Judicial Review Under the Superfund Amendments: Will Parties Have Meaningful Input to the Remedy Selection Process? William J. Friedman* I. THE SARA STANDARD OF REVIEW In October, 1986, Congress enacted the Superfund Amend- ments and Reauthorization Act ("SARA")', substantially revising and expanding the federal hazardous substance site cleanup stat- ute, the Comprehensive Environmental Response, Compensation and Liability Act ("CERCLA," also known as "Superfund")2. The many lawsuits which arose under CERCLA since its adoption in 1980 had highlighted significant ambiguities in the Act and had framed many questions for which no legislative guidance had been provided. SARA (among other purposes) attempted to fill in these gaps and to clarify the workings of the CERCLA pro- gram. Among the more significant matters addressed by SARA is the scope and the standard of judicial review to be employed when parties wish to challenge actions carried out by the U.S. En- vironmental Protection Agency ("EPA") under the CER- CLA/SARA Program.3 SARA attempts to cover this area squarely by specifying both the scope of judicial review and the procedures for establishing the administrative record which is to form the basis for such re- view.4 Nevertheless, as explained below, SARA does not neces- * Mr. Friedman is an attorney with Brach, Eichler, Rosenberg, Silver, Bernstein, Ham- mer & Gladstone, P.C., Roseland, New Jersey, who has been involved in the remedy selec- tion process for numerous Superfund sites. Mr. Friedman wishes to thank William H. Hyatt, Jr. and Betsy L. Weiss, attorneys with Pitney, Hardin, Kipp & Szuch of Morristown, New Jersey for their assistance and advice in preparing this article. 1. SARA, Pub. L. 99-499, 100 Stat. 1613 (1986). 2. 42 U.S.C. §§ 9601 - 9657 (Supp. IV 1986). 3. SARA § ll3(c)(2), 42 U.S.C. § 9613(j) and (k) (Supp. IV 1986) (adding new CER- CLA §§ 1l3(j) and (k)). 4. Id. 187 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 sarily provide parties wishing to challenge EPA action with a meaningful opportunity to do so. This article will assess the potential impact of the judicial re- view standards included in SARA and will attempt to offer advice to current and potential defendants in CERCLA/SARA actions (these parties are commonly known as "potentially responsible parties" or "PRPs") as to how they can maximize their influence on the formulation of the administrative record. The article will also analyze instances where reviewing courts have found agency action to be arbitrary and capricious and will offer suggestions as to the form and content of the future regulations which the EPA must promulgate, pursuant to SARA, setting out a mechanism for establishing an administrative record. Judicial review/administrative record issues under SARA will relate primarily to the selection of a remedial response alternative for individual hazardous substance sites,5 under powers granted by SARA to the President and largely delegated to the EPA pur- suant to Executive Order 12580 ofJanuary 23, 1987.6 The selec- tion of an appropriate remedial action for a site containing hazardous substances is of particular concern to PRPs, because the remedy chosen will significantly impact the total cost of site cleanup, thus, in essence, fixing the EPA's "damages." It is not at all uncommon for a high-cost remedial alternative under consid- eration by the EPA to be several times more expensive than a low- cost alternative. Consequently, by choosing a remedial alterna- tive, the EPA will typically be taking a significant step in determin- ing the potential liability of PRPs. EPA regulations governing the development and selection of a remedial plan by the EPA are set out as a portion of the National Oil and Hazardous Substances Pollution Contingency Plan ("NCP") 7 , that is known as the Hazardous Substances Plan.8 The NCP establishes procedures and standards for responding to re- leases of hazardous substances, pollutants and contaminants. 5. Remedial action selection is mandated by SARA § 121(a), 42 U.S.C. § 9621(a) (Supp. IV 1986) (which creates a new CERCLA § 121(a)), where it is stated that: "[tihe President shall select appropriate remedial action determined to be necessary to be carried out under section 104 or secured under section 106 which are in accordance with this section and, to the extent practicable, the national contingency plan, and which provide for cost- effective response." 6. 52 Fed. Reg. 2,923, (1987). 7. 40 C.F.R. § 300 (1987). 8. Id. 188 Judicial Review Under SARA The procedures include methods for discovering and investigat- ing facilities at which hazardous substances have been disposed of, methods for evaluating and remedying releases or threats of releases from facilities posing a substantial danger to public health or the environment and methods and criteria for determin- ing the appropriate extent of remedial measures.9 Any removal or remedial action chosen or implemented under CERCLA must be consistent with the NCP.' 0 A. Limitation of Review to the Administrative Record As mentioned briefly above, SARA mandates that the EPA es- tablish an administrative record upon which to base its selection of any response action.II SARA limits judicial review of any issues concerning the adequacy of any response action to that adminis- trative record'2 and goes on to state that "[o]therwise applicable principles of administrative law shall govern whether any supple- mental materials may be considered by the court."'3 SARA further provides that the remedial choice is to be upheld unless it can be shown, on the administrative record, that it is ar- bitrary and capricious: "In considering objections raised in any judicial action under this chapter, the court shall uphold the Pres- ident's decision in selecting the response action unless the ob- jecting party can demonstrate, on the administrative record, that the decision was arbitrary and capricious or otherwise not in ac- cordance with law."'4 The EPA is not required to hold an adjudi- catory hearing when it develops an administrative record and selects a response.'5 B. Procedures for Establishing the Administrative Record In establishing the administrative record, the EPA is required to utilize a variety of participation procedures when it is choosing 9. See 40 C.F.R. H§ 300.63 - 300.70 (1987). 10. See CERCLA H§ 107(a)(4)(A) and (B), 42 U.S.C. §§ 9607(a)(4)(A) and (B) (1982 and Supp. IV 1986). 11. SARA § 113(c)(2), 42 U.S.C. § 9613(k) (Supp. IV 1986). 12. Id., 42 U.S.C. § 9613(j)(1). 13. Id. 14. Id., 42 U.S.C. § 9613(j)(2). The Administrative Procedure Act ("APA") similarly limits judicial review of informal agency decisions to the administrative record and pro- vides for overturning agency rulings which are "arbitrary, capricious, an abuse of discre- tion, or otherwise not in accordance with law;" 5 U.S.C. § 706(2)(A) (1982). 15. SARA §I 13(c)(2), 42 U.S.C. § 9613(k)(2)(C) (Supp. IV 1986). 1989]1 189 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 a "remedial" action.16 Remedial actions are defined in Section 101(24) of CERCLA' as actions which are consistent with a per- manent remedy. Thus, they are the major, long-term actions con- ducted to clean up a hazardous substance site. The minimum procedures required for the selection of a reme- dial action include: (i) Notice to potentially affected persons and the public, which shall be accompanied by a brief analysis of the plan and alternative plans that were considered. (ii) A reasonable opportunity to comment and provide in- formation regarding the plan. (iii) An opportunity for a public meeting in the affected area, in accordance with 9617(a)(2) of this title (relating to pub- lic participation). (iv) A response to each of the significant comments, criti- cisms and new data submitted in written or oral presentations. (v) A statement of the bases and purpose of the selected action. 18 SARA is far less specific in setting out required procedures when EPA is choosing a "removal" action.'9 Removal actions (which are usually considered to be in the nature of emergency actions) are defined in Section 101(23) of CERCLA20 as actions which are deemed necessary to respond to the threat of the imme- diate release of hazardous substances into the environment. Ex- amples of removal actions, as set out in CERCLA Section 101(23), are construction of security fencing and the provision of an alternative water supply.21 For removal actions, SARA calls for the EPA to establish procedures for the appropriate participa- tion of interested persons in the development of an administra- tive record, but the Act does not in any way specify what the participation procedures are to be.2 2 Thus, by implication, SARA may be granting EPA the latitude to utilize less extensive partici- pation procedures for removal actions than those set out for re- medial actions. Section 117 of SARA, which adds a new Section 117(d) to CER- CLA, further provides that prior to adoption of any remedial plan 16. Id., 42 U.S.C. §§ 9613(k)(2)(B). 17. 42 U.S.C. § 9601(24) (1982 and Supp. IV 1986). 18. SARA § 13(c)(2), 42 U.S.C. § 9613(k)(2)(B) (Supp. IV 1986). 19. See SARA § 113(c)(2), 42 U.S.C. § 9613(k)(2)(A) (Supp. IV 1986). 20. See 42 U.S.C. § 9601(23) (Supp. IV 1986). 21. See id. 22. SARA § 113(c)(2), 42 U.S.C. § 9613(k)(2)(A) (Supp. IV 1986). 190 Judicial Review Under SARA a notice and brief analysis of the proposed plan must be pub- lished, at a minimum, in a major local newspaper of general circu- lation.23 Furthermore, each item developed, received, published or made available to the public is required to be made available for public inspection and copying at or near the facility in ques- tion.24 Also, if the EPA selects a remedy which is not classified as a "preferred remedy" under CERCLA Section 121(b)(1), 25 then the President must publish an explanation of the failure to select the preferred remedy.26 Close examination of the SARA provisions covering judicial re- view, the establishment of an administrative record and proce- dures for public participation, indicate that these new statutory sections closely track traditional standards delineating the param- eters ofjudicial review as set out and discussed in the APA and in administrative case law. In particular, SARA calls for the same "arbitrary and capricious" standard of review to which agency ac- tion is typically held, and to which CERCLA action was often held prior to SARA.27 Furthermore, the Act sets forth participation procedures which closely follow those that have been enunciated in leading cases applying the arbitrary and capricious standard to agency decision- making.28 Nevertheless, by setting out specific participation pro- cedures and a specific standard of review, and by requiring the promulgation of regulations on the development of an adminis- trative record, SARA allows the EPA and the regulated commu- nity to focus on specific methods of improving EPA decision- making on remedial actions. The challenge will be for the EPA to develop a participation process that will at the same time be com- prehensive, fair, effective, expeditious and protective of PRP rights. 23. 42 U.S.C. § 9617(d) (Supp. IV 1986). 24. In order to carry out all of the public participation procedures outlined above, the President is required to promulgate regulations, which have not yet appeared. SARA § 113(c)(2), 42 U.S.C. §§ 9613(k)(2)(A) and (B). 25. 42 U.S.C. § 9621(b)(1) (Supp. IV 1986) (this term was added by SARA § 121(a), and encompasses remedies where treatment permanently and significantly reduces the volume, toxicity or mobility of hazardous substances, pollutants or contaminants). 26. Id. 27. See, e.g., United States v. Northeastern Pharmaceutical & Chemical Co., Inc., 810 F.2d 726, 748 (8th Cir. 1986) and United States v. Ward, 618 F. Supp. 884, 900 (E.D.N.C. 1986) (in both cases the courts said they would defer to the EPA unless the agency acted arbitrarily and capriciously in determining removal and remedial actions). 