Rulemaking in the Shadows: The Rise of OMB and Cost-Benefit Analysis in Environmental Decisionmaking Jeffrey H. Howard and Linda E. Benfieldt I. INTRODUCTION The goal of achieving the best for the most has permeated Western political thought for more than 250 years. In 1720, Francis Hutcheson declared: "That action is best which procures the greatest happiness for the greatest number."' In the 1800s, Jeremy Bentham rediscovered the principle and gave it wide rec- ognition among political philosophers as "the foundation of morals and legislation."2 This utilitarian philosophy has manifested itself in modern ad- ministrative law through cost-benefit analysis. In theory, such analysis assists decision-makers in achieving the greatest good for the greatest number by promoting cost efficiency. In the Flood Control Act of 1936,3 for example, Congress directed that flood control projects be constructed only "if the benefits to whomsoever they may accrue are in excess of the estimated costs .... -"4 Under this standard, following a detailed evaluation of costs and benefits, new water projects have been approved only when the numerical ratio tipped in favor of the benefits. t Mr. Howard is a partner in the Washington, D.C., law firm of Crowell & Moring and has practiced in the health and environmental field for many years. He received his law degree in 1969 from the Law School at the University of Virginia, where he was a member of the Law Review and the Order of the Coif. In the mid-1970s, he served as Associate General Counsel of the United States Environmental Protection Agency in charge of Tox- ics, Pesticides and Solid Waste Management. Ms. Benfield is an associate in the Milwaukee, Wisconsin, law firm of Foley & Lardner and previously worked with Mr. Howard in Washington, D.C. Ms. Benfield received her law degree in 1985 from the University of Chicago and practices in the environmental field. 1. F. HUTCHESON, INQUIRY CONCERNING MORAL GOOD AND EVIL (1720). 2. 10J. BENTHAM, WORKS 142 (1843). 3. 33 U.S.C. §§ 701-09 (1988). 4. 33 U.S.C. § 701a (1988) (emphasis added). 144 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 During the 1970s the number of ambitious, new social pro- grams exploded. Among these new initiatives were a panoply of environmental laws. After the establishment of the Environmen- tal Protection Agency ("EPA") in 1970, Congress also created more than 14 new environmental statutes regulating air, water, drinking water, coastal zones, hazardous waste, ocean disposal, pesticides, noise, wildlife and toxic substances.5 This explosion of social legislation produced a concomitant in- crease in federal regulatory activity. The number of Federal agency employees almost trebled from 1970 to 1979.6 Federal budgetary expenditures for such agencies swelled by over 400%'.7 The implementation of these statutes also produced a deluge of new, complex administrative regulations. From 1970 to 1979, the number of pages of new regulations published in the Federal Regis- ter more than tripled from 20,036 in 1970 to 77,497 in 1979.8 During this same period, the U.S. economy was experiencing serious problems. The continued proliferation of costly new agency regulations in the face of troubling national economic news produced sharp political reactions. Critics of the growth in regulation objected that "[w]ithout some countervailing restraint, EPA and OSHA [Occupational Health and Safety Administration] 5. See, e.g., the Federal Insecticide, Fungicide and Rodenticide Act of 1972, 7 U.S.C. §§ 136-36y (1988) (environmental and health regulation of pesticides); the Toxic Sub- stances Control Act, 15 U.S.C. §§ 2601-29 (1988) (regulating the production and sale of new toxic substances); the Marine Mammal Protection Act, 16 U.S.C. §§ 1361-07 (1988); the Coastal Zone Management Act of 1972, 16 U.S.C. §§ 1451-64 (1988) (Federal grants and standards for managing coastal waters); the Endangered Species Act of 1973, 16 U.S.C. 99 1531-43 (1988) (protecting wildlife species designated as endangered or threatened); the Surface Mining Control and Reclamation Act of 1977, 30 U.S.C. §§ 1201- 28 (1988) (establishing environmental controls over surface mining); the Federal Water Pollution Control Act, 33 U.S.C. §§ 1251-76 (1988) (comprehensive controls over dis- charges of pollutants into U.S. waters); the Marine Protection, Research and Sanctuaries Act of 1972, 33 U.S.C. §§ 1401-45 (1988) (regulating ocean disposal of wastes); the Safe Drinking Water Act, 42 U.S.C. §§ 300f-300j (1988) (establishing national drinking water standards); the National Environmental Policy Act, 42 U.S.C. §§ 4321-61 (1988) (requir- ing environmental impact statements for major Federal actions); the Noise Control Act of 1972, 42 U.S.C. §§ 4901-18 (1988) (setting national noise limits for products and aircraft); the Resource Conservation and Recovery Act of 1976, 42 U.S.C. §§ 6901-87 (1988) (cra- dle-to-grave controls over hazardous solid waste and upgraded controls over non-hazard- ous solid waste); the Clean Air Act, 42 U.S.C. §§ 7401-42 (1988) (comprehensive controls over mobile and stationary sources of air pollution); and the Comprehensive Environmen- tal Response Compensation and Liability Act of 1980, 42 U.S.C. §§ 9601-57 (1988) (regu- lating inactive hazardous waste dump sites and providing a "superfund" for cleanup). 6. S. REP. No. 284, 97th Cong., Ist Sess. 10 (1981). 7. Id. at 11. 8. Id. 11MB and Cost-Benefit Analysis would 'spend' - through regulations that spend society's re- sources but do not appear in the federal government's fiscal budget - 'too much' on pollution control and workplace safety." 9 In addition, they maintained that cost-benefit analysis should be required for all regulations because it would "force regulators to confront problems of covert redistribution and overzealous pursuit of agency goals, which experience has shown to be common in regulatory programs."' 0 Finally, they claimed that the regulatory process needed a "traffic cop" to coordinate national policy. This "traffic cop," they argued, must coordinate inter-agency rule making and take into account national budget- ary constraints, as well as the President's policy priorities." It was in this political setting that Ronald Reagan made regula- tory reform a national issue during the presidential campaign of 1980. Following his victory in the election, Reagan took action on this issue both by sponsoring legislation to reform the regula- tory process and by issuing Executive Order 12,291. Executive Order 12,291 appointed the Office of Management and Budget ("OMB") as the President's regulatory "traffic cop" and required cost-benefit analysis as a predicate to all regulatory action.' 2 These developments have raised serious legal and policy ques- tions. Under United States administrative law, may OMB play this role, and may agencies engage in cost-benefit analysis? Some environmental statutes permit such an analysis, while others ex- plicitly preclude it. In both situations, OMB has reportedly ap- plied cost-benefit analysis. Further, is cost-benefit analysis helpful in formulating complex environmental policy? As this ar- ticle will discuss, cost-benefit analysis in regulatory decisionmak- ing yields some benefits in rulemaking, in terms of streamlining the process and creating a structure for decisionmaking, but it is not clear that such analysis complies with administrative law prin- ciples in all situations. Part I of this article explains the origin and operation of former President Reagan's regulatory reform. Part II analyzes the case law relating to this reform. The limitations of cost-benefit analysis in environmental decisionmaking are examined in Part III. Fi- 9. DeMuth and Ginsburg, White House Review of Agency Rulemaking, 99 HARV. L. REV. 1075, 1081 (1986). 10. Id. at 1081-82. 11. Id. at 1081-84. 12. Exec. Order No. 12,291, 46 Fed. Reg. 13,193 (1981). 1991] 146 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 nally, Part IV discusses suggestions and pending legislative efforts to refine this regulatory process. II. THE REAGAN REVOLUTION IN THE REGULATORY PROCESS A. The Background of Regulatory Reform Several recent Presidents have adopted programs to reform the regulatory process by exercising greater White House control and promoting cost efficiency. This process began under President Nixon through the so-called "Quality of Life Review" by which OMB reviewed proposed agency regulations on an informal ba- sis. s According to one commentator, the Quality of Life Review had its origin in a memorandum by George Schultz, then Director of 0MB, creating a review process for proposed rules " 'pertain- ing to environmental quality, consumer protection, and occupa- tional and public health and safety.' '14 In 1974, President Ford issued executive orders requiring all agencies to prepare "inflation impact statements" for regulations likely to have a $100 million impact nationwide.' 5 A similar pro- cedure called "regulatory analysis review" was substituted by the Carter Administration. Carter adopted his predecessor's $100 million threshold, adding a requirement that a regulatory analysis be performed if major inflationary impacts were expected.'16 Gen- erally speaking, both programs were designed only to illuminate the dark corners of regulatory analysis, embodying the kind of stop-and-think approach characteristic of an environmental im- pact statement. 13. See Morrison, OMB Interference With Agency Rulemaking: The Wrong Way To Write A Reg- ulation, 99 HARV. L. REV. 1059, 1061 (1986). 