Association of Irritated Residents v. Environmental Protection Agency

ROGERS, Circuit Judge,

dissenting:

This case involves the intersection of two doctrines. The first involves an agency’s unreviewable enforcement discretion, and the second relates to agency rulemak-ing power. The initial question for the court is whether the scope of enforcement discretion is expansive enough to cover the animal feeding operation (“AFO”) protocol formally announced by the Environmental Protection Agency (“EPA”) in the Federal Register on January 31, 2005, Animal Feeding Operations Consent Agreement and Final Order, 70 Fed.Reg. 4958, 4958, 4962-68 (Jan. 31, 2005) (“Initial Notice”). The court concludes that the enforcement protocol is an exercise of enforcement discretion that falls within the scope of the exception to judicial review set forth in Heckler v. Chaney, 470 U.S. 821, 832-33, 105 S.Ct. 1649, 84 L.Ed.2d 714 (1985), see Op. at 1028, 1031-32, and that EPA has not promulgated a legislative rule subject to the notice and comment requirements of the Administrative Procedure Act (“APA”), 5 U.S.C. § 553, see Op. at 1033-35. Undoubtedly there is some conceptual overlap between the doctrines to the extent that policies adopted by agencies often reflect discretionary determinations about how to enforce statutes that Congress has entrusted them to implement. However, by imposing a civil penalty on AFOs in the absence of individualized determinations of statutory violations, EPA has attempted to secure the benefits of legislative rulemaking without the burdens of its statutory duties. Our precedent does not permit the boundless stretching of Chaney to undercut the purposes of notice-and-comment rulemaking.

Additionally, even if Chaney creates a presumption of unreviewability of the enforcement protocol, Chaney further instructs that the presumption disappears when an agency veers far afield of Congress’s enforcement regime. See Chaney, 470 U.S. at 832-33, 105 S.Ct. 1649. By replacing the enforcement scheme in three congressional statutes with an unauthorized system of nominal taxation of regulated entities, EPA has promulgated a reviewable regulation. EPA cannot avoid the regulatory responsibilities imposed by Congress by trading nominal sanctions for amnesty to the regulated industry. However much enforcement discretion EPA may have in determining whether or not to file enforcement actions and whether to settle and on what terms, Congress has not authorized EPA to allow the regulated community to buy its way out of compliance with the statutes. For a minimum penalty plus $2,500, an AFO can, under the enforcement' protocol, avoid liability for any potential and ongoing violations of three statutes for at least a two-year period while EPA gathers and studies emissions data and for an indeterminate period thereafter while EPA develops and publishes new estimation methodologies, see Initial Notice, 70 Fed.Reg. at 4959, 4963-67; at no point are there repercussions beyond a possible future enforcement action if an AFO opts out of the agreement to be bound by the methodology regula*1038tions that EPA develops. Assuming no glitches, EPA’s endeavor to develop reliable methodologies could, according to the recommendations it has followed, take five, twenty, or even thirty, years. This is not an enforcement scheme at all, and is not a decision that Congress committed to agency discretion. See Chaney, 470 U.S at 832-33 & n. 4, 105 S.Ct. 1649.

Accordingly, I respectfully dissent.

I.