28. See Section II.C. of the text. 1989] 191 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 C. Scope of Review Under SARA Several provisions in SARA shed some light on the question of how a reviewing court is to proceed if it should determine that the selection of a response action was either arbitrary and capricious or contained procedural errors. Under new CERCLA Section 113(j)(3),2 9 a reviewing court which holds a response action to be arbitrary and capricious or not in accordance with law is required to decide which, if any, of the selected response costs and the requested damages are in accordance with the NCP and which are not. The court is then to reward only those response costs or damages which are "not inconsistent" with the NCP.3 0 Finally, as a substitute for the portion of the selected response action which it finds to be arbitrary and capricious or not in accordance with law, the court is permitted to fashion "such other relief as is con- sistent with the National Contingency Plan."3 Presumably, this gives the court some discretion and latitude to either fashion ele- ments of a remedy itself or to remand to EPA with orders for the Agency to correct the invalidated portion of the remedial selec- tion process (with the possible result that the remedy could be changed). However, under new CERCLA Section 113(j)(4),32 reviewing courts may disallow costs or damages caused by procedural errors only if the errors were so serious and related to matters of such central relevance to the remedial action that the action would have been significantly changed had such errors not been made. It is somewhat difficult to understand how this latter provision would apply as a practical matter. It appears that a serious proce- dural error by the EPA (e.g., failure to review comments that cor- rectly questioned the efficacy of a chosen remedy) would be likely to necessitate new proceedings to correct the procedural error (by considering the overlooked comments). This could only be accomplished through consideration of supplemental materials by the reviewing court or by calling for the disallowance of partic- ular costs or damages. In contrast, a Senate Committee Report accompanying SARA provides "[ilf major deficiencies are shown to exist in the admin- 29. SARA § 113(c)(2), 42 U.S.C. § 9613(j)(3) (Supp. IV 1986). 30. Id., 42 U.S.C. § 9613(j)(3)(A). 31. Id., 42 U.S.C. § 9613(j)(3)(B). 32. 42 U.S.C. § 9613(j)(4) (Supp. IV 1986). 192 Judicial Review Under SARA istrative record that has been assembled, judicial review of the response in an enforcement or cost recovery action may be de novo, i.e. open to the introduction of evidence by all parties."33 Thus, Congress has made it clear that at the very least, major defi- ciencies in an EPA administrative record (perhaps including sub- stantive errors and "major" procedural errors) will justify de novo judicial review. It remains to be seen how courts will define "ma- jor deficiencies." D. A Potential Limitation on Administrative Record Review Under SARA A recent case appears to have significantly limited the applica- bility of the SARA requirement that judicial review be limited to the administrative record.34 This case distinguishes, between in- stances where the EPA selects a remedy and instances where the EPA asks a court to require other parties to implement a pro- posed remedy. Under CERCLA and SARA, the EPA has available to it several different alternate methods for pursuing site cleanup. It may (1) undertake cleanup itself under CERCLA Section 104(a)3 5 (and later can seek reimbursement for its costs from potentially re- sponsible parties);36 (2) it may administratively order responsible parties to conduct remedial activities under CERCLA Section 106(a);37 or (3) it may seek a court-ordered injunction to have remedial work carried out by responsible parties, also under the provisions of CERCLA Section 106(a).38 In United States v. Hardage,39 the court held that when the EPA chooses the third method listed above, then judicial review would not be limited to the administrative record. The Hardage court noted that SARA limits judicial review to the administrative record only if a response action is taken or ordered by the President.40 The court reasoned that Section 106 actions requesting court-ordered injunctive relief do not deal with re- 33. S. Rep. No. 11, 99th Cong., Ist Sess. 57 (1985). 34. See note 39, infra, and accompanying text. 35. 42 U.S.C. § 9604(a) (Supp. IV 1986). 36. 42 U.S.C. § 9607(a) (procedures and standards for reimbursement actions). 37. 42 U.S.C. § 9606(a) (1982). 38. Id. 39. 663 F. Supp. 1280 (W.D. Okla. 1987). 40. 663 F. Supp. at 1284. 1931989] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 sponse actions taken or ordered by the President.4 1 Furthermore, the court held that because such response actions would not be reviewed on the administrative record, the arbitrary and capri- cious standard would not apply.42 The Hardage court noted that CERCLA Section 106(a) gives courts the power to "grant such relief as the public interest and the equities of the case may require,"4 3 and interpreted this to allow courts deciding injunctive actions under section 106 to util- ize traditional equitable discretion by fashioning remedies with- out being limited in any manner by the record established by the EPA. 4 4 The court found that CERCLA and SARA do not contain any clear and valid legislative command, which is necessary in or- der to divest courts of their traditional equitable jurisdiction.4 5 Hardage went on to interpret SARA as: afford[ing] greater deference to remedies which the govern- ment chooses to implement and finance itself, by restricting re- view thereof to the administrative record, and favor[ing] defendants when the government seeks to force its own remedy upon them, to implement and to finance, by allowing de novo review in such cases. In the former, the government has more incentive accurately to research and to analyze all available remedies, initially, and to propose the most cost effective rem- edy, since the government's own monies will finance the imple- mentation thereof. In the latter, the government has less incentive to inquire into cost-effectiveness issues, since private defendants (as opposed to the government) initially will finance costs of clean-up. De novo review of any recommended remedy which the government seeks to impose on private parties effec- tively ensures its diligence in consideration of all issues, includ- ing cost-effectiveness, since the government knows the higher standard of review of such recommendations (de novo) will be applied by the Court.4 6 The concept of de novo review has been supported by at least one commentator,4 7 although it has been rejected by one court.4 8 If the Section 106 injunctive action distinction is upheld by courts 41. Id. 42. Id. 43. 42 U.S.C. § 9606(a). 44. 663 F. Supp. at 1284-85. 45. Id. 46. Id. 47. See Di Leva, Record Review Under SARA, 14 CHEMICAL WASTE LITIGATION REPORTER 234 (1987). 48. United States v. Seymour Recycling Corp., 679 F. Supp. 859 (S.D. Ind. 1987). 194 Judicial Review Under SARA in the future, it will impose a significant limitation on the applica- bility of the SARA record review provisions. In fact, it could in- duce the EPA to stop bringing Section 106(a) judicial actions for injunctive relief, and instead utilize other enforcement mecha- nisms in its arsenal, by either carrying out cleanup actions itself, or negotiating PRP-led cleanups under administrative consent orders. Another means of obtaining de novo review occurs when CER- CLA actions are brought in suits which contain counts praying for injunctive relief under common law or statutes such as Section 7003 of the Federal Resource Conservation and Recovery Act ("RCRA").49 Under such counts, there would be no provision for record review, and judicial review would be de novo.50 E. The Timing of judicial Review Under SARA An agency decision will be ripe for judicial review only when it is "final" under the APA. 5 1 Several courts, confronted with deter- mining the appropriate time for judicial review of EPA decisions under CERCLA (pre-SARA) determined that such review is not appropriate until the EPA initiates an enforcement action on the remedial or removal plan.52 SARA has now apparently ended the controversy on the proper timing of review by specifically stating in Section 113(c)5 3 that ju- dicial review of EPA remedial or removal plans is not proper until the EPA begins enforcement proceedings. PRPs who disagree with the EPA's final selection of a remedial plan will be forced to wait until a cost recovery action is initiated to seek judicial review. As this may occur years after remedy selection, PRPs may be straight-jacketed while awaiting the initiation of a court action. Therefore, objecting parties should be careful to include as much 49. 42 U.S.C. § 6973 (Supp. IV 1986). 50. See United States v. Hardage, 663 F. Supp. 1280, 1286 (W.D. Okla. 1987) and cases cited therein. 51. D'Imperio v. United States, 575 F. Supp. 248 (D.N.J. 1983). 52. See Aminoil Inc. v. United States Envtl. Protection Agency, 599 F. Supp. 69 (C.D. Calif. 1984); Lone Pine Steering Comm. v. United States Envtl. Protection Agency, 777 F.2d 882 (3d Cir. 1985), cert. denied, 476 U.S. 1115 (1986); Earthline Co. v. Kin-Buc., Inc., 21 E.R.C. 2161 (D.N.J. 1984). As stated in Aminoil, "Allowing an alleged responsible party to challenge the merits of the § 106(a) administrative order prior to an enforcement or recovery action would handcuff the . . . [EPA] by delaying effective responses to emer- gency situations." 599 F. Supp. at 71. 53. 42 U.S.C. § 9613(c) (Supp. IV 1986). 1989] 195 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 pertinent information as possible in comments and other written submissions during the administrative record-making process, so that the record will fully protect their positions regardless of how long the time lapse may be between remedy selection and enforcement. F. Should the SARA Provisions Limiting fudicial Review Be Applied Retroactively? Several recent cases have considered whether SARA's record review restrictions should be applied retroactively to cases that had already been initiated - and in some instances had already advanced through numerous procedural steps - at the time of SARA's enactment. The decisions have been split, with three courts holding that SARA provisions limiting judicial review to the record would not be applied retroactively.5 4 However, in United States v. Nicolet, Inc.5 5 the court characterized the SARA re- view provision as a procedural change not affecting substantive or vested rights and found retroactive application to be proper. Other courts have held that Congress plainly intended that the judicial review provisions under CERCLA Section 113(j),5 6 as ad- ded by SARA, apply to ongoing cases.57 In part, the decisions on retroactivity have been based upon the equitable consideration of whether it would be fair to use SARA's record review mechanism in ongoing litigation where parties had been operating under the assumption that de novo review would occur.5 8 Intertwined with this question is the issue of whether CERCLA has always implicitly limited review to the administra- tive record under an arbitrary and capricious standard, or whether SARA substantially changed the rules for judicial review. The EPA and the Department of Justice have taken the position that SARA's judicial review provisions merely confirm previous 54. Hardage, 663 F. Supp. at 1283; Conservation Chemical Co., 661 F. Supp. at 1426- 31; United States v. Ottati & Goss, No. 82-225-L (D.N.H. Nov. 14, 1986), petitionfor manda- mus dism'd, No. 87-1003 (1st Cir. Feb. 4, 1987) (as cited in United States v. Conservation Chemical Co., 661 F. Supp. at 1426 n. 19 and discussed at length in Standard of Review, 13 CHEMICAL WASTE LITIGATION REPORTER 381 (1987)). 55. Slip op., No. 85-3060 (E.D. Pa. May 12, 1987). 56. 42 U.S.C. § 9613(j) (Supp. IV 1986). 57. See United States v. Rohm and Haas Co., Inc., 669 F. Supp. 672 (D.N.J. 1987); United States v. Seymour Recycling Corp., 679 F. Supp. 859 (S.D. Ind. 1987). 58. See, e.g., Seymour Recycling Corp., 679 F. Supp. at 862-65; Conservation Chemical Co., 661 F. Supp. at 1431. 196 Judicial Review Under SARA case law determinations on the scope of review that existed under CERCLA. The Senate Report also takes this position.59 II. THE ARBITRARY AND CAPRICIOUS STANDARD A. Its Elusive Meaning SARA's use of the arbitrary and capricious standard holds EPA action to a test that has been cited in untold judicial opinions. Nevertheless, a precise meaning of the arbitrary and capricious standard remains elusive and difficult to set out with specificity. As Professor Davis has explained, this state of affairs exists be- cause the Supreme Court, after writing a multitude of opinions which discuss the substantial evidence, clearly erroneous and clear error of judgment standards of review, as well as the arbi- trary and capricious standard, has been unable to delineate which standard is the most stringent.60 Ostensibly, the authoritative Supreme Court language on the arbitrary and capricious standard appears in Citizens to Preserve Overton Park v. Volpe,6' which states that in order to find agency action to be arbitrary and capricious: [T]he court must consider whether the decision was based on a consideration of the relevant factors and whether there has been a clear error of judgment. Although this inquiry into the facts is to be searching and careful, the ultimate standard of review is a narrow one. The court is not empowered to substi- tute its judgment for that of the agency. (citations omitted) However, the use by the Supreme Court of the term "a clear error ofjudgment" in attempting to establish a parameter for arbitrary and capricious conduct has created particular confusion amongst scholars attempting to practically apply the standard, because it blurs the distinction between the "arbitrary and capricious" and the "clearly erroneous" standards. Furthermore, a series of Supreme Court decisions over the past twenty years has failed to 59. See Senate Report, supra, note 32, at 57, which notes that SARA "clarifies and con- firms that judicial review of a response action is limited to the administrative record and that the action shall be upheld . .. unless the action was arbitrary and capricious or other- wise not in accordance with law." 60. 