concurring in part and dissenting in part:
The core issue of this case is the right of the government to have access to information pertaining to government contractors. Because it is the government, and because of its size, the role of the federal government as the purchaser of goods and services is hard to square with the free market concept of a willing purchaser doing business with a willing seller. Aware of these difficulties, Congress sought to make the government’s role easier by creating the right of access to records of government *1238contractors.1 I recognize that the access-to-records provision has received thorough consideration in the cases cited by the majority, and that conflicts in those decisions must be resolved by the Supreme Court. But I also believe that we must fulfill our obligation to provide guidance on this issue until the Supreme Court has spoken, should it decide to do so.2 I too shall try to avoid “replowing the field,” but I cannot agree that the district court paid sufficient heed to the congressional purpose in requiring these access provisions in government contracts negotiated without advertising. The legislative history establishes that Congress intended to give the Comptroller General exactly the sort of authority he seeks to exert here. To the extent the district court found to the contrary, it should be reversed.
This litigation3 poses two questions of statutory interpretation. First, for what purposes may the government exercise its rights under the access-to-records clause? Second, what is the scope of the access provision when it is properly invoked? The district court’s opinion, which we now affirm almost casually, holds that the government properly invoked the access clause with regard to the four contracts it entered with Merck & Co., Inc. (Merck) in 1974. Memorandum and Order, Aug. 12, 1977, Joint Appendix (J.A.) at 511. That opinion limited governmental access considerably, however, and held that Merck’s data concerning such costs as research, development, marketing, promotion, distribution, and general administration do not come within the scope of the access provision. Id., J.A. at 513. I agree with the first holding, which is supported by precedent and legislative history. I disagree with the second holding, however, which disendows the usefulness of the information that the court allows the government to gain. Such an incongruous result does not comport with the manifest congressional purpose in enacting the access-to-records statutes.
I. PROPER APPLICATIONS OF THE ACCESS CLAUSE
At issue here are four fixed-price contracts for the sale of drugs negotiated in 1973 between Merck and the United States Government. Each of the contracts incorporated by reference a standard form access clause:
The Contractor agrees that the Comptroller General of the United States or any of his duly authorized representatives shall, until the expiration of three years after final payment under this contract, have access to and the' right to examine any directly pertinent books, documents, papers and records of the Contractor involving transactions related to this contract.
J.A. 24. Under the access-to-records provision, such a clause is required by law to be *1239added to all government contracts for more than $10,000 which are negotiated without advertising.4 It is important to realize at the outset, then, that we interpret a clause in contracts entered freely by Merck, and not the validity of some administrative agency’s demand for involuntary access to private business records.
The legislative history of the access-to-records statutes provides compelling evidence that Congress expected the General Accounting Office (GAO) to use the statutes to improve government procurement methods, and not merely to combat and deter fraud. Every circuit to examine this question has come to this conclusion, as did the lower court here, and Merck’s argument to the contrary reopens a question that ought to be considered closed. Nevertheless, because the basic premise of the statutes sheds light on my disagreement with the court below, it may be useful to review Merck’s contentions and explain why they lack merit.
Merck opposes governmental access to its records under the statutes on three main grounds. The first can be disposed of quickly. Merck claims that the GAO demand for Merck’s records was prompted by the prodding of individual Senators who sought this “headline-grabbing” information for their own ends.5 Speculation about Senatorial prodding or GAO’s plans to leak confidential information to the media,6 however, is irrelevant if the Comptroller General otherwise has a legitimate purpose for making the demand for access at issue here.7 Our inquiry must focus on whether the Comptroller General’s request was within the scope of his authority under the statutes. It is to that question we now turn.
A. The Basis for GAO’s Request
Merck argues that GAO has no legitimate claim for access to Merck’s records because GAO has not alleged any possibility of fraud, corruption, or the need to audit *1240Merck’s books. The Comptroller General is not authorized to demand access to contractor records simply to gather data for an industry-wide economic study, Merck claims, and this is said to be the aim of the GAO’s demand.8 In Merck’s analysis, the access statutes were only meant to apply in situations where allegations of fraud or illegality suggested that it would be desirable to conduct an audit of a specific contract. But the access demands may not be made merely to let GAO “conduct broad-ranging and ill-defined studies of the general adequacy of government procurement procedures.” Merck Brief at 53.
This contention is not new.9 A variety of sources are said to show that the access provision was intended “solely to provide a method to discover fraudulent activities,”10 or to serve “as a deterrent to improper conduct by government procurement officers,”11 or to apply only to situations in which there was “reason to suspect fraud or bad faith or illegality.”12 Merck emphasizes repeatedly that Congress had a “single and limited purpose,” which was “to protect the government from fraud and other abuses.” Merck Reply Brief at 26; see id. at 4, 14, 28. “[0]nly the suspicion of fraud or other improprieties” could trigger the access clause, id. at 27, because “the debates leave no room for doubt as to the limited anti-fraud purpose of the access statute.” Id. at 29. These would be persuasive claims if they were supported by the legislative history. They grossly overstate that history, however, and it is time to dispatch the contention once and for all.
It is true that Representative Hardy, sponsor of the access-to-records legislation,13 explained that his “main purpose” in introducing the bill was to deter “improprieties in the negotiation of Government contracts.” 97 Cong.Rec. 13198 (1951). Representative Hardy, however, intended the clause to do far more than combat or deter fraud. He spoke of “excessive” contract prices, “wastefulness in the negotiation of contracts,” and the need for “every reasonable safeguard against waste and extravagance.” Id. These comments convey concerns that go far beyond the “limited anti-fraud purpose” asserted by Merck. They show that Congress was particularly concerned about the general adequacy of *1241government procurement techniques and not simply illegalities or frauds. Early in his explanation of the benefits of the access legislation, for example, Representative Hardy spelled out precisely the goal of improving procurement methods that Merck rejects:
Under conditions as they now exist, competitive bidding has little or no effect upon contracts which are negotiated without advertising. As a result, when a contract is being negotiated, here is a typical illustration of what usually happens: A contractor with years of experience comes to the conference table accompanied by a highly competent accountant and equally competent lawyer. The Government representative on the other side of the table will, in a great majority of cases, be at a tremendous disadvantage, from the standpoint of both training and experience, no matter how conscientious and honest he may be. So, aside from any intentional liberality on the part of the Government contracting officer, there is every chance in the world that the Government will come out on the short end of the deal. This bill would at least enable the agent of the Congress to check the transaction, both from the Government records and the contractors’ books.