14. Olson, The Quiet Shift of Power: Office of Management & Budget Supervision of Environmen- tal Protection Agency Rulemaking Under Executive Order 12,291, 4 VA. J. NAT. RES. L. 1, 9 n.20 (1984). 15. Exec. Order No. 11,821, 39 Fed. Reg. 41,501 (1974), as amended by Exec. Order No. 11,949, 42 Fed. Reg. 1017 (1977). See Liroff, Cost-Benefit Analysis in Federal Environmen- tal Programs, [hereinafter Liroff] in CONSERVATION FOUNDATION, COST-BENEFIT ANALYSIS AND ENVIRONMENTAL REGULATIONS 35, 37 (1982) [hereinafter CONSERVATION FOUNDATION REPORT]. 16. Exec. Order No. 12,044, 43 Fed. Reg. 12,661 (1978). See Liroff, supra note 15, at 37- 38. See generally Rodgers, Benefits, Costs, And Risks: Oversight Of Health And Environmental Decisionmaking, 4 HARV. ENVrL. L. REV. 191 (1980). 11MB and Cost-Benefit Analysis B. The Reagan Administration's Legislative Effort No recent Chief Executive was more concerned with the bur- den of regulation than President Reagan and his program for reg- ulatory reform was the most extensive yet devised. First, Reagan sought legislation to reform the entire federal regulatory process. Second, he adopted a far-reaching and unprecedented Executive Order to revolutionize rule making by executive decree. At President Reagan's request, in November 1981, Senator Laxalt organized 81 Senators to co-sponsor the Administration's bill (S. 1080) with the laudable goals of requiring "[flederal agen- cies to analyze the effects of rules to improve their effectiveness and to decrease their compliance costs."' 7 Among other features, S. 1080 would have directed all regulatory agencies to conduct a thorough "regulatory analysis" before adopting any "major" reg- ulation.' 8 "Regulatory analysis" was defined to include a detailed analysis of the costs and benefits of the proposed rule.' 9 Public comment on the proposed rule and on the regulatory analysis were then required. If a final rule were subsequently adopted, S. 1080 directed that it be accompanied by a final report of the cost- benefit analysis including reasonable determination "that the benefits justified the costs .... 20 "Costs"21 and "benefits" 22 were defined broadly and included both quantifiable and unquan- tifiable elements. New rules were deemed to be "major" if they were "likely to have an annual effect on the economy of $100,000,000 or more in direct or indirect enforcement and com- pliance costs." 2 3 On November 30, 1981, the SenateJudiciary Committee issued its report on the bill. The Report concluded that, although S. 1080 would delegate lawmaking powers like Congress's own to an administrative agency, the delegation was not impermissible be- 17. S. REP. No. 284, supra note 6, at 1. 18. Regulatory Reform Bill Stalled in House, 1982 CONG. Q. ALMANAC 523. 19. Id. 20. Id. at 525. 21. " 'Cost' or 'costs' means the reasonably identifiable significant costs and adverse effects, including social and economic costs and effects, expected to result directly or indi- rectly from implementation of a rule or an alternative to a rule." Id., adding § 621(5) to 5 U.S.C. Chapter 6. 22. " 'Benefit' or 'benefits' means the reasonably identifiable significant benefits and beneficial effects, including social and economic benefits and effects, expected to result directly or indirectly from implementation of a rule or an alternative to a rule." Id., adding § 621(4) to 5 U.S.C. Chapter 6. 23. S. 1080, 97th Cong., 1st Sess. § 4(a) (1981). 1991] 148 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 cause Congress remained "the primary overseer" of the agency. 2 4 With Bentham-like prose, the Committee declared that "'all in- telligent decision-making has to be based on some sort of implicit cost-benefit analysis' "25 because the real objective of health and safety regulation is not simply to save lives but " 'to save the most lives for the level of resources society is willing to spend.' "26 The Senate Report further explained that the Committee in- tended a "rational weighing of regulatory proposals" - "but by no means a cost-benefit analysis in any strict sense .. ."27 The Report emphasized that it merely required that the costs be "jus- tified" in terms of the benefits. "[P]recise numerical quantifica- tion of costs and benefits is neither required nor anticipated in all cases ... ,"28 A few months after the Senate Committee reported the bill, on March 24, 1982, the full Senate unanimously adopted it. 24. S. REP. No. 284, supra note 6, at 53. In 1981, both the House and Senate Commit- tees on the Judiciary held hearings on the need for regulatory reform, including require- ments for new cost-benefit analysis. Regulatory Reform Act - S. 1080: Hearing before the Subcomm. on Regulatory Reform of the Senate Comm. on the Judiciary, 97th Cong., 1st Sess. (1981). Similar hearings had been held in earlier years. See S. REP. No. 284, supra note 6, at 1-2; H.R. REP. No. 435, 97th Cong., 2d Sess. 17-18 (1982). In July and October 1979, two subcommittees of the House Committee on Interstate and Foreign Commerce held a series ofjoint legislative hearings "on the use and abuse of cost-benefit analysis by regula- tory agencies." Use of Cost-Benefit Analysis by Regulatory Agencies: Hearings before the Subcomms. on Oversight and Investigations and on Consumer Protection and Finance of the House Comm. on Interstate and Foreign Commerce, 96th Cong., 1st Sess. 2 (1979). In February 1979, the Senate Subcommittee on Environmental Pollution also held extensive hearings on cost-benefit analysis which focused on the "review of environmental regulations by economists in the Executive Office of the President ...." Executive Branch Review of Environmental Regulations: Hearings before the Subcomm. on Environmental Pollution of the Senate Comm. on Environment and Public Works, 96th Cong., 1st Sess. 1 (1979). 25. S. REP. No. 284, supra note 6, at 73. 26. Id. at 72 (citing to response of Lester Lave during hearings on bill). "Our scarce resources and limited knowledge mean that often the decision is a matter of choosing where regulatory effort will produce the most good." Id. at 80. 27. Id. at 62. "The regulatory analysis of S. 1080 can be equated with cost-benefit anal- ysis only in this broad sense. S. 1080's regulatory analysis does not require that a numeri- cal value be applied to every effect of a proposed regulation and does not mandate the use of a mathematical formula to determine the ultimate merit of a proposed regulation. Thus, regulatory analysis in S. 1080 is what one commentator has called 'a kind of organ- ized common sense.'" Id. at 65. 28. Id. at 149. Although one important objective of this new requirement would be to reduce the costs of regulation, the Committee estimated that "regulatory analysis" would itself cost about $100,000 for each regulation to which it is applied. Id. at 90. 11MB and Cost-Benefit Analysis Meanwhile, on February 25, 1982, the House Judiciary Com- mittee reported a similar bill,29 which ultimately died on Decem- ber 9, 1982, when the House Rules Committee recessed without clearing the bill for floor action. In the meantime, the bill had attracted a great deal of opposition from consumer and environ- mental groups.30 John Dingell, Chairman of the House Energy and Commerce Committee, also opposed the bill, on the ground that it would erode existing environmental statutes and elevate the power of OMB over that of Congress itself.31 The resulting inaction on the bill in the House brought to a close President Reagan's efforts at regulatory reform through legislative action.3 2 C. Executive Order 12,291 In order to begin regulatory reform while Congress considered the Administration's legislative proposal, on February 17, 1981, Reagan issued Executive Order 12,291. 33 This Order sought to achieve by executive directive what S. 1080 would have achieved by statute had it been adopted. First, it establishes a new standard for the development of regu- lations. It requires that in adopting or reviewing regulations, or developing legislative proposals, each agency shall "to the extent permitted by law" assure that "[r]egulatory action shall not be undertaken unless the potential benefits to society for the regula- tion outweigh the potential costs to society . . . ."34 The Order requires all agencies to prepare a detailed "Regulatory Impact 29. H.R. REP. No. 435, 97th Cong., 2d Sess. 43-45 (1982). "This legislation does not require a 'cost-benefit test' that compares two columns of dollars. Such tests tend to place artificial values on such things as human life and a clean environment, and they tend to overemphasize costs and to imply accuracy in the estimates that is not present." Id. at 45. 30. 1982 CONG. Q. ALMANAC, supra note 18, at 523. 31. Id. at 527. 32. Other bills have been introduced in subsequent Congresses, but none has achieved the prominence or likelihood of success of the 1981-82 legislative efforts. See Note, Execu- tive Orders 12,291 and 12,498 Usurpation of Legislative Power or Blueprint for Legislative Reform?, 54 GEO. WASH. L. REV. 512, 516-19 (1986). 33. Exec. Order No. 12,291, 3 C.F.R. 127 (1981), reprinted in 5 U.S.C.A § 601 app. at 322-25 (West Supp. 1990). 34. Id. at § 2(b). Section 2 also requires that: such administrative decisions "be based on adequate information" concerning need and consequences, id. at § 2(a); regulatory objectives be designed to "maximize net benefits to society," id. at § 2(c); the least net cost alternative be chosen, id. at § 2(d); and the regulatory aim be "maximizing the aggregate net benefits" taking into account the condition of the regulated industries, the condition of the national economy and cumulative regulatory impacts in the future, id. at § 2(e). 