5 DAVIS, ADMINISTRATIVE LAw TREATISE § 29:7 (2d ed. 1984). 61. 401 U.S. 402, 416 (1971) (citations omitted). 1989] 197 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 clarify the comparative breadth of a court's review under the above-referenced or the "substantial evidence" standards.62 The confusion has resulted in Professor Davis remarking on the distinction between the "arbitrary and capricious" and "substan- tial evidence" standards that, "[tlhe law is, then, all at one time, that the one test requires more than the other, that the other re- quires more than the one, and that the difference between the two tests is largely semantic!"63 As a result, he notes in frustration that, "[a]ny conclusion about what the law is must take account of the fundamental and long-continuing inconsistency."6 Davis is forced to conclude that, rather than applying a particular stan- dard, courts as a practical matter "will go on substituting judg- ment on the kind of questions of law that are within their special competence and using a reasonableness test on other questions."65 Although the real meaning of the arbitrary and capricious stan- dard remains elusive, one commentator66 has attempted to inte- grate the many different legal rationales used by various courts into a general proposition that is somewhat different, but more comprehensive than any of the Supreme Court's "tests". That proposition states: [I]t is arbitrary for an agency to establish an internal organiza- tional decisionmaking process that creates an undue risk of a capricious result. An agency that blurs the issues in its initial notice, truncates opportunities for participation, fails to iden- tify or explain intermediate choices, or fails to respond to criti- cism is more likely to reach an arbitrary result than is one that approaches the issues in a more thoughtful way. A court is en- titled to consider these inadequacies in its review.6 7 62. In particular, the Overton Park decision has raised questions as to whether the clearly erroneous test is equivalent to the examination by a court of whether there has been a "clear error of judgment." 401 U.S. at 416. If these two concepts are the same, then different Supreme Court decisions have made the "clearly erroneous" test afford both more and less restricted review than the "substantial evidence" test, and have also cast doubt as to the relative stringency of the "arbitrary and capricious" test. See Abbott Labo- ratories v. Gardner, 387 U.S. 136, 143 (1967); Camp v. Pitts, 411 U.S. 138, 141-142 (1973); Bowman Transportation v. Arkansas-Best Freight System, Inc., 419 U.S. 281, 284 (1974). 63. 5 DAVIS, supra, note 60, § 29:7, at 359. 64. Id. 65. Id. 66. DeLong, Informal Rulemaking and the Integration of Law and Policy, 65 VA. L. REV. 257, 306 (1979). 67. Id. at 306. 198 Judicial Review Under SARA Whatever the arbitrary and capricious standard may really mean when put to practical use by a reviewing court, some of the basic actions required of an administrative agency under the stan- dard appear to be well established and non-controversial. These actions, comprising both procedural requirements calling for public participation by interested parties and substantive require- ments calling for an agency to demonstrate that it has engaged in reasoned decision-making, are explored below with respect to ju- dicial review under SARA. B. Procedural Requirements Agencies operating under the arbitrary and capricious standard are held to a procedural requirement mandating public participa- tion by interested parties in the decision-making process.68 Nec- essary components of this procedural requirement, as stated in the APA, are that proper notice of proposed or contemplated agency action be provided to interested parties either through ac- tual notice or Federal Register publication69 and that such parties be given an opportunity in some manner to comment on the pro- posed or contemplated action through submission of written data, views or arguments.70 SARA sets standards for public participation by adding CER- CLA Section 1 13(k)(2)(B)(i), (ii), and (iii),7' providing that poten- tially-affected persons in the public are to: (1) receive notice, accompanied by a brief analysis of the EPA remedial plan and the alternative plans that were considered; (2) have a reasonable op- portunity to comment and provide information regarding the plan; and (3) have an opportunity for a public meeting in the af- fected area. Of course, the opportunity provided to the public by SARA to offer comments places a corresponding obligation on objectors to proposed EPA actions to raise their objections through the com- ment process, so as to preserve them for a later judicial challenge. As one commentator has noted, "an arguably troubling critique of an agency action should be given less weight if the challenger 68. See Buckeye Power, Inc. v. United States Envtl. Protection Agency, 481 F.2d 162 (6th Cir. 1973), cert. denied, 425 U.S. 934 (1975). 69. 5 U.S.C. § 553(b) (1982). 70. Id. at § 553(c). 71. 42 U.S.C. § 9613(k)(2)(B)(i)-(iii) (Supp. IV 1986). 1989] 199 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 did not clearly articulate it or support it at the agency level."72 Or, as the Eleventh Circuit stated in Taft v. Alabama By-Products Corp. , "an appellate court should not overrule an administrative decision unless the administrative body erred against objections presented to it." Thus, objectors to EPA remedial plans must be diligent in placing their objections on the record if they wish to have any meaningful opportunity to later mount an argument that the agency acted in an arbitrary and capricious manner. 1. The Use of Trial-Type Hearings SARA calls for a public meeting in the vicinity of a site for which a remedial plan is to be chosen and states that this meeting shall be recorded and the transcript made available to the public (CERCLA § 117(a)(2)).74 However, SARA specifically states that an adjudicatory hearing is not to be held as part of the remedy selection process.75 Although the EPA could presumably include trial-type provisions in the regulations which it must promulgate covering public participation, such a course of action is unlikely, because the EPA has never permitted trial-type hearings to be uti- lized in the selection process for the remedies which it has chosen since CERCLA's inception in 1980. Thus, as will be seen below, there is a substantial question as to whether SARA's public partic- ipation procedures provide objecting parties, particularly ob- jecting PRPs, with a forum which will allow even diligent parties to present meaningful evidence challenging EPA's remedy selec- tion process. Only through the use of such a forum will objecting parties be able to take full advantage of subsequent arbitrary and capricious judicial review. a. Does Remedy Selection Involve Adjudicative Facts? In general, where agencies have provided public participation through notice and an opportunity to comment on the record, but have not provided an adjudicatory hearing, courts have re- fused to mandate that a trial-type hearing be held. If the adminis- trative record is sufficiently developed, the courts have not found 72. Levin, Federal Scope-of-Review Standards: A Preliminary Restatement, 37 AD. L. REV. 95, 118 (1985). 73. 733 F.2d 1518, 1523 (11th Cir. 1984) (citations omitted). 74. SARA § 117, 42 U.S.C. § 9617 (Supp. IV 1986). 75. SARA § 113(c)(2), 42 U.S.C. § 9613(k)(2)(c) (Supp. IV 1986). 200 Judicial Review Under SARA the agency action to be arbitrary and capricious.76 It has been noted that "[t]he right to a full trial-type hearing in administrative proceedings is generally limited to the situation where adjudica- tory facts-that is, facts pertaining to a particular party-are in issue."" Thus, the question of whether an adjudicatory hearing is re- quired in any particular administrative context will often hinge on whether courts consider the facts in question to be adjudicatory in nature. Adjudicative facts have been defined by Professor Davis as "facts pertaining to the parties and their businesses and activi- ties. Adjudicative facts usually answer the questions of who did what, where, when, how, why, with what motive or intent; adjudi- cative facts are roughly the kind of facts that go to a jury in a jury case."78 In contrast, legislative facts, the determination of which will not normally require a trial-type hearing, "do not usually con- cern the immediate parties, but are the general facts which help the tribunal decide questions of law and policy and discretion."79 As Professor Davis has explained, "[a]n agency should engage in formal factfinding when, regardless of the role it is playing, the need for factual accuracy outweighs other considerations and trial-type procedures will effectively decrease uncertainty."s0 For example, in Alaska Airlines, Inc. v. Civil Aeronautics Board,8 ' the court held that the Civil Aeronautics Board's orders which precluded the plaintiff from operating intra-Alaska charter flights involved adjudicatory facts which entitled plaintiff to a hearing. Although the orders affected a class of carriers, the court noted that "the concern of the Board with . . . [plaintiff's] particular operation rather than with other similar past operations or condi- tions in other air markets indicates that the Board was dealing ... individually and not in a class context."82 At first glance, the selection of a remedial action at a hazardous site might appear to involve non-adjudicatory facts, because the 76. See. e.g., Camp v. Pitts, 411 U.S. 138, 142 (1973) (inadequate agency explanation of its decision was not enough to merit de novo review). 77. Project, Federal Administrative Law Developments - 1971, 1972 DUKE L.J. 115, 171 n.24 (1972) (citations omitted). 78. 2 DAVIS, ADMINISTRATIVE LAw TREATISE § 12:3 (2d ed. 1984). 79. Id. 80. McGarity, Substantive and Procedural Discretion in Administrative Resolution of Science Policy Questions: Regulating Carcinogens in EPA and OSHA, 67 GEo. L.J. 729, 775 (1979). 81. 545 F.2d 194, 200-201 (D.C. Cir. 1976). 82. Id. at 201. 1989] 201 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 remedial selection process does not typically address the rights of specific individual parties. However, a more reasonable view is that the selection of a remedy and determinations of the attend- ant type, magnitude and cost of that remedy will, as a practical matter, ultimately have a substantial bearing on the magnitude of the liabilities of all parties who are held to be liable for remedial costs under CERCLA. Thus, remedy selection, because it in- volves particularized site-specific facts rather than generally-ap- plicable policy determinations, should be considered an adjudication.83 Several commentators on the subject have agreed that CERCLA remedy selection is adjudicatory in nature.84 b. When Does Due Process Require an Adjudicatory Hearing? A fundamental question that courts must ultimately answer is whether CERCLA public participation and record-making proce- dures, as fleshed out in forthcoming EPA regulations, will provide an opportunity for concerned parties to be heard "at a meaning- ful time and in a meaningful manner" in compliance with due process guarantees.85 Due process will be afforded only if PRPs are given "an effective chance to respond to crucial facts" in an appropriate forum.8 6 The Hardage decision extensively analyzed this issue and, under the facts of that case, held that due process could only be satisfied through de novo review of an already completed SARA remedial selection process.8 7 The Hardage court reached its decision through analysis of the three-pronged test set out by the Supreme Court in Mathews v. Eldridge88 for determining whether adminis- trative procedures afford due process to affected parties. These three factors are: (1) the private interest at stake; (2) the risk of erroneous deprivation of that interest through the use of the agency procedures, and the probable.value of additional or sub- stitute safeguards; and (3) the governmental interest, when con- sideration is given to the burdens entailed in additional 83. See United States v. Florida East Coast Ry. Co., 410 U.S. 224, 244-45 (1973) (where the Supreme Court set out the distinction between the promulgation of policy-type rules and proceedings designed to adjudicate disputed facts in particular cases). 84. Di Leva, supra note 47, at 246; Bode and Leone, 11 CHEMICAL WASTE LITIGATION REPORTER 654, 656 (1986) (footnote omitted). 85. Armstrong v. Manzo, 380 U.S. 544, 552 (1965). 86. National Org. for Women v. Social Security Admin., 736 F.2d 727, 739 (D.C. Cir. 1984). 