Sunroc Refrigeration Co. v. United States, 104 F. Supp. 131 (ED Pa. 1952), which, following Leeds & Northrup, also applied the Wunderlich scope of review against the Government.
II
The Wunderlich opinion concluded, “If the standard of fraud that we adhere to is too limited, that is a matter for Congress.” 342 U. S., at 100. Almost immediately after the decision was issued, congressional legislation was sought to expand the scope of judicial review limited by Wunderlich to “fraud” in a narrow sense. I have attached an Appendix detailing the legislative history and shall only summarize that history here.
Although several bills were introduced in the 82d Congress, congressional attention focused upon S. 2487. In its original form, S. 2487 provided:
“That no provision of any [Government] contract . . . relating to the finality or conclusiveness of any decision of the Government [official], in a dispute involving a question of fact arising under such contract, shall be construed to limit judicial review of any such decision only to cases in which fraud by such Government [official] is alleged.”
Wunderlich, of course, construed the standard disputes clause, which purported to make disputes decisions final, to limit judicial review to instances of fraudulent decisions. S. 2487, then, was simply an acceptance of the invitation extended in Wunderlich itself. S. 2487, however, did not specify what the scope of judicial review would be, but merely directed that judicial review could not be limited to fraud. Moreover, there was no indication in the language of S. 2487 that it was over*48ruling Wunderlich only as to disputes decisions unfavorable to contractors. It obviously applied to the judicial review of “any such decision.” (Emphasis added.)
The Comptroller General’s initial report of GAO’s views on S. 2487 made that abundantly clear. The report criticized Wunderlich as contrary to the interests of both the Government and contractors. Indeed, as a representative of the Government, the Comptroller General stressed Wunderlich’s undesirable impact upon the Government’s interest, for administrative “officials can make just as arbitrary determinations in favor of contractors, at the expense of the taxpayers.” 10 And, as the Assistant Comptroller General put it in his testimony at the Senate hearings, Wunderlich “means that the decision of the administrative officials nearly always will be final because of the extreme difficulty of proving fraud.” 11 Because the restricted scope of judicial review prescribed in Wunderlich applied to the Government no less than to contractors, GAO had good reason for its concern.12
GAO then offered a substitute bill that it believed would protect the Government’s interests. The bill provided that a disputes clause decision
“shall not be treated as binding if the General Accounting Office or a court finds that the action of [the Government official] is fraudulent, arbitrary, *49capricious, grossly erroneous, or that it is not supported by substantial evidence.”
GAO’s substitute bill thus differed from S. 2487 in two respects. First, rather than merely reversing Wunder-lich, it explicitly defined the expanded scope of review by specifying five grounds upon which a disputes decision could be set aside. Clearly this expanded review was to operate for both contractors and the Government, just as the “fraud” standard of review always had. It would be absurd to suppose that GAO defined the expanded scope of review only for contractors.
Second, GAO’s substitute bill authorized GAO review in addition to judicial review. More precisely, it empowered GAO as well as the courts to set aside any disputes decision, whether favorable to the contractor or favorable to the Government. That was a significant expansion of S. 2487. GAO never previously was empowered to upset a disputes decision. Rather, GAO authority was always limited to refusing to sanction payment on a decision favorable to a contractor, thereby forcing him into court. At that point, of course, GAO’s view of the merits of the disputes decision was irrelevant. Consequently, GAO’s substitute bill, if enacted, would have increased GAO’s power enormously, for it effectively authorized GAO to oust the courts of all jurisdiction to review disputes decisions that GAO considered unacceptable. Not surprisingly, this part of GAO’s proposal became highly controversial.
Extended hearings on S. 2487 were held in the Senate. Although most of the witnesses and statements concerned themselves solely with urging expanded judicial review for contractors, without adverting to such review for the Government, there were notable exceptions. The Associated General Contractors took the position that judicial review must be available to both parties, as did *50several attorneys who specialized in the representation of contractors.13 Opponents of that view proposed bills that would have expressly limited the right of judicial review to contractors.14 The Comptroller General subsequently submitted another report objecting to these bills because their adoption would deprive the Government of the defense of administrative finality while permitting contractors “to utilize such defense should the accounting officers of the Government attempt to question the validity of a payment.” 15 It is significant that no one ever suggested during the Senate hearings that the expanded scope of review provided in S. 2487 and GAO’s substitute bill was to be available only for contractors and not also for the Government.
An amended S. 2487 was reported out of Committee following the hearings.16 It provided that no disputes clause
“shall be pleaded as limiting judicial review of any [disputes] decision to cases in which fraud by [the Government] official ... is alleged.”
Thus, amended S. 2487, like the bill in its original form, contained an explicit reversal of the Wunderlich standard of judicial review. Like the original bill, moreover, amended S. 2487 gave not the slightest indication that it was a command solely to the Government not to “plead” the disputes clause as limiting the contractor’s right to judicial review. Amended S. 2487 plainly di*51rected that no disputes clause could be pleaded to limit judicial review of any disputes decisions. Neither party, under amended S. 2487, could rely upon a disputes clause to limit the other party’s right to judicial review to instances of fraudulent disputes decisions.