In announcing the new enforcement protocol, EPA advised AFOs in the egg, broiler chicken, turkey, dairy, and swine industries through the Initial Notice that they could avoid liability for “certain past and ongoing” violations of the Clean Air Act (“CAA”), the Comprehensive Environmental Response Compensation and Liability Act (“CERCLA”), and the Emergency Planning and Community Right to Know Act (“Right to Know Act”). See Initial Notice, 70 Fed.Reg. at 4959. All they had to do was agree to pay a minimal “civil penalty,” calibrated by the number of “farms,”1 plus (approximately) $2,500 per farm to fund the collection and study of nationwide emissions data over an (estimated) two-year period, and if selected, to allow EPA to monitor their operations’ emissions during that period. See id. at 4959, 4963-67. Based on the collected data, EPA would develop emission estimation methodologies, loosely commencing approximately eighteen months after completion of the study. See id. at 4960. Publication of the new methodologies “will trigger the obligation of participating AFOs to determine their emissions” and to come into compliance with the CAA, CERCLA, and the Right to Know Act. Id. at 4959. AFOs signing the identical form agreements, appended to the Initial Notice, see id. at 4962-68, would receive from EPA “a limited release and covenant not to sue” until after the AFO uses the new estimation methodologies and reports its releases under CERCLA and the Right to Know Act and applies for and receives the requisite CAA permits.2 Id. Any AFO would, however, be able to opt out of the agreement at any time without suffering any repercussions beyond the possibility of enforcement actions for past violations. Id. at 4959, 4963-67. The Initial Notice advised that the protocol would become effective only if a “sufficient” number of AFOs signed up. Id. at 4962. EPA requested public comment, “with particular emphasis on implementation,” within thirty days, id. at 4961, subsequently reopening *1039the comment period for thirty-three days, see Animal Feeding Operations Consent Agreement and Final Order, 70 Fed.Reg. 16266,16266 (Mar. 30, 2005).

The Initial Notice has all the earmarks of a legislative rule subject to APA notice and comment requirements, 5 U.S.C. § 553. Because the proposed' enforcement protocol is of “general ... applicability,” will have “future effect,” and defines the rights and obligations of members of the regulated community, thereby constraining EPA’s enforcement authority, it is a rule. See id. § 551(4); Indus. Safety Equipment Ass’n. v. EPA, 837 F.2d 1115, 1120 (D.C.Cir.1988); Batterton v. Marshall, 648 F.2d 694, 701-02 (D.C.Cir.1980). EPA’s enforcement protocol is not unlike the enforcement protocol in CropLife America v. EPA, 329 F.3d 876, 878 (D.C.Cir.2003), where EPA announced in the Federal Register that it would no longer consider human studies in its regulatory decision-making on the safety of pesticides under the Federal Insecticide, Fungicide and Ro-denticide Act (“FIFRA”), 7 U.S.C. § 136 et seq., and the Federal Food, Drug and Cosmetic Act (“Food and Drug Act”), 21 U.S.C. § 301 et seq. The court held that the new enforcement policy was a legislative rule because “it create[d] a ‘binding norm’ that [wa]s ‘finally determinative of the issues or rights to which it [wa]s addressed.’ ” CropLife, 329 F.3d at 881 (quoting Chamber of Commerce v. U.S. Dep’t of Labor, 174 F.3d 206, 212 (D.C.Cir.1999)). The policy was binding not only on the individuals challenging the rule, but also “on the agency because EPA ha[d] made it clear that it simply ‘will not consider’ human studies.” Id. Similarly here, EPA has announced a new general approach to carrying out its responsibilities under three statutes — provided a “sufficient” number of AFOs sign up and the Environmental Appeals Board (“EAB”) approves the individual AFO form' agreements. See Initial Notice, 70 Fed.Reg. at 4962, 4967. The new protocol is binding on both the AFOs who sign up and the agency, and under its terms, see id. at 4962, it will bind most of the regulated AFO industry.3 Under the circumstances, EPA’s new enforcement protocol is a legislative rule subject to notice and comment requirements under the APA.