1991] 150 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 Analysis" for each "major" rulemaking.3 5 As in S. 1080, the Or- der defines "major" rules to be those with an "annual effect on the economy of $100 million or more."3 6 Second, the Order in effect gives OMB veto power over any agency regulation. Under the Order, the agencies are required to submit all regulatory proposals to OMB for "review." 37 The Or- der directs that the agency "shall... refrain from publishing" the action until the OMB review is "concluded," for preliminary Reg- ulatory Impact Analyses or notices of proposed rule making, or "until the agency has responded to [OMB's] views," in the case of final agency action.3 8 In addition, OMB is expressly empowered, inter alia, to "[r]equire an agency to obtain and evaluate.., any additional relevant data from any appropriate source," to over- rule any agency on its determination of whether the rule is deemed to be "major," to waive any requirements of the Order, to eliminate any "duplicative overlapping and conflicting rules," to develop procedures for assessing costs and benefits, to pro- pose changes to the agencies' governing statutes, and to "[m]onitor agency compliance with the requirements of this Order and advise the Presi- dent with respect to such compliance."39 Finally, the Order purports to cut off judicial review of compli- ance with the Executive Order, by providing that it does not "cre- ate any right or benefit, substantive or procedural, enforceable at law by a party against the United States .... -40 On January 4, 1985, the President supplemented Order 12,291 with Executive Order 12,49841 which requires all agencies to sub- mit annually to OMB, in advance of any rulemaking, a "Draft Regulatory Program," in order that OMB can review all of the agencies' contemplated regulatory actions in advance to assure "consistency . . . with the Administration's policies and priori- 35. Exec. Order No. 12,291, supra note 33, at § 3. 36. Id. at § 1 (b). 37. Id. at § 3(c). 38. Id. at § 3(). 39. Id. at § 6 (a)(1)-(8) (emphasis added). As a former Deputy Administrator of OMB stated: "The Government works using three things: money, people, and regulations; the agency must get all three from OMB." Olson, supra note 14, at 6 (quotingJ. Tozzi). 40. Exec. Order No. 12,291, supra note 33, at § 9. At least one court has held that this section of the Executive Order is effective to cut offjudicial review of an agency's compli- ance with the other terms of the Order. See Michigan v. Thomas, 805 F.2d 176 (6th Cir. 1986). 41. 50 Fed. Reg. 1036 (1985). OMB and Cost-Benefit Analysis ties."' 42 This Order further empowers OMB to control and effec- tively veto any agency rulemaking, subject only to review by the President or his Cabinet Council. 43 D. Regulatory Reform in Practice A significant number of regulations have been reviewed by OMB under Order 12,291. In 1989, for example, OMB reviewed 2220 proposed rules,44 of which only 3.5% were "major." 45 Of the 2220,73.8% were found by OMB to meet the requirements of the Order. Slightly over 19% were determined to be consistent with the Order once the rule had been modified by the agency in response to OMB's comments. 46 Just under 5% were withdrawn by the submitting agency, returned to the agency for further re- consideration, or suspended.47 In 1989, EPA submitted 201 rules to OMB for review.48 Among these 201 rules were the largest number of major rules submitted by any agency. 49 Yet, OMB reviewed many fewer EPA rules by 1989 than it had in the early days of Order 12,291. From 1981 to 1989, the number of EPA rules reviewed declined by 72.6%.50 On the other hand, the amount of time taken by OMB to review EPA rules lengthened during the same period. In 1981, OMB took 12 days to review major rules and 9 days to review nonmajor rules. 5' In 1989, OMB took on average 104 days to review major rules and 49 days to review nonmajor rules.52 Because the regulatory review process mandated by these exec- utive orders usually occurs in private, there are few public sources of information about how the process has operated in practice. However, in March 1989, the Subcommittee on Oversight and In- vestigations of the House Committee on Energy and Commerce issued a report in which it disclosed detailed information about 42. Exec. Order No. 12,291, supra note 33, at § 3(a). 43. Id. 44. OFFICE OF THE PRESIDENT, REGULATORY PROGRAM OF THE UNITED STATES GOVERN- MENT, app. at 624 (April 1, 1990 - March 31, 1991). 45. Id. at 634. 46. Id. 47. Id. 48. Id. at 627. 49. Id. at 624. 50. Id. at 627. 51. Id. at 647. 52. Id. 1991] 152 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 OMB's involvement in reviewing EPA's proposed revisions to the Superfund National Contingency Plan. 53 The Report also con- tained copies of EPA communications with OMB and EPA's pro- posed revisions to the National Contingency Plan made in response to what are reported to be OMB comments. 54 The Sub- committee further identified four areas of congressional concern with respect to OMB's review of agency rulemaking under Order 12,291 including "delays beyond statutory and judicial deadlines, substantive inteference with Agency decisionmaking, the undue leverage of OMB officials and, most importantly, the secret nature of OMB's communications with agency officials and outside parties. ' '55 The Subcommittee summarized concerns which had been voiced by congressional committees over the years in which Or- der 12,291 was in effect. For example, it cited a 1985 report in which the same subcommittee reported that OMB had interfered with EPA's development of asbestos regulations by means of " 'secret, undisclosed, and unreviewable communications and contacts by parties interested in influencing the substance of agency rulemaking actions [which] are anathema to meaningful public involvement and effective judicial review.' "56 Based on these and similar concerns, the Report concluded that "a full re- view of OMB's rulemaking role is necessary." 57 In contrast, the minority statement in the Report was highly critical both of the Report's self-characterization as a Subcommit- tee Report and of the Report's condemnation of OMB's role, as well as its release of confidential deliberative process docu- ments. 58 The minority makes a strong case for the appropriate- ness of OMB's role in this matter.5 9 53. SUBCOMM. ON OVERSIGHT AND INVESTIGATIONS OF THE HOUSE COMM. ON ENERGY AND COMMERCE, 101ST CONG., lST SESs., THE SUPERFUND NATIONAL CONTINGENCY PLAN: RE- PORT ON A CASE STUDY OF OMB INVOLVEMENT IN AGENCY RULEMAKING (Comm. Print 101- B, 1989). 54. See id. at 27-138. 55. Id. at 7. 56. Id. at 10, quoting from SUBCOMM. ON OVERSIGHT AND INVESTIGATIONS OF THE HOUSE COMM. ON ENERGY AND COMMERCE, 99TH CONG., lST SESS., EPA's ASBESTOS REGULATIONS: CASE STUDY ON OMB INTERFERENCE IN AGENCY RULEMAKING 101 (Comm. Print 99-V, 1985). 57. Id. at 11. 58. Id. at 13-19. 59. Id. OMB and Cost-Benefit Analysis Another commentator has combined what is available from public sources with information obtained from a series of inter- views with OMB and EPA officials, most of whom are not identi- fied except by code letters, to produce what the author characterizes as a "guide to this elusive practice."60 The legal problems potentially posed by Order 12,291 were foreseen in the Order itself and by the Administration that issued the Order. The Order's directive to balance costs and benefits is expressly conditioned by the phrase "to the extent permitted by law." On its face, this qualification is supposed to assure that OMB review will abide by the standards set forth by Congress. 6' In 1981, James C. Miller, III, the administrator of OMB's Office of Information and Regulatory Affairs, testified before the House Oversight Committee on the importance of this provision of the Order, pointing to the inclusion of the provision and stating that "[t]he limited application of [Order 12,291] is a crucial point, one that insures [its] legality and the legality of actions pursuant to [it]."62 In addition, the Justice Department has also noted that Order 12,291 must be construed narrowly to survive legal chal- lenge. A 1981 Office of Legal Counsel Opinion states that, "it is clear that the President's exercise of supervisory powers must conform to legislation enacted by Congress. In issuing directives to govern the Executive Branch, the President may not, as a gen- eral proposition, require or permit agencies to transgress bound- aries set by Congress." 63 60. See Olson, supra note 14, at 40-73. Recently, an EPA official reported that OMB's rejection of hazardous waste incineration regulations under RCRA in March 1989 would also delay the similar regulations proposed for burning of hazardous waste fuels in boilers and industrial furnaces (52 Fed. Reg. 16,981 (1987)). The statutory deadline for the latter regulations was November 1986 (42 U.S.C. § 6924(q) (1990 Supp.)). Remarks of S. Silverman, Attorney, EPA Office of General Counsel, During ABA Satellite Seminar, Haz- ardous Waste and Superfund (April 27, 1989). 61. In testimony before the House Subcommittee on Oversight and Investigations on June 18, 1981, James C. Miller III, OMB Administrator for Information and Regulatory Affairs, testified, "If a statute expressly or by necessary implication prohibits the consider- ation of benefits or costs or alternatives by an agency during in its rulemaking, then those provisions of Executive Order 12,291 imposing them would not apply." Roe of OMB in Regulation: Hearings Before the Subcomm. on Oversight and Investigations of the House Comm. on Energy and Commerce, 97th Cong., 1st Sess. 46 (1981) [hereinafter House Oversight Hearings]. See also Olson, supra note 14, at 51 & n.255. 