Amended S. 2487, however, went beyond the original bill by incorporating GAO’s substitute bill:
“[A]nd any such provision shall be void with respect to any such decision which the General Accounting Office or a court, having jurisdiction, finds fraudulent, grossly erroneous, so mistaken as necessarily to imply bad faith, or not supported by reliable, probative, and substantial evidence.”
Thus, amended S. 2487 reversed Wunderlich, adopted GAO’s definition of the expanded scope of review, and authorized GAO as well as the courts to apply that expanded review and set aside any disputes decisions.
The Committee Report on amended S. 2487 expressly noted “that to the same extent [the Wunderlich] decision would operate to the disadvantage of an aggrieved contractor, it would also operate to the disadvantage of the Government in those cases, as sometimes happens, when the contracting officer makes a decision detrimental to the Government interest in the claim.” 17 The reversal of Wunderlich, then, was clearly seen as an expansion of judicial review that would apply no matter which party, the Government or the contractor, challenged the disputes decision.
The report then explained that the addition of GAO’s proposal meant that amended S. 2487 would
“have the effect of permitting review in the General Accounting Office or a court with respect to any decision of a contracting officer or a head of an agency which is found to be fraudulent, grossly erroneous, *52so mistaken as necessarily to imply bad faith, or not supported by reliable, probative, and substantial evidence. In other words, in those instances where a contracting officer has made a mistaken decision, either wittingly or unwittingly, it will not be necessary for the aggrieved party to, in effect, charge him with being a fraud or a cheat in order to affect [sic] collection of what is rightfully due.” 18
Thus, the expanded scope of review, explicitly defined, would be available to both parties before either GAO or a court. In short, amended S. 2487 empowered a court to set aside a disputes decision at the behest of either the Government or the contractor, and, likewise, it empowered GAO to set aside a decision challenged by either party. Although the report asserted that amended S. 2487 was intended “simply to recognize the jurisdiction which the General Accounting Office already has,” 19 in fact amended S. 2487 would have given GAO the entirely new power to make a binding review of disputes decisions. It would have made GAO, as was later charged, into a second court of claims.
Although the Senate passed amended S. 2487, the 82d Congress expired without House action. When it was reintroduced in the Senate of the 83d Congress,20 Senator McCarran, the bill’s sponsor, observed that the Wunder-lich decision “cuts two ways” and, as an example, cited a case I have already discussed, Leeds & Northrup Co. v. United States, 101 E. Supp. 999 (ED Pa. 1951), in which “[t]he Comptroller General . . . attempted to recover on behalf of the Government, because the mistake was against the Government. The contractor interposed a *53defense based on . . . the Wunderlich case. . . . [T]he result was a failure of recovery on behalf of the Government.” 21 Thus, Senator McCarran, like GAO, recognized that the narrow review permissible under Wunder-lich bound both the Government and the contractor, and, like GAO, he considered that reversal of Wunderlich would also apply equally to both parties. A month later during floor debate, Senator McCarran again emphasized that while Wunderlich could “operate greatly to the disadvantage of contractors,” it could also “operate to the disadvantage of the Government.” 22 The Senate then passed the bill, obviously with the understanding that the expanded scope of judicial review provided would be available to both the Government and contractors.
Amended S. 2487 was also introduced in the House of the 83d Congress.23 At the initial House hearing in July 1953, several witnesses asserted that enactment of the bill was essential to enable both the Government and contractors to obtain effective judicial review of disputes decisions.24 Opposition then developed to the provision empowering GAO to invalidate such decisions. The objection was, quite predictably, that “[t]he effect of the provision is to set up the General Accounting Office as a 'court of claims.’ . . . [A]n agency of the legislative branch . . . should not be used to perform functions intended for the judicial branch.” 25
Understanding the precise nature of this objection is important. No one suggested that amended S. 2487 *54did not grant the Government the same scope of judicial review that it granted contractors. Obviously, since amended S. 2487 authorized both GAO and the courts to exercise the same review, and since the objection was that GAO should not be able to set aside disputes decisions favorable to contractors, it would have been absurd to suggest that amended S. 2487 did not likewise authorize the courts to set aside such decisions. Nor did anyone question the ability of GAO to obtain judicial review for the Government through its power to refuse to approve payment on disputes decisions. All agreed that the purpose of the proposed legislation was to overturn the standard of review set by Wunderlich; the narrow scope of judicial review permissible under that case was to be done away with in favor of a broader, specifically defined'review. The purpose was to expand judicial review, not to insert further administrative review into the disputes process. Thus, the opposition urged, not unreasonably, that the avowed purpose of overruling Wunderlich would not be served by expanding GAO’s power to transform it into another court. Hence, deletion of GAO from amended S. 2487 would leave the power of binding review exclusively with the courts.
