The Dow Chemical Company v. United States Environmental Protection Agency and Douglas Costle, Administrator

OPINION OF THE COURT ADAMS, Circuit Judge. One of the assumptions upon which our society is premised is that technological advancement will be encouraged and material comfort will be increased by rewarding innovation with profit. Long ago, Sir Edward Coke noted that “everyone thirsteth after gaine,” 1 and, by and large, the law recognizes that experimentation and innovation will be persisted in only to the degree that an opportunity for “gaine” is presented. 2 This may be especially so today when our continued efforts in scientific research depend almost as much on those who are willing to fund them as on the ingenuity of the researchers themselves. But the recognition of the desirability of encouraging technological advancement does not mean that the Congress, if it chooses to do so, may not diminish innovation to some degree in order to further some other purpose. Nor does it mean that courts may refuse to give effect to such a legislative decision even if they suspect that the “other purpose” guiding the legislature will be disserved rather than furthered by the statute in question. This is so because Congress is entitled to make its own judgments in this area and once they are made the judiciary must respect them even if its own assessments are to the contrary. 1. DOW’S CHALLENGE TO THE EPA’S INTERPRETATION OF SECTION 8(d) OF TOXIC SUBSTANCES CONTROL ACT. The present appeal is by Dow Chemical Company, which charges that the Environmental Protection Agency (EPA) is misread its statutory mandate in seeking, under § 8(d) of the Toxic Substances Control Act, 15 U.S.C. § 2607(d) (1976), to obtain information and studies regarding research and development projects undertaken by a commercial enterprise, when such research and development are ultimately expected to produce a profitable product. Section 8(d) provides: (d) Health and safety studies. — The Administrator shall promulgate rules under which the Administrator shall require any person who manufactures, processes, or distributes in commerce or who proposes to manufacture, process, or distribute in commerce any chemical substance or mixture (or with respect to paragraph (2), any person who has possession of a study) to submit to the Administrator— (1) lists of health and safety studies (A) conducted or initiated by or for such person with respect to such substance or mixture at any time, (B) known to such person, or (C) reasonably ascertainable by such person, except that the Administrator may exclude certain types or categories of studies from the requirements of this subsection if the Administrator finds that submission of lists of such studies are *676 unnecessary to carry out the purposes of this Act; and (2) copies of any study contained on a list submitted pursuant to paragraph (1) or otherwise known by such person. Guidance in interpreting § 8 is provided by the following definition in subsection (f): “For purposes of this section, the terms ‘manufacture’ and ‘process’ mean manufacture or process for commercial purposes.” The dispute in the case at hand concerns the scope of the authority given the Administrator under § 8. Congress passed the Toxic Substances Control Act, 15 U.S.C. §§ 2601-2629 (1976), in order to prevent the general environment from becoming the laboratory in which harmful effects of chemicals are discovered. The Act establishes an intricate and at times unfathomable system of regulation administered by the EPA. It places the burden of ensuring adequate research and testing of toxic substances squarely on those companies who seek to profit from the use or sale of such chemicals. One of the most important provisions of the Act is § 4, which gives the Administrator the authority to require that certain companies — those that manufacture, distribute in commerce, process, use, or dispose of chemical substances or mixtures that the EPA suspects may be hazardous — conduct tests on such materials. 15 U.S.C. § 2603(a). Section 4(e) calls for the establishment of a committee that is expected to make recommendations to the EPA regarding what chemicals or mixtures should be tested pursuant to § 4(a). This committee, known as the Interagency Testing Committee (I.T.C.), is required to publish in the Federal Register and transmit to the EPA a list of chemicals and proposed areas of study. In October 1977 the I.T.C. published such a list, including ten chemicals and categories of chemicals. 42 Fed.Reg. 55026. According to the I.T.C. list, the specified chemicals presented serious potential for public exposure and there was a substantial possibility that they could be hazardous. The Committee recommended testing in six general areas: carcinogenicity, mutagenicity, teratogenicity, other chronic effects, environmental effects, and epidemiology. 3 To determine what tests would be necessary, the EPA, pursuant to § 8(d), proposed a rule designed to obtain the submission of already existing health and safety studies. 4 This proposed regulation stipulated that all manufacturers, processors and distributors of any of the named chemicals must submit to the EPA lists of studies initiated by or conducted for them and copies of any such studies in their possession. Section 1(a)(4) of EPA’s proposed rule defined the terms “manufacture or process” to include all manufacture or process of chemicals even if only for research purposes. 5 Several interested parties commented on § 1(a)(4) of the proposed rule, complaining that submission to the EPA of information gleaned from tests for new uses of chemicals might discourage product research and development inasmuch as competitors would be apt to gain valuable information about a company’s future plans. 6 Notwithstanding these complaints, the changes made in the final rule gave no relief to *677 those troubled by the possible chilling effects of the regulation on product research and development and innovation in the chemical field generally. Instead, the Agency specifically included a “Note” in the final regulation stating that the definition of manufacturing “for commercial purposes” includes “product research and development.” 43 Fed.Reg. 39086. And the two principal changes made to the proposed rule broadened rather than narrowed the scope of the regulation. Most important for our purposes, the final rule required that the companies subject to the rule submit copies of health and safety studies in their possession even if the studies were of chemicals that that company did not manufacture, process, or distribute. The listing requirement itself, however, remained limited to those substances actually manufactured, processed or distributed by the reporting company. See 40 C.F.R. § 730.5(a), 43 Fed.Reg. 30986. A second change substantially expanded the scope of the type of studies sought beyond the six areas suggested by the I.T.C. Dow, believing the changes in the proposed rule to be significant, charged that, the EPA had violated the Administrative Procedure Act by enlarging the scope of the regulation without adequate notice and fair opportunity to comment. See 5 U.S.C. § 553 (1976). Dow also challenged the Administrator’s substantive authority under the Act in two respects. First, it contended that the manufacture of small quantities of a chemical solely for purposes of research and development is not manufacture “for commercial purposes” and, therefore, is beyond the reach of § 8(d). Second, Dow urged that the EPA’s effort to obtain copies of studies under § 8(d)(2) should be limited to those studies that a company was required to list under § 8(d)(1). The second paragraph, Dow