Id. (emphasis added). Representative Hardy obviously sought to improve the governmental procurement system as a whole by equalizing the relationship between more experienced contractors and government representatives, who were otherwise “at a tremendous disadvantage,” no matter how conscientious and honest they might be.14 The access-to-records clause, by shedding light where none had been cast before, would reduce the number of times that the United States “will come out on the short end of the deal.”
Other statements also demonstrate the true tenor of congressional concerns. Representative Hardy continued his explanation of the access bill by suggesting two hypothetical situations to which the clause would apply. In the first, a contractor charged the government for a much greater amount of “overhead” than it asked under a private contract for the same work. “[I]t would then obviously be desirable for the General Accounting Office to look behind the rate which had been established.” Id. Should GAO find that “the rates were excessive,” Representative Hardy explained, it would then be able to bring the facts to the attention of contracting authorities. Id. In the second “vivid example of how this authority would enable the GAO to do an effective job,” Representative Hardy referred to an earlier congressional investigation into government procurement of automobile parts:
We found one situation where the Government was buying parts from an automobile dealer who, in turn, was getting them from a parts distributor who, in turn, was getting them from a small tool shop. Naturally, the price paid by the Government included profits upon profits and completely wasteful administrative and handling costs. It would be difficult, if not impossible, for GAO to detect such a situation without the right afforded in this bill.
Id. (emphasis added). There is no implication in the statements about either hypothetical that the private contractors had acted fraudulently or illegally. Representative Hardy’s access legislation was designed to uncover inefficient procurement methods as well as misfeasance, and Congress expected the access clause to reveal much more than “fraud.” Moreover, the second example effectively refutes Merck’s suggestion that the government must allege or *1242suspect procurement abuses before the access clause can be invoked. Merck Brief at 14, 33. Representative Hardy’s example is one in which it would be impossible for GAO to detect waste without the authority to examine contractor records at will. Had Congress intended GAO to use the access clause only when it had suspicions of wrongdoing, Representative Hardy would surely have spoken of the difficulty for GAO “to investigate” that situation rather than the ability of GAO “to detect” it.
Several exchanges in the debates which followed Representative Hardy’s introduction of the access-to-records legislation also indicate that Merck’s reading of that legislation is woefully narrow. When Representative Mills suggested how useful the bill might be in investigating potentially fraudulent situations, Representative Hardy responded emphatically:
The gentleman is right, with this exception that I think should be clearly understood, that there are a lot of other situations besides those involving fraud which might be uncovered.
97 Cong.Rec. 13199 (1951) (emphasis added). Representative Hardy also had the following exchange with Representative Hoffman:
Mr. HOFFMAN. The whole thing arises out of the belief ... that the Federal employees who negotiate these contracts are no match for the people on the other end of the deal, those who negotiate for private enterprise. It that not right? That is what gives rise to this?
Mr. HARDY. What is the gentleman’s question?
Mr. HOFFMAN. This bill is based upon the thought that the Federal employees who negotiate these contracts are no match for the people who represented private corporations or individuals?
Mr. HARDY. I think that is one factor.
Id. at 13377.
Congress thus anticipated that the access-to-records legislation would lead to general improvement in the government’s procurement system simply because it would equalize the relationship between agents of the United States and private contractors. Not surprisingly, this is also a long-standing interpretation by GAO:
The right of access by GAO to records of costs of production is not limited to that necessary to determine whether the contracts were performed in accordance with their terms and whether the contracting officer had been defrauded or misled in entering into the contracts.
115 Cong.Rec. 25801 (1969) (GAO memorandum reprinted at request of Senator Prox-mire).15 In the instant case, the Comptroller General has explained that the request for Merck’s records is necessary “to determine the adequacy of the protection afforded the Government by the negotiating techniques used by the procuring agencies.” Memorandum and Order, Aug. 12, 1977, J.A. 509. This purpose is within the scope of his lawful authority, for it puts the access legislation to exactly the sort of use for which Congress intended it.
B. The Focus of GAO’s Request
Merck’s third challenge to the applicability of the access clause is based on the nature of the contracts at issue. Because these contracts are not “cost-based,” and were negotiated without any reference to Merck’s costs, Merck argues that its cost records cannot be relevant to the prices paid by the government or “directly pertinent” to the negotiation or performance of the contracts. Merck Brief at 63. The *1243United States did not ask for any cost data or estimates from Merck when it negotiated these fixed-price contracts, and Merck did not volunteer or submit any such data or estimates during the negotiations. As a result, the company says, its cost data simply do not involve “transactions related to such contracts.” Id.
This contention is deceptively appealing, but the legislative history provides no support for limiting the access-to-records clause to contracts that are “cost-based” or negotiated with reference to costs. It is true that the access-to-records legislation originated during the Korean War, when inflation and changes in demand for raw materials threatened to force many government contractors with fixed-price contracts into default. These pressures convinced Congress to give the president emergency authority to modify existing contracts in January, 1951,16 and the access-to-records clause was a concomitant measure to ensure that contractors not take advantage of modifications under the emergency authority.17 But the fact that a specific concern provides the initial impetus for the passage of reform legislation will not necessarily defeat the broader application of that legislation simply because the initial stimulus disappears. Although Congress clearly regarded contracts negotiated without advertising as extraordinary measures, often necessitated by wartime pressures,18 Congress explicitly wanted the clause to apply to all contracts negotiated without advertising. It is a far leap from this history to the conclusion that application of the access clause depends on whether representations about costs are made during the negotiation, or to the claim that the clause may apply to cost-plus contracts but not those negotiated in other ways. See Merck Brief at 64, 72 n.28.