87. 663 F. Supp. at 1288-90. 88. 424 U.S. 319 (1976). 202 Judicial Review Under SARA procedural requirements.89 In particular, the Hardage court was concerned that PRPs were given only 45 days by the EPA to inves- tigate conditions at the Hardage site and develop data that might disagree with the conclusions in EPA's feasibility study.90 The Hardage PRPs stated that they expended almost six months and more than $1.5 million to investigate and gather sufficient data to support their charge of alleged shortcomings in the EPA's feasi- bility study.9 ' Furthermore, the Court noted the strong private interest of the PRPs (the first factor in the Mathews test) because defendants found liable would be ordered to implement a remedy which might exceed $70 million in costs.9 2 .The court went on to em- phasize that the use of non-adjudicatory procedures by the EPA created a high risk that PRP interests would be erroneously de- rived (under the second prong of the Mathews test).93 Finally, the court found that the "most flagrant denial of due process" oc- curred because of the EPA's failure to establish an independent staff or tribunal to evaluate issues raised by the PRP-defendants. As a result, the same EPA staff persons who performed the reme- dial investigation, considered the options, formed the feasibility study and selected the preferred remedy were also given respon- sibility for evaluating the comments and issues raised by other parties, including the PRP-defendants. The court remarked that due process dictated separation of the prosecutorial function from the decision-making function.94 In contrast, the court in Rohm and Haas95 applied the same Ma- thews v. Eldridge analysis to the informal record review procedures mandated by SARA and reached an opposite conclusion. That court was swayed by what it characterized as "an overwhelming countervailing public interest, as evinced in CERCLA, in effecting the expeditious clean-up of potentially health and life threatening 89. Id. at 335 (citation omitted). 90. 663 F. Supp. at 1289. 91. Id. 92. Id. at 1290. 93. Id. 94. Id. (citing Morrissey v. Brewer, 408 U.S. 471, 485-86 (1972)). See also Porter v. Califano, 592 F.2d 770 (5th Cir. 1979), which criticized agency procedures where the same person whose actions were being challenged was allowed to act in a reviewing capacity when the challenge was raised. 95. 669 F. Supp. 672 (D.N.J. 1987). 1989] 203 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 hazardous waste sites."96 The court went on to note that "[t]he imposition of long, drawn-out, and costly trial-type procedures, either at the agency level or in a de novo proceeding in district court, could greatly hinder this effort."97 The Rohm and Haas decision is also grounded upon the assump- tion that a non-adjudicatory remedy selection process reflecting "the contemporaneous analyses and criticisms of all interested parties,"98 but without live testimony or cross-examination of ex- pert witnesses would reveal all relevant technical information. Furthermore, the Rohm and Haas court held that "an administra- tive record built on such an exchange of opinions and comments by experts and informed citizens and containing an explanation by the agency of its reasons for accepting or rejecting the various proposals . ... provides a comprehensive framework from which the Court can scrutinize the agency's action."99 However, in upholding the non-adjudicatory hearing proce- dures set out in SARA, the Rohm and Haas court betrays a lack of experience and sensitivity to the practical intricacies of the EPA remedy selection process for so-called "Superfund sites." Under the mandates of the NCP, EPA remedial selection decisions can occur only after technically sophisticated and often voluminous remedial investigations and feasibility studies ("RI/FSs") have been completed.'0 0 Remedial investigations, designed to accu- rately characterize the type, extent and movement of pollution at a site and the human and environmental receptors that may be impacted by the pollution, will often draw conclusions about hid- den and controversial phenomena such as the speed and direc- tion of contaminated groundwater flow or the permeability of underground rock strata. The conclusions in a remedial investi- gation will often be based upon various educated technical as- sumptions or will draw upon theoretical environmental models devised by experts. 96. Id. at 680. 97. Id. Similar reasoning under the Mathews standard was utilized to reach a similar conclusion in Seymour Recycling Corp., 679 F. Supp. at 864-65, where the court said that defendants would not be erroneously deprived of their interest in the remedy because they would be afforded judicial review of the EPA's remedy selection on the basis of the admin- istrative record. 98. 669 F. Supp. at 681. 99. Id. 100. See 40 C.F.R. § 300.68(d) (1987). 204 Judicial Review Under SARA Feasibility studies, which are based upon the results of remedial investigations, evaluate a range of remedial alternatives and make recommendations on a preferred alternative based on factors such as cost, reliability and efficacy of the remedies.'0' Because the hazardous substance site remedy field is still new and devel- oping, remedies that are recommended frequently have only a short history of prior use. Thus, their efficacy for a particular site application may be subject to substantial disagreement among experts. Consequently, RI/FS and remedy selection processes have fre- quently been marked by disputes among experts as to the reliabil- ity of investigative data obtained in remedial investigations, the correctness of environmental models utilized by experts to pre- dict pollution movement and, most importantly, the appropriate- ness of particular proposed remedies. Despite the controversial nature of the process, any particular expert who has conducted all or a part of an RI/FS will in all probability be able to produce an extensive report with reams of data and numerous technical footnotes supporting his findings, models or remedial rec- ommendations. Where the EPA relies upon its own experts to conduct an RI/FS and recommend a remedial alternative, it will similarly have extensive backup information supporting the conclusions of its expert. Even if objectors comment on the evidence presented by the EPA expert and present alternative reports and studies supporting an alternative remedy, in many (or most) instances, the objectors' comments and alternative remedial plans will still not be able to effectively question the assumptions and bases for the conclusions of the EPA experts. Such questioning of complex technical information can only be effectively carried out if EPA experts are subject to cross-examination and other detailed fact- finding, such as the taking of depositions. Only then will there be an in-depth exploration of the value or accuracy of scientific mod- els -and the efficacy of little-tested remedies. Absent a procedure for cross-examination, the EPA will simply have before it two or more alternative remedial plans, including its own, each of which will typically be supported by a detailed report and/or expert testimony. If it chooses the plan recom- mended by its own experts, which it presumably will in the vast 101. See 40 C.F.R. §§ 300.68(f)-(h) (1987). 1989] 205 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 majority of cases, then a reviewing court applying the arbitrary and capricious standard may have very little basis for overturning the agency remedial choice. Nevertheless, that remedial choice may very well be excessively expensive compared to other effica- cious remedies, or the remedial choice may simply not work. In effect, without cross-examination at a hearing on its choice of a remedy, EPA will go a long way toward insulating its remedial decision from meaningful further review. Two further questions inevitably follow: (a) can cross-examination be a part of the rem- edy-selection process without unduly delaying or interfering with that process and (b) does a remedy selection process without cross-examination afford PRPs and other interested parties due process of law? 2. The Value of Cross-Examination in the Remedy Selection Process The past experience of administrative decision-makers indi- cates that permitting cross-examination in an EPA hearing on re- medial choices may very well not unduly burden the remedy selection process. In his perceptive article, Rulemaking and the Myth of Cross-Examination,1 0 2 William D. Dixon argues persuasively that cross-examination in administrative proceedings can be an invaluable aid to decision-makers and can in fact be conducted under controlled conditions which do not significantly add to the costs or length of hearings. As an example, he cites the Magnuson-Moss Warranty-Federal Trade Commission Improve- ment Act of 1974,103 which provides that upon a determination by the Federal Trade Commission ("FTC") that there are dis- puted issues of material fact in a rulemaking proceeding, parties in a hearing are entitled to conduct such cross-examination as the FTC determines to be appropriate and required for a full and true disclosure with respect to the disputed issues. One of the powers of the FTC under the Act is to group parties with the same or similar interests and to require them to select a single representative for purposes of cross-examination. If they cannot agree on such a representative, the FTC can make the selection for them. 102. 34 AD. L. REv. 389 (1982). 103. 15 U.S.C. § 57a(c) (1982). 206 Judicial Review Under SARA In order to prevent cross-examination in FTC proceedings from dragging on interminably, FTC presiding officers under the Magnuson-Moss Act developed a concept known as freedom-for- time where group representatives were allowed to cross-examine witnesses without specific restrictions as to the issues they could cover; however, the cross-examination was subject to strict time limits for each group. "In that way group representatives were afforded the freedom to develop their questioning along lines suitable to the witness on the stand and the presiding officer was able to keep the hearings on schedule."0 4 The result was that "hearings were taking very little longer than previous hearings without cross-examination."0 5 Dixon goes on to note that "[w]hile a great deal of the questioning was pure waste, seldom could that judgment be made until after the fact, and much infor- mation and clarification which would otherwise have escaped found its way into the record." 0 6 He concludes that complaints were rarely registered about the conduct of the hearings. Dixon cites many positive contributions arising from cross-ex- amination. He notes that credibility attacks sometimes show "the extent to which a witness ha[s] based his opinion on policy or value judgments rather than on facts" and also sometimes estab- lish that "witness[es] made sweeping statements beyond their area of expertise.0 7 He goes on to cite findings of studies stating that cross-examination has a unique ability to expose method- ological shortcomings or problems of interpretation,0 8 and that direct questioning of experts on assumptions or test procedures is more efficient than filing another expert's critique.'09 He cites numerous examples of the positive effects which can arise from cross-examination." 0 104. Dixon, supra note 102, at 400. 105. Id. 106. Id. 107. Id. at 407. 108. Id. at 407-408. 109. Id. at 415. 110. E.g., Dixon references a proceeding on the appropriateness of a particular site for a natural gas terminal, where a wiiness had maintained that there was an active seismic fault underlying the site. Cross-examination exposed both that the witness had limited education and experience in seismology and that his opinion was formed solely on the basis of an airplane flight over the site on a cloudy day. Id. at 436 n.146 (citing Pierce, The Choice Between Adjudication and Rulemaking for Formulating and Implementing Energy Policy, 31 HASTINGs L.J. 1, 49-50 (1979)). 1989] 207 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 Another commentator analyzes the pros and cons of utilizing cross-examination of scientific experts by noting: Unfortunately, cross-examination usually reveals only the depth of the disagreement among the experts; it rarely reveals any basis for choosing one expert's interpretation of the data over another's. Cross-examination, however, can be of great value in probing inferences. Direct oral testimony and cross- examination provide an opportunity to examine the assump- tions upon which an expert bases his interpretation. The as- sumptions an expert uses can be result-oriented and highly policy-dominated. The validity of the scientist's assumptions will obviously affect the accuracy of his inferences. The deci- sionmaker should be aware of those assumptions to be able to draw his own conclusions." In short, allowing cross-examination will not always result in a better, more useful record. In a highly technical area such as remedy selection for hazardous substance sites, however, the EPA will typically rely upon experts in drawing its conclusions, and there is substantial room for experts to disagree as to acceptable assumptions, methods and technologies. Thus, cross-examina- tion becomes virtually the only reliable method of probing into the minds of experts so as to explore the substantive basis for remedial decisions. Undoubtedly, without cross-examination, a proper remedy will be chosen in some instances. However, the use of cross-examination will greatly increase the chances that faulty reasoning of experts can be exposed in those instances where unworkable or non-cost-efficient remedies are recom- mended to the EPA. If steps are taken to carefully circumscribe the extent of cross-examination, then the relatively minor amount of additional time involved will certainly be worth the benefit of better remedies. 