argued, cannot be read to give the Agency a power broader than that given in the first paragraph. In apparent response to Dow’s filing of a petition and brief on December 15,1978, the EPA, on January 26, 1979, withdrew the rule. But the EPA still insists that the regulation was within its statutory authority and that it was revoked solely because of procedural irregularities. Given the revocation, however, the EPA maintains that Dow’s appeal is now moot inasmuch as it challenges a rule that is no longer in effect. Before we may consider the merits of Dow’s petition, then, we must first address the question of mootness — a question that implicates the jurisdiction of the Court. II. MOOTNESS. At first reading, the case for dismissing the present petition as moot has considerable appeal. Simply put, the government urges that there is at this time no rule extant with which Dow must comply. Moreover, the government asserts there is no suggestion that Dow failed to comply in any material respect with the now-withdrawn rule when that rule was in effect. Thus, it may fairly be questioned whether this litigation retains the characteristics of a live case or controversy necessary for adjudication in a federal court. A careful analysis persuades us, however, that the petition is not moot. In recent years the Supreme Court has evolved a mootness doctrine that includes both constitutional 7 and policy 8 elements. Under Article III, section 2 of the Constitution, the power of the judiciary is limited to “cases or controversies.” When a dispute ceases to be sufficiently alive or concrete so as to be a genuine “controversy” it is no longer a fit subject for federal court adjudication. The judgment of a federal court must resolve a real and substantial controversy admitting of specific relief through a de *678 cree of a conclusive character, as distinguished from an opinion advising what the law would be upon a hypothetical set of facts. 9 We have interpreted Supreme Court pronouncements on this subject to require (1) a legal controversy that is real and not hypothetical, (2) a legal controversy that affects an individual in a concrete manner so as to provide the factual predicate for reasoned adjudication, and (3) a legal controversy with sufficiently adverse parties so as to sharpen the issues for judicial resolution. 10 Here Dow and the EPA are undoubtedly adverse parties. They offer contrary interpretations of a specific section of the statute. The dispute is thus real, not hypothetical; concrete not generalized. To resolve the question, this Court need not anticipate certain contingencies. Had the EPA, for example, withdrawn the rule because it was uncertain as to its statutory authority under the Act, the case might have been rendered non-justiciable — either because it lacks sufficient adversary quality or because the controversy would become to some extent hypothetical' inasmuch as it would depend on the mere possibility of future action by the EPA. The agency, however, withdrew the rule only because Dow’s petition raised questions regarding the EPA’s compliance with the procedural requirements of the Administrative Procedure Act. In the notice revoking the regulation, the EPA clearly stated that it did not intend “to change its interpretation of the scope of its statutory authority” about which “it entertains no doubt.” 44 Fed. Reg. 6099. In fact, the agency is in the process of proposing a new § 8(d) rule that will take precisely the same substantive position adopted in the withdrawn rule. Id. Under these circumstances, we believe that the Article III prerequisites for the exercise of federal court jurisdiction are satisfied. An' argument may be made, of course, that even if the constitutional prerequisites for federal court jurisdiction exist, the policy aspects of the mootness doctrine should not be lightly dismissed. But in this case policy considerations favor resolution of the petition at this time. First it is the EPA, not Dow, that urges us to treat the appeal as moot as a result of its own action in withdrawing the rule. Courts are understandably reluctant to permit agencies to avoid judicial review, whenever they choose, simply by withdrawing the challenged rule. As Judge Skelly Wright has noted: Where a court is asked to adjudicate the legality of an agency order, it is not compelled to dismiss the case as moot whenever the order expires or is withdrawn. Consideration of important legal issues “ought not to be, as they might be, defeated, by short term orders, capable of repetition but evading review » 11 Although we do not believe that this case may be classified as one involving official action “ ‘capable of repetition but evading review,’ ” 12 we are at least to some degree *679 persuaded by the policy concerns that underlie that exception to the mootness doctrine. The EPA has not altered its substantive stance, it has merely withdrawn its regulation for technical reasons with the declaration that it will be resubmitted. If this action by the EPA were alone sufficient to render a live dispute moot, the timing and venue of judicial review could be effectively controlled by the agency. We are reluctant, then, to dismiss a genuine and concrete controversy for what in this case amounts to a technical reason, brought about by the party seeking such a dismissal. 13 As the Supreme Court has .noted: [Voluntary cessation of allegedly illegal conduct does not deprive the tribunal of power to hear and determine the case, i. e., does not make the case moot. A controversy may remain to be settled in such circumstances, e. g., a dispute over the legality of the challenged practices. The defendant is free to return to his old ways. This, together with a public interest in having the legality of the practices settled, militates against a mootness conclusion. 14 A second policy reason also supports retention of jurisdiction over this case. Although the precise regulation under scrutiny has been withdrawn, the previous promulgation of the regulation, combined with EPA’s adherence to the interpretation of the statute included therein, may continue to have a present impact on Dow. Because the EPA has unequivocally taken the public posture that it will seek health and safety studies resulting from research and development work, Dow’s present decisions regarding research and development projects are necessarily colored by the very real possibility of having to provide the EPA with confidential information of considerable value to Dow’s competitors. The resulting “chill” on Dow’s research and development is thus a present harm that merits our attention, provided that the jurisdictional requirements have been met. To delay adjudication here would not leave the parties in the same position they occupied before the EPA took any action — rather it would leave Dow under a non-speculative threat of agency action while delaying any decision on the legality of that action. As in Super Tire Engineering Co. v. McCorkle, 416 U.S. 115, 94 S.Ct. 1694, 40 L.Ed.2d 1 (1974), where the Supreme Court declared that it was not deprived of jurisdiction because of mootness, the challenged governmental activity in the present case is not contingent has not evaporated or disappeared, and, by its continuing and brooding presence, casts what may well be a substantial adverse *680 effect on the interests of the petitioning parties. 15 Finally, it bears emphasis that there appear to be no persuasive policy reasons supporting a conclusion of mootness here. There is no possibility of “passing prematurely on constitutional questions.” 