Moreover, the legislative saga of the access-to-records clause did not end with the emergency grant of wartime renegotiation authority. The January statute was limited by its terms to the duration of the national emergency declared on December 16, 1950, and was to expire in any event after eighteen months. 64 Stat. at 1258. But the present and more permanent version of the access-to-records clause was proposed by Representative Hardy in October, 1951, after he learned that procurement officers were negotiating contract modifications under the permanent procurement statutes rather than the January amendments to the First War Powers Act, in order “to plug this loophole.” 97 Cong.Rec. 13198 (1951). Merck argues that the temporary duration of the January legislation shows that the October legislation could not have been intended to allow “generalized studies of the operation of the procurement system.” Merck Brief at 26. But this argument fails to take account of the fact that the October legislation had an indefinite duration and was to apply to two permanent procurement statutes.
Merck’s confusion about the relevance of costs to the applicability of the access clause is understandable because it can be traced to Merck’s argument that the access legislation was designed merely to achieve “a limited anti-fraud purpose.” Were that correct, it might follow that legislation to deter fraudulent reporting of costs should not apply where no representation about costs *1244are made at all. But the government’s purpose behind this access legislation was much broader; Congress was concerned about general improvements in the procurement system. There is compelling reason to think that Congress saw as much reason to have access to cost records for contracts negotiated without regard to costs as for contracts based expressly on cost, perhaps even more.
This interpretation of the access legislation is hardly new.19 Every court to hear the question has held that the aceess-to-records clause gives the government the right to examine cost data even when the contracts are negotiated without party representations concerning costs. Hewlett-Packard Co. v. United States, 385 F.2d 1013 (9th Cir. 1967), cert. denied, 390 U.S. 988, 88 S.Ct. 1184, 19 L.Ed.2d 1292 (1968), involved GAO efforts to examine cost data for equipment sold under several fixed-price contracts. Although production costs were not taken into consideration in negotiating the contracts in question, the court said, it was clear that the subject matter of the contracts was the procurement of described property by the government. Id. at 1016. In Eli Lilly & Co. v. Staats, 574 F.2d 904 (7th Cir.), cert. denied, 439 U.S. 959, 99 S.Ct. 362, 58 L.Ed.2d 351 (1978), the court rejected the drug company’s argument to the contrary:
Production costs directly pertain to that subject matter, because if out of line with the contract price the contract may have been an inappropriate means of meeting this particular procurement need of the Government. While this appraisal could not affect these particular contracts, it could lead to the use of other methods of meeting future procurement needs.
Such an interpretation might be reasonable if the statutory standard were that the information must be “directly pertinent” to the negotiation, but the standard is “directly pertinent” to the contract and plaintiff offers no reason to suppose that there would be a reason to negotiate about each item that might have a significant effect on the seller’s cost or performance.
Id. at 914.20
The foregoing discussion also demonstrates the unpersuasiveness of Merck’s arguments by analogy to other procurement acts. Merck asks that we interpret the access clause in light of the “broader structure of government procurement fashioned by Congress.” Merck Brief at 64. This structure includes the Renegotiation Act of 1951, Pub.L.No. 82-9, 65 Stat. 7, as amended in 1954, Pub.L.No. 83-764, 68 Stat. 1116, and the 1962 Truth-in-Negotiations Act, Pub.L.No. 87-653, 76 Stat. 528. Under the Renegotiation Act, contractors are generally required to permit audits of their books and records so that the government can detect and recover excess profits. The 1954 amendments exempted from this process those contracts involving “standard commercial items,” 50 U.S.C.App. §§ 1216(e)(1)(A) and (4)(B). The government concedes that the Merck contracts in *1245this case involve standard commercial items and thus are exempt from the reporting requirements of the Renegotiation Act, and Merck argues that the access clause should be read no more broadly. Merck Brief at 67. Similarly, under the Truth-in-Negotiation Act, contractors need not submit actual or estimated cost and pricing data during the negotiation process when the contract price is based on the established catalog price for a standard commercial item sold in substantial quantities to the general public. 10 U.S.C. § 2306(f).21 Merck argues that the products sold under the contracts at issue here are standard commercial items sold at or below the established catalog price. It urges that the cost data that Congress has held to be irrelevant under the Truth-in-Negotiation Act should be found equally irrelevant under the access-to-records clause.
These arguments prove too much. If the subsequent procurement legislation is of the broad sweep Merck supposes, it renders the 1951 access-to-records legislation redundant. Yet it is a canon of statutory construction that we are not to assume the redundancy of statutes, if they can be construed to be independently meaningful.22 Furthermore, the later provisions illustrate that Congress could easily have drafted a more narrow access provision in 1951, or any later time, had it chosen to do so. See Hewlett-Packard v. United States, 385 F.2d at 1016 (“if such a limited scope of examination had been intended by Congress it would have found the means to so indicate”); Eli Lilly & Co. v. Staats, 574 F.2d at 916 (“Since Congress in the Renegotiation Act demonstrated its ability to draft a specific exemption provision, its failure to do so in the access-to-records legislation indicates that no such exemption was intended.”).
Most importantly, however, Merck’s arguments from these unrelated procurement statutes are unconvincing because their premise is erroneous. Essentially, Merck argues that it is pointless for the United States to inquire about Merck’s costs because Merck is “market-oriented” and sells its products to the government and the public at the same “market price,” thus proving that the instant contracts are ones “in which the reasonableness of prices has been assured by competitive market forces.” Merck Reply Brief at 41. Merck’s assumptions about the market may indeed be true, but then again they may not. The situation can easily be imagined in which a number of companies do business with the government at an established catalog price, and yet in which the government is badly overcharged because an absence of truly competitive market forces permits oligopo-listic pricing.23 The fact that products are “standard commercial items,” or even that they are sold in substantial quantities to the public, does not necessarily determine whether the market for them is truly competitive. No single purchaser of drugs is anywhere near as large or as non-competitively situated as is the United States government. Moreover, even if the market is competitive, why should not the United States (which is probably the largest single purchaser of some of these drug items) receive a measure for quantity discounting?