3. Due Process Requirements Call for Cross-Examination in the Remedy Selection Process As discussed briefly above, any determination as to whether an administrative process affords due process must be made through analysis of the three factors set out by the Supreme Court in Ma- thews v. Eldridge."2 If, as explained above, limited cross-examina- tion in a controlled manner can become a non-burdensome part of the remedy selection process, then each of the three Mathews 111. McGarity, supra note 80, at 777-78. 112. 424 U.S. at 335 (1976). 208 Judicial Review Under SARA factors would appear to heavily tilt the constitutional balance to- ward requiring the use of cross-examination. Because the cost difference between different remedies for the same hazardous substance site can often run into the tens of mil- lions of dollars, the private interests in choosing the least expen- sive efficient remedy will be very high, thus satisfying the first Mathews factor. Second, because expert theories on the move- ment of contaminants and experts' recommendations on the use of often newly-devised and untested remedies are frequently sub- jects of great controversy in the technical community, the risk of erroneous deprivation of private interests can be very high, unless carefully chosen procedures are utilized to test the assumptions and the methods of EPA's experts. Thus, the second Mathews fac- tor is similarly fulfilled when remedy selection is taking place. The third and final Mathews factor, examining the governmental interest under the burdens that additional procedural require- ments would entail, also weighs in favor of cross-examination. Although there is a strong governmental interest in expeditious remediation of hazardous waste sites, a circumscribed, controlled cross-examination and discovery process would not trample upon that governmental interest. In fact, as long as cross-examination and other discovery procedures can be handled in a manner that would not significantly slow down the remedy selection process, there would not appear to be any policy reason for not undertak- ing such measures. Furthermore, Mathews v. Eldridge analysis in- structs that the use of such procedural steps is constitutionally required. Cross-examination of experts who present complex technical testimony in administrative proceedings is of particular impor- tance where courts will later be asked to study the record and re- view the agency's decision. Without cross-examination, most reviewing judges simply will not have a sufficient technical back- ground to'conduct a meaningful review of the expert testimony offered. Case law illustrates this point. In Bunker Hill Co. v. Environmental Protection Agency,' is the court ordered that an EPA determination on the allowable level of air emissions from a smelter should be remanded to the Agency and that cross-examination should be utilized on remand, because the record then existing provided an insufficient basis for judicial re- 113. 572 F.2d 1286 (9th Cir. 1977). 1989]1 209 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 view. The court noted that the complexity of the question before it called for cross-examination in order to crystalize varying con- tentions of experts and thereby to facilitate judicial review.14 It noted that when a " 'proceeding involves specific issues of critical importance that cannot be adequately ventilated' by normal pro- cedures," the use of cross-examination by an administrative agency is proper."5 Similarly, in another EPA emissions standards case, International Harvester Co. v. Ruckeshaus,"6 the court noted that cross-examina- tion possesses unique potential as an "engine of truth,"" 7 which can serve a particularly useful purpose "on critical points where the general procedure proved inadequate to probe 'soft' and sen- sitive subjects and witnesses.""8 In addition, at least two com- mentators have noted that the absence of cross-examination contributes to making EPA fact-finding in CERCLA cases inade- quate, and by implication short of due process, because it does not afford PRPs a meaningful opportunity to participate."9 4. The Proper Forum for Adjudication If, for the reasons set forth previously, it is held that adjudica- tion with cross-examination is necessary in order to select an EPA remedial response, a determination must be made as to the proper forum in which the adjudication should occur. At least two of the courts considering the retroactivity of SARA judicial review provisions determined that the adjudicatory record should be made on a de novo basis in district court, rather than in an EPA proceeding.120 In fact, the court in Hardage held that remedy selection under CERCLA and SARA constitutes "informal agency action," thus placing it outside of the arbitrary and capricious review standard of the APA and mandating traditional review on a de novo basis. The Hardage court reasoned that since SARA expressly rejects the 114. Id. at 1305. 115. Id. (quoting Thompson v. Washington, 497 F.2d 626, 641 n.48 (1973)). 116. 478 F.2d 615 (D.C. Cir. 1973). 117. Id. at 631. 118. Id. 119. Bode and Leone, supra note 84, at 656-58. 120. See Hardage, 663 F. Supp. at 1280, Ottati & Goss, No. 82-225-L, slip op. (D.N.H. Nov. 14, 1986) (as discussed in Standard of Review, supra note 55). Cf Conservation Chemi- cal Co., 661 F. Supp. at 1431 (de novo review issue deemed beyond the scope of issues noticed to the court). 210 Judicial Review Under SARA concept of a formal adjudicatory hearing, the "substantial evi- dence" standard of review for agency actions covered by the APA also is inapplicable, noting that "informal adjudication is non- APA adjudication in which ad hoc agency rules apply and due process is to be tested by the courts." 2 1 The Hardage view is sup- ported by one commentator, who additionally notes that the right to judicial review under the APA does not address suits brought for money damages.122 In contrast, the court in Rohm and Haas decided that in light of the unique circumstances present in that case, and what it called "inadequacies in an administrative record [which would] frustrate proper judicial review of the agency's actions under the arbitrary and capricious standard," the proper course of action was to re- mand the case to the EPA, to enable development of a record with the full panoply of procedures afforded by SARA.' 23 5. The Restrictive Effect of The Vermont Yankee Decision Despite the analysis above providing substantial due process reasons calling for cross-examination and detailed discovery as part of the remedy selection process, the chance of having a court order that cross-examination should be required where an agency has chosen minimal public participation procedures was greatly reduced by the Supreme Court's 1978 decision in Vermont Yankee Nuclear Power Corp. 124 Vermont Yankee involved objections to both rulemaking and licensing procedures for nuclear power plants. One petitioner had been granted a license to operate a nuclear power plant and the other was issued a permit to construct a plant. Following each of these grants, the Atomic Energy Com- mission adopted rules which mandated the consideration of fac- tors not dealt with in the earlier proceedings. Intervenors then sought court orders to require the Commission to reconsider both grants on the basis of the newly promulgated rules, at the same time that they challenged the rule-making procedures involved. After the circuit court remanded the cases to the Commission, the Supreme Court reversed, holding both that rule-making pro- cedure was within the agency's discretion, and that the appeals 121. 663 F. Supp. at 1287. 122. Di Leva, supra note 47, at 250 (footnote omitted). 123. 669 F. Supp. at 683-84. 124. 435 U.S. 519 (1978). 1989) 211 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 court should have limited its review of the merits to the adminis- trative record, instead of imposing "upon the agency its own no- tion of which procedures are 'best' or most likely to further some vague, undefined public good." 25 The Supreme Court stated that "Congress intended that the discretion of the agencies and not that of the courts be exercised in determining when extra pro- cedural devices should be employed." 26 Procedural discretion should be left to the agency, the Court reasoned, because if re- viewing courts are allowed to impose additional procedures on agencies, the agencies would require full formal procedures, thereby losing all of the advantages inherent in informal proceedings.12 7 Under the Vermont Yankee reasoning, agencies are free to grant additional procedural rights in the exercise of their discretion, but reviewing courts are generally not free to impose them if the agencies have not chosen to grant them. This is not to say neces- sarily that there are no circumstances which would ever justify a court in overturning agency action because of failure to employ procedures beyond those required by statute. In fact, the Supreme Court noted in passing that the rule might not apply if constitutional constraints existed or if there were "extremely compelling circumstances."1 2 8 But such circumstances, if they exist, are apparently rare.129 Despite the limitations on judicial review imposed by Vermont Yankee, the decision obviously could not roll back the basic proce- dural rights contained in the APA. Thus, Vermont Yankee contains a proscription only against courts imposing additional procedural devices on administrative agencies beyond those required by the APA or by other due process analyses. Certainly, for instance, the Mathews v. Eldridge test as to whether particular administrative 125. 435 U.S. at 549 (footnote omitted). 126. Id. at 546. 127. Id. at 547. 128. Id. at 543. 129. Id. at 524. One commentator has described the method that courts may use under Vermont Yankee principles to overturn agency action in the following way: Vermont Yankee reaffirms the court's power to determine that the overall processes fol- lowed were arbitrary in the sense that they produced an arbitrary result and that something more is needed. There are often many possible paths through a rulemak- ing jungle. After Vermont Yankee, a court cannot tell an agency which path to choose; the court can, however, tell the agency that it has not found any of the proper routes and require the agency to make a good map. DeLong, supra note 66, at 316. 212 Judicial Review Under SARA procedures meet the demands of due process is in no way affected by Vermont Yankee. While Vermont Yankee has not yet been reconsidered by the Supreme Court, it "has been criticized for its chilling effects on judicial development of new review techniques for administrative law." 3 0 The decision is contrary to many earlier decisions which questioned the adequacy of notice and comment procedure as a means to produce a sufficient record for judicial review. Another commentator notes that the Vermont Yankee decision in- structs courts to restrict their review of informal agency decisions on the basis of the administrative record, but recognizes that sometimes remand may be necessary because the APA does not mandate appropriate procedures for all situations.'3 1 As a practi- cal matter, the commentator notes that courts tend to exercise more exacting scrutiny of the factual and analytical bases of deci- sions, because of concern about agency discretion. 132 6. Elements to Place on the Record Under SARA The fact that, in this era of Vermont Yankee, the EPA is unlikely to provide for (or to be required to provide for) a formal adjudica- tive hearing for the selection of a remedial alternative highlights the necessity for parties objecting to a proposed EPA remedy to place as much relevant information as possible on the record that is established as part of the EPA's public participation proce- dures. Although objecting parties who cannot cross-examine EPA witnesses will face an uphill battle in attempting to challenge remedies based on technical work supervised by EPA experts, such objecting parties will need to thoroughly document their po- sition on the record if they are to stand any chance of either per- suading the EPA of their position or causing a reviewing court to substantively question an EPA decision on remedial selection. Of course, the best way for objecting PRPs to exercise control over the remedial selection process and to assure that their ex- perts' opinions are heard is for PRPs to conduct an RI/FS for a CERCLA site themselves, under the EPA's oversight, and pursu- 130. Marcel, The Role of the Courts in a Legislative and Administrative Legal System-The Use of Hard Look Review in Federal Environmental Litigation, 62 OR. L. REV. 403, 436 (1983) (foot- note omitted). 