62. House Oversight Hearings, supra note 61, at 46. 63. 5 Op. Off. Legal Counsel 59, 61 (1981)(footnote omitted). 1991] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 However, in practice, OMB's review of agency rulemaking is di- rected not only by the statutory language, but also by politics. In- deed, former OMB officials have stated: The tension between an agency head's statutory responsibili- ties and his accountability to the president is not resolved in Executive Order 12,291 or in the earlier regulatory review or- ders. Nor is it resolved in any statute or in the Constitution itself. It is a political question that can be "answered" only through the tension and balance between the president and Congress - that is, the political branches - in overseeing the work of the agencies. 64 Considering its view of the relationship between the President and Congress, it is no surprise that OMB reportedly applies cost- benefit regulatory analysis even where the governing statute pre- cludes such considerations. 65 Moreover, there are indications that OMB has tried to shape regulations under such statutes by urging that the agency consider the cost impacts, albeit sub silentio.6 6 There is also evidence that OMB has, in practice, influenced the development of the regulations based on political considerations, which are neither cognizable under the statutory standard, nor under the cost-benefit analysis called for by the Executive Order. As some OMB officials have admitted, regulatory review can serve primarily as a vehicle for imposing the "cosmic presidential poli- cies," as OMB staff see them, on the rulemaking process. A key OMB official revealed that the real controversies surrounding OMB's review could not be resolved by debating the validity of cost-benefit analysis as an economic tool; "where OMB has budg- etary, philosophical or political problems with a rule, the regula- tory analysis is used as 'a key' in holding up or changing the EPA action." 67 Under the influence of OMB, EPA may be forced to reshape a regulation to fit with these political considerations and then re- write its regulatory statement of "basis and purpose," as required by the Administrative Procedure Act, to fit the rewritten regula- tion. Indeed, the Supreme Court may have foreshadowed just 64. DeMuth and Ginsburg, supra note 9, at 1083. 65. Olson, supra note 14, at 52 & nn. 261-62. 66. See testimony of former EPA Chief of Staff, John Daniel, quoted in Olson, supra note 14, at 51. 67. Olson, supra note 14, at 53 (footnotes omitted). OMB and Cost-Benefit Analysis such an exercise in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council 68 when it held, We have made it abundantly clear before that when there is a contemporaneous explanation of the agency decision, the va- lidity of that action must 'stand or fall on the propriety of that finding, judged, of course, by the appropriate standard of re- view .....' 69 In these circumstances, the agency's statement of basis and pur- pose may take on a surrealistic quality, since the written explana- tion is designed to justify the rule under statutory standards which did not form the true basis for decision. As Justice (then Judge) Scalia pointed out during a recent D.C. Circuit Judicial Conference: [O]ne way to look at the statement of basis and purpose is not as the real reason, it [sic] is heart of hearts, why the agency adopted the rule; but rather as a plausible justification that would make it reasonable to adopt the rule if that were the agency's genuine motivation. If that is a correct view of the matter, then what is the difference whether it is the agency itself or OMB that provides the unexpressed "heart of hearts" justification? The reviewing court can still determine whether the formally stated (nonheart-of-hearts) reason is adequate. 70 OMB's participation in rulemaking activity has also changed prior agency practices designed to assure an open public discus- sion of the regulatory issues. The process of OMB review and negotiation takes place, for the most part, out of view of the pub- lic.71 This non-public activity is necessary according to OMB "be- cause, like any other deliberative process, it can flourish only if the agency head or his delegate, and OMB as the president's dele- gate, are free to discuss frankly the merits of a regulatory proposal." 72 This practice has allowed interested parties a non-public "back- door" to influence the regulatory result and prompted EPA Gen- eral Counsel Joan Bernstein to suggest in testimony before a House Committee that a private practitioner would be seriously remiss in not lobbying OMB directly to influence the rulemak- 68. 435 U.S. 519 (1978). 69. Id. at 549. 70. A. Scalia, Remarks at the Proceedings of the Forty-Fifth Judicial Conference of the District of Columbia Circuit (May 21-22, 1984), reprinted in 105 F.R.D. 251, 335-36 (1984) (emphasis added) (hereinafter D.C. Circuit Judicial Conference). 71. DeMuth and Ginsburg, supra note 9, at 1085. 72. Id. 1991] 156 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 ing.73 As one successful practitioner has stated: "One group of your troops are in the agency working up a rulemaking record. The other set of your troops are over at OMB talking to David Stockman ... to get things arranged at the White House .... ,74 As a result of the secretive nature of the process, parties may not know what information has been provided EPA by OMB and hence do not have an opportunity to contest that information. 75 Finally, OMB's participation in agency rulemaking under the Order raises questions as to the scope of the President's authority to control agency rulemaking. On the one hand, the Constitu- tion, as interpreted by the courts, grants extensive powers, in- cluding the powers of appointment and removal, the power to require executive officers to supply written opinions, and the power to invoke executive privilege to protect the privacy of com- munications. 76 On the other hand, it is not clear that the Presi- dent has the authority to dictate the outcome of a rulemaking, particularly where Congress has "peculiarly and specifically com- mitted [duties] to the discretion of a particular officer. . . "77 In addition, OMB's exercise of the President's supervisory preroga- tive may be more troublesome since OMB officials may not have the scientific and technical expertise expected of agency person- nel, and the appearance of improper influence by third parties and secrecy may be exacerbated. The "corrective action" rule proposed by the EPA is a good example of the regulatory review process concerning OMB and EPA. 78 OMB's involvement in the progress of this proposed rule has been described as "a classic example of OMB exerting it's full force" under Order 12,291. 79 The proposed rule was drafted under the Resource Conservation Recovery Act to guide the clean up of active hazardous waste facilities and was intended to apply 73. Cited in Olson, supra note 14, at 56 & n.282. 74. D.C. Circuit Judicial Conference, supra note 70, at 334. 75. Id. at 337; see also Olson, supra note 14, at 55-62; Note, supra note 32, at 530-31; Morrison, supra note 13, at 1067-68. 76. Sierra Club v. Costle, 657 F.2d 298, 405 (D.C. Cir. 1981). 77. Myers v. United States, 272 U.S. 52, 135 (1926). Several Attorney General opinions emphasize that the President may not displace or interfere with subordinate officials exer- cising their statutory duties. See 1 Op. Att'y Gen. 624, 625-29 (1823); 18 Op. Att'y Gen. 31, 32-33 (1884); 19 Op. Att'y Gen. 685, 686-87 (1890). See also Public Citizen Health Research Group v. Tyson, 796 F.2d 1479, 1507 (D.C. Cir. 1986), clarified, 823 F.2d 626 (D.C. Cir. 1987). 78. 55 Fed. Reg. 30,798 (1990). 79. 21 Env't Rep. (BNA) 187 (1990). OMB and Cost-Benefit Analysis to approximately 5,700 facilities nationwide, including some facil- ities operated by federal agencies.8 0 Cost estimates of the cleanup that would be required by the proposed rule range from nearly $250 billion to $1 trillion,8' with federal facilities incurring a majority of the costs.8 2 EPA sent the draft proposed rule to OMB on October 7, 1988, but OMB did not release it until the summer of 1990, 21 months later. There was no statutory or judicial deadline for promulga- tion of the rule and it is not projected to be finalized until 1992 or later.85 Review of the rule was reportedly delayed by disagree- ment between EPA and OMB on a number of issues, including the cost estimates used by EPA to analyze the potential costs of the rule, the "point of compliance" at which clean up would be required, the availability of conditional remedies, and the appro- priate residual risk range.8 4 The practice that has evolved under Order 12,291 has funda- mentally changed the agency rule making process. While many of these changes are undoubtedly advantageous, some of them may undermine the standards set by Congress and the effectiveness of judicial review. In addition, they raise questions about the pub- lic's confidence in the rulemaking system and its appearance of fairness and openness. These issues are more fully discussed in Part IV. III. ADMINISTRATIVE LAW ISSUES RAISED BY EXECUTIVE ORDER 12,291 A. Case Law Interpretation of the Role of OMB OMB's role as the President's regulatory "traffic cop" has been examined in two recent cases. In Environmental Defense Fund v. Thomas,8 5 the Environmental Defense Fund ("EDF") challenged EPA's tardiness in promulgating regulations for underground tanks under the Resource Conservation and Recovery Act. The court determined that EPA was 16 months behind the statutory 80. 