The court would avoid our precedent on legislative rules on two grounds. First, the court concludes that it lacks subject matter jurisdiction to consider Petitioners’ challenges because. EPA’s decision to adopt the enforcement protocol is committed to its discretion by law and therefore falls within the scope of Chaney’s, exception to judicial review, see 470 U.S. at 832-33, 105 S.Ct. 1649. See Op. at 1028, 1031-32. However, EPA’s decision is hardly the type of particularized discretionary enforcement determination that confronted the Supreme Court in Chaney. In Chaney, death row inmates challenged the refusal of the Food and Drug Administration (“FDA”) to take various enforcement actions to prevent the use of lethal drugs for capital punishment. Chaney, 470 U.S. at 823, 105 S.Ct. 1649. The inmates alleged that use of these drugs for executions vio*1040lated the Food and Drug Act because the drugs had not been tested or approved for this purpose,-they were likely to be administered by untrained personnel, and they were unlikely to induce quick and painless death as intended. Id. The Court held that the FDA’s decision not to bring an enforcement action against particular members of the regulated community was “committed to agency discretion by law” under 5 U.S.C. § 701(a)(2) and therefore judicially unreviewable. Id. at 832, 837-38, 105 S.Ct. 1649. In announcing that such decisions were unsuitable for judicial review, the Court offered three rationales: (1) such decisions “involve [ ] a complicated balancing of a number of factors which are peculiarly within [the agency’s] expertise,” id. at 831, 105 S.Ct. 1649; (2) “when an agency refuses to act it generally does not exercise its coercive power over an individual’s liberty or property rights, and thus does not infringe upon areas that courts often are called upon to protect,” id. at 832, 105 S.Ct. 1649;' and (3) “an agency’s refusal to institute proceedings shares to some extent the characteristics of the decision of a prosecutor in the Executive Branch not to indict — a decision which has long been regarded as the special province of the Executive Branch, inasmuch as it is the Executive who is charged by the Constitution to ‘take Care that the Laws be faithfully executed,’ ” id. (quoting U.S. Const., art. II, § 3).

The enforcement .protocol set forth in the Initial Notice involves neither EPA’s decision to bring or not to bring an enforcement action based on an investigation giving rise to a belief that a regulated party has failed to comply with statutory requirements, nor is it a decision by EPA to settle an enforcement action that it has brought against a particular entity or is prepared to file in view of evidence of a violation. Indeed, the Initial Notice expressly contrasts the enforcement protocol with the type of enforcement action discussed in Chaney. See Initial Notice, 70 Fed.Reg. at 4958. This court has applied Chaney to such traditional enforcement actions, based on the statutory scheme for enforcement, where an agency has made a particularized enforcement determination following an investigation of a regulated party’s activities showing evidence of a statutory violation. For example, in Baltimore Gas & Electric Co. v. Federal Energy Regulatory Commission, 252 F.3d 456, 457, 460 (D.C.Cir.2001), the court held it lacked jurisdiction over a challenge to the settlement of an enforcement action brought by FERC against two vendors of natural gas. Similarly, in Schering Corp. v. Heckler, 779 F.2d 683, 684-86 (D.C.Cir.1985), the court held it lacked jurisdiction over a challenge seeking to invalidate the settlement of a lawsuit brought by a drug manufacturer against the FDA.

Although EPA could opt to forego bringing enforcement actions entirely, and its decision might be presumptively unre-viewable, see Chaney, 470 U.S. at 832-33, 105 S.Ct. 1649, that is not what happened here. Instead, EPA adopted a new generalized approach toward enforcing three environmental statutes in the future by means of an enforcement protocol unrelated to particularized findings of past or ongoing statutory violations and untethered to the enforcement regimes established by Congress,4 which EPA has previously utilized. Unlike the particularized actions which EPA has brought against AFOs in the past, see Initial Notice, 70 Fed.Reg. at 4958; Sandra S. Howland Aff. ¶¶ 8-11 (Jan. 5, 2006); Brief of Interve-nors for Respondents National Pork Producers Council and Roe Farm, Inc. at 5 (“Industry Intervenors’ Br.”), the enforcement protocol neither tracks the statutory *1041enforcement mechanisms of the three statutes nor purports to proceed on the basis of particularized information causing the agency to believe a statutory violation has occurred. Rather, EPA has used its coercive power to the extent that, upon entering into the identical form agreements, AFOs must commit, in return for a release from liability for any potential past or ongoing violations, to paying a civil penalty, to submitting to possible emissions monitoring, and to coming into compliance with new measurement methodologies or, upon opting out, to possible enforcement action, see Initial Notice, 70 Fed.Reg. at 4959. The statutory schemes for enforcement as previously interpreted by EPA have not included bringing its coercive power to bear absent some particularized factual basis for concluding the subject is in violation of the statutes.