The Comptroller General bowed to this opposition. Stating (erroneously, I think) that GAO “has not asked for authority which it did not have before the decision in the Wunderlich case,” he offered another substitute bill deleting the objectionable provision. He asserted that “this substitute language will accomplish what we have been striving for all along and will place the General Accounting Office in precisely the same situation it was in before” Wunderlich,26 This bill, in the form submitted by GAO with one minor addition, was enacted as the Wunderlich Act.
*55Thus, the result of GAO’s attempt to obtain the power of binding review over disputes decisions was failure. That power was left where it was before the Act, solely with the courts. GAO simply retained the power it had always had, the power to force the contractor into court where the Government would get judicial review of the disputes decision in his favor.
The hearings resumed in January 1954. In urging passage of GAO’s revised substitute bill, GAO’s General Counsel stated that, despite deletion of the provision for binding GAO review, the bill would not only protect contractors but would also protect the Government “against decisions adverse to the interests of the United States. Certainly the rights of contract [ors] and the Government to review or appeal should be coextensive.” 27 Similarly, the Associate General Counsel of the General Services Administration asserted that GAO’s revised substitute bill was adequate to “insure an opportunity to protect the Government against excessive generosity,” since GAO, under the bill, “could seek a court review by a set-off or by applying to the Department of Justice for recovery in a case where they felt that the action of the contracting officer was grossly erroneous as against the Government.” 28
Many witnesses who opposed GAO’s original substitute bill, and thus opposed amended S. 2487, now supported GAO’s revised substitute bill because it made clear that the power to set aside disputes decisions was vested exclusively in the courts and not shared by the courts with GAO. There was no suggestion from anyone that deletion of GAO from amended S. 2487 also had the effect of precluding the Government from obtaining judicial review under the standards available to contractors. Any *56such suggestion would have been absurd, for, as noted above, amended S. 2487 granted the courts and GAO exactly the same power. In fact, at one point in the hearings, a witness objected that GAO’s revised substitute bill did “not say specifically that an appeal can be taken by an aggrieved contractor.” The ensuing colloquy with Committee members made plain that the language of the bill “necessarily include [d] both parties.”29 Moreover, as in the case of the Senate Committee, the House Committee was presented with a proposed bill that would have expressly limited the right of judicial review to contractors.30 As with the Senate, that suggestion was not adopted. Instead, the Committee reported out the bill, submitted by GAO, that is now the Wunderlich Act.
The Act expanded the scope of judicial review, and that was all it did. The Committee report made that plain. “The committee foresees no possibility of the proposed legislation creating any new rights that a contractor may not have had prior to its enactment, with the exception of the standards of review therein prescribed.” 31 Nor did the Act grant GAO new power, for, as the report said, “there is no intention of setting up the General Accounting Office as a 'court of claims.’ ” On the other hand, the Act did not diminish GAO’s existing authority to hold up payment and force the contractor to bring suit, as the report also stressed. “The elimination of the specific mention of the General Accounting Office from the provisions of the bill as amended should not be construed as taking away any of the jurisdiction of that Office.” 32 Thus GAO authority was left exactly where it was.
*57A point I have already made about deletion of the reference to GAO bears repeating. Amended S. 2487, by incorporating GAO’s original substitute bill, granted GAO precisely the same binding power of review that it granted the courts. Contractors did not object to that provision because it authorized GAO to set aside disputes decisions unfavorable to contractors. They objected because amended S. 2487 authorized GAO to set aside disputes decisions favorable to contractors. That power, opponents of amended S. 2487 urged, must be vested solely in the courts. They prevailed, and the reference to GAO was deleted. Deletion of the authority granted to GAO obviously could have no effect whatever on the identical authority granted to the courts.38
The Senate originally passed amended S. 2487 upon the clear understanding that the expanded scope of judicial review it contained would be available to both the *58Government and contractors. When the House bill came to the Senate after deletion of the GAO provision, Senator McCarran, who had previously stressed that Wunderlich hurt both the Government and contractors, explained that while the House bill differed from the bill passed by the Senate, since it deleted the authority to GAO, it was “designed to accomplish the same purpose.” 34 That purpose, of course, was to overturn Wunderlich and to provide the courts with grounds of review in addition to fraud. The two bills could not, of course, “accomplish the same purpose” if the House bill authorized expanded judicial review only for contractors, leaving the Government either with the Wunderlich standard or with no review at all. After Senator McCarran responded affirmatively to the statement that the difference was only “a modification of the language in the Senate bill, and the two bills agree in their effect,” 35 the Senate passed the House bill.
The text of the Act is its own witness to the congressional purpose. It provides that no clause in a Government contract purporting to make final an administrative determination of a dispute arising under the contract “shall be pleaded in any suit ... as limiting judicial review.” The proviso then defines the applicable scope of review.
It is impossible to read the plain words of this statute as directing that judicial review is available only for disputes decisions unfavorable to contractors. Indeed, the language is so clear that there should be no need to search through the legislative history for a contrary meaning.36 That history, in any event, demonstrates that the Act means exactly what it says.
*59Two significant considerations buttress my conclusion that the Court’s construction of the Act is patently and grievously erroneous.