55 Fed. Reg. 30,798, 30,861 (to be codified at 40 C.F.R. §§ 264, 265, 270, 271) (proposed July 27, 1990). 81. 21 Env't Rep. (BNA), supra note 79, at 187-88. 82. Id. at 189. 83. 21 Env't Rep. (BNA) 653 (1990). 84. 21 Env't Rep. (BNA), supra note 79, at 188. 85. 627 F. Supp. 566 (D.D.C. 1986). 1991] 158 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 deadline for these regulations. EDF asked the court to restrict OMB's review of the regulations to the statutory time table. The court concluded that although a degree of deference should be given the President in the control and supervision of executive policy: the use of EO 12291 to create delays and to impose substantive changes raises some constitutional concerns. Congress enacts environmental legislation after years of study and deliberation, and then delegates to the expert judgment of the EPA Adminis- trator the authority to issue regulations carrying out the aims of the law. Under EO 12291, if used improperly, OMB could withhold approval until the acceptance of certain content ... thereby encroaching upon the independence and expertise of EPA. Further, unsuccessful executive lobbying on Capitol Hill can still be pursued administratively by delaying the enactment of regulations beyond the date of a statutory deadline. This is incompatible with the will of Congress and cannot be sustained as a valid exercise of the President's Article II powers. 86 Concerned that OMB might again delay the promulgation of the regulations beyond the deadline imposed by Congress, the court held that when a deadline for promulgation of regulations has been included in a statute, OMB has to carry out any regulatory review under Order 12,291 within this deadline. Likewise, if the deadline has already expired before OMB receives the proposed regulation, then no regulatory review is possible.8 7 The court concluded by noting that while its holding would necessarily in- terfere with the workings of OMB, that degree of interference was needed to ensure that the deadlines set by Congress were not abused. In the only other case to consider the role of OMB in reviewing EPA regulations under Order 12,291, the court found that, under the statute at issue, OMB's participation was not improper. In Sierra Club v. Costle,8 8 both environmental groups and industry pe- titioned for review of EPA's standards-under the Clean Air Act for coal-fired steam generators. In addition to numerous substantive issues, the environmental groups complained about EPA's proce- dures for dealing with post-comment period written and oral communications from the President and OMB. Relying on unique provisions of the Clean Air Act prescribing the process by 86. Id. at 570. 87. Id. at 571. 88. 657 F.2d 298, 396-408 (D.C. Cir. 1981). OMB and Cost-Benefit Analysis which the agency develops regulations, the environmental groups argued that the rule should be invalidated because EPA had met with the President and his representatives to review the proposal before it was promulgated and that these meetings had not been reflected in the agency record. Among the Clean Air Act provisions is the requirement that EPA prepare and maintain a "regulatory docket" containing all material received during the comment period and all other docu- ments which EPA receives later which it determines "are of cen- tral relevance to the rulemaking."8 9 The Act also expressly provides that "[t]he promulgated rule may not be based (in part or whole) on any information or data which has not been placed in the docket as of the date of such promulgation.- 90 The Court of Appeals noted that executive branch review of agency action plays a role in rulemaking, and rejected the argu- ment made by the environmental groups, stating that: [t]he authority of the President to control and supervise execu- tive policymaking is derived from the Constitution; the desira- bility of such control is demonstrable from the practical realities of administrative rulemaking. Regulations such as those involved here demand a careful weighing of cost, envi- ronmental, and energy considerations. They also have broad implications for national economic policy. Our form of govern- ment simply could not function effectively or rationally if key executive policymakers were isolated from each other and from the Chief Executive. Single mission agencies do not always have the answers to complex regulatory problems. An over- worked administrator exposed on a 24-hour basis to a dedi- cated but zealous staff needs to know the arguments and ideas of policymakers in other agencies as well as in the White House.91 The Court went on to note, however, that in some situations, even discussions between the President and the agency might have to be included in the administrative record. For example, a statute may, like the Clean Air Act, require that information or data the agency receives from the President or OMB and on which it relies in adopting the final regulation be included in the record. 92 89. 42 U.S.C. § 7607(d)(4)(B)(i) (1990). 90. 42 U.S.C. § 7607(d)(6)(C) (1990). 91. 657 F.2d at 406. 92. Id. at 406-07. 1991] 160 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 The lesson of these two cases is that the courts will defer to the President and his "traffic cop" - OMB - unless a particular stat- utory requirement would be violated. 93 Thus, where the statute requires either the promulgation of the regulation within a speci- fied time frame, or the inclusion in the record of all material in- formation on which the agency relied in adopting the rule, OMB's conduct will in effect be open to judicial review. B. The Need to Adhere to Statutory Standards The courts have generally restricted regulatory agencies in rulemaking to the standards established by Congress in the gov- erning statutes. Moreover, absent some clear indication that Congress intended that health and environmental impacts be balanced against costs, the courts have precluded such consideration. In American Textile Manufacturers Institute, Inc. v. Donovan,94 the Supreme Court first faced the question of whether a specific regu- latory statute required cost-benefit balancing in formulating in- dustry-wide health standards. Occupational exposure to cotton dust (an airborne particle byproduct of manufacturing cotton products) had been demonstrated to cause respiratory disease.9 5 In 1978, the Occupational Safety and Health Administration adopted final standards for worker exposure to cotton dust.9 6 The cotton industry challenged these rules principally on the ground that the agency had failed to show that the costs of the standards bore a reasonable relationship to their benefits. The industry's basic position was that it is not enough that some worker health risks would be reduced by a particular numerical standard. Rather, the agency must find that a particular level of reduced exposure would produce significant health benefits 93. However, the Supreme Court recently refused to defer to OMB's interpretation of its authority under the Paperwork Reduction Act of 1980. In Dole v. United Steelworkers of America, 110 S. Ct. 929 (1990), the Court held that OMB's authority under the Act to review agency "information collection requests" did not authorize OMB to review and countermand Department of Labor regulations mandating disclosure by regulated entities to third parties. The regulations at issue - requiring employers to disclose potential hazards posed by chemicals in the workplace - did not result in information being made available for agency use and therefore did not constitute an information collection request. 94. 452 U.S. 490 (1981). 95. 43 Fed. Reg. 27,352-54 (1978). 96. Id. at 27,350. 1991] OMB and Cost-Benefit Analysis 161 which were proportionate to, and justified in terms of, their cost.97 Obviously, the industry hoped that if such analysis were required, it could prove that the costs of the final standards were disproportionate. Although neither the statute nor its legislative history spoke in terms of "cost-benefit" analysis, the statute did expressly require the protection of worker health " 'to the extent feasible.' " The industry contended that "feasibility" in effect required the agency to balance costs and benefits. This interpretation had been sup- ported by several decisions of lower federal courts and by an ear- lier concurring opinion of at least one member of the Supreme Court. 98 The Supreme Court rejected the industry's argument. Accord- ing to the Court, the common meaning of "feasible" was "capa- ble of being done." Though the requirement of feasibility might place an outer limit on some rules, it did not compel cost-benefit balancing in the Court's view. Moreover, the Court concluded that when Congress intended cost-benefit balancing, it indicated its intent much more clearly - as in the Flood Control Act of 1936, for example. The Court, therefore, held that the agency was not required to strike a substantive balance between costs and benefits. One very significant and troubling question remained in dis- pute following the American Textile decision: Did the Court intend that absent a clear Congressional directive to engage in cost-ben- efit analysis, such analysis is precluded, or did it mean merely that cost-benefit analysis is not required in that situation? Justice Rehnquist's dissenting opinion states that the holding of the case meant that cost-benefit analysis is permitted, but not compelled. 