Neither EPA’s view that attaining compliance through case-by-case enforcement actions “is difficult and time consuming,” id. at 4958; Op. at 1035, nor the novelty of the enforcement protocol means that it is not a legislative rule. Although EPA styles the Initial Notice as involving a consent decree, it is a consent decree only in the sense that any regulated party has a choice whether or not to proceed in accordance with an agency rule; in CropLife, for example, if using human studies was important to a regulated party, it had a choice either to proceed or not under EPA rule’s that it would not consider such studies. Further, contrary to the Industry Intervenors’ view, the enforcement protocol does not propose “informal adjudication[s],” Industry Intervenors’ Br. at 17, because the AFOs who sign up admit to no liability (or even that the three statutes apply to their farms), see Initial Notice, 70 Fed.Reg. at 4962 (Appendix 1), and EPA has not determined the liability of any AFO, see id., as would occur in an informal adjudication. See Ass’n of Nat’l Advertisers, Inc. v. FTC, 627 F.2d 1151, 1160-61 & n. 17 (D.C.Cir.1979).

Second, the court draws distinctions with our rulemaking precedent that begin with mischaracterizing the challenge before the court. The petition does not only challenge the individual form agreements between AFOs and EPA, see Op. at 1028, but includes a challenge to the enforcement protocol in the Initial Notice on the grounds' that EPA violated the APA’s notice and comment requirements and exceeded its statutory authority, see Petitioners’ Br. at 26-27, 44-53. From there the court invokes distinctions with our precedent that do not align with the terms of the protocol. The court states that the enforcement protocol is not a legislative rule because rather than “prescribing law,” 5 U.S.C. § 551(4), the protocol “merely defers enforcement of the statutory requirements, and makes that deferral subject to enforcement conditions that will ultimately result in compliance.” Op. at 1033. In fact, the enforcement protocol forever absolves AFOs who sign up (and do not opt out) from liability for any potential past and ongoing violations. In the period of time that the court characterizes as a deferral, signing AFOs will face no prospect of liability through EPA enforcement actions for whatever they do or do not do to comply with the three statutes because EPA has committed not to exercise its enforcement authority. See Initial Notice, 70 Fed.Reg. at 4959, 4963. Any liability imposed by EPA for statutory violations will arise only sometime after EPA develops and publishes, and AFOs apply, new measurement methodologies, based on the two-year nationwide emissions study, and will apply only to violations found at that time. See id. at 4959.

Additionally, the court struggles to distinguish precedent such as National Ass’n of Home Builders v. U.S. Army Corps of *1042Engineers, 417 F.3d 1272, 1279-80 (D.C.Cir.2005), where the court held that the agency’s authorization of a right to bypass certain statutory requirements constituted a rule. See Op. at 1033-34. Under the enforcement protocol EPA has standardized minimum civil penalties based on farm size, bypassing those that would otherwise apply had EPA conducted site investigations as contemplated under the statutory enforcement schemes, supra notes 1, 4. The court’s attempt to distinguish CropLife, 329 F.3d at 881, likewise fails. See Op. at 1033-34. It is unclear why the court views the enforcement protocol as any different from EPA’s announcement that it would no longer consider human studies in enforcing FIFRA and the Food and Drug Act, see CropLife, 329 F.3d at 878-79. So too here, EPA announced that for AFOs who sign up, it will no longer consider evidence of statutory violations by individual AFOs but instead will impose a farm-size minimum civil penalty, subject to a cap, and require a $2,500 payment in exchange for which AFOs will be absolved of liability for all past and ongoing statutory violations. See Initial Notice, 70 Fed.Reg. at 4959.

The court concludes, notwithstanding EPA’s stated intention to impose “a civil penalty” and to require a payment to fund a nationwide emissions study, that the identical form agreement for AFOs to sign “does not express the agency’s implementation of any provision of the [three statutes].” Op. at 1033-34. The court is correct in the sense that EPA has made “no determination of an AFO’s compliance with the Acts,” id., after a particularized investigation of a regulated party’s activities as contemplated under the enforcement regimes of the statutes, supra note 4, but it is incorrect to suggest that EPA has not made a “definitive statement of enforcement or interpretive practices,” Op. at 1034. Just as with its deferral characterization, the court ignores the reality of EPA’s enforcement protocol, which absolves AFOs who sign up of liability for any potential past and ongoing violations.