First. The bill that became the Wunderlich Act was a Government bill. As the Committee report said, the Act, with a minor exception, “is exactly the same legislation suggested by the Comptroller General.” 37 GAO offered it as a substitute for the original S. 2487 because of Government concern that administrative “officials can make just as arbitrary determinations in favor of contractors, at the expense of the taxpayers.” 38 The bill explicitly stated that the expanded scope of review would add to “fraudulent” the grounds that the disputes decision was “arbitrary,” “capricious,” “grossly erroneous,” or “not supported by substantial evidence.” After GAO modified the bill to delete the provision authorizing GAO review, in addition to court review, on those grounds, Government procurement agencies joined forces with GAO in strong support of passage. It is absurd to suppose that the Government pressed for a bill that granted contractors an expanded scope of judicial review, inserted in the bill by the Government, yet denied the Government judicial review on those same grounds.
Second. That absurdity is compounded by the consequences that result from interpreting the Act to deny *60the Government judicial review of disputes decisions. Before Wunderlich, the Government could challenge the finality of those decisions at least on the ground of fraud. If the Act affords only contractors judicial review and denies review to the Government, it follows that the Government has been deprived even of the right it had under Wunderlich to challenge “fraudulent” disputes decisions. The principal Government procurement agencies, now including the Atomic Energy Commission, have. created contract appeals boards as the final level of agency review of disputes decisions. Because the Act expressly provides for judicial review of such “board” decisions, interpreting it to deny the Government review means that however “fraudulent,” however “arbitrary,” however “capricious,” however “grossly erroneous,” however clearly “not supported by substantial evidence” the board’s determination, the procurement agency and the Government itself are helpless to redress the wrong. In this case, that might mean the loss of more than one million dollars to American taxpayers. But at stake are countless millions. To say that Government wrote and secured passage of a bill to work that result is preposterous.39
Ill
So far as I can penetrate the Court’s opinion, its primary premise is exposed by such sentences as these: “The purpose of avoiding Vexatious litigation’ would *61not be served, however, by substituting the action of officials acting in derogation of the contract.” Ante, at 8.40 “Neither the Wunderlich Act nor the disputes clause empowers any other administrative agency to have a veto of AEC’s ‘final’ decision or authority to review it.” Id., at 9. “In other words, we cannot infer that by some legerdemain the disputes clause submitted the dispute to further administrative challenge or approval....” Ibid. “Here, the AEC spoke for the United States and its decision, absent fraud or bad faith, should be honored.” Id., at 10.41 “Since the AEC withheld payment solely because of the views of the Comptroller General and since he had been given no authority to function as another tier of administrative review, there was no valid reason for AEC not to settle with petitioner according to its earlier decision.” Ibid,42 “That action by the *62Comptroller General was a form of additional administrative oversight foreclosed by the disputes clause.” Id,., at 12. “[The Act] should not be construed to require a citizen to perform the Herculean task of beheading the Hydra in order to obtain justice from his Government.” Id., at 14. “We are reluctant to construe a statute enacted to free citizens from a form of administrative tyranny so as to subject them to additional bureaucratic oversight, where there is no evidence of fraud or overreaching.” Id., at 14.43 “This objective [preventing the inflating of bids] would be ill served if Government contractors — having won a favorable decision before the agencies with whom they contracted- — had also to run the gantlet of the General Accounting Office and the Department of Justice.” Id., at 14-15.
The Court’s bite noire, then, is primarily the General Accounting Office, with a sideswipe at the Department of Justice. We are left to infer, I gather, that Congress shared the Court’s distaste for the activities of those agencies in these cases and enacted the Wunderlich Act, not only to arm contractors with expanded grounds of judicial review of disputes decisions favorable to the Government, but also, by the device of denying judicial review to the Government, to abolish the authority of GAO to disapprove payments to contractors under disputes decisions, thus forcing contractors to sue, and, by that device, to relieve the Department of Justice of any suits *63to defend on behalf of the United States. There are three dispositive answers to the Court’s supposition.
First. The notion that Congress enacted the Wunder-lich Act to abolish the authority of GAO and the Department of Justice is completely a figment of the Court’s own imagination. As the judicial history shows, both agencies have exercised for decades powers identical to those exercised in this case, with no prior complaints that I can discover and with complete congressional approval. I need only quote from the Committee report that accompanied the bill that is now the Wunderlich Act.
“The proposed legislation, as amended, will not add to, narrow, restrict, or change in any way the present jurisdiction of the General Accounting Office either in the course of a settlement or upon audit, and the language used is not intended either to change the jurisdiction of the General Accounting Office or to grant any new jurisdiction, but simply to recognize the jurisdiction which the General Accounting Office already has.
“The elimination of the specific mention of the General Accounting Office from the provisions of the bill as amended should not be construed as taking away any of the jurisdiction of that Office. It is intended that the General Accounting Office, as was its 'practice, in reviewing a contract and change orders for the purpose of payment, shall apply the standards of review that are granted to the courts under this bill. At the same time there is no intention of setting up the General Accounting Office as a ‘court of claims.’ Nor should the elimination of the specific mention of the General Accounting Office in the bill be construed as limiting its review to the .fraudulent intent standard prescribed by the Wunderlich decision.