131. Stewart, Vermont Yankee and The Evolution ofAdministrative Procedure, 91 HARv. L. REV. 1805, 1816 (1978). 132. Id. at 1811 (footnote omitted). 1989] 213 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 ant to the provisions of an administrative consent order. This is often permitted by EPA.' 33 With a PRP-conducted RI/FS, the PRPs' experts will be able to choose the various remedial alterna- tives that are included in the feasibility study, and PRPs will thus insure that their experts' opinions are included in the official rec- ord. However, for a variety of reasons, the EPA may either not give PRPs the opportunity to conduct an RI/FS, or it may impose such onerous conditions on a PRP-conducted RI/FS that PRPs will decide to let the EPA conduct the study. In such instances, the need to have meaningful PRP input into the remedy selection process becomes greatest. Even when PRPs do not conduct an RI/FS, they (and other ob- jecting parties) should endeavor to submit well-documented and complete reports from their own experts on workable cost-effec- tive remedial alternatives, and such objecting parties should at the same time submit comprehensive comments criticizing weak- nesses in-the EPA's expert reports and remedy selections. In or- der for objecting parties' experts to submit a convincing case for their position, it may be necessary for objecting parties' contrac- tors to conduct their own tests of the hazardous site in question and to carefully examine the site hydrogeology. At the very least, PRP representatives should request access to the site (if they do not already have it), should closely examine site conditions and, if possible, should observe the work conducted by the EPA or its contractor. Where PRPs are attempting to argue that a less-ex- tensive and less-expensive remedy will be adequate for the site, they should submit as much evidence as possible demonstrating that there are no pollution pathways emanating from the site which could affect residents in the area or affect the environment in the site vicinity. Off-site testing by PRPs will often be helpful in establishing such facts. Because cross-examination of EPA experts will probably not be allowed during the course of public hearings, objecting parties would be well advised to submit a list of written questions to the EPA during the comment period seeking to flesh out the basis for EPA remedial decisions and to evaluate whether all relevant fac- tors were considered by the agency. The scope and number of 133. Such actions are carried out under the authority of CERCLA § 106(a), 42 U.S.C. § 9606(a) (1982 & Supp. IV 1986). See, e.g., United States E.P.A. Memorandum, Office of Solid Waste and Emergency Response, Interim Guidance on Potentially Responsible Party Participation in Remedial Investigations and Feasibility Studies, May 16, 1988 at 6. 214 Judicial Review Under SARA these questions should be sufficiently extensive so as to probe any controversial or questionable steps that the EPA takes in proceed- ing on its path to the choice of a remedy. The EPA may see fit to provide written answers to these questions as part of the agency's mandated response to comments.134 If the agency does not do so, this may be one of the bases for asserting that the EPA acted arbitrarily and capriciously or contrary to law.'35 Objecting parties would also be well advised to question the EPA as to the basis for any assumptions which it employs in reme- dial selection. The Agency could additionally be asked to explain every alternative remedy which it has explored, so that parties can determine whether the EPA in fact evaluated all relevant factors in reaching its remedial decision. Objectors should also explore whether the EPA has considered any of the growing number of innovative technologies that are available for site remediation. In essence, objecting parties should strive to insure that sufficient questions are posed to the EPA so that the record "reflects the contemporaneous analyses and criticisms of all interested parties, and therefore provides a comprehensive framework from which the court can scrutinize the Agency's action."'36 Furthermore, unless the EPA regulations provide that all corre- spondence involving a particular site is to be included in the rec- ord established under SARA, objecting parties may want to submit all such correspondence for inclusion in the record during the public comment period. Such correspondence may later prove useful in a judicial challenge in documenting EPA delays in carrying out the remedial process and in demonstrating the EPA's failure to consider relevant factors, despite being confronted with such factors by other parties.S? 134. See SARA § 1l3(c)(2), 42 U.S.C. § 9613(k)(2)(B)(iv) (Supp. IV 1986). Probing into the intermediate steps undertaken by the EPA in its remedy selection process is particu- larly important because, as stated by Judge J. Skelly Wright, "the 'arbitrary, capricious' standard seems to require not an evaluation of the rulemaker's empirical conclusions, but rather an inquiry into the basic orderliness of the process by which evidence and alterna- tive rulings were considered. Wright, The Courts and the Rulemaking Process: The Limits of judicial Review, 59 CORNELL L. REv. 375, 392 (1974). 135. See also Section C.6., infra, at 538. 136. Rohm and Haas, 669 F. Supp. at 681. Of course, as discussed above, although the Rohm and Haas court believed that such a complete record could be achieved without re- sort to cross-examination or detailed discovery, in the typical remedy selection process, this appears doubtful. 137. In Lone Pine Steering Comm., 777 F.2d at 887, Third Circuit, in discussing the EPA remedy selection process, noted that "the courts are not unaware of bureaucratic ex- 1989] 215 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 C. The Requirement that the EPA Undertake a Reasoned Decision- Making Process. 1. Administrative Record Requirements in General In addition to mandating public participation, the arbitrary and capricious standard mandates that an agency carry out a careful and reasoned decision-making process, and that it provide docu- mentation of that process. Traditionally, the courts look to the administrative record in order to examine the substantive basis of the agency's deliberations, and the Supreme Court has noted in Camp v. Pitts that "the focal point for judicial review should be the administrative record already in existence, not some new record made initially in a reviewing court."138 SARA provides for the establishment of an EPA administrative record by noting that selection of a response action shall be based upon such administrative record.139 In addition, SARA mandates that the EPA provide a response to each of the significant com- ments, criticisms, and new data submitted in written or oral presentations as part of the public participation process, and fur- ther provides that the EPA include "a statement of the basis and purpose" of the remedial action which it selects.14 0 Once again, the SARA requirements closely track the record-making require- ments that have previously been set out in the APA and case law. An extensive body of case law has expanded and defined the various elements that must be included in any administrative rec- ord which comports with due process. In summary, an adequate administrative record must: (1) disclose all relevant data and in- formation relied upon by the agency in reaching a final decision; (2) include an explanation of the agency's decision, including its basis and purpose; (3) demonstrate that the agency has consid- ered all relevant factors in arriving at its decision; (4) discuss and evaluate alternatives which were considered; and (5) contain all cesses," and warned the EPA that CERCLA requires the agency to observe cost-effective- ness, describing that mandate as "a limitation, not a license to squander." The court further noted in this pre-SARA case that PRPs can contribute to an EPA record by observ- ing the remedial project, submitting pertinent comments or suggestions and continually monitoring the EPA. 138. 411 U.S. at 142. 139. SARA § 13(c)(2), 42 U.S.C. § 9613(k)(1) (Supp. IV 1986). 140. Id., 42 U.S.C. H§ 9613(k)(2)(B)(iv) and (v). 216 Judicial Review Under SARA relevant public comments and agency responses to such comments.14I Each of these administrative record elements is necessary be- cause, as stated by one commentator, "[w]ithout such a record courts could decide only whether the choice made was plainly un- reasonable on the basis of the agency's explanation of its action and the materials submitted by opponents in comments to the agency or court briefs." 42 A complete administrative record is necessary because "in order for a court to make a critical evalua- tion of the agency's action and to determine whether it acted 'per- functorily or arbitrarily,' the agency must in its decision 'explicate fully its course of inquiry, its analysis and its reasoning.' "" There have been numerous cases where courts have held that the administrative record failed to demonstrate that an adminis- trative agency had complied with the requirements of due process for one or more of the reasons cited above. Examples of cases rejecting administrative decision-making for failure to render their decisions based on a clear, reviewable record are discussed below. Each of the arguments that has been successfully made in the cited cases might also provide the basis for challenging a CERCLA Record of Decision as being inadequate in the face of due process scrutiny. 2. Disclosure of Data Relied Upon The Food and Drug Administration failed to disclose the scien- tific data on which it based its decision in U.S. v. Nova Scotia Food Products Corp.,'44 resulting in the Agency's failure to meet its bur- den of connecting the data to its ultimate findings. The Second Circuit consequently noted that "when the basis for a proposed rule is a scientific decision, the scientific material which is be- lieved to support the rule should be exposed to the view of inter- ested parties for their comment."'45 The court recognized that in some instances the agency may rely on its expertise, which is not 141. See the detailed discussion of these requirements infra, at text accompanying notes 147-73. 142. Stewart, The Development of Administrative and Quasi-Constitutional Law in judicial Re- view of Environmental Decisionmaking: Lessons from the Clean Air Act, 62 IOWA L. REV. 713, 738 (1977). 143. Appalachian Power Co. v. Envtl. Protection Agency, 477 F.2d 495, 507 (4th Cir. 1973), (quoting Ely v. Velde, 451 F.2d 1130, 1138-39 (4th Cir. 1971)). 144. 568 F.2d 240 (2d Cir. 1977). 145. Id. at 252. 1989] 217 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 part of the administrative record, but the court made clear that when research data is available it should be disclosed.'46 "Scien- tific research is sometimes rejected for diverse inadequacies ... [or] because of a lack of adequate gathering technique or of sup- portable extrapolation.... To suppress meaningful comment by failure to disclose the basic data relied upon is akin to rejecting comment altogether."' 4 7 The court went on to note that "[i]f the failure to notify interested persons of the scientific research upon which the agency was relying actually prevented the presentation of relevant comment, the agency may be held not to have consid- ered all 'the relevant factors'." 4 8 In turn, this can result in a de- cision being held to be arbitrary and capricious.149 In Portland Cement Association v. Ruckelshaus,15 0 the EPA failed to make available its findings and procedures, which in turn resulted in a denial of the right to make meaningful comments. The court noted that "[i]t is not consonant with the purpose of a rulemaking proceeding to promulgate rules on the basis of inadequate data, or on data that, in critical degree, is known only to the agency."1i5 Similarly, in Natural Resources Defense Council, Inc. v. Train,152 the court allowed discovery to occur because it found that decision-makers directly or indirectly considered information in documents that were not included in the record, and thus were not before the reviewing court. In a remedy selection proceeding under CERCLA and SARA, a challenge might lie against the EPA if, for instance, the Agency failed to reveal all of the groundwater sampling data which it ob- tained during its remedial investigation of a particular site. In such a situation, challenging parties would have no opportunity 146. Id. at 251. 