The same flaw is evident in the court’s statement that “EPA has not bound itself in a way that reflects ‘cabining’ of its pros-ecutorial discretion because it imposed no limit on its general enforcement discretion if the substantive standards are violated.” Op. at 1034. In fact, the enforcement protocol has done precisely that. In Community Nutrition Institute v. Young, 818 F.2d 943, 948 (D.C.Cir.1987), the court explained that “cabining of an agency’s pros-ecutorial discretion can in fact rise to the level of a substantive, legislative rule.” Under the enforcement protocol EPA fore-goes its enforcement discretion to file actions against signing AFOs, regardless of whether EPA may attain information during the life of the agreement that an AFO has violated or is violating one of the statutes; only the AFOs that do not sign the agreements — approximately eight percent of AFOs, see supra note 3 — “will be subject to enforcement actions in which significant penalties and injunctive relief could be sought for violations of the CAA, section 103 of CERCLA, and section 304 of [the Right to Know Act].” Initial Notice, 70 Fed.Reg. at 4961. Until EPA begins to publish the new methodologies for estimating emissions “on a rolling basis as work is completed,” id. at 4960, following the nationwide two-year emissions study, EPA has released the signing AFOs from any liability prior to the time the AFO applies the new measurement methodologies and reports its releases under CERCLA and the Right to Know Act and applies for and receives the requisite CAA permits, see id. at 4959. Tellingly, the court concedes that “to the extent EPA has limited its enforcement discretion, it has done so only with regard to those AFOs who have signed Agreements,” Op. at 1034, but chooses to ignore that EPA conditioned its enforce*1043ment protocol on a “sufficient” number of AFOs signing up, see Initial Notice, 70 FecLReg. at 4962, 4965, and that, given the terms of the enforcement protocol, the response to the sign-up invitation in the Initial Notice has been overwhelming, see supra note 3.

Finally, detaching Chaney from its moorings undercuts one of the key purposes of the APA. The legislative history of the APA states that “due to the unrepresentative nature of an administrative agency, ‘public participation ... in the rulemaking process is essential in order to permit administrative agencies to inform themselves and to afford safeguards to private interests.’ ”5 Batterton v. Marshall, 648 F.2d 694, 704 n. 47 (D.C.Cir.1980) (quoting S. Comm, on The JudiCIARy, 79th Cong., Senate JudiciaRy Committee Print (1945), as reprinted in S. Doo. No. 248, 79th Cong., 2d Sess. 11, 19-20 (1946)); see Nat’l Ass’n of Home Health Agencies v. Schweiker, 690 F.2d 932, 949 (D.C.Cir.1982); Home Box Office, Inc. v. FCC, 567 F.2d 9, 35 (D.C.Cir.1977). The Supreme Court has observed that “[i]n enacting the APA, Congress made a judgment that notions of fairness and informed administrative decisionmaking require that agency decisions be made only after affording interested persons notice and an opportunity to comment.” Chrysler Corp. v. Brown, 441 U.S. 281, 316, 99 S.Ct. 1705, 60 L.Ed.2d 208 (1979). “[Pjublic participation assures that the agency will have before it the facts and information relevant to a particular administrative problem ... [and] increase^] the likelihood of administrative responsiveness to the needs and concerns of those affected.” Am. Hosp. Ass’n v. Bowen, 834 F.2d 1037, 1061 (D.C.Cir.1987) (omission in original) (alterations in original) (quoting Guardian Fed. Sav. & Loan v. Fed. Sav. & Loan Ins. Corp., 589 F.2d 658, 662 (D.C.Cir.1978)); see Nat’l Welfare Rights Org. v. Mathews, 533 F.2d 637, 648 n. 17 (D.C.Cir.1976).