*64 “The specific intent of this legislation, insofar as it affects the General Accounting Office, is explicitly stated in the letter . . . from the Comptroller General himself . . .
The report then quoted from the Comptroller General's letter, in which he said that GAO “has not asked for authority which it did not have before the decision in the Wunderlich case,” and in which he quoted from the Senate Committee’s report on amended S. 2487:
“[I]t is not intended to narrow or restrict or change in any way the present jurisdiction of the General Accounting Office, either in the course of a settlement or upon audit; [it] is not intended either to change the jurisdiction of the General Accounting Office or to grant any new jurisdiction, but simply to recognize the jurisdiction which the General Accounting Office already has.” 44
Second. The case law detailed earlier in this opinion, including Eaton, Brown & Simpson, Inc. v. United States, 62 Ct. Cl. 668 (1926), in which GAO disagreed with a disputes decision in favor of the Government and paid the contractor, establishes without question that GAO has no power to overturn a disputes decision. The limit of its authority is to refuse to sanction payment to the contractor and thus force him to bring suit. The judicial precedents in this Court, the Court of Claims, and the district *65courts are explicit that only a court can determine the merits of the dispute within the grounds of review specified by the Wunderlich Act. It is therefore completely irrelevant that “the AEC withheld payment solely because of the views of the Comptroller General.” Ante, at 10. Indeed, the Court exposes the fallacy of its own position when it states that “the disputes clause in the contract says that the decision of the AEC is 'final and conclusive,’ unless a court determines that the award is vulnerable under §§ 1 and 2 of the Act.” Id., at 3-4 (emphasis added). See also id., at 9: “By the disputes clause the decision of AEC is 'final and conclusive’ unless ‘a court of competent jurisdiction’ decides otherwise for the enumerated reasons.” (Emphasis added.)
Third. Similarly, the Court states, in response to the Government’s nonexistent contention that the Department of Justice has “the power to overturn decisions of coordinate offices of the Executive Department,” id., at 12, “That power [of the Department of Justice to defend suits against the United States] is pervasive but it does not appear how under the Wunderlich Act it gives the Department of Justice the right to appeal from a decision of the Atomic Energy Commission,” id., at 12-13 (emphasis added). See also id., at 13: “The power to appeal to the Court of Claims a decision of the federal agency under a disputes clause in a contract which the agency is authorized to make is not to be found in the Wunderlich Act and its underlying legislative history.” (Emphasis added.) No one suggests that the Department of Justice has a “right to appeal.” It is involved in this case only because GAO’s refusal to sanction payment forced petitioner to sue the United States, thus creating a lawsuit that the Department of Justice, as the Government’s lawyer, had a duty to defend. It would be strange if the Department had a duty to confess judgment.
In support of its construction of the Act, the Court *66makes a statement, which I have already quoted, that invites a further comment:
“[JJudicial review was provided so that contractors would not inflate their bids to take into account the uncertainties of administrative action. This objective would be ill served if Government contractors— having won a favorable decision before the agencies with whom they contracted' — had also to run the gantlet of the General Accounting Office and the Department of Justice.” Id., at 14-15.
Contractor witnesses at the committee hearings asserted that contractors would have to inflate their bids if they could attack a disputes decision only on the ground that it was fraudulent. As the Court says, the Act resolved this problem by expanding the scope of judicial review, so that contractors can attack a disputes decision on grounds in addition to fraud. That was the protection Congress gave contractors so that they would not have to inflate their bids.
After recognizing this, the Court says that because contractors got expanded judicial review to prevent the necessity of inflating bids, they also got the benefit of not having decisions in their favor subject to judicial review at all, since otherwise the objective of preventing inflated bids “would be ill-served.” It would be difficult to imagine a more obvious non sequitur. The Court could as easily say that “[t]his objective would be ill served” if the contractors ever lost a disputes decision.
I might add that the Court does not say that the “objective would be ill served” if favorable contractor decisions were subject to judicial review; it says that the “objective would be ill served” if contractors “had also to run the gantlet of the General Accounting Office and the Department of Justice.” Yet what the Court means, of course, is judicial review, for neither GAO nor *67the Department of Justice can take a favorable decision away from a contractor. Only a court can do that.
The Court is forced to go to extreme lengths to assert that the Government still may have relief for fraud. That is because the Court concedes, as it must, that its construction of the Act denying the Government judicial review forecloses review of disputes decisions that are “fraudulent,” just as it forecloses judicial review of decisions that are “arbitrary,” “capricious,” “grossly erroneous,” or “not supported by substantial evidence.” The Court’s attempted escape is to suggest that the Government may have relief for fraud under the statutes in which “Congress has made elaborate provisions for dealing with fraudulent claims of contractors.” Id., at 16. Apart from the absence of any explanation why, if statutory remedies were always available, this Court found it necessary to fashion, for Government and contractor alike, a judicial exception to the finality of disputes decisions, the point is frivolous.45 Obviously the fraud statutes the Court mentions have no application whatever to the fraud we are discussing in this case.