*1246Recognition of this fact is singularly important. It is too little appreciated that government, in a number of ways, may often require far greater protection in the marketplace than private individuals. The manufacturer who sells to a retailer must hold down his prices or risk the possibility that the retailer will be forced out of business, thereby killing the proverbial goose with the golden eggs. Unfortunately, when government is the buyer, no such scruples need depress the prices that contractors can charge. Seldom, and almost never for financial reasons, do governments ‘go out of business.’ Opportunities for fraud and corruption in government contracting have long been obvious; the inherent non-corrupt, non-fraudulent disadvantages that government often has in dealing with private contractors are only beginning to be understood. The very size of government purchases, and the necessity thereof, may make the traditional marketplace shields ineffective. In this era of cost-overrun scandals, it is sometimes forgotten that Congress devised the cost-plus contract as a major reform of its procurement methods, which once depended on negotiations without regard to costs that left the government “at a tremendous disadvantage” vis-a-vis the private contractor. The Hardy bill was designed to reduce this inequality by providing government with data that could enhance its procurement practices, thereby fighting wastefulness as well as fraud. There is no reason to think that Congress decided the reforms should not apply simply because private contractors were so much more sophisticated than the government that they could manage to avoid discussing their costs entirely. Congress expected to protect the government in exactly the sort of situation now presented, and the district court correctly.found that the access provision applies to the Merck contracts at issue here.
II. PROPER SCOPE OF THE ACCESS CLAUSE
The second question of statutory interpretation raised by this dispute, and the one on which Merck prevailed below, is the scope of the access-to-records clause when that clause is properly invoked. The trial court held that the Comptroller General had a right of access to data on Merck’s
manufacturing costs (including raw and packaging materials, labor and fringe benefits, quality control and supervision), manufacturing overhead (including plant administration, production planning, warehousing, utilities and security), royalty expenses, and delivery costs ....
Judgment and Order of January 24, 1979, J.A. 518. The court limited the government’s access, however, by ordering
that defendant Comptroller General shall not demand access to books, documents, papers, or records with respect to research and development, marketing and promotion, distribution, and administration (except to the extent such data may be included in the cost items listed above)
Id. For the reasons stated in Part I supra, and in recognition of the special economic structure of the pharmaceutical industry, I think we have an obligation to review this holding and to reverse it.
The trial court’s ruling on this point can be said to differentiate between Merck’s “direct costs” and “indirect costs.” The trial court granted access to data on “direct costs,” such as manufacturing and labor expenses, but denied access to data on “indirect costs” such as research, development, promotion, and distribution.24 Initially, this holding appears entirely reasonable; it is an attempt, in the government’s words, to “split the baby.” Government Brief at 10, 28. On closer analysis, however, it seems inescapable that this holding fails to give *1247adequate consideration to the congressional purpose in enacting the aecess-to-records clause. If the access legislation was in fact meant to equalize the relationship between the government and private contractors, see Part I supra, then the trial court’s narrow reading of the clause in this case utterly fails to advance that purpose because of the special characteristics of the pharmaceutical industry.
The Comptroller General asserts that the “direct costs” of manufacture in the pharmaceutical industry, such as labor and raw materials, may constitute an extraordinarily small part of the product price — less than nine percent. Government Brief at 25; see Eli Lilly & Co. v. Staats, 574 F.2d at 913-14. The remainder of drug revenues may go to “indirect costs,” such as research, development, promotion, and distribution, as well as taxes and profits. In short, the “indirect costs” of the pharmaceutical industry may be ten times the size of its “direct costs,” 25 and denial of the government’s access to the preponderance of this cost data effectively defeats the purpose of the access provision. Improvement of procurement techniques is impossible if the government cannot get a fair idea of the industry with which it negotiates.
Merck argues that giving the United States access to data on its “indirect costs” would constitute a failure to apply the “directly pertinent” limitation in the access clause. It argues, in effect, that although “direct costs” might be “directly pertinent,” “indirect costs” cannot be directly pertinent by definition. The similarity of these phrases, though seductive, cannot substitute for analysis. The lower court’s decision approving this distinction26 was erroneous on two main grounds.
First, the narrow reading of the clause defies reason. The four contracts involved here between Merck and the United States concerned the supply of pharmaceutical products for a total price of more than $2,800,000. J.A. 34-37. If Merck’s “direct costs” are only nine percent of this figure, and if Congress’ grant of access is limited to direct costs, the lower court’s decision precludes governmental access to cost data behind the remaining charges of $2,548,000. The implications of such a limited interpretation of the access clause become even more significant when one considers the extent of the federal government’s contractual business with drug companies. Federal expenditures for pharmaceuticals exceed one billion dollars annually. J.A. 366. The narrow reading of the district court, if generally followed, would preclude government access to data concerning the underlying costs of more than $900,000,000 in pharmaceutical purchases annually. It should be obvious that Merck’s “indirect costs” are “directly pertinent” to procurement transactions concerning these government contracts, for otherwise the legislative purpose of reducing waste in government procurement would be undermined severely.
Fortunately for the public fisc, the district court’s reasoning has been followed by only one circuit. Staats v. Bristol Laboratories, 620 F.2d 17 (2d Cir. 1980). The Seventh Circuit, on the other hand, has consistently held that such costs as research and development “have a logical, causal and consequential relationship to a portion of the price of the products purchased by the Government.” Eli Lilly & Co. v. Staats, 574 F.2d at 914.