EPA’s new enforcement protocol will have significant and immediate negative consequences. As EPA acknowledges, it affects both members of the regulated industry and those whom Congress intended to protect under the statutes as well as the safety and health of the environment. See Initial Notice, 70 Fed.Reg. at 4959. Emissions from AFOs not only have negative impacts on nearby residents by causing odors and other nuisances, but they emit pollutants, including ammonia and hydrogen sulfide, which are classified as hazardous substances under CERCLA and the Right to Know Act, as well as particulate matter and volatile organic compounds, which, along with hydrogen sulfide, are regulated under the CAA. See id. Ensuring accountability and informed decision-making means an agency needs to hear from those who are affected before it adopts an enforcement policy that eliminates enforcement of several statutes for years for a significant part of the AFO *1044industry, including “most or a lot of the largest farms,” supra note 3, and perhaps for one hundred percent of the AFOs in light of EPA’s view that its current methodology needs improvement, see Initial Notice, 70 Fed.Reg. at 4958. Congress determined in enacting the APA that closed door meetings and informal discussions in the absence of public notice of the agency’s view of a final proposal are not sufficient to the task. See Chrysler Corp., 441 U.S. at 316, 99 S.Ct. 1705; Batterton, 648 F.2d at 704 n. 47.

II.

The presumption of unreviewability announced in Chaney may be overcome “where the substantive statute has provided guidelines for the agency to follow in exercising its enforcement powers.” Chaney, 470 U.S. at 832-33, 105 S.Ct. 1649. It also is possible that it may be rebutted where the agency refuses to act based solely on the belief it lacks jurisdiction or “the agency has ‘consciously and expressly adopted a general policy’ that is so extreme as to amount to an abdication of its statutory responsibilities.” Id. at 833 n. 4, 105 S.Ct. 1649 (citing Adams v. Richardson, 480 F.2d 1159 (D.C.Cir.1973) (en banc)).

Each of the three statutes at issue prescribes monetary and other penalties for violations, including criminal penalties in some instances, when EPA undertakes its enforcement responsibilities. See supra note 4. For example, under the CAA, upon finding a violation of a State Implementation Plan or permit, the EPA Administrator is required to notify the violator and the State and, after thirty days, may issue an order of compliance, impose an administrative penalty, or bring a civil action. 42 U.S.C. § 7413(a)(1). In other circumstances, the Administrator is required to give public notice, id. § 7413(a)(2), and for other violations the Administrator may issue compliance orders, impose administrative penalties, bring a civil action, or request the Attorney General to commence a criminal action, id. § 7413(a)(3). Although Congress vested discretion in the Administrator over whether to bring civil enforcement actions, Congress expressly required that such actions be based upon a determination that the subject “has violated, or is in violation of, any requirement or prohibition of’ a State plan or permit or certain statutory provisions. Id. §§ 7413(a)(1)(C), (a)(2)(C), (b)(1), (b)(2). Criminal penalties are likewise predicated upon findings of knowing violations. Id. § 7413(c). CERCLA similarly provides for civil and administrative penalties based on found violations. Id. §§ 9609(a)(1), (b), (c). Even the Right to Know Act includes criminal as well as civil penalties for found violations. Id. §§ 11045(b), (c). Notwithstanding this broad delegation of enforcement authority, however, Congress did not authorize EPA to tax members of the regulated industry, no matter what its motives. Instead, Congress authorized EPA to impose penalties based on particularized evidence of statutory violations. Further evidence that Congress intended EPA’s imposition of penalties to be based on found violations is apparent from the authority it delegated to EPA, under section 114 of the CAA, for example, to require AFOs to monitor and report emissions from their facilities, id. § 7414(a).