16 Nor are we likely to gain any additional information or insights by awaiting EPA’s re-promulgation of the contested regulation. Furthermore, until this issue is resolved there is a substantial possibility that product research and development as well as the EPA’s investigation of toxic chemicals will suffer: the.former from the chill discussed above, the latter from further litigation regarding the scope of the agency’s power. Because it is in the public interest to encourage both innovation and the control of toxic substances, we believe that it is the better policy to resolve the issue and thereby proffer Congress the opportunity quickly to rectify the situation if we have misread their intent. The only advantage in delaying adjudication here is that the EPA might change its mind and thus eliminate the controversy. But this possibility has been strongly and explicitly negatived by the agency, and must be considered remote at best. To refuse to pass on the questions presented, then, would require both the parties and the Court to undergo considerable additional expense and effort for no valid reason. Moreover, it would leave unresolved a dispute that it is in the interests of both parties as well as the public at large to have resolved. Accordingly, we conclude that this matter is not moot and proceed to address the merits. III. THE EPA’S STATUTORY AUTHORITY UNDER § 8(d) OF THE TOXIC SUBSTANCES CONTROL ACT. Two issues of statutory construction are presented by Dow’s challenge to the EPA’s regulation. First, does § 8(d) permit the Administrator to include in his definition of “manufacture” the production of small amounts of a chemical for purposes of research and development? Given the definition of “manufacture” provided in § 8(f)— “for commercial purposes” — this issue may be reduced to the question whether the production of limited quantities of chemical for research and development purposes may be considered a manufacture for commercial purposes. Second, does § 8(d)(2) permit the Administrator to seek copies of studies in the possession of a company even if that company is not required to list the studies pursuant to § 8(d)(1) and is not itself a manufacturer, processor, or distributor of the chemical that is the subject of the studies? Each of these questions is a matter of first impression. Not only has this Court not previously confronted them, but neither counsel nor our research has yielded any judicial construction of the section in question. Before attempting an analysis of the statutory language and legislative history of the sections in question, however, we must first consider what deference should be accorded the agency’s interpretation of these sections. A. The Deference to be Accorded the EPA’s Interpretation of the Act. It is now a commonplace of administrative law that courts in construing a statute will give “considerable weight” to a “consistent and longstanding interpretation by the agency charged with administration of [that statute].” United States v. National Association of Securities Dealers, Inc., 422 U.S. 694, 719, 95 S.Ct. 2427, 2442, 45 L.Ed.2d 486 (1975). 17 This rule of deference has been repeated frequently by the Su *681 preme Court 18 and by this Court, 19 and remains the standard for interpreting federal statutes. ' Like any rule, however, this precept of deference must be tailored to those situations that occasioned its creation. It should not be used, for example, as a shield for permitting agencies to extend their statutory authority beyond that delegated to them by Congress. Deference is generally accorded administrative agencies because of their experience and expertise. Although it has been said that the nation’s faith in government by experts has recently declined, 20 it is only logical that courts of general jurisdiction should give deference on matters that are more fully understood by specialist agencies. Agencies may be more qualified in certain fields because of their understanding of the actual area that is being regulated. Thus in labor law, the N.L.R.B. may ascertain what in fact chills workers organizational rights better than would a tribunal that is less experienced in labor matters. 21 Alternatively, agencies may be more qualified because of their sensitivity to the practicalities of the administrative process, and because of the need to supervise complex and often confusing statutory programs in a realistic and pragmatic manner. Consequently, the EPA has often been given broad deference in its attempts to produce a workable administrative system from obscure statutory language in highly technical statutes. 22 Were we dealing here with an agency interpretation stemming either from its experience with the subject matter being regulated or with administrative practicalities we would be inclined to accept EPA’s suggestion that its interpretation of § 8(d) is entitled to great or controlling deference. The statutory problems for resolution in the present case, however, are largely unrelated to the agency’s expertise in regulating possibly toxic substances or the nature of its administrative process. Rather, the problems we address here involve the limits to be placed upon the EPA’s information-gathering authority under the Act. These are issues of statutory construction requiring careful analysis of the language used by Congress in drafting the statute, as well as the legislative history of the Act. 23 A federal court is as qualified, as an agency in assaying such a task, and is more disinterested. The words at issue do not relate directly to the subject matter that EPA is regulating: we are not construing “hazardous,” “mixture,” or “chemical” as those terms are used in the Act. *682 Such words may properly depend on expert interpretation and standards. Instead we are construing an alleged limitation placed on the authority of the agency to obtain certain information. That an agency may adopt an interpretation maximizing its own authority is not surprising, but we do not believe we can fairly accord such an interpretation the same degree of deference that is generally given to the interpretation of those portions of a statute directly related to the expertise of the agency involved. Congress has delegated considerable authority to the EPA under this and other acts. The agency must be given deference in determining how to exercise that authority. But it need not and should not be given deference in determining how much authority has in fact been delegated to it. Were courts required to defer to an agency’s determination of the extent of its authority, the agency could proceed to exercise as much power as “reasonable,” even though a more plausible reading of its mandate suggests the power it seeks to exercise was never delegated by Congress. Accordingly, pursuant to Dow’s request that we set aside agency action “in excess of statutory jurisdiction, authority, or limitations,” 24 we shall objectively construe § 8(d) of the Toxic Substances Control Act, and sustain the agency’s interpretation of the authority given the Administrator under the Act only if it appears that such interpretation comports with the statutory language and the intent of Congress. B. Research and Development as a Commercial Purpose. Under § 8(d) the Administrator is authorized to seek certain information from those who “manufacture, process or distribute in commerce” or who “propose to manufacture, process or distribute in commerce” any chemical substance or mixture. “Manufacture” is defined in § 3 of the Act: The term “manufacture” means to import into the customs territory of the United States (as defined in general headnote 2 of the Tariff Schedules of the United States), produce, or manufacture. 