Although it might be possible to argue in other industries that costs such as research and development have such a small impact that they are not “directly pertinent” to the contract and its price, research and other costs not immediately attributable to one product form such a large portion of the costs of a pharmaceu*1248tical product that their import to such a contract seems inescapable.
Id. at 913-14. The same conclusion was reached in SmithKline Corp. v. Staats, 483 F.Supp. at 722 (E.D.Pa.1980), and arguably by the Ninth Circuit in Hewlett-Packard Co. v. United States, 385 F.2d at 1016 (9th Cir. 1967).27
The lower court’s decision is erroneous for a second reason. Its ruling should have been restricted to an analysis of the congressional purpose in enacting the access legislation, but embarked instead on a misguided attempt to find “a realistic interpretation of the intent of the parties.... ” J.A. at 512. This emphasis on the parties’ intent misapprehends the applicable law, for although the access clause was included in Merck’s contracts, it was required in those contracts by statute. When a clause is included by reason of congressional command, rather than as the result of contractual negotiations, the proper focus of inquiry is the intent of Congress.28 Merck did business with the government voluntarily, and thereby accepted the access clause as a precondition to doing business with the government.29 The access clause is of standardized, nationwide application; there was no discussion or negotiation as to inclusion of the provision or its terms. Merck acknowledges:
Indeed the standard-form clauses were among the various provisions that were simply incorporated by reference and not even made a physical part of the contracts. J.A. 74, 84-5. Not surprisingly, there was never any actual discussion or negotiation about the access clauses or their scope. See J.A. 88-9.
Merck Reply Brief at 2. Attempts to analyze the “intent” of the parties to such contracts must be considered dubious. “Obligatory Congressional enactments are held to govern federal contracts because there is a need to guard the dominant legislative policy against ad hoc encroachment or dispensation by the executive . ...” G.L. Christian & Associates v. United States, 320 F.2d 345 (Ct.Cl.), cert. denied, 375 U.S. 954, 84 S.Ct. 444, 11 L.Ed.2d 314 (1963). This rule of deference to congressional intent is especially important in this case, in which the legislative history demonstrates the expectation of Congress that GAO access to cost data would equalize the relationship between government and contractor.30
*1249Ultimately, Merck’s position on this issue can be reduced to two rather plaintive arguments. First, Congress must have meant something when it added the word “directly” in front of the phrase “pertinent books, documents, papers, and records.” This observation, while doubtless sensible, does not help Merck’s case. The amendment to the access legislation was made on the House floor by Representative Hoffman, a critic of the measure, who explained that its purpose was “to limit the ‘snooping’ that may be carried on under this bill which we do not have the votes to defeat” 97 Cong.Rec. 13377 (emphasis added). But this word of limitation must be read in the context of the congressional debates and Representative Hoffman’s understanding of the bill before his amendment was added. To Representative Hoffman, the access legislation “would let the GAO go into everybody’s business and look it over if they just wanted to take a look at it.... ” 97 Cong.Rec. 13373. For example:
If someone out in the country makes a contract for, as he said, an item which may be 10 percent of the total production, that under this bill, if it goes through, the GAO can send someone out there and go through the books, not only with reference to that one item of 10 percent which went into production for the Government, but into all of his other business where that same item was used.
97 Cong.Rec. 13377 (emphasis added). In other words, Representative Hoffman feared not that the GAO would inspect the indirect costs behind production for the government, but that GAO would inspect the direct costs of production of all items that were ever used by the government, even when the production was not relevant to government procurement at all. Representative Hoffman’s opinion of what was proper governmental access under the clause merely restated Representative Hardy’s position, and thus the sponsor of the legislation said he had no objection to the proposed amendment offered by one of the bill’s critics. 97 Cong.Rec. 13377. The demand by the Comptroller General here for access to indirect cost data is thus entirely consistent with Representative Hoffman’s amendment. The Comptroller General does not exceed the scope of the access clause envisioned by Congress, for he seeks only the data relevant to procurement by the government — but all of that data rather than a mere portion of it.
Merck’s final argument is that it is unfair to let the definition of “directly pertinent” costs depend on the structure of different industries. Merck Reply Brief at 42 — 13. I believe Congress concluded exactly the opposite, and expected such open-ended phrases as “directly pertinent” to be applied in light of the particular facts and conditions of each government contract. Merck suggests that such flexibility would lead to “an absurd result” in the instant case, for GAO “would be entitled to examine a massively greater number of documents because fewer are evidently pertinent.” Merck Brief at 72. But this result is not absurd at all. The test of “directly pertinent” is not the number of documents an industry must furnish, or the relative burden of doing so, but whether those documents will assist the government to determine whether its negotiating practices sufficiently protect it from wasteful, fraudulent, or inefficient procurement practices. Merck notes that “there simply are no generally accepted accounting principles permitting allocation of these indirect costs in an accurate and meaningful manner to particular products or contracts, including those at issue in this case.” Id. at 71. But this is exactly why Congress gave the Comptroller General a broad grant of authority. As the Ninth Circuit noted in Eli Lilly & Co. v. Staats:
The mere fact that, as plaintiff asserts, it does not or even could not allocate costs such as research and development to an individual contract does not undercut the proposition that those costs are directly pertinent to an individual contract but merely indicates that they are directly pertinent to more than one individual contract.
*1250574 F.2d at 914.31 It will not do to let the difficulty of the task before GAO stand in the way of the congressional mandate that it be attempted.
Obviously, the government has no right to examine costs recovered exclusively from non-governmental contracts.32 Moreover, governmental access should also be denied where Merck can show, through generally accepted accounting principles, that even its unallocated costs are minimally relevant to government contracts. But the burden of showing that such costs are irrelevant must fall on Merck because only Merck has all the data necessary to make such a showing. Indeed, Merck acknowledges that it has already undertaken a number of cost studies of these unallocated costs, an undertaking that seems somewhat surprising if these studies are as inaccurate and meaningless as Merck claims.33 Unless the burden of showing that unallocated costs are not relevant to government contracts is imposed on the party that has the data, companies might avoid compliance under the access clause simply by refusing to allocate any of their costs to any of their products. Such a result would abort the purpose of the access-to-records legislation as adopted by Congress. The district court’s order limiting the government’s access to Merck’s books and records should therefore be reversed.