The “fraud” that is an issue in a disputes clause case is not contractor fraud. Not one case construing a disputes clause, from 1878 to the present day, ever mentions “fraud” by the contractor. Nor has anyone ever sug*68gested that the Government needs judicial review of disputes decisions to guard against fraud by the contractor. The “fraud” that is involved is a fraudulent decision. The disputes clause and the Act itself provide judicial review to determine whether the “decision . . . is fraudulent.” (Emphasis added.) When a disputes decision is challenged, the only questions concern that decision: was it “fraudulent”? was it “capricious”? was it “arbitrary”? was it “grossly erroneous”? was it “not supported by substantial evidence”?46 The Court is absolutely right that “[a] contractor’s fraud is of course a wholly different genus than the case now before us.” Id., at 15.
IV
The time-tested standards of statutory construction require interpretation of the statutory wording to effect the congressional purpose as revealed by legislative history. The Court totally discards those standards in construing the Wunderlich Act. Instead, the Court purports to discover a nonexistent hostility of Congress toward the “intermeddling,” id., at 19, of GAO and the Department of Justice in the disputes process and for that reason a congressional purpose to prevent the subjection of “citizens ... to additional bureaucratic oversight,” id., at 14. The virtually century-long judicial history that forms the background of the Act, its explicit language, and its clear legislative history completely refute the proposition. I dissent and would affirm the judgment of the Court of Claims.
The concurring opinion seems to read the judicial-review provision out of the disputes clause: “And if the contractor accepts a decision of the contracting officer, and does not appeal to the Commission, that decision, by the specific provisions of the dis*27putes clause, is final and conclusive as to questions of fact. Under the Government’s position, however, the decision at the agency head would enjoy no such preferred and conclusive status.” Ante, a,t 21 (emphasis added). The Commission’s disputes decision does not have “conclusive status” under the disputes clause, of course, because of a “specific provision” of the clause. That provision directs that the Commission’s decision is “final and conclusive unless” (emphasis added) a court determines that it was “fraudulent,” etc. It does not direct that the Commission’s decision is final and conclusive unless the contractor appeals to the courts. That is the language of the earlier provision, referred to by the concurring opinion, under which the contracting officer’s decision is final and conclusive unless the contractor appeals to the Commission. If “the specific provisions of the disputes clause” apply after the contracting officer’s decision, surely they also apply after the Commission’s decision.
The Court’s opening sentence appears to say that we are dealing with a pre-Wunderlich Act disputes clause that “provides that the decision of AEC shall be ‘final and conclusive.’ ” Ante, at 2-3. The Court later recognizes the obvious: “By the Disputes Clause the decision of AEC is ‘final and conclusive’ unless ‘a court of competent jurisdiction’ decides otherwise for the enumerated reasons.” Id., at 9.
It was suggested at oral argument that the procurement agency might pay the contractor in accordance with a disputes decision in his favor and that subsequently, prompted by GAO’s post-audit, the Department of Justice might sue the contractor to recoup the payment on the ground that the agency’s decision was improper under the disputes clause and the Wunderlich Act. The Court’s holding today, of course, prohibits the Government from obtaining judicial review of disputes decisions by that method. Indeed, that would be an a fortiori case, for the agency not only would have decided in favor of the contractor, but also would have paid him in accord-*29anee with its decision. If a disputes decision is final when the agency refuses to implement it by payment, certainly it is final when the agency pays.
The Court’s constant repetition of the phrase “fraud or bad faith” might suggest to the casual reader that the Court is holding that the Government may challenge the finality of disputes decisions on those grounds. That, however, is not true, for fraud and bad faith are two of the grounds specified in the disputes clause and the Wunderlich Act: a disputes decision may be set aside if it is “fraudulent” or if it is “so grossly erroneous as necessarily to imply bad faith.” In contrast to the disputes clause and the Act, the Court is not referring to disputes decisions resulting from the fraud or bad faith of the disputes decisionmaker. Rather the Court is referring to fraud or bad faith on the part of the contractor, as the Court’s statement of facts makes clear: “The defenses tendered raised no issue of any fraud or bad faith of the contractor against the United States.” Ante, at 7. “The Commissioner did not base his opinion on any issue of fraud or bad faith of the contractor against the United States, nor did the Court of Claims.” Ibid. See also id., at 9-10, n. 8 and Part IV of the Court’s opinion. The concurring opinion also refers to “fraud” and “bad faith.” Ante, at 19. Again, however, the reference is not to fraud and bad faith as used in the disputes clause and the Act.
Goltra v. Weeks, 271 U. S. 536 (1926), which involved a contractor’s challenge to the finality of a disputes decision, by a Government official, also demonstrates that the rule was the same no matter which party challenged the decision. The Court there held that the official’s decision was binding “unless there is an absence of good faith in the exercise of the judgment.” Id., at 548. Significantly, the Court cited as authority, not only Kihlberg, Sweeney, March, and Gleason, all cases in which the contractor challenged and the Government (in March, the party whose official decided the dispute) relied upon the disputes decision, but also Mason & Hanger, in which the Government challenged the finality of a disputes decision upon which the contractor relied.