25 This definition would appear to do little more than include importing as a type of manufacture. It offers little guidance in the resolution of the problem posed in this appeal. In two sections of the Act, however — § 5 and § 8 — a specific limitation is put on the term manufacture. For purposes of those sections, manufacture means manufacture “for commercial purposes.” 26 Dow contends that the manufacture of small quantities of a chemical that it does not propose to offer in commerce, but that it will use solely for purposes of product research and development, is not a manufacture “for commercial purposes” and, therefore, the EPA may not seek information about such a limited “manufacture.” There is a certain plausibility to this reading of the statute, and Dow adds to its appeal by emphasizing the harm that would result from a contrary reading. In enunciating the policy of the United States with regard to possibly toxic substances, Congress specifically stated that: authority over chemical substances and mixtures should be exercised in such a manner as not to impede unduly or create unnecessary economic barriers to technological innovation . . . 27 Dow argues with considerable force that the obligation to report to the EPA on tests done at the preliminary stages of product development will effectively eliminate the competitive advantage that is a prime spur to innovation. If competing concerns are able to learn of the type of research being performed by Dow — or any other similar company — those competitors would be able to ascertain what the reporting company’s future course of marketing and product promotion will be. Companies such as Dow, accordingly, are likely to eschew any re *683 search regarding those chemicals that the EPA is interested in learning about. The result of such a policy, as Dow sees it, is to discourage research on the very materials that EPA fears may be hazardous to our citizens. The EPA counters this argument by pointing to § 14 of the Act, 15 U.S.C. § 2613, which provides for some measure of protection for confidential information. Just because Dow must share its secrets with the EPA, the agency contends, it does not have to share them with competitors. Whether or not one could fairly expect Dow to be reassured by this position, 28 it, as the EPÁ freely admits, is largely beside the point. This case concerns Dow’s duty to submit information to the EPA, not the EPA’s duty, or ability, to keep that information confidential — as important as this may be. It is the language of the statute itself, in our view, that defeats Dow’s interpretation of the Act. For Dow’s construction to be correct, the use of “for commercial purposes” in limiting §§ 5 and 8 must have excluded manufacture for purposes of research and development. But this it clearly did not do. In both section, § 5 and § 8, specific language is used to exempt research and development from certain of the requirements set out in those sections. Thus § 8(a)(1) reads: (a) Reports — (1) The Administrator shall promulgate rules under which— (A) each person (other than a small manufacturer or processor) who manufactures or processes or proposes to manufacture or process a chemical substance (other than a chemical substance described in subparagraph (B)(ii)) shall maintain such records, and shall submit to the Administrator such reports, as the Administrator may reasonably require, and (B) each person (other than a small manufacturer or processor) who manufactures or processes or proposes to manufacture or process— (i) a mixture, or (ii) a chemical substance in small quantities (as defined by the Administrator by rule) solely for purposes of scientific experimentation or analysis or chemical research on, or analysis of, such substance, or another substance, including any such research or analysis for the development of a product, shall maintain records and submit to the Administrator reports but only to the extent the Administrator determines the maintenance of records or submissions of reports, or both, is necessary for the effective enforcement of this chapter. 29 Similarly § 8(b)(1) includes an exception from the Administrator’s duty to compile an inventory: (b) Inventory. — (1) The Administrator shall compile, keep current, and publish a list of each chemical substance which is manufactured or processed in the United States. . . . The Administrator shall not include in such list any chemical substance which is manufactured or processed only in small quantities (as defined by the Administrator by rule) solely for purposes of scientific experimentation or analysis or chemical research on, or analysis of such substance or another substance, including such research or analysis for the development of a product. 30 A review of § 5, the other section using the “commercial purposes” language, also reveals a specific exception for research and development. Section 5(h) reads: (3) The requirements of subsections (a) and (b) of this section do not apply with respect to the manufacturing or processing of any chemical substance which is *684 manufactured or processed, or proposed to be manufactured or processed, only in small quantities (as defined by the Administrator by rule) solely for purposes of— (A) scientific experimentation or analysis, or (B) chemical research on, or analysis of such substance or another substance, including such research or analysis for the development of a product, if all persons engaged in such experimentation, research, or analysis for a manufacturer or processor are notified (in such form and manner as the Administrator may prescribe) of the risk to health which the manufacturer, processor, or the Administrator has reason to believe may be assodated with such chemical substance. 31 Were we to adopt Dow’s construction of “commercial purposes” the above provisions of the Act would seem to be rendered almost meaningless. If the manufacture or process of a chemical in small quantities for purposes of research and development does not constitute a commercial purpose, there would be no need to provide for any special treatment for such chemicals in either of the two sections subject to the “commercial purpose” limitation. They would be entirely exempt from both sections. An objective reading of these sections reveals that Congress thought that the special exemptions for reporting, inventory, and notice were necessary, and that they did not believe that research and development had been generally excluded by the addition of a “commercial purposes” limitation. Dow attempts to meet this problem with two arguments. First it contends that the exceptions found in §§ 8(a)(1), 8(b)(1), and 5(h)(3) are not necessarily deprived of meaning under Dow’s interpretation. Second, it argues that the EPA’s interpretation may also be criticized on the ground that it conflicts with certain other language in the statute. Neither of these contentions is persuasive. The alternative explanation offered for the three specific exceptions provided for research and development is that although such exceptions were not strictly necessary for “in-house” research and development, they were needed to protect research chemicals that are actually sold commercially. According to Dow, “in-house” research and development is wholly exempt from both §§ 5 and 8 because of the “commercial purposes” limitation, whereas manufacturers of research and development chemicals produced for sale are offered more limited protection by §§ 8(a)(1), 8(b)(1), and 5(h)(3). We find this reading of the statute an unlikely one, both in terms of the language of the statute itself and the legislative history. Nowhere in the congressional reports and debates is there any indication that the