CONCLUSION
One must come to statutes such as this with a prodisclosure bias. The public business ought to be public; the government’s right to know the costs incurred by its contractors is as important as the public’s right to know the costs incurred by its government. As Representative Hardy said in 1951, noting that nine of every ten companies voluntarily opened their cost records to the government,
this bill would affect that comparative few contractors who now refuse the agent of the Congress permission to see their records. It would be compulsory upon them. It seems logical to me to suppose that there is greater need for the GAO to see the records of those few who do not wish to make them available on a voluntary basis than it is to see the records of those who readily comply.
97 Cong.Rec. 13197. Representative Hardy’s view of the private sector may be unworldly, but it underlies the design of Congress in granting the GAO access to the books and records of those seeking to sell goods to the government. Moreover, there is absolutely no legally compelling basis on which to conclude that such access extends to the manufacturing but not to the research and development costs of pharmaceutical companies such as Merck.
The majority opinion embraces far too constrained an interpretation of the aecess-to-records clause. I believe that when the Supreme Court ultimately resolves the present split among the circuits, it will not impose a reading of the statute which makes the intent of Congress so diminutive. I dissent.
. The clause, granting the Comptroller General access to “directly pertinent” cost data of contractors whose contracts are negotiated without advertising, is based on the Act of October 31, 1951, 65 Stat. 700, codified at 10 U.S.C. § 2313(b) (1976) and 41 U.S.C. § 254(c) (1976).
. The Supreme Court has denied certiorari or conditional petitions for certiorari before judgment in four cases arising under the access clause, including this one. 449 U.S. 1038, 101 S.Ct. 618, 66 L.Ed.2d 501 (1981). When the Court did hear Staats v. Bristol Laboratories Division of Bristol-Myers Co., 451 U.S. 400, 101 S.Ct. 2037, 68 L.Ed.2d 343 (1981), the decision was affirmed without comment by an equally divided Court. Hence there has been no dispositive word from the Court.
. Identical demands by GAO for the cost records of other drug companies have led to extensive litigation in other circuits. See majority opinion at 1237. These demands grew out of a two-phase study of the economics of the pharmaceutical industry initiated by GAO in 1972. Phase I of the study, in which a number of companies participated voluntarily, was designed to give GAO “an understanding of the drug manufacturing industry,” and Phase II was to be a more detailed study that would develop data for a comprehensive presentation to Congress. Voluntary participation by the companies had broken down by August, 1974, leading the Comptroller General to make a formal demand for access to Merck’s cost data under the contract clause. Merck filed an action for declaratory relief in October, 1974, and the United States counterclaimed seeking access. After cross-motions for summary judgment, the district court issued an order on August 12, 1977, and modified that order in January, 1979. Both parties appeal from that order.
. By their terms, 10 U.S.C. § 2313(b) and 41 U.S.C. § 254(c) apply to negotiated contracts of any dollar amount. The government’s standard form contract, however, requires insertion of the access clause only in negotiated supply contracts which exceed or may exceed $10,000. 41 C.F.R. § 1-16.901-32 (1978).
. Reply Brief for Appellant Merck (Merck Reply Brief) at 8; see Brief for Appellant Merck (Merck Brief) at 18. It is undisputed that the impetus for GAO’s inquiry originated in January, 1971, during hearings on the status of competition in the pharmaceutical industry before the Senate Select Committee on Small Business, Monopoly Subcommittee. At hearings in May, 1972, Senator Nelson, subcommittee chair, urged the Comptroller General to obtain “production costs respecting drugs” from pharmaceutical manufacturers. Competitive Problems in the Drug Industry: Hearings Before the Subcommittee on Monopoly of the Senate Select Committee on Small Business, 92d Cong., 2d Sess. 8573 (1972). The Comptroller General responded that he doubted whether meaningful conclusions could be drawn on the basis of individual drugs, and would prefer an industry-wide study “of total sales to the Government on negotiated contracts.” Id. at 8577. In late 1972, GAO approached the Pharmaceutical Manufacturers Association and suggested the voluntary study discussed in note 3 supra. The voluntary participation of the drug companies allegedly broke down because GAO sought access to “sensitive cost data” and refused to give adequate assurances of confidentiality. Merck Brief at 4, 9. But see note 6 infra.
. Merck concedes that GAO “was prepared to give written assurances of confidentiality to the companies” in 1974. Merck Brief at 10. The district court added protective conditions to its order granting the government access to certain Merck-cost data. J.A. 518-19.
. See, e.g., United States v. LaSalle National Bank, 437 U.S. 298, 316, 98 S.Ct. 2357, 2367, 57 L.Ed.2d 221 (1978) (error to deny enforcement of IRS summons simply on basis of agent’s personal intent); McCray v. United States, 195 U.S. 27, 56, 24 S.Ct. 769, 776, 49 L.Ed. 78 (1904) (judiciary may not restrain the exercise of lawful power on the assumption that a wrongful purpose has caused the power to be exerted); Lynn v. Biderman, 536 F.2d 820, 826 (9th Cir.), cert. denied, 429 U.S. 920, 97 S.Ct. 316, 50 L.Ed.2d 287 (1976) (“It is not ... a ground to deny enforcement of a subpoena that it is being employed for a wrongful purpose if there is also a legitimate purpose for the subpoena.”). Cf. United States v. O’Brien, 391 U.S. 367, 383, 88 S.Ct. 1673, 1682, 20 L.Ed.2d 672 (1968) (Supreme Court will not strike down an otherwise constitutional statute on the basis of allegedly illicit legislative motives for passing it).