The Court cites McShain Co. for the proposition that “[t]he cases deny review” by GAO “absent fraud or overreaching.” Ante, at 10. Since McShain Co. is simply another example of the application of the Kihlberg rule against the Government, I am at a loss to understand the Court’s statement. As the excerpt I have quoted in the text demonstrates, McShain Co. did not “deny review” by GAO; rather, like the other cases, it held that GAO’s view of the merits of the disputes decision was irrelevant in court and that the Government could upset the finality of that decision only by proving in court that it was fraudulent.
The Court states, ante, at 11, that the District Court in James Graham, by referring to “fraud or overreaching,” referred to instances “where the Comptroller General’s power was founded upon specific statutory provisions such as 41 U. S. C. § 53,” a statute relating to “kickbacks by Government contractors,” id., at 9 n. 8. In *44fact, however, the District Court not only did not refer to that statute, it did not refer to any statute, nor even intimate that a statute might be relevant. What the District Court did was use the phrase “fraud or overreaching” as shorthand for the Kihlberg rule, the judicially created fraud exception to the finality of disputes decisions. That usage is readily apparent from a glance at the District Court’s citations: Mason & Hanger and Moorman from this Court, and Penn Bridge, Carroll, and McShain Co. from the Court of Claims.
The Court also says, id., at 11, that in James Graham “summary judgment was entered by the court, which said, ‘Since the Navy Department has determined that plaintiff contractor is entitled to the payment sought, this Court must adjudge accordingly.’ ” The Court omits to quote the immediately preceding sentence in the James Graham opinion: “And the Navy Department’s decision that these particular dues and contributions are reimbursable is not arbitrary or unconscionable.” 91 F. Supp., at 717 (emphasis added). Thus, again, the District Court was referring to the disputes decision, and not, as the Court today would have it, to “fraud or overreaching” by the contractor.
The concurring opinion asserts that “[t]he contractor here, according to the long-term understanding of the disputes clause, consented to the disposition of disputes by the contracting officer and by the AEC on appeal, and to the finality of decision at those points.” Ante, at 21. If the concurring opinion is speaking of pre-Wunderlich Act disputes clauses, the authorities I have cited establish the utter inaccuracy of the assertion. Indeed, the concurring opinion also asserts that “for years, with the specified exceptions, [the disputes] clause itself has been regarded as conferring no right of judicial review on the part of the Government.” Id., at 20 (emphasis added). The italicized words can only refer to the judicially created exception for fraudulent decisions. The concurring opinion gives no indication that, in either of the assertions, it is referring to the current disputes clause.
Hearings on S. 2487 before a Subcommittee of the Senate Committee on the Judiciary, 82d Cong., 2d Sess., 6.
Id., at 8.
It is misleading to assert, as does the Court, that Wunderlich “closed the courthouse doors to certain citizens." Ante, at 14 (emphasis added). Similarly, the concurring opinion asserts that Wun-derlich “restricted cowiracior-instigated judicial review” and that the Government “prevailed” in Wunderlich with “a narrow judicial review standard for the contractor." Ante, at 22 (emphasis added). The concurring opinion’s assertions are the more surprising in view of its apparent recognition that the Government was subject to the same standard of judicial review as contractors. See n. 9, supra.
Hearings on S. 2487, supra, n. 10, at 29-32, 68, 83-84, 107, 114.
Id., at 59, 107. Moreover, H. R. 6301, also introduced in the 82d Congress, provided for judicial review only in those instances “in which the contractor shall seek to set aside a decision on a disputed question between the United States and such contractor, made by an officer, board, or other representative of the United States . . . .” Neither House supported this bill.
Hearings on S. 2487, supra, n. 10, at 119.
See S. Rep. No. 1670, 82d Cong., 2d Sess.
Id., at 2.
Ibid.
Id., at 3.
Amended S. 2487 was reintroduced as S. 24, but for ease of reference I will continue to refer to it as amended S. 2487.
99 Cong. Reo. 4573.
99 Cong. Rec. 6170.
Amended S. 2487 was introduced as H. R. 1839, but for ease of reference I will continue to refer to it as amended S. 2487.
Hearings on H. R. 1839 et al. before Subcommittee No. 1 of the House Committee on the Judiciary, 83d Cong., 1st and 2d Sess., ser. 12, at 3-20.
Id., at 26.
Id., at 136.
Id., at 39.
Id., at 59.
Id., at 110.
Id., at 89.
H. R. Rep. No. 1380, 83d Cong., 2d Sess., 6.
Id., at 7.