drafters of the legislation perceived the distinction drawn by Dow and intended to craft an exception specifically for those in the business of manufacturing research chemicals for sale. In fact, it appears that Congress was under the impression that the exemptions in §§ 5 and 8 were necessary to protect both purchased and “in-house” research chemicals. Thus in explaining the exemption provided for research and development chemicals in § 5 — which requires that a prospective manufacturer give notice to the EPA ninety days before manufacturing a new chemical substance — Congressman McCollister of Nebraska made clear that both types of research were to be protected:. Section 5 is probably one of the most onerous provisions of the legislation with its requirement for premarket notification of all new chemical substances. However, this section does contain exemptions. For example, chemicals which are manufactured in small quantities for sale to laboratories for research purposes would not be subject to the premarket notification provisions of the bill. Further, research and analysis being performed during the developmental stages of chemicals which may ultimately be produced commercially would also be exempt from the premarket notification *685 provisions of the bill. This exception for research and analysis related to product development would exist regardless of whether the manufacturer were evaluating the product within its own plant or had made the product available to a potential customer with or without the payment of a fee. Section 5 must not be interpreted in such a way as to stifle product development and innovation, and we expect the EPA will implement Section 5 and the other provisions of the bill so that this result will not occur. 32 This view was echoed by other Congressmen as well, including the authors of the bill. Indeed, it would appear that there was more certainty that the exemption was designed to protect “in-house” research than that it covered the activity for which Dow claims it was intended. Congressman Murphy of New York made a point of emphasizing its dual purpose: Mr. Chairman, I think it is important to point out to my colleagues that research chemicals are exempted from the requirements of section 5. It has come to my attention that some people have interpreted this exemption as applying only to research chemicals which are manufactured and used in-house by a company. While the exemption does not include such research chemicals, it is not limited to them. If a chemical is manufactured by one person in small quantities for use as a research chemical and sold to another person to use for research purposes, then the exemption covers such a chemical. In other words, one company can make chemicals to be used for research by another company or by another person and the exemption covers that chemical. I would like to ask the authors of the legislation, the gentleman from Texas (Mr. Eckhardt) and the gentleman from North Carolina (Mr. Broyhill), if that is their understanding of the exemption? Mr. Eckhardt. Mr. Chairman, that is my understanding of the exemption. Mr. Broyhill. Mr. Chairman, I agree with the gentleman from New York. 33 These statements of the congressional understanding indicate that the drafters of the Act did not believe that the exemption they had fashioned in § 5(h)(3) was necessary only to protect research chemicals manufactured for sale, although it served to protect that manufacture as fully as “in-house” manufacture. Rather, they believed that the exception would benefit both types of research chemicals, and drew no apparent distinction between them. The same reasoning would logically apply to the exemptions appearing in § 8 and the use of the term “commercial purposes” in that section. The statute accordingly does not appear to bear the distinction urged by Dow to explain the inclusion of the exceptions in §§ 5(h)(3), 8(a), and 8(b) when no such exemption is found in § 8(d). Dow’s second reason for rejecting the EPA’s interpretation is that that construction, too, appears to render other language in the statute close to meaningless. Although it would not be unheard of to discover that a statute is internally inconsistent, we do not agree that the language referred to by Dow is deprived of meaning by the EPA’s construction of the legislation. There are two principal portions of the statutory language of § 8(d) that Dow relies upon as being incongruent with the EPA’s interpretation. The first is “for commercial purposes” — the very language upon which Dow has based its interpretation of the statute. If “commercial purposes” does not serve to exclude research projects, Dow suggests, it is difficult to see what it does do. The second is the inclusion of “proposes to manufacture, process or distribute” in the statement of those from whom the Administrator may require lists of health and safety studies. If a company’s initial research on a not-yet-developed product is a “commercial purpose,” Dow asks, why is the “proposes to” language necessary inasmuch *686 as the EPA can reach even the pre-proposal stage of product development? We agree that Dow’s explanation of these terms is internally consistent, but we do not believe that the EPA’s explanation is any less so. The EPA interprets any activity engaged in for profit as commercial activity. Inasmuch as Dow engages in product development in the hope of making profit and not out of pure scientific interest, its purpose may be fairly described as “commercial.” In this sense the “for commercial purposes” language distinguishes any profit-making enterprise from a non-profit institution, such as a university or foundation. Moreover, the “for commercial purposes” language does provide commercial companies with some protection — albeit less than Dow would like. The House Committee explained the meaning of this clause in its report referring to § 5: By use of the term “for commercial purposes” the Committee does not intend to restrict coverage to substances manufactured or processed “for sale.” Any commercial purpose, such as use as a chemical intermediate in a manufacturing process, is sufficient to bring the manufacture or processing of a substance within the ambit of section 5. The Committee realizes that there are certain minor reactions occurring incidental to the mixing process or upon storage of a mixture, such as the cross-linking of polymers. Such a minor reaction may result in what would technically be considered a “new” chemical substance. However, since the “new” substance is not manufactured for commercial purposes per se it would not be subject to the notification provisions of this section. 34 Although this explanation does not specifically contradict Dow’s construction, neither does it gainsay the EPA’s interpretation. The language is not deprived of meaning under either view; it merely is given a narrower scope under the EPA’s reading. As we have previously stated, this interpretation, although narrower, is, in our view, more in keeping with other subsections of § 8 and the statute as a whole. . Nor do we find the EPA’s interpretation to be at odds with the inclusion of “proposes to” in § 8(d). 