. Merck notes the admission by Gregory J. Ahart, Director, Human Resources Division, that GAO has no specific reason to believe that Merck made excessive or unreasonable profits on any of the four contracts in question. J.A. 216. But the same witness also explained that “it is impossible to describe the types of recommendations that might be made” as a result of GAO’s study, J.A. 242, and that “there is a range of possibilities of some modification on the basic procurement statutes.” J.A. 243. The government’s stated purpose for the access demand is “to review contract pricing by the Government’s suppliers, including plaintiff, of drug items to determine the adequacy of the protection afforded the Government by the negotiation techniques used by the procuring agencies.” Defendant’s Response to Plaintiff’s Second Set of Interrogatories, J.A. at 146.
. See, e.g., United States v. Abbott Laboratories, 597 F.2d 672, 674 (7th Cir. 1979); Eli Lilly & Co. v. Staats, 574 F.2d 904, 908 (7th Cir.), cert. denied, 439 U.S. 959, 99 S.Ct. 362, 58 L.Ed.2d 351 (1978); Smith-Kline Corp. v. Staats, 483 F.Supp. 712, 718 (E.D.Pa.), appeal pending No. 80-1464 (3d Cir. Mar. 19, 1980), conditional petition for cert. before judgment denied, 449 U.S. 1038, 101 S.Ct. 619, 66 L.Ed.2d 502 (1980); Bristol Laboratories Div. of Bristol-Myers Co. v. Staats, 428 F.Supp. 1388, 1390 (S.D.N.Y.1977), aff’d, 620 F.2d 17 (2d Cir. 1980), aff’d by evenly divided court, 451 U.S. 400, 101 S.Ct. 2037, 68 L.Ed.2d 343 (1981).
. Note, Government Contracts: Contractor’s Obligation to Allow Examination of Records Under 10 U.S.C. § 2313(b), 72 Dick L.Rev. 687, 691 (1968) (emphasis added in Merck’s Brief at 53). But see id. at 691 (possibility that Congress sought to improve general procurement techniques and “there is nothing in the Act specifically negating this idea ....”)
. Merck Brief at 31 (analyzing the remarks of Rep. Hardy discussed infra).
. 97 Cong.Rec. 13198 (Rep. Hardy) (citing one example in which GAO could profitably use proposed access authority).
. A sponsor’s statement of purpose is generally regarded as a persuasive indication of congressional intent. See, e.g., Federal Energy Adm. v. Algonquin SNG, Inc., 426 U.S. 548, 564, 96 S.Ct. 2295, 2304, 49 L.Ed.2d 49 (1976); Schwegmann Bros. v. Calvert Distillers Corp., 341 U.S. 384, 394-95, 71 S.Ct. 745, 750-51, 95 L.Ed. 1035 (1951).
. Indeed, Merck’s own account of the negotiations here suggests the possibility that the government operated under “a tremendous disadvantage.” Merck admits that “[it]here was no actual negotiation of the price prior to any of the awards,” Merck Brief at 14, and that “no terms in any of the contracts were the result of negotiations between the parties.... In each case, Merck was asked to submit a bid for specified pharmaceuticals, and in each case the Merck bid was accepted by the government purchasing officer, without negotiation.” Merck Reply Brief at 11 (citation omitted).
. Cf. Comptroller General Reports to Congress on Audits of Defense Contracts: Hearings Before the Military Operations Subcommittee of the House Committee on Government Operations, 89th Cong., 1st Sess. 147 (1965) (statement of Robert Keller) (“Congress intended the GAO to examine the books and records of the contractor to make certain that the Government was getting a fair deal out of negotiated contracts”). Merck’s narrower reading of the legislative history stems in part from understanding the word “audit” in the narrow, technical sense of to “check the books.” Merck Brief at 32 (quoting 97 Cong.Rec. 13200 (1951) (remarks of Rep. Riehlman)). Representative Hardy, however, used the term in the more general sense of “to inspect the pertinent books and records .. . . ” 97 Cong.Rec. 13197.
. Act of January 12, 1951, 64 Stat. 1257. Similar authority had been granted the president during World War II by the First War Powers Act of 1941, Pub.L.No. 77-593, 55 Stat. 838.
. Representative Celler described the Hardy amendment as “a Damoclean sword that will hang over the heads of the contractors whose contracts are changed. The amendment will give power to the General Accounting Office to go into the books and delve into the records of these contractors who have been relieved to determine whether or not there is fraud or overreaching or whether they have done anything untoward, in which event reports will be made and suitable action may be taken.” 96 Cong.Rec. 17123 (1951).
. See 97 Cong.Rec. 13198 (Rep. Hardy) (“Authority for negotiating all types of contracts without advertising has not always been available to Government procurement officers. Normal peacetime procedures generally require competitive bidding, but during times of emergency procurement by negotiation becomes necessary”); id. at 13199 (Rep. Hardy) (“In normal times most contracts are advertised.”).
. Moreover, the government argues that Merck must have realized the relevance of its • costs, even though the contract was negotiated without regard to costs, from the nature of the executed contracts. Brief for Appellee United States (Government Brief) at 18-20. The contracts also included a clause requiring Merck to include in all subcontracts a provision to the effect that the subcontractor also agreed to give access to the Comptroller General. J.A. 23-24. The express requirement concerning access to subcontractors’ records is manifestly a cost and pricing item, and should have made Merck aware that cost records would be considered pertinent to its negotiated, fixed-price contract.
. The Seventh Circuit followed its Lilly holding in United States v. Abbott Laboratories, 597 F.2d 672, 674 (7th Cir. 1979). This issue was not involved in Bristol Laboratories Division of Bristol-Myers Co. v. Staats, 428 F.Supp. 1388, 1391 (S.D.N.Y.1977), aff’d, 620 F.2d 17 (2d. Cir. 1980), aff’d by an evenly divided court, 451 U.S. 400, 101 S.Ct. 2037, 68 L.Ed.2d 343 (1981), because the company conceded the applicability of the access clause to its fixed-price contracts. The issue was also not directly presented in SmithKline Corp. v. Staats, 483 F.Supp. 712 (E.D.Pa.1980), appeal pending No. 80-1464 (3d Cir. Mar. 19, 1980), conditional petition for certiorari before judgment denied, December 8, 1980, 449 U.S. 1038, 101 S.Ct. 619, 66 L.Ed.2d 501 (1980).