99 Some commentators have reached the same interpretation.O° Others have interpreted the holding to preclude cost-benefit 97. See 452 U.S. at 506-07 n.26. 98. Texas Indep. Ginners Ass'n v. Marshall, 630 F.2d 398 (5th Cir. 1980); American Petroleum Inst. v. OSHA, 581 F.2d 493 (5th Cir. 1978), aff'd on other grounds sub norm., Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607, 663-65 (1980) (Powell J., concurring). While the majority in the latter case never reached the issue, the concur- ring opinion explicitly construed the statute to compel cost-benefit balancing. Justice Powell did not participate in the American Textile decision, however. 99. 452 U.S. at 545. 100. See, e.g., Hadley & Richman, The Impact of Benzene and Cotton Dust: Restraints on the Regulation of Toxic Substances, 1981 SuP. CT. REV. 291, 302-05, reprinted in 34 ADMIN. L. REV. 59, 68 (1982). 162 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 analysis absent Congressional intent that it be applied.' 0 ' How- ever, it seems fairly clear that the Court meant to preclude cost- benefit balancing under the particular statute involved in American Textile, at least to the extent that it would alter the regulatory re- sult. As the Court explained its decision: Congress itself defined the basic relationship between costs and benefits, by placing the "benefit" of worker health above all other considerations save those making attainment of this "benefit" unachievable. Any standard based on a balancing of costs and benefits by the Secretary that strikes a different balance than that struck by Congress would be inconsistent with the command set forth in § 6(b)(5). Thus, cost-benefit analysis by OSHA is not required by the statute because feasibility analysis is. 10 2 The D.C. Circuit has addressed some of the issues remaining open after American Textile in a variety of contexts. In Lead Indus- tries Association, Inc. v. EPA,'03 the Court of Appeals held that under the Clean Air Act, EPA is not required to consider eco- nomic feasibility in setting ambient air quality standards for lead. Industry argued that EPA exceeded its statutory authority by promulgating a standard "more stringent than ... necessary to protect the public health."1o4 They argued that since "Congress only authorized" EPA to set standards "aimed at protecting the public against health effects which are known to be clearly harmful ... [Congress] was concerned that excessively stringent air qual- ity standards could cause massive economic dislocation."' 10 5 Thus, "EPA erred by refusing to consider the issues of economic and technological feasibility in setting the air quality standards for lead." 0 6 The court rejected this argument as "totally without merit," since the "statute and its legislative history make clear that economic considerations play no part in the promulgation of 101. See The Supreme Court, 1980 Term, 95 HARV. L. REV. 91, 319-24 (1981). 102. 452 U.S. at 509 (emphasis added). In Aqua Slide "N" Dive Corp. v. Consumer Prod. Safety Comm'n, 569 F.2d 831 (1978), the U.S. Court of Appeals for the Fifth Circuit held that the Consumer Product Safety Act required at least a generalized cost-benefit balancing. Although the Act does not expressly mandate "cost-benefit" balancing, it re- fers to the need to regulate "unreasonable risk" which is explained in the legislative his- tory to require such balancing. The Senate Report accompanying the Act explains that it requires "balancing the probability that risk will result in harm and the gravity of such harm against the effect on the product's utility, costs, and availability to the consumer." S. REP. No. 749, 92nd Cong., 2d Sess. 14-15 (1972). 103. 647 F.2d 1130 (D.C. Cir.), cert. denied, 449 U.S. 1042 (1980). 104. Id. at 1148. 105. Id. (emphasis in original). 106. Id. OMB and Cost-Benefit Analysis ambient air quality standards .. .. "107 The court concluded that there was no indication of any Congressional intent that EPA take economic or technological factors into account in setting air qual- ity standards and that the agency was limited to considering only those factors specified by Congress in determining standards.' 0 In a recent decision, Public Citizen v. Young, 10 9 the Court of Ap- peals rejected the Food and Drug Administration's attempt to avoid the impact of the Delaney Clause through reliance on an asserted de minimis level of carcinogenic dyes. The "Delaney Clause" in the Food, Drug and Cosmetic Act" 0 specifies that "no additive shall be deemed to be safe if it is found to induce cancer when ingested by man or animal. . . ." The Court reviewed the extensive legislative history of the clause and concluded that Con- gress had intended an " 'extraordinarily rigid' " statutory rule which did not allow for an exemption for small quantities or triv- ial effects. " I Based on this reasoning, the Delaney Clause would obviously not permit cost-benefit balancing. In another recent decision, Natural Resources Defense Council v. EPA," 2 the Court of Appeals found that the standard for air tox- ics must initially be set based on safety considerations only. Sec- tion 112 of the Clean Air Act provides for the regulation of "hazardous air pollutants," which it defines as pollutants "which may reasonably be anticipated to result in an increase in mortality .... ", The statute directs the Administrator of EPA to set an emission standard under § 112 "at the level which in his judg- ment provides an ample margin of safety to protect the public health ... ."114 EPA determined that vinyl chloride was carcino- genic and proposed that it would either set a zero tolerance for 107. Id. In Union Electric Co. v. EPA, 427 U.S. 246, 253 (1976), the Supreme Court rejected a challenge to a state implementation plan on the grounds that "economic and technological difficulties ... made compliance with the emission limitations impossible." The Court found that "Congress intended claims of economic and technological in- feasibility to be wholly foreign to the Administrator's consideration of a state implementa- tion plan." Id. at 256. 108. 647 F.2d at 1150. See also EPA v. National Crushed Stone Ass'n, 449 U.S. 64, 78 (1980) (although the Clean Water Act directs EPA "to consider the benefits of effluent reductions as compared to the costs of pollution control in determining BPT limitations," the statute does not allow such consideration in granting variances from those limitations). 109. 831 F.2d 1108 (D.C. Cir. 1987). 110. 21 U.S.C. §§ 301-92, 348(c)(3)(A) (1989). 111. 831 F.2d at 1117. 112. Natural Resources Defense Council v. EPA, 824 F.2d 1146 (D.C. Cir. 1987). 113. 42 U.S.C. § 7412(a)(1) (1988). 114. 42 U.S.C. § 7412(b)(l)(B) (1988). 1991] 164 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 emissions, or, if the costs of such a zero tolerance were found to be "grossly disproportionate to the benefits," would authorize "the lowest level achievable by use of the best available control technology." NRDC petitioned for review on the ground that EPA was required by the statute to adopt a zero tolerance if it was unable to determine any "safe" threshold. EPA, on the other hand, claimed that the statute allowed it to balance cost and tech- nological factors against risk, relying on the decision of a panel of the D.C. Circuit holding that EPA could reasonably consider "economic and technological feasibility" in setting vinyl chloride standards."15 On rehearing en banc, the full Court of Appeals de- termined that the statute supported neither the argument of the petitioner nor of the agency. Rather, by focusing on an "ample margin of safety," the Court held that Congress intended that the agency first establish a "safe" level taking into account risk factors only. After this "safe" level was established, the Administrator could take into account cost and technological factors in deter- mining whether an "ample margin of safety" requires still further reduction of the emission level."l 6 In Center for Science in the Public Interest v. Department of the Treas- ury," i7 the district court found that cost-benefit balancing man- dated by Executive Order 12,291 could not lawfully be applied under the Federal Alcohol Administration Act."18 Pursuant to that Act, the Secretary of the Treasury is directed to prescribe regulations requiring ingredient disclosure on labels of wine, dis- tilled spirits, and malt beverages. Regulations were proposed in February of 1979, and a final rule was issued on June 13, 1980. Prompted by Order 12291, the agency reviewed the regulations in 1981 and on November 6, 1981, rescinded them, stating that (1) "ingredient labeling regulations would result in increased costs to consumers and burdens on industry which are not com- mensurate with the benefits which might flow from the additional label information," and (2) "ingredient labeling would not result in an appreciable benefit to consumers when compared to the ex- 115. Natural Resources Defense Council v. EPA, 804 F.2d 710 (D.C. Cir. 1986). 116. EPA has recently interpreted and applied the holding of the Vinyl Chloride decision in proposals to establish emission levels for another hazardous air pollutant - benzene. See 53 Fed. Reg. 28,496 (1988) (to be codified at 40 C.F.R. Part 61) (proposed July 28, 1988). 117. 573 F. Supp. 1168 (D.D.C. 1983), appeal dismissed, 727 F.2d 1161 (D.C. Cir. 1984). 118. 