35 One who does not “manufacture” any chemicals could “propose to” do so, or could “propose to” distribute chemicals in commerce even if it were not the manufacturer. A company proposing to engage in the production industrial use, or distribution of chemicals is still subject to the requirements of § 8(d) even if it has not yet begun actual manufacture, use or distribution. The inclusion of such companies— which are not as yet manufacturers, processors, or distributors — is not inconsistent with an intention to secure research studies from those who are manufacturers, processors, or distributors. Because we find EPA’s reading of its authority under § 8(d) to be in accord with the statute as a whole, and because we believe Dow’s interpretation does not adequately explain the specific exemptions provided for research and development elsewhere in the Act, we believe that the EPA’s inclusion of manufacture for research purposes in its final rule promulgated to effect the purposes of § 8(d) was within its statutory authority under that section. *687 C. The EPA’s Authority Under § 8(d)(2) to Seek Copies of Health and Safety Studies From Persons Not Subject to the Listing Requirement of § 8(d)(1). Dow’s second challenge to the substantive authority of the Administrator also presents a difficult question. Section 8(d)(1) requires manufacturers, processors and distributors of possibly hazardous chemicals to submit lists of existing health and safety studies known to them. 36 Section 8(d)(2) then requires the submission of copies of any studies sought by the Administrator. The question presented is whether § 8(d)(2) authorizes the Administrator to seek copies of studies not appearing on the submitted lists. Specifically, the EPA contends that whereas § 8(d)(1) applies only to studies regarding those I.T.C. specified chemicals that are actually manufactured, processed, or distributed by the reporting company, § 8(d)(2) permits the Administrator to obtain copies of studies in the possession of a reporting company regarding all the I.T.C. recommended substances, even if the reporting company does not manufacture, process, or distribute the chemical that is the subject of the study. 37 Dow, on the other hand, urges that a reporting compahy may be required to submit only copies of those studies that it is required to list pursuant to § 8(d)(1). Dow’s interpretation of § 8(d)(2) is certainly an understandable one. It is not illogical to reason that the requirement of submitting copies of studies is narrower than that of submitting lists. Indeed, the legislative history contains several statements by legislators assuming this to be the normal procedure. Thus, the Senate Report states: [Manufacturers must maintain with the Administrator lists of health and safety studies conducted, whether or not they have been conducted as a result of this legislation. The Administrator is authorized to require the submission of any study on the list. 38 Similarly, the House Committee Report does not appear to have contemplated that companies would be required to submit copies of studies regarding chemicals that they did not manufacture, process or distribute: Briefly, the bill will. . . . [authorize the Administrator to require manufacturers and processors to submit reports and maintain records respecting their commercially produced chemical substances and mixtures, to maintain records respecting adverse health or environmental effects of such substances and mixtures, and to provide available health and safety data on them. 39 The legislative history, therefore, may be fairly said to give some support to Dow’s two-step reading of the statute. But the legislative history may also be read to support the EPA’s interpretation of § 8(d)(2). Some of the legislative language indicating that the submission of copies of studies was to be confined to listed studies reflects an earlier Senate draft of the Toxic Substances Bill which explicitly stated: “The Administrator, on the basis of the lists submitted, may request submission of any study appearing on such list.” 40 This is not the language used in the final version of § 8(d), however. A possible reason is suggested by the Conference Committee Report, which stated, with respect to § 8(d), that “the conferees emphasize the importance of gaining information that errs on *688 the side of too much rather than too little.” 41 Thus, it cannot be said that the legislative history is all, or even predominantly, on Dow’s side. Where the legislative history is confused and the statutory language is clear, the latter must certainly prevail. 42 Indeed, the language actually used in a statute is the most determinative guide to the meaning of any piece of legislation. Here the language of the statute itself appears to comport more closely with the EPA’s approach. Section 8(d) refers to “any person who manufactures, processes, or distributes in commerce” or who “proposes to” do so. But immediately following this language there is a parenthetical clause, which reads: “or with respect to paragraph (2), any person who has possession of a study.” 43 This parenthetical expression can be viewed only as broadening, at least to some degree, the coverage of § 8(d)(2) — the paragraph requiring the submission of copies. Although § 8(d)(1) is limited to manufacturers, processors or distributors,-» § 8(d)(2) includes “any person who has possession of a study.” This latter group may be required to submit to the Administrator “copies of any study contained on a list submitted pursuant to paragraph (1) or otherwise known by such person.” 44 We believe this language plainly contradicts Dow’s theory that the Administrator may require only the submission of studies appearing on lists prepared pursuant to § 8(d)(1), even though that theory finds support in certain portions of the legislative history. It is possible that some members of Congress envisioned the two-step process described by Dow, but the language used in the statute unquestionably points the other way in two respects. First, it includes persons other than manufacturers, processors, or distributors — i. e. those who have possession of studies. The parenthetical expression can hardly be read any other way. Second, it includes studies other than those listed pursuant to § 8(d)(1). That is the only possible reading of the addition of “or otherwise known to such person” in § 8(d)(2). Given this statutory language, we are inclined to believe that Congress did intend that the EPA possess a wider authority in securing copies of studies than in securing lists. As in contract law, where “there is no surer way to find out what parties meant, than to see what they have done,” 45 so in construing statutes the intent of Congress is best determined from the words used in the legislation itself. Dow recognizes the apparent inclusiveness of the statutory language and attempts to meet it by offering a more narrow interpretation. First, Dow explains the apparently broad parenthetical clause by suggesting that it was added for the limited purpose of permitting the EPA to secure copies of listed studies in the possession of a third party. Otherwise manufacturers could insulate studies performed for them by leaving such studies in the hands of another company or individual. And, according to Dow, the “otherwise known by such person” language, appearing in § 8(d)(2) “serves the function of insuring that copies are submitted even if a study was inadvertently left off the list, or was initiated after the list was submitted . In this manner, the language ‘otherwise known to’ serves a valuable ‘mop-up’ function, guaranteeing that EPA receives all listable studies, whether or not included on the previously submitted lists.” 