The enforcement protocol takes an entirely different tack. Relying on a study by the National Academy of Sciences in 2003,6 EPA has determined that its meas*1045urement methodologies require updating and improvement. See Initial Notice, 70 Fed.Reg. at 4958. According to the study, emission estimating methodologies can be improved through research in the short term, i.e., five years, but research is merely the first step in developing the comprehensive process-based model that the study recommends, which may involve twenty to thirty years.7 See NAS study on AFO Air Emissions at 11, 12, 16, 154-55. EPA has nonetheless estimated that it will begin publishing new methodologies within approximately three and a half years, see Initial Notice, 70 Fed.Reg. at 4960, assuming, contrary to the reality that Industry Intervenors suggest regarding monitoring methods, see Industry Intervenors’ Br. at 4-5, that the two-year nationwide emissions study commences promptly upon EPA’s determination of the appropriate monitoring methods and procedures.

In announcing the new enforcement protocol, EPA purports to respond to a problem of “uncertainty regarding emissions from AFOs” arising from lack of data involving large operations that would require a commitment of substantial resources and time to gather the data. Initial Notice, 70 Fed.Reg. at 4958; see Respondent’s Br. at 21-22; Industry Intervenors’ Br. at 1, 3-5. AFOs “have no ‘smokestacks’ from which air emissions can be measured.” Industry Intervenors’ Br. at 2. EPA states that it believes that the enforcement protocol, in comparison to filing individual enforcement actions, is “the quickest and most effective way to address the current uncertainty regarding emissions from AFOs and to bring all participating AFOs into compliance with all applicable regulatory requirements.” Initial Notice, 70 Fed. Reg. at 4958 (citing NAS study on AFO Air Emissions). Although EPA plans to establish and begin to publish the new methodologies “on a rolling basis as work is completed,” id. at 4960, within approximately eighteen months after completion of the (approximately) two-year nationwide monitoring study, how long it will actually take for EPA to begin monitoring emissions and then to establish the new methodologies is unclear; in the past, disputes regarding emission monitoring methods and procedures have caused delays, see Howland Decl. ¶¶ 6, 9. The recommendation of the National Academy of Sciences, which EPA claims to follow, see Initial Notice, 70 Fed.Reg. at 4960, calls for extensive air emissions testing over a wide variety of facilities and locations in order to have a more widely accepted standard for determining whether AFOs are in compliance with the statutes. See Howland Decl. at ¶ 6; NAS study on AFO Air Emissions, at 71-72, 152-68, 172-75.

Neither EPA’s “uncertainty” nor the NAS study on AFO Air Emissions is dis-*1046positive of EPA’s authority to proceed in the proposed manner. In the meantime, while EPA is gathering data and developing emission estimating methodologies, the National Academy of Sciences did not suggest that EPA could do nothing by way of enforcement. The study’s basic criticism was that EPA had too little data but this is not the same as concluding that relevant and sufficient data could not be collected, as Congress provided in section 114 of the CAA, 42 U.S.C. § 7414(a), for example, by requiring AFOs to monitor and report their emissions so that EPA could determine whether there is evidence of a statutory violation and if so, decide whether to bring enforcement actions. Assuming the scientific merit of the National Academy of Sciences’ recommendations, the study does not demonstrate, notwithstanding the industry’s objection to the high cost of self-monitoring, see Industry Intervenors’ Br. at 3-4; Tr. of EAB Hearing at 43-44 (Dec. 13, 2005), that Congress’s enforcement schemes were not designed to achieve the same results, namely enforcement based on reliable data on AFO emissions.

EPA concludes that by imposing “a civil penalty” and requiring a $2,500 study payment as the price for escaping liability for any potential past or ongoing statutory violations until EPA develops and publishes — and AFOs apply — the new methodologies, it is likely to bring the AFO industry into compliance with the statutes more quickly than would traditional, individual enforcement actions, which may involve extensive and difficult data collection efforts and lengthy enforcement actions. See Initial Notice, 70 Fed.Reg. at 4958; Howland Decl. at ¶¶ 8-13. Maybe so, although the forty-two month period EPA identifies is uncertain at both ends. Regardless, Congress has made a different choice about how to achieve statutory compliance and until Congress determines that evidence of violations is not the basis for imposition of penalties, EPA is bound to adhere to the enforcement regimes Congress has established, including directing the regulated community to monitor and report emissions data. EPA only has such authority as Congress has delegated to it: “As the Supreme Court has recognized, ‘an agency literally has no power to act ... unless and until Congress confers power upon it.’ ” Cal. Indep. Sys. Operator Corp. v. Fed. Energy Regulatory Comm’n, 372 F.3d 395, 398 (D.C.Cir.2004) (omission in original) (quoting La. Pub. Serv. Comm’n v. FCC, 476 U.S. 355, 374, 106 S.Ct. 1890, 90 L.Ed.2d 369 (1986)).