27 U.S.C. § 201 (1989). OMB and Cost-Benefit Analysis isting label information requirements and standards of iden- tity." 119 Consumer groups challenged the agency's decision to rescind the regulations, and the district court invalidated the deci- sion. The court held that the Federal Alcohol Administration Act precludes the consideration of costs mandated by Order 12,291 and engaged in by the agency: Whatever the consideration given to costs and benefits, how- ever, an agency may not substitute its policy judgment for the judgment that has already been articulated by Congress. In this case, Congress announced that the Department had the au- thority to issue regulations requiring producers of alcoholic beverages to adequately inform the consumer of the identity and quality of the products, the alcoholic quantity thereof, the net contents and the name of the bottler. It did not condition such a grant of authority with a proviso that the regulations could be withdrawn if the costs to the industry turned out to be too high.120 The message of these cases is that Congress, not the agencies, has the role of determining federal policy, and that it does this through the adoption of statutes. Through statutes, it is Con- gress which strikes the balance between costs and benefits and sets the overall standards that must be applied in rulemaking. Agencies may not recast the balance struck by Congress in a stat- ute. Yet, the message of these cases may not be very far-reaching since there are many statutes, such as that involved in the Vinyl Chloride case, in which Congress has been less than articulate in defining the applicable standards. C. The Need to Compile an Accurate Administrative Record Since many OMB contacts with EPA are not included in the public record of a rulemaking, issues of accuracy of the record as a result of the OMB's regulatory review may arise in some cases even though most EPA rules do not have to be made "on the record." Generally, the courts have recognized an important dif- ference between informal rulemaking and rulemaking "on the 119. 46 Fed. Reg. 55,094 (1981). 120. 573 F. Supp. at 1174. But cf American Pilots' Ass'n, Inc. v. Gracey, 631 F. Supp. 827 (D.D.C. 1986) (Coast Guard may take into account regulatory costs in designing li- censing regulations). 1991] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 record." 12 1 The latter is usually thought of as involving trial-type proceedings and findings of fact on the basis of the trial-type rec- ord. The administrative record for purposes of informal rulemak- ing is often not easy to define. Adding to this complexity is the plethora of different provisions in the statutes which EPA administers. The Clean Air Act, as noted above in the discussion of the Sierra Club case, has detailed provisions which require that EPA compile a record consisting of all the documents of importance to its rulemaking decisions and that the agency base its decisions on the record. 22 The Toxic Substances Control Act has similar provi- sions which require the agency to base its rules on the "rulemak- ing record."'' 23 Like the Clean Air Act, however, this Act does not require formal rulemaking "on the record" within the meaning of the Administrative Procedure Act. In addition, CERCLA now re- stricts judicial review of the adequacy of response actions to the "administrative record"' 24 and defines the record in terms which are similar to those used in the Clean Air Act.' 25 These provisions bolster the general requirements of adminis- trative law and seem to create a trend toward a more formal type of "record." However, provisions which restrict judicial review to the agency "record" are a two-edged sword. On the one hand, they assist the agency by cutting off the opportunity for trial de novo, or for the introduction of evidence in most cases. On the other hand, however, they place greater emphasis on the develop- ment of the administrative record and necessarily tend to give in- terested parties greater access to that record. Even aside from the effects of cases involving incorporation of cost-benefit analysis in rulemaking processes, the agency's rulemaking record has taken on increasing significance. The Fed- eral Rules of Appellate Procedure expressly provide that, "The order sought to be reviewed or enforced, the findings or report on which it is based, and the pleadings, evidence and proceedings before the agency shall constitute the record on review in pro- 121. The Administrative Procedure Act recognizes a formal distinction between infor- mal rule making governed by 5 U.S.C. § 553 and formal rule making based "on the rec- ord" before the agency and governed by 5 U.S.C. § 556. 122. 42 U.S.C. § 7607(d)(4) (1988). 123. 15 U.S.C. § 2618(a)(3) (1988). 124. 42 U.S.C. § 9613(j)(2) (1988). 125. Id. at § 9613(k). OMB and Cost-Benefit Analysis ceedings to review or enforce the order of an agency."' 26 These Rules also authorize the courts to supplement an agency record as needed to correct omissions or misstatements.127 Courts have not hesitated to invoke this power by requiring that a supplement be filed when a supplementary record is necessary in order for the court to effectively review an agency's actions. 128 In addition, there are established public expectations that an agency, particularly EPA, will provide for full public participation in its rulemaking process. Several environmental statutes ex- pressly provide for public participation 2 9 and EPA's general reg- ulations governing rulemaking proceedings provide that the agency include comments from any interested party plus any agency responses to comments in the public record and allow public inspection of the record.' 30 Both statutory provisions re- quiring a more formal record and efforts made by an agency on its own initiative to increase the inclusiveness and openness of the record produce an inherent tension between an agency's desire to cut off evidentiary proceedings on judicial review and its desire to maintain a somewhat elastic concept of the administrative record. The more that interested parties are cut-off from presenting evi- dentiary material on judicial review, the more likely courts are to demand a more formal record. That tension is the inevitable consequence of the Supreme Court's decision in Camp v. Pitts.'3 ' In Camp, a bank challenged the decision of the Comptroller of the Currency to deny the issu- ance of a national bank charter. The district court affirmed the Comptroller and the Court of Appeals reversed, holding that de novo review was required because the Comptroller had not suffi- ciently stated the basis for his decision. The Supreme Court re- versed the Court of Appeals on the ground that where the agency decision was not required to be made on the basis of a trial-type hearing, "the focal point for judicial review should be the admin- 126. Fed. R. App. P. 16(a). 127. Fed. R. App. P. 16(b). 128. Bethlehem Steel Corp. v. EPA, 638 F.2d 994, 1000 (7th Cir. 1980). 129. See, e.g., Resource Conservation and Recovery Act § 7004(b), 42 U.S.C. § 6974(b) (1988); Comprehensive Environmental Response, Compensation, and Liability Act of 1980 § 117, 42 U.S.C. § 9617 (1988). 130. 40 C.F.R. § 25.10(a) (1990). 131. 411 U.S. 138 (1973). 1991] 168 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 istrative record already in existence, not some new record made initially in the reviewing court."' 132 As further insurance that a rulemaking record accurately re- flects the information or data considered by the agency, the Ad- ministrative Procedure Act, and some statutes, require disclosure of ex parte communications in certain proceedings.13 3 As the Dis- trict of Columbia Court of Appeals noted in Professional Air Traffic Controllers Organization v. Federal Labor Relations Authority, disclosure of ex parte communications: serves two distinct interests. Disclosure is important in its own right to prevent the appearance of impropriety from secret communications in a proceeding that is required to be decided on the record. Disclosure is also important as an instrument of fair decision making; only if a party knows the arguments presented to a decisionmaker can the party respond effectively and ensure that its position is fairly considered.13 4 However, the APA ex parte disclosure requirements have limited application since they do not apply to informal rulemaking, such as that at issue in Sierra Club v. Costle.' 35 The court in Sierra Club found that neither general administrative law principles nor the Clean Air Act required disclosure of intra-executive meetings be- tween OMB and EPA. Although the APA requires disclosure of information on which the agency relied in making its decision, the meetings did not have to be disclosed "since EPA makes no effort to base the rule on any 'information or data' arising from that meeting." 36 An even more difficult problem arises if OMB officials act as "conduits" for private parties in reviewing and formulating agency regulations. In Sierra Club, the court noted that it was not addressing the disclosure of "so-called 'conduit' communications, in which administration or inter-agency contacts serve as mere conduits for private parties in order to get the latter's off-the-rec- ord views into the proceeding."'13 7 Under the current practice, it 132. Id. at 142. 133. See, e.g., 5 U.S.C. §§ 554(d), 557(d) (1988); Action for Children's Television v. FCC, 564 F.2d 458, 469 (D.C. Cir. 1977); Orangetown v. Ruckelshaus, 740 F.2d 185, 188 (2d Cir. 1984). 134. Professional Air Traffic Controllers Org. v. Federal Labor Relations Auth., 685 F.2d 547, 563 (D.C. Cir. 1982). 135. 657 F.2d 298 (D.C. Cir. 1981). See also Center for Science in the Public Interest v. Treasury, 797 F.2d 995 (D.C. Cir. 1986). 136. 657 F.2d at 407. 