46 *689 Again this interpretation is plausible. But it concedes the point that the EPA, in seeking copies of studies, may reach parties other than manufacturers, processors, and distributors. And it admits that the EPA may seek copies of studies not appearing on any submitted list. Once these points are conceded, the legislative history referring to submitting copies of any study on a list is no longer helpful to Dow because it appears to conflict as much with Dow’s narrower reading as it does with the EPA’s broader construction. Deprived of the support provided in the legislative history noted above, Dow’s interpretation seems no more likely than the EPA’s construction. Indeed, we believe it to be less likely. The “otherwise known” language unquestionably contemplates the existence of studies in the possession of a company, which studies that company is not under an obligation to list. Such studies would necessarily concern chemicals that the company does not manufacture, process, or distribute, inasmuch as they must list all those chemicals that they do produce. When the section is read as a whole, and the parenthetical clause combined with the “otherwise known” language of paragraph two, this reading seems more likely than Dow’s attempt to confine § 8(d)(2) to a “mop-up” role of including “listable” studies that are, as yet, unlisted. Consequently, on this point, too, we believe that the EPA’s final rule did not go beyond the authority delegated to it under § 8(d) of the Toxic Substances Control Act. IV. CONCLUSION. The result we reach today may understandably cause concern to those troubled about the relative decline of technological innovation in the United States. If companies are required to submit to a federal administrative agency the results of tests they perform in the process of developing new products their chances of realizing a substantial competitive advantage may be measurably reduced. With the opportunity for “gaine” diminished in this fashion, corporate research may concentrate on substances that are not presently subject to the agency’s scrutiny. The result may be a net reduction of general research on the very substances — hazardous chemicals — on which research is greatly needed. 47 Were this to come about, it would presumably conflict with one of Congress’ purposes in passing the Act. Alternatively, companies may simply reduce their research and development spending altogether, particularly since the Toxic Substances Control Act is not the only current disincentive to innovation. 48 Of course, in the present case good arguments may be made for the Act as drafted. It is fair to doubt whether the primary commitment of large corporations is to the health of our citizenry — and reasonable for the government to seek to learn about and more carefully control toxic substances. The drafting of legislation often entails difficult policy choices, and the statute at issue here is no exception. But the issue for this Court cannot be whether we would have drafted § 8(d) so as to provide greater protection for product research and development. Nor is the question whether the regulation actually promulgated is a desirable one. We recognize the potentially unfortunate consequences of the EPA’s regulation and of our reading of the Act. Our role, however, is confined to construing the statute so as to give effect to the legislation *690 as written and to the intent of Congress in enacting such legislation. In so doing we have determined that Congress delegated to the EPA the authority to promulgate the regulation that is under challenge here. Perhaps the public may have reason to regret this result, but the possibility that the Act as drafted may inhibit technological innovation may not be relied upon as a justification for ignoring the apparent congressional decision in the drafting of the statute. Rather, any change in this regard is for the Congress to consider. The petition for review will be denied; each side to pay its own costs. Footnotes [1] Coke, The Third Part of the Institutes of the Laws of England, 195 (E. & R. Brooke eds. 1797). [2] See, e. g., the power given Congress in Article I, Section viii, clause 8 of the Constitution: “To promote the Progress of Science and useful Acts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries.” [3] Not all chemicals were recommended for all areas of testing. [4] The purpose of the proposed regulation was to help EPA to “ascertain whether, and if so what, testing rules are necessary.” 43 Fed. Reg. 4074. [5] § 730.1(a) of the proposed rule reads: (a) “Manufacture or process” means to manufacture or process for commercial purposes, which includes (1) for distribution in commerce, including for test marketing purposes; (2) for use as a catalyst or intermediate; (3) for the exclusive use by the manufacturer or processor, or (4) for product research or development. 43 Fed.Reg. 4075. [6] One of the comments illustrates the problem: For example, the fact that a consumer goods firm was testing a compound formerly used only in limited factory applications for its skin immersion effects might indicate to the competing firms in its industry that a new cleaning application was being considered for that compound. A-43. [7] The first mootness holding explicitly relying on Article III of the Constitution was Sibron v. New York, 392 U.S. 40, 88 S.Ct. 1889, 20 L.Ed.2d 917 (1968), although there were references to Article III limitations in earlier cases. See generally Note, The Mootness Doctrine in the Supreme Court, 88 Harv.L.Rev. 373 (1974). [8] See Franks v. Bowman Transportation Co., 424 U.S. 747, 756 n.8, 96 S.Ct. 1251, n.8, 47 L.Ed.2d 444 (1976). [9] Preiser v. Newkirk, 422 U.S. 395, 401, 95 S.Ct. 2330, 2334, 45 L.Ed.2d 272 (1975), quoting North Carolina v. Rice, 404 U.S. 244, 246, 92 S.Ct. 402, 30 L.Ed.2d 413 (1971), quoting Aetna Life Ins. Co. v. Haworth, 300 U.S. 227, 241, 57 S.Ct. 461, 81 L.Ed. 617 (1937). [10] See Geraghty v. United States Parole Comm’n, 579 F.2d 238, 246 (3d Cir. 1978) cert. granted 440 U.S. 945, 99 S.Ct. 1420, 59 L.Ed.2d 632 (1979). [11] Nader v. Volpe, 154 U.S.App.D.C. 332, 333, 475 F.2d 916, 917 (D.C. Cir. 1973) quoting Southern Pacific Terminal Co. v. I. C. C., 219 U.S. 498, 515, 31 S.Ct. 279, 283, 55 L.Ed. 310 (1911). [12] Dow urges that the present case does, in fact, constitute an example of official action capable of repetition yet evading review. We agree that EPA’s regulation is capable of repetition — indeed it is almost certain of repetition given the EPA’s continued adherence to the statutory construction it advanced in the regulation. The use of the capable of repetition but evading review exception must turn therefore on whether the Administrator’s interpretation will “evade review.” We are disinclined to read this traditional exception to the mootness doctrine as broadly as Dow does. Most cases utilizing this approach have involved official action that by its very nature could not, or probably would not be able to be adjudicated while fully “live.” Thus Roe v. Wade, 410 U.S. 113, 93 S.Ct. 705, *679 35 L.Ed.2d 147 (1973) concerned a woman’s claim of a right to abort her unborn child. Given the nine month gestation period, her pregnancy was certain to terminate before the case was adjudicated. Accord, First National Bank of Boston v. Bellotti, 435 U.S. 765, 98 S.Ct. 1407, 55 L.Ed,2d 707 (1978) (election campaign would terminate before bar to corporate participation therein could be reviewed); Sosna v. Iowa, 419 U.S. 393, 95 S.Ct. 553, 42 L.Ed.2d 532 (1975) (one year residency requirement for obtaining a divorce found to be of insufficient duration to permit complete review while still applicable to any particular plaintiff.). Here, in contrast, the new rule that EPA proposes to advance would be subject to review. We thus cannot say with certainty that EPA’s actions, if repeated, would “evade review.” At worst they would appear to do no more than delay review. We, therefore, rest our rejection of mootness here not on the repetitious evasion of review exception to the mootness doctrine but on the conclusion that that doctrine is not