147. Id. at 252. 148. Id. at 251. 149. "[A]lthough we recognize that an agency may resort to its own expertise outside the record in an informal rulemaking procedure, we do not believe that when the pertinent research material is readily available and the agency has no special expertise on the precise parameters involved, there is any reason to conceal the scientific data relied upon from the interested parties.. [W]e can think of no sound reasons for secrecy or reluctance to expose to public view (with an exception for trade secrets or national security) the ingredients of the deliberative process." Id. at 251 (citation omitted). 150. 486 F.2d 375, 402 (D.C. Cir. 1973), cert. denied, 417 U.S 921 (1974). 151. Id. at 393. 152. 519 F.2d 287 (D.C. Cir. 1975). 218 Judicial Review Under SARA to question the quality of the withheld data or its predictive value. If the withheld data could be used to dispute or question EPA claims that off-site contamination existed or that contamination was moving toward a critical area, then the data would effectively prevent potential challengers from disputing the efficacy of or need for a particular EPA-selected remedy. Under such a scena- rio, the Agency would presumably be acting arbitrarily and capriciously. 3. Agency Explanation of Basis and Purpose of Decision Courts frequently find administrative decisions arbitrary and capricious when the agency has not explained the basis for its de- cision or set forth a reasoned analysis of the decision. As one commentator has noted: The recent cases require the agency to describe its decision- making processes in detail, and the court will examine the ra- tionality of the processes as a part of its analysis of the final result. ... If the agency has made choices irrationally, the taint is presumed to carry over into the final action, and the final action will be held arbitrary.153 In National Nutritional Foods Association v. Weinberger,15 4 the court stated that although 5 U.S.C. section 553, governing informal rulemaking procedure, does not require conclusions that are as detailed as those for a formal proceeding, the agency must still publish its reasons in enough detail to allow a reviewing court to be able to make a meaningful evaluation of their sufficiency. The Second Circuit noted that, "[i]ndeed the very absence of a de- tailed record of the type that would be made if an evidentiary hearing were held makes it advisable for the agency, in lieu thereof, to provide a thorough and comprehensive statement of the reasons for its decision." 55 In Kennecott Copper Corp. v. Environmental Protection Agency,'5 6 the D. C. Circuit discussed the inadequate reasoning included on the record by the EPA. The EPA had failed to provide a statement of the basis of the regulation promulgated. The court pointed out that although in environmental rule-making proceedings it is 153. DeLong, supra note 66, at 285 (footnote omitted). 154. 512 F.2d 688, 701 (2d Cir.) cert. denied sub nom. National Nutritional Foods Ass'n v. Mathews, 423 U.S. 827 (1975). 155. Id. at 701. When an agency bases its decision on past experience, it "must be a matter of record in order to qualify for consideration by a reviewing court." Id. at 701 n.11. 156. 462 F.2d 846, 849 (D.C. Cir. 1972). 1989] 219 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 often impractical to require the agency to include a response to all issues raised in comments submitted, "[t]here are . .. contexts of fact, statutory framework and nature of action, in which the minimum requirements of the Administrative Procedure Act may not be sufficient."'5 7 Although, in the wake of Vermont Yankee, it is doubtful that a court would today speak so glibly about requiring an agency to exceed the minimum standards of the APA, one sus- pects that the basic reasoning espoused by the court on the inade- quacy of the agency's response to comments would not be substantially different today. Because complex technical facts will often be necessary to sup- port a CERCLA remedy selection, there is always the substantial potential that the record submitted to a reviewing court will be unclear to a judge with no technical background. In such a cir- cumstance, courts may be forced to engage in de novo review in order to make sense of proceedings at the administrative level and to understand and meaningfully evaluate the basis for an EPA decision. As one court has noted: "in the often difficult task of reviewing administrative regulations, the courts are not straight jacketed to the original record in trying to make sense of complex technical testimony, which is often presented in administrative proceedings without ultimate review by non-expert judges in mind."s58 Furthermore, it has been noted that because a review- ing court normally gives deference to an agency's substantive conclusions in complex regulatory matters, courts should insist that required recordmaking procedures be strictly followed.'59 157. Id. at 850 (footnote omitted). 158. Bunker Hill Co. v. Envtl. Protection Agency, 572 F.2d at 1286, 1292 (9th Cir. 1977). This raises the pertinent question as to the degree to which judges are actually able to review and evaluate complex scientific and technical issues. Differing judicial views are illustrated by the different opinions in Ethyl Corp. v. United States Entil. Protection Agency, 541 F.2d I (D.C. Cir. 1976), cert. denied, 426 U.S. 941 (1976). Judge Bazelon, in a concurring opinion, states that lay judges are not competent to probe the factual basis for highly technical agency decisions, leading to the result that "substantive review of mathematical and scientific evidence by technically illiterate judges is dangerously unreliable." Id. at 67. He continues, "[die novo evaluation of the scientific evidence inevitably invites judges of opposing views to make plausible-sounding, but simplistic, judgments of the relative weight to be afforded various pieces of technical data." Id. at 66. In contrast, Judge Leventhal, who submitted a separate statement on the case, takes the position that judges have the obligation to "acquire whatever technical knowledge is necessary as background for decision of the legal questions." Id. at 68. 159. Natural Resources Defense Council v. United States Envtl. Protection Agency, 824 F.2d 1258 (1st Cir. 1987). However, contrast the statement in Amoco Oil Co. v. United States Envtl. Protection Agency, 501 F.2d 722, 741 (D.C. Cir. 1974), where the court noted that 220 Judicial Review Under SARA In making the record in a remedial selection process, the EPA could be faulted under a "nondisclosure of basis and purpose" theory if, in picking a particular preferred remedy, it failed to ade- quately explain its basis for distinguishing between various alter- native remedies included in its feasibility study for the site. For instance, if the Agency did not explain on the record why a cheaper alternative remedy with seemingly similar protective qualities was not chosen, then an arbitrary and capricious attack might be mounted. Similarly, under Section 107(a)(4)(B)o60 of CERCLA, costs of response must be deemed to be "necessary and consistent with the NCP." If the EPA failed to demonstrate on the record that designated response costs were both necessary and consistent with the NCP, then the EPA decision to expend or approve such response costs could be subject to challenge on arbitrary and ca- pricious grounds. Furthermore, if the EPA fails to choose a "preferred" remedy under CERCLA Section 121 (b)(1),16 1 that section requires EPA to publish an explanation of why it failed to do so. If no such expla- nation is forthcoming, this would be a cause to challenge the EPA decision. According to CERCLA Section 121(2)(d)(A)1 62 the EPA's basis for and purpose in reaching its decision are also important when the EPA chooses standards and criteria which it determines are "applicable or relevant and appropriate" requirements to be ap- plied to remediation for a particular hazardous substance site. The selection of these so called "ARARs" will often determine the degree of cleanup which must be accomplished and conse- quently the overall cost of cleanup. For instance, the choice of a particular ARAR may determine the level to which soil, surface water or groundwater must be remedied at a site. Obviously, the basis behind the EPA's choice of an ARAR will be of great interest to parties who feel that a particular chosen standard is either too stringent or too lenient, and any failure by the EPA to explain when agency decisions "turn on choices of policy, on an assessment of risks, or on predic- tions dealing with matters on the frontiers of scientific knowledge, we will demand ade- quate reasons and explanations, but not 'findings' of the sort familiar from the world of adjudication." 160. 42 U.S.C. § 9607(a)(4)(B) (1982 & Supp. IV 1986). 161. Id. at § 9621(b)(1). 162. Id. at § 9621(d)(2)(A). 1989] 221 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 why it chose an ARAR would provide grounds for claiming that the decision was arbitrary and capricious. 4. The Requirement that an Agency Consider All Relevant Factors A concise discussion of the type of substantive record review that is to be conducted by courts under the arbitrary and capri- cious standard was set out in Motor Vehicle Manufacturers Association v. State Farm Mutual Automobile Insurance Company.163 The Supreme Court there stated: Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not in- tended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. Or, as more simply stated in Hanly v. Mitchell: "it is 'arbitrary and capricious' for an agency not to take into account all relevant fac- tors in making its determination."' " In the EPA remedy selection process, there are many potential relevant factors which could be ignored by the Agency, thus form- ing the basis for legal challenge. For example, EPA may choose a remedy designed to return contaminated groundwater to drink- ing water standards. However, if that groundwater could never in any event be utilized for drinking purposes, e.g., because of a salt intrusion problem in the vicinity, then EPA would appear to be open to challenge on its remedy selection. Similarly, if EPA were to give no consideration to the cost-effectiveness of a particular selected remedy, or ignore potential engineering problems asso- ciated with the remedy, then the charge could be made that all relevant factors were not appropriately considered. 5. Analysis of Alternatives The requirement that an agency consider and analyze alterna- tives to its chosen course of action is particularly relevant to the EPA remedy selection process because, under the NCP, remedy 163. 463 U.S. 29, 43 (1983). 164. 460 F.2d 640, 648 (2d Cir. 1972), cert. denied, 409 U.S. 990 (1972). This language is derived from the discussion on the arbitrary and capricious standard in Overton Park, 401 U.S. at 416. 222 Judicial Review Under SARA selection is mandated to entail the initial screening of alterna- tives,' 6 5 followed by a detailed analysis of alternatives,1 6 6 prior to the selection of a remedy. Thus, in the context of arbitrary and capricious review of EPA remedy selection, analysis of remedial alternatives holds particular importance. In Motor Vehicle Manufacturers Association, 167 the Secretary of Transportation's decision to abandon a proposed requirement for automatic automobile seat belts and airbags was held to be arbitrary and capricious because the record did not include an analysis of a technologically feasible alternative. The notice given had not apprised parties that if one of the aforementioned alter- natives was infeasible, the other would not be approved. The agency in turn decided not to approve either the seat belt or airbag requirement, but only explained why it rejected one pro- posal without discussing the other. The agency's deficient notice resulted in the decision being arbitrary and capricious, because the, parties had not been informed that rejection of one alterna- tive would justify rejection of the other, and the agency failed to explain on the record how it arrived at the decision. In a highly technical matter such as remedy selection, it is of course vital that parties wishing to bring alternate factors to the attention of the agency do so during the establishment of the ad- ministrative record. Otherwise, it may not be possible for a re- viewing court which is not skilled in the minutiae of groundwater hydrology and remedy design to determine through its record re- view whether the EPA in fact considered all the factors which it should have in arriving at its selection of a remedy. As the Court stated in Asarco, Inc. v. United States Environmental Protection Agency:' 6 8 It will often be impossible, especially when highly technical matters are involved, for the court to determine whether the agency took into consideration all relevant factors unless it looks outside the record to what matters the agency should have considered but did not. The court cannot adequately dis- charge its duty to engage in a "substantial inquiry" if it is re- quired to take the agency's word that it considered all relevant matters. 