The Court’s only foray into the legislative history is its assertion that “Congress contemplated giving the General Accounting Office such powers and, indeed, the Senate twice passed — in the form of the McCarran bill — a provision which would have allowed the Comptroller to review disputes decisions to determine if they” satisfied the standards of the Act. Ante, at 11. The Court therefore concludes that the Act cannot be construed “to give the Comptroller General powers which Congress has plainly denied.” Id., at 12. Similarly, the concurring opinion asserts that “[t]he flat rejection by Congress of the proposed provision for GAO review is significant. There would be no point in that rejection if GAO has the power to defeat the finality of the disputes decision anyway.” Ante, at 22-23. Unfortunately, the Court and the concurring opinion overlook that the proposed provision was not simply “for GAO review.” It was for binding GAO review. Because it was not enacted, GAO does not “have a veto of AEC’s ‘final’ decision,” ante, at 9 (opinion of the Court); GAO does not have “power to defeat the finality of the disputes decision,” ante, at 23 (concurring opinion). Both the Act and the disputes clause specifically provide that only a court can set aside a disputes decision. And that is precisely the point the legislative history makes clear.
100 Cong. Rec. 5717.
100 Cong. Rec. 5718.
The need arises in this ease only because petitioner argues that, despite the clear language of the Act, the legislative history reveals that Congress meant to reserve the right of judicial review solely *59to contractors. It is thus somewhat odd that the Court considers it worthwhile to assert “that the Act’s legislative history ‘has something for everyone’” and that the Court “find[s] the Act’s history at best ambiguous.” Ante, at 13 n. 9. The concurring opinion likewise professes to find the legislative history “decidedly ambiguous at best,” ante, at- 22, yet nevertheless goes on to assert that Congress “intended to relieve contractors” and “opened the door to the contractor,” ibid, (emphasis added). These comments are all the more inexplicable because neither the Court nor the concurring opinion attempts even the most cursory analysis of the text of the Act itself.
H. R. Rep. No. 1380, supra, n. 31, at 6.
See n. 10, supra.
The concurring opinion asserts that “[i]n the exercise of its legislative judgment, Congress has determined that in this area the Government,” unlike contractors, does not need the Act’s protection “against fraud, capriciousness, arbitrariness, bad faith, and absence of evidence.” Ante, at 23. As the concurring opinion never refers to the language of the Act, and finds the legislative history “not at all that clear,” “decidedly ambiguous at best,” id., at 22, and “inconclusive,” id., at 23, it is difficult to understand the basis for this statement.
This statement, albeit obscurely, may mean that the purpose of avoiding litigation would not be served by subjecting a disputes decision in favor of the contractor to judicial review, for that would be litigation. Yet just as obviously the purpose of avoiding litigation would not be served by subjecting a disputes decision against the contractor to judicial review.
See n. 4, supra, n. 43, infra.
This is a difficult statement to understand. Assume that the Commission had “no valid reason” not to pay petitioner. Was the Commission’s nonpayment in violation of the contract? Was it in violation of the Wunderlich Act? The Court does not say. If nonpayment violated neither the contract nor the Act, it seems rather strange that this Court should order the Commission to pay. The Court’s statement appears to be connected with its later statement that “[t]he AEC has not, to this day, repudiated the merits of its decisions in favor of petitioner.” Ante, at 19. Again, however, the Court does not say how or even whether the Commission’s “non-repudiation” violated the contract or the Act.
In the same vein, the concurring opinion asserts that there is “a possible breach of contract” in this case: “When the United States then disavows the Commission’s decision — a decision which, as the Court notes, to this day has never been withdrawn or repudiated by the AEC — it seems to me that the Government imposes *62something to which the contractor has not agreed.” Ante, at 21, 22. The concurring opinion, however, does not say how the Government’s “disavowal” violated the contract.
If this statement implies that a contractor is "subject ... to additional bureaucratic oversight, where there is . . . evidence of fraud or overreaching” (emphasis added), one might well ask why that is so. Fraud is only one of the five grounds of judicial review specified in the Act and the disputes clause. Obviously either all or none are available. See n. 4, supra.
H. R. Rep. No. 1380, swpra, n. 31, at 6-7 (emphasis added). This detailed refutation that GAO authority was being curtailed was necessary to allay the fears expressed by the attorney who argued Wunderlich for the contractor. He testified during the House hearings that deletion of GAO from amended S. 2487, passed by the Senate, might be misconstrued as depriving GAO of its prior authority to refuse to sanction payment and thereby “throw the matter into court.” See Appendix, infra, at 78-80. Today’s decision fulfills his prophecy.
The Court asserts that “[i]f the Comptroller General has the broad, roving, investigatory powers that are asserted, specific statutory grants of authority such as this provision [41 U. S. C. § 53] relating to kickbacks would be superfluous.” Ante, at 10 n. 8. The GAO authority asserted here, however, is simply the authority to refuse to sanction payment under a disputes decision on the ground that the decision does not satisfy the standards of the Wunderlich Act. The Act, of course, has nothing whatever to do with illegal activities of contractors. It concerns only the finality of administrative disputes decisions. Enforcement of the Act obviously would not make the statutory prohibition of kickbacks “superfluous.”
Even my Brother Douglas once recognized this:
“We should allow the Court of Claims, the agency close to these disputes, to reverse an official whose conduct is plainly out of bounds whether he is fraudulent,, perverse, captious, incompetent, or just palpably wrong.” United States v. Wunderlich, supra, at 102 (dissenting opinion) (emphasis added).