properly applicable to this, a live controversy, virtually certain of repetition, and with some present effect on the appellant. [13] See Natural Resources Defense Council, Inc. v. EPA, 489 F.2d 390 (5th Cir. 1974) reversed on other grounds 421 U.S. 60, 95 S.Ct. 1470, 43 L.Ed.2d 731 (1975). There, Judge Wisdom refused to allow the EPA’s disapproval of the particular regulations under challenge to moot plaintiffs’ argument, inasmuch as the EPA did not disapprove the underlying theory behind those regulations. A live dispute still existed, he reasoned, regarding the underlying strategy despite the EPA’s removal of specific regulations. The subsequent Supreme Court decision does not appear to have addressed this question. [14] United States v. W. T. Grant Co., 345 U.S. 629, 632, 73 S.Ct. 894, 897, 97 L.Ed. 1303 (1953) (citations omitted). [15] 416 U.S. at 122, 94 S.Ct. 1694, 1698. [16] Franks v. Bowman Transportation Co., 424 U.S. 747, 756 n. 8, 96 S.Ct. 1251 n. 8, 47 L.Ed.2d 444 (1976), quoting Flast v. Cohen, 392 U.S. 83, 97, 88 S.Ct. 1942, 20 L.Ed.2d 947 (1968). [17] One of the reasons for such deference, suggested in Saxbe v. Bustos, 419 U.S. 65, 95 S.Ct. 272, 42 L.Ed.2d 231 (1974), is the opportunity— over time — for Congress to correct a serious misinterpretation. See id. at 74, 95 S.Ct. 272. [18] See e. g., International Brotherhood of Teamsters v. Daniel, 439 U.S. 551, 566, & n. 20, 99 S.Ct. 790, 800 & n. 20, 58 L.Ed.2d 808, 820 & n. 20 (1979); DuPont v. Train, 430 U.S. 112, 134-35, 97 S.Ct. 965, 51 L.Ed.2d 204 (1977). The reasons for this deference and, more im- ■ portantly, limits to it were stated by the Supreme Court in Daniel: “This deference is a product both of an awareness of the practical expertise which an agency normally develops, and of a willingness to accord some measure of flexibility to such an agency as it encounters new and unforseen problems over time. But this deference is constrained by our obligation to honor the clear meaning of a stature, as revealed by its language, purpose and history.” 439 U.S. at 566 n. 20, 99 S.Ct. at 800 n. 20, 58 L.Ed.2d at 808 n. 20. [19] E. g., American Iron & Steel Institute v. EPA, 526 F.2d 1027, 1041 (3d Cir. 1975); Budd Co. v. OSHA, 513 F.2d 201, 203-05 (3d Cir. 1975). [20] See J. Freedman, Crisis and Legitimacy; The Administrative Process and American Government, 44-57 (1978). [21] See, e. g., Unemployment Compensation Commission v. Aragon, 329 U.S. 143, 153, 67 S.Ct. 245, 91 L.Ed. 136 (1946) (agency determination of what constitutes labor dispute “in active progress”). But see Getman & Goldberg, The Myth of Labor Board Expertise, 39 U.Chi.L.Rev. 681 (1972) (criticizing courts’ assumption that NLRB has special expertise to determine impact of employer conduct on exercise of employee rights). [22] See DuPont v. Train, 430 U.S. 112, 134-35 & n. 25, 97 S.Ct. 965, 51 L.Ed.2d 204 (1977); Power Reactor Co. v. Electricians, 367 U.S. 396, 408, 81 S.Ct. 1529, 6 L.Ed.2d 924 (1961). [23] “The question . is . solely one of statutory interpretation. The resolution of that issue does not require any expertise on the part of the [agency]; the proper interpretation is certainly not a matter of discretion.” McKart v. United States, 395 U.S. 185, 198-99, 89 S.Ct. 1657, 1665, 23 L.Ed.2d 194 (1969). See United States ex rel. Sanders v. Arnold, 535 F.2d 848, 853-55 (3d Cir. 1976) (Adams, J., dissenting). [24] See the scope of review set out in the Administrative Procedure Act, 5 U.S.C. § 706(2)(C). [25] 15 U.S.C. § 2602(7). [26] 15 U.S.C. §§ 2604(i) and 2607(f). [27] 15 U.S.C. § 2601(b)(3). [28] Dow questions the protection offered by § 14, noting that § 14(b)(1)(A) authorizes public disclosure of all health and safety studies once a testing rule is issued, regardless of whether a particular chemical has been offered for commercial distribution. Even were this problem avoided, Dow questions whether a large government agency is equipped to guarantee absolute confidentiality in circumstances such as these. [29] 15 U.S.C. § 2607(a)(1) (emphasis supplied). [30] 15 U.S.C. § 2607(b)(1) (emphasis supplied). [31] 15 U.S.C. § 2604(h)(3). [32] 122 Cong.Rec. 48805 (Aug. 23, 1976), reprinted in Legislative History of the Toxic Substances Control Act at 520-21 (emphasis added). [33] Id. 574-75 (emphasis added). [34] H.Rep.No.94-1341, 94th Cong., 2d Sess. at 30-31, reprinted in Legislative History, supra, • at 437-38. Nor does the legislative history reveal any Congressional intent to exclude research and development in its entirety: Mr. Broyhill. I want to compliment that gentleman for giving a very adequate explanation of the many provisions in a bill that do exempt research chemicals from provisions of this bill. But would the gentleman not agree with me that striking the reporting requirements altogether would not be wise, because there is one thing the EPA should have, and that is adequate knowledge of the new chemicals that are being developed? Mr. Murphy of New York. That is one of the thrusts of this bill. Mr. Broyhill. That is not certainly an onerous burden to know the formulas of new chemicals that are being developed, is it? Mr. Murphy of New York. It is not. Cong.Rec., Aug. 23, 1976, reprinted in Legislative History, supra, at 557. [35] The use of this term would appear to be tied to the § 5 requirement of a formal proposal of new chemical manufacture or sale. See 15 U.S.C. § 2604. [36] As is noted, supra, a company must list studies (1) conducted by or for it; (2) known to it; or (3) reasonably ascertainable by it. See 15 U.S.C. § 2607(d)(1). [37] EPA Internal Memorandum Submitting Final Regulation to the Administration. A-228. See § 730.5(a)(2) of the final rule, 43 Fed.Reg. 30986. [38] S.Rep.No.94-698, 94th Cong., 2d Sess., at 2, reprinted in Legislative History, supra, at 158 (emphasis added), U.S.Code Cong. & Admin. News 1976, pp. 4491, 4492. [39] H.R.Rep.No.94-1341, 94th Cong., 2d Sess., at 4, reprinted in Legislative History, supra, at 410 (emphasis added). [40] Hearings Before the Subcommittee on the Environment of the Committee on Commerce, S. 776, 94th Cong., 1st Sess., 94-24 (1975) at 27. [41] Conference Report, H.R.Rep.No.94-1679, 94th Cong., 2d Sess. at 81, U.S.Code Cong. & Admin.News 1976, p. 4566, reprinted in Legislative History, supra at 694. [42] In Greenwood v. United States, 350 U.S. 366, 374, 76 S.Ct. 410, 415, 100 L.Ed. 412 (1956), Justice Frankfurter puckishly declared that in such a situation it is prudent to apply “the canon of construction of the wag who said, when the legislative history is doubtful, go to the statute.” [43] 15 U.S.C. § 2607(d) (emphasis added). [44] 15 U.S.C. § 2607(d) (emphasis added). [45] Insurance Co. v. Dutcher, 95 U.S. 269, 273, 24 L.Ed. 410 (1877). [46] Brief for the Petitioner at 37. [47] The EPA may hope to remedy such decline through the use of testing requirements pursuant to § 4 of the Act. Whether government ordered research is an adequate substitute for individual initiative is at least open to debate. [48] Spending for research and development as a percentage of the gross national product is lower now than it was fifteen years ago. Indeed, private spending for research and development as a percentage of gross national product has now fallen behind that of both Germany and Japan. And it has been suggested that governmental actions are part of the problem. See I. Shapiro, Technology’s Decline: America’s Self-Made Paradox (Speech Before the Economic Club of Detroit, Jan. 22, 1979) reprinted in 45 Vital Speeches of the Day 360 (1979); J. Harley, The Day Innovation Died: Defensive R. & D., (Speech Before the Houston Club, Sept. 26, 1978) reprinted in 45 Vital Speeches of the Day 55 (1978).