165. 40 C.F.R. § 300.68(g) (1987). 166. Id. at § 300.68(h). 167. 463 U.S. at 43. 168. 616 F.2d 1153, 1160 (9th Cir. 1980). 2231989] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:187 In an EPA remedy selection process, a challenge might be mounted by objecting parties if the range of alternatives consid- ered by EPA were not sufficiently broad. For example, such a challenge could be made if certain available innovative technolo- gies were not considered. Or, if the Agency could be shown to have inadequately evaluated the environmental efficiency of one or more alternatives, then a legal challenge might lie. 6. The Agency's Duty to Provide an Opportunity for Comments and to Offer Responses Often, the public participation and notice requirements under the arbitrary and capricious standard and the requirements for substantive agency consideration of various factors will be inter- twined. Frequently, substantive deficiencies in the record will re- sult from the procedural deficiency of failing to provide interested parties with adequate opportunity to submit all relevant com- ments. Conversely, substantive deficiencies in an initial agency record may prevent parties from being aware of, and thus re- sponding to, important aspects of an agency decision. For example, a party may be denied a meaningful opportunity to participate and comment in an agency decision-making process if the agency has failed to disclose all of the data upon which it relied. Similarly, an agency failing to provide adequate notice to interested parties may be subject to the charge that it failed to consider all relevant factors, including factors that would have been raised by the party or parties not receiving notice. For in- stance, in Rodway v. United States Department of Agriculture, 169 the agency failed to provide notice to interested parties, which in turn resulted in the denial of their right to make comments. This then resulted in the agency giving an insufficient explanation of its de- cision. The court found the agency's decision to be arbitrary and capricious, because its post hoc explanations, without comment from parties, did not result in the creation of a "record" on which the court could review the basis of the decision.70 Similarly, in Natural Resources Defense Council v. United States Envi- ronmental Protection Agency,'7' the court held that the EPA's expla- nation of why certain groundwater protection requirements were 169. 514 F.2d 809 (D.C. Cir. 1975). 170. Id. at 817. 171. 824 F.2d 1258 (1st Cir. 1987). 224 Judicial Review Under SARA adopted was not very helpful. This lack of a cogent explanation was tied to the fact that there had been no comment period on this issue, with the court noting that: "Because the petitioners never had a comment period in which to express their concerns, of course the Agency's explanation is going to be vague and can- not address petitioners' complaints." 7 2 In the EPA remedy selection process, SARA requires not only that there be adequate opportunity for public comment, but that EPA offer, on the record, a response to each of the significant comments, criticisms and new data submitted.'73 A definition of what is a "significant" comment is not provided by the statute, but any comment questioning the Agency's methodologies, its data collection, its engineering analysis or the qualification of its experts would appear to be significant. Thus, EPA's failure to ad- dress such comments would provide a basis for an arbitrary and capricious challenge. Also, if commentors' pleas for additional testing or arguments supporting alternative remedies are not squarely addressed on the record, the EPA deliberative process, and consequently the final EPA remedy selection, may become arbitrary and capricious. III. OTHER POTENTIAL AVENUES OF CHALLENGE TO EPA DECISION-MAKING UNDER SARA A. Challenges Under APA Principles In addition to the relatively narrow standard of review provided under the arbitrary and capricious standard, there are a number of other avenues by which protesting parties can attempt to chal- lenge EPA remedial action decision-making under CERCLA or SARA. For instance, the APA provides'74 that in addition to arbi- trary and capricious challenges, agency action can be challenged if it is: (1) not in accordance with law; (2) contrary to constitu- tional right, power, privilege or immunity; (3) in excess of statu- tory jurisdiction, authority, or limitations, or short of statutory right; or (4) without observance of procedure required by law. Similarly, new CERCLA Section 113(j)(2)175 allows for challenge to EPA action in selecting a response action if the EPA decision 172. Id. at 1286. 173. SARA § 113(c)(2), 42 U.S.C. § 9613(k)(2)(B)(iv). 174. 5 U.S.C. § 706(2) (1982). 175. 42 U.S.C. § 9613(j)(2) (Supp. IV 1986) (added by SARA § 113(c)(2)). 1989]1 225 COLUMBIA JOURNAL oF ENVIRONMENTAL LAW [Vol. 14:187 can be demonstrated to be not in accordance with law. Thus, EPA action can be challenged if the agency fails to follow any of the many requirements under SARA, or if it fails to follow the EPA's own regulations or its own internal procedures and guide- lines.'76 As a practical matter, most challenges claiming that the EPA did not follow its own regulations in choosing a remedial action will focus upon compliance with the NCP's procedures to be followed and standards to be employed in choosing and carry- ing out remedial measures under CERCLA.' 77 A related area of attack of EPA decision-making could claim that the agency's ac- tion is not consistent with the purpose of CERCLA and SARA.178 B. New Parties Identified and Issues Arising After the Making of the Record The fact that EPA actions under CERCLA often identify and add PRPs after initiation of the action may create an opportunity for claiming a need for de novo review. This circumstance oc- curred in the Rohm and Haas case,'79 where PRPs were identified but had not been formally joined as parties at the time of the deci- sion. The court noted that this situation "raises a potential con- stitutional question concerning the adequacy of SARA's judicial review procedures as applied to parties who had no opportunity to participate in the process before the agency." 8 0 Similarly, there will sometimes be instances where additional information regarding the progress of contamination or the characteristics of geological formations at a hazardous site will become apparent only after a remedy has been chosen through the procedures pro- 176. See, e.g., Allegheny County Sanitary Auth. v. United States Envtl. Protection Agency, 557 F. Supp. 419 (W.D. Pa. 1983), aff'd., 732 F.2d 1167 (3d Cir. 1984) (The EPA failed to perform mandatory duties under the Federal Water Pollution Control Act and its implementing regulations); Donner Hanna Coke Corp. v. Costle, 464 F. Supp. 1295 (W.D.N.Y. 1979)(reviewing court must determine whether agency follows lawful procedures). 177. See, e.g., the district court's discussion of the plaintiff's claims in Lone Pine Steer- ing Comm. v. United States Envtl. Protection Agency, 600 F. Supp. 1487 (D.N.J.), aff'd, 777 F.2d. 882 (3d Cir. 1985), and United States v. Northeastern Pharmaceutical and Chemical Co., Inc. 810 F.2d 726, 747-48 (8th Cir. 1986), cert. denied, 108 S. Ct. 146 (1987). 178. See, e.g., Chrysler Corp. v. United States Envtl. Protection Agency, 631 F.2d 865 (D.C. Cir. 1980), cert. denied, 449 U.S. 1021 (reviewing court should give deference to the EPA's judgment if it is demonstrated that the judgment is inter alia consistent with the purpose of the Clean Air Act). 179. 669 F. Supp. at 684. 180. Id. 226 Judicial Review Under SARA vided by SARA. Such new information, if it proves to be signifi- cant and relevant to the choice of remedy or the conduct of the remedial action, may very well merit de novo review or at the very least remand by a court for additional procedures. Certainly, "when evidence either confirming or denying agency predictions made in the original decision becomes available,"18' the record should be reopened. As one court stated in a non-CERCLA context: "when the question is one which the Agency may never have fully con- fronted and which may deserve further input both from Agency and outside sources, only a remand for further hearings and an extended record seems adequate."82 C. Other Challenges Furthermore, an agency determination on the record can be challenged, and either record supplementation or de novo review can be ordered, upon a showing that an agency has exercised bad faith or improper behavior.'83 Also, because virtually all powers under CERCLA and SARA have been delegated from the President to the EPA administra- tor1 8 4 it may be possible to mount challenges on the theory that powers have been delegated improperly or that a particular exer- cise of power is not supported by existing delegations. IV. CONCLUSION By setting an arbitrary and capricious standard of review, and by limiting review of EPA response action decisions to the admin- istrative record, SARA has placed great focus on EPA proceed- ings that will establish the administrative record upon which 181. Conservation Law Found. of New England, Inc. v. Clark, 590 F. Supp. 1467, 1474 (D. Mass. 1984) (citation omitted). 182. South Terminal Corp. v. United States Envtl. Protection Agency, 504 F.2d 646, 655 (1st Cir. 1974) (citing Portland Cement Ass'n v. Ruckelshaus, 486 F.2d 375, 393-94 (1973), cert. denied, 417 U.S. 921 (1974)). 183. Overton Park, Citizens to Preserve Overton Park v. Brinegar, 401 U.S. 402, 420 (1971); Bethlehem Steel Corp. v. United States Envtl. Protection Agency, 638 F.2d 994, 1000 (7th Cir. 1980). See also Roberts v. Morton, 549 F.2d 158, 164 (10th Cir. 1976), cert. denied, sub nom. Roberts v. Andrus, 434 U.S. 834 (1985): "An adjudicatory hearing before an administrative tribunal must afford a fair trial in a fair tribunal as a basic requirement of due process" (citations omitted) and Gibson v. Berryhill, 411 U.S. 564, 579 (1973) (hold- ing no need to wait for exhaustion of state law issues before proceeding to federal court where agency bias can be shown). 184. 52 Fed. Reg. 2923 (1987). 1989] 227 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:187 remedial decisions will be made. The forthcoming EPA regula- tions governing the compilation of the administrative record in the remedy selection process will determine whether remedy se- lection is a dynamic proceeding drawing upon the best available technical expertise or whether it is merely a rubber stamp of the remedy choice of EPA experts. Unless the regulations on remedy selection allow cross-examination of EPA's experts, the process will fail to flesh out issues which should be explored in order to assure selection of the most efficient, cost-effective remedy available. In fact, the due process rights of PRPs can only be guaranteed if cross-examination of EPA witnesses is provided. Furthermore, because EPA remedy selection demonstrates the basic character- istics of an adjudicative fact determination, precedent states that a full adjudicatory proceeding should be provided in order to de- termine such facts: Experience has demonstrated that limited, controlled access to cross-examination will not unduly burden ad- ministrative proceedings such as the remedy selection process. However, despite the compelling due process arguments in favor of cross-examination, it is unlikely that EPA will provide any cross-examination rights when it formulates its regulations for remedy selection hearing procedures. Furthermore, in the wake of the Vermont Yankee decision, it is somewhat doubtful that a court would overturn the. EPA's choice of remedy selection procedures, unless a due process violation under Mathews v. Eldridge principles is demonstrated. Consequently, potentially responsible parties and other parties with an interest in remedial choices for a hazardous waste site should pursue two avenues of input to the remedy selection pro- cess. First, these parties should take particular care to establish as complete an administrative record as possible and to clearly high- light on the record all objections to EPA's proposed remedial se- lections and all weaknesses underlying EPA's analysis. Second, if EPA's ultimate remedy selection is still contrary to the position of interested parties, PRPs can avail themselves of significant oppor- tunities for attacking the EPA remedy selection as arbitrary and capricious on the basis of either public participation deficiencies or the lack of a complete, well-documented deliberative process. Ultimately, it is possible that the CERCLA/SARA judicial re- view standards will provide the battleground upon which the ten- 228 1989] Judicial Review Under SARA 229 sion between the principles of Vermont Yankee and the guarantees of due process are finally resolved.