. This section was amended in 1968 to give the Defense Department authority to examine “all books, records, documents, and other data of the contractor or subcontractor related to the negotiation, pricing or performance of the contract .... ” Pub.L.No. 90-512, 82 Stat. 863 (1968), codified at 10 U.S.C. § 2306(f) (1980).
. Moreover, the Truth-in-Negotiation Act exemption is merely permissive, not mandatory. Sperry Flight Systems Div. of Sperry Rand Corp. v. United States, 548 F.2d 915, 920-21 (Ct.C1.1977) (statute says cost data sometimes “need not” be submitted, not “shall not” be submitted, so demand for data upheld).
. See, e.g., 2 P. Areeda & D. Turner, Antitrust Law ¶ 404 (1980) (basic economic model of oligopoly as collective but non-collusive “monoplization”). See note 5 supra (original interest in pharmaceutical industry’s costs grew out of hearings on competitive problems before Subcommittee on Monopoly of the Senate Select Committee on Small Business). Cf. S.Rep.No. 643, 83 Cong., 1st Sess. 4 (1953) (exemption in Renegotiation Act for “standard commercial items” based on fact that cost and pricing experience had already been acquired and prices made in a competitive market, “in most cases”).
. This framework appears to correspond to Merck’s distinction between its “allocated” and “unallocated” costs. Merck Brief at 70-71. There may be a crucial difference between “indirect” and “unallocated” costs, of course, if some indirect costs can be allocated. Merck implies that all indirect costs are unallocable, id. at 72, and claims that no generally accepted accounting principles permit allocation of indirect costs such as research. But see notes 31-33 and accompanying text infra.
. Merck appears to concede as much. E.g., Merck Brief at 71 (“costs of research and development . .. tend to be the largest items of overall expenses at pharmaceutical companies”); Merck Reply Brief at 6 (“probably true that research and development costs ... tend to be a higher percentage of total expenses in the pharmaceutical industry than in most other industries”).
. J.A. 512. The lower court explicitly followed the holding of Bristol, 428 F.Supp. at 1390.
. Later decisions have disagreed over the Hewlett-Packard holding on this point. The contract in that case concerned electronic equipment for which direct labor and material costs undoubtedly made up the bulk of the price, but the trial court found the government entitled to inspect data on Hewlett-Packard’s “costs of direct material, direct labor and overhead costs" in producing the items furnished under the contracts. 385 F.2d at 1016 (emphasis added). The Seventh Circuit subscribed to a broad reading of this order because it “included overhead costs, a type of cost that is not always assignable to individual products, and because the Ninth Circuit’s opinion, which emphasized the appropriateness of inquiry into the procurement process as a whole, was not limited to disclosure of the specific costs requested by the Comptroller General in that particular case.” Lilly, 574 F.2d at 913. The Second Circuit in Bristol, on the other hand, read the decision in Hewlett-Packard as limited to production costs only and therefore as “of no positive aid in resolving the instant controversy,” 428 F.Supp. at 1391.
. The parties do not dispute this. Merck agrees that the contract clause “merely mirrored the access statute,” and thus “it is congressional intent that is determinative .... [I]t remains Congress’s purpose that is controlling, at least where the parties to the contract did not expressly agree to alter the statutorily prescribed right of access.” Merck Reply Brief at 10. The government also agrees: “Since the contract here involved is required by statute, there can, of course, be no issue as to bargaining as to the terms of the clauses . .. . ” Government Brief at 20-21 n.12.
. Cf. Almeida-Sanchez v. United States, 413 U.S. 266, 271, 93 S.Ct. 2535, 2538, 37 L.Ed.2d 596 (1973) (“The businessman in a regulated industry in effect consents to the restrictions placed upon him.”).
. Even if the intent of the parties were relevant, Merck must be charged with knowledge that the government read the access clause broadly in light of Hewlett-Packard. Merck asks that we refuse to follow Hewlett-Packard, Merck Reply Brief at 4, but it is clear that Congress too knew the holding in that case and yet did not amend the access statute. See 115 Cong.Rec. 25800 (1969) (GAO memorandum reporting Hewlett-Packard decision, reprinted at request of Senator Proxmire) (words “directly *1249pertinent” in clause “were intended only to limit GAO’s right of access to records pertaining to Government work as distinguished from non-government work.”).
. See, Case Comment, The Comptroller General’s Authority to Examine the Private Business Records of Government Contractors: Eli Lilly & Co. v. Staats, 92 Harv.L.Rev. 1148, 1157 & nn. 67, 68 (1979) (Lilly’s interpretation of “directly pertinent” is supported by the legislative history).
. Eli Lilly & Co. v. Staats, 574 F.2d at 918: “[N]o information need be produced relating to classes of costs that have no more than a de minimis impact on the price of a product purchased by the Government.”
. “Merck has, on occasion, performed ‘profitability’ studies of some of its product lines.” Merck Reply Brief at 7. Merck refers to these studies as “the essentially arbitrary assumptions made by one company,” and emphasizes that “there are no generally accepted accounting principles for compiling such studies.” Id. But the fact that there is “no consistent methodology for these ad hoc internal studies” and “ ‘broad allocations,’ ” id. at 48, hardly shows that there is no point in trying to make such a study, as Merck’s own efforts demonstrate. But see Note, Eli Lilly & Co. v. Staats: An Undue Expansion of the GAO's Investigatory Power Under the Access to Records Statutes, 74 Nw.L.Rev. 122, 132 (1979) (no need to allocate insignificant research costs).