. Although the release of civil liability is described as a "limited release and covenant not to sue,” Initial Notice, 70 Fed.Reg. at 4959, the form agreement, set forth in Appendix 1 of the Initial Notice, states that the release is complete for farms that an AFO lists in Attachment A to the form agreement, except for emergency situations due to accidental releases. Id. at 4963. Further, it is complete even as to violations that may be uncovered using the new methodologies: "The release and covenant not to sue found in paragraph 26 [of the form agreement] resolves only violations identified and quantified by applying the Emissions-Estimating Methodologies developed using data from the national air emissions monitoring study described herein.” Id.

. Unsurprisingly, given its virtual free pass for statutory violations, see supra notes 1, 2, approximately ninety-two percent (92%) of AFOs signed up. EPA informed the EAB that it had reports that there are 15,000 or more concentrated animal feeding operations nationwide and that the identical form agreements signed by AFOs "captured most or a lot of the largest farms.” According to EPA’s brief, "[a]pproximately 13,908 farms'are covered under the [] agreements.” Brief for Respondent EPA at 3 n. 2. Before the EAB, EPA also acknowledged that, for the most part, industry trade associations, and not the individual AFOs who sign up, will actually pay the (approximately) $2,500 to fund the emissions study.

. See 42 U.S.C. § 7413(CAA); id. § 9609 (CERCLA); id. § 11045 (Right to Know Act).

. With regard to public access to agency records, the legislative history of the APA explains that:

The public information requirements of section [5 U.S.C. § 552] are in many ways among the most important, far-reaching, and useful provisions of the bill. For the information and protection of the public wherever located, these provisions require agencies to take the mystery out of administrative procedure by stating it. The section has been drawn upon the theory that administrative operations and procedures are public property which the general public, rather than a few specialists or lobbyists, is entitled to know or to have the ready means of knowing with definiteness and assurance.

S. Rep. No. 79-752 (1945), as reprinted in Legislative History of the Administrative Procedure Act, S. Doc. No. 248, 79th Cong., 2d Sess. 185, 198 (1946); see also H. Rep. 79-1980 (1946), as reprinted in Legislative History of the Administrative Procedure Act, S. Doc. No. 248, 79th Cong., 2d Sess. 233, 252, 256 (1946).

. Nat’l Acad, of Scl, Ad Hoc Comm, on Air Emissions From Animal Feeding Operations et al„ Nat’l Research Council, Air Emissions From Animal Feeding Operations: Current Knowledge, Future Needs (2003), available at http:// www.nap.edu/catalog/10586.html ("NAS study on AFO Air Emissions"); see also Initial Notice, 70 Fed.Reg. at 4960 (citing Nat’l Acad. of Sci, The Scientific Basis For Estimating Air *1045Emissions From Animal Feeding Operations, Interim Report, National Research Council (2002)).

. The National Academy of Sciences study concluded that “available estimates of emissions factors, rates, and concentrations are sufficiently uncertain that they provide a poor basis for regulating or managing air emissions from AFOs,” NAS study on AFO Air Emissions at 6, and that ”[s]cientifically sound and practical protocols for measuring air concentrations, emission rates, and fates are needed,” id. at 8. The study identified short term (five years or less) and long term (twenty to thirty years) research priorities. Under the five year model, EPA would research concentration measurements, dispersion modeling, odor measurement and characterization, and abatement and management strategies. Id. at 154-61, 168. Under the twenty-plus year model, EPA would focus on researching an integrated program to reduce the losses of materials from AFOs to the environment by more efficient use of input materials and increased recycling of materials containing certain chemicals within the AFOs. Id. at 161-68.