137. Id. at 405 n.520. 9OMB and Cost-Benefit Analysis is clear that the administrative record may not be an accurate re- pository of all agency contacts. IV. ISSUES RAISED BY THE REGULATORY PROCESS UNDER EXECUTIVE ORDER 12,291 REGARDING THE VALUE OF COST-BENEFIT ANALYSIS IN ENVIRONMENTAL DECISIONMAKING "Cost-benefit analysis" has become a term of art with two alter- native meanings. First, the phrase can refer to a formal decision making process in which costs and benefits are identified, quanti- fied, reduced to monetary terms and balanced against each other. The regulatory decision is then made by the mere tilt of the scale - that is, if the benefits are greater than the costs, then the pro- posed regulation should proceed, but if the costs are greater than the benefits, the regulation is not justified. This has been re- ferred to as "technical cost-benefit analysis." Second, cost-bene- fit analysis has also come to refer to a much less rigid framework for decisional analysis in which relevant information is organized according to the costs and benefits of particular decisions, but no formal, mathematical balancing takes place. This has been re- ferred to as "cost-benefit organization"'' 38 and resembles the ap- proach which would have been required by S. 1080. It satisfies the need for some reasonable assurance that regulatory compli- ance money is being well spent while avoiding the rigidity and other problems associated with technical cost-benefit analysis. Society as a whole, and the regulated industry in particular, should be able to insist that a cost-benefit organizational ap- proach be employed to determine that there will be identifiable benefits which bear some rough relationship to the costs of ob- taining them. The present implementation of the Superfund pro- gram, for example, starkly demonstrates the need for this type of analysis. Under that program, EPA is constantly being con- fronted with cleanup plans which pose questions such as whether a further 10% reduction in contaminant levels at a site is worth a 50% increase in the cost of cleanup. Technical cost-benefit analysis has been the subject of wide- spread commentary, scientific discussion and criticism. Indeed, 138. The United States Court of Appeals for the Tenth Circuit has acknowledged this distinction in the types of cost-benefit analysis. See Quivira Mining Co. v. Nuclear Regula- tory Comm'n, 866 F.2d 1246, 1250 (10th Cir. 1989). 1991] 170 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 the recent Congressional proposals to require a cost-benefit anal- ysis deliberately steered away from the more rigid balancing tech- nique precisely because of these criticisms.' 3 9 Technical cost- benefit analysis suffers from especially serious limitations when applied to environmental decisionmaking. These limitations can be classified as problems relating to (a) distribution, (b) identifica- tion, (c) quantification, (d) valuation, and (e) present value calculation. Technical cost-benefit analysis measures only economic effi- ciency as a whole and ignores the distribution of those efficien- cies. Many, if not all, regulatory decisions of "major" proportions have costs which fall with disproportionate impact on one particular industry or plant while the benefits accrue to a dif- ferent particular segment of society, often located in a different geographical area. For example, the regulatory decision to remove a chemical pes- ticide from the market because of a cancer risk may force the clo- sure of one plant in one town. Although that cost is small on a national basis, it could be devastating to the particular town. Likewise, the benefits of removing a pesticide from the market if it is non-persistent or non-mobile would be felt principally in the areas where it is applied. The cost could be incurred most harshly in one town in New Jersey while the benefit might occur among migrant farm workers in Florida. Thus, the problem with technical cost-benefit analysis is that it can "often obscure important distributional considerations."'140 While the method can "identify which proposals offer the most pie for the money, [o]rdinarily, it does not consider the sizes of the slices that get passed around, or who receives them."' 4 1 In addition to this inherent limitation in the technical cost-ben- efit approach, the approach has serious practical limitations. To begin with, it requires identification of all costs and benefits. This is, at the least, difficult, and with respect to the benefits of many environmental and health decisions, it may even be impossible. The classic example is the European clearance for pre-natal use of the drug thalidomide. Animal tests had failed to identify the birth defects which the drug ultimately produced in humans. This type 139. See, e.g., H.R. REP. No. 435, 97th Cong., 2d Sess. 45 (1982). 140. NATIONAL ACADEMY OF SCIENCES, DECISION MAKING FOR REGULATING CHEMICAL IN THE ENVIRONMENT 42 (1975) [hereinafter 1975 NAS REPORT]. 141. Rodgers, supra note 16, at 194. 11MB and Cost-Benefit Analysis of uncertainty continues to plague decisionmaking with respect to many chemicals in the environment. 142 Moreover, since "more study appears to have been done of the costs of regulation than of its benefits,"' 143 it is likely that more costs will be identified by the cost-benefit approach. Even in the case of costs, however, identi- fication is difficult, particularly for more complex environmental decisions. Equally troubling is the absolute dependence of technical cost- benefit analysis on precise quantification of costs and benefits. As the Senate Report accompanying S. 1080 concluded: By almost all accounts, giving the benefits of a proposed rule expression in numbers is the hardest task in any effort to quan- tify the effects of regulation. Giving a precise estimate of some benefits may not even be possible. For example, even if the Administrator of EPA knew that a particular hazardous waste regulation would reduce human exposure to a given carcino- gen by a certain amount, she could not be sure how much the actual incidence of cancer would be reduced.' 44 The chronic human health hazards may be the most difficult to quantify. Hazard data are typically extrapolated from high-dose animal feeding tests. Various mathematical models have been de- veloped to estimate the potential human health hazard from these data, but basic uncertainties exist about the mechanisms of these chronic diseases and their natural history in animals and humans, and thus their potential incidence among large, genetically di- verse communities of humans, who are exposed to differing con- centrations of other active compounds, precludes precise quantification. Moreover, the available techniques for estimating such effects produce widely divergent predictions. One commen- tator stated that: In assessments of the risks posed by saccharin, for example, the multi-stage model predicted five cancer cases per million per- sons exposed, the probit model predicted 450, and the one-hit model predicted 1,200. A fourth model offered by the Food Safety Council, an industry group, predicted only one death per billion people exposed, a million-fold less than the one-hit model prediction. To some extent the regulator can avoid the risk of catastrophic error in choosing among models by using the "conservative" 142. 1975 NAS REPORT, supra note 140, at 43-44; Rodgers, supra note 16, at 194. 143. S. REP. No. 284, supra note 6, at 19. 144. Id. at 83. 1991] 172 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 one-hit model. However, the huge range of estimates which the models produce is compelling evidence of the magnitude of scientific ignorance about the true relationships which quantita- tive risk assessment methodologies attempt to approximate. 145 Quantification of costs also remains troublesome, since in these volatile economic times, quantifying future economic effects can be extremely difficult. In addition to quantification, technical cost-benefit analysis also depends on reducing all costs and benefits to dollars and cents. The insurmountable problem with such "monetization" is that "[m]any factors cannot be satisfactorily expressed in dollar terms because they involve important values on which there is no agree- ment.' ' 146 The most important example of this problem is how to value a human life in dollars. Although many of the difficulties should be obvious, this has not prevented regulatory agencies from jumping into the fray. The EPA Office of Radiation Pro- grams, for example, once decided to set human radiation stan- dards to require no more than the expenditure of $500,000 per life saved. 147 Similarly, the Consumer Product Safety Commis- sion has variously used figures of $200,000 to $2 million per life.'14 Huge amounts of money and research time have been de- voted to devising various techniques for valuing life. Under one popular method, life is valued based on the individ- ual's lifetime earnings less anticipated consumption. This is somewhat piously called the "human capital - net value" ap- proach. 149 A variation looks to the "gross value," ignoring the individual's consumption. The valuation techniques based on lifetime earnings have obvious and devastating flaws. They con- tain the implicit political judgment that high wage earners are worth more than lower wage earners. Equally repugnant, how- ever, is that they would view the death of non-income producers (e.g., retired people, disabled people or the unemployed) as a benefit rather than a cost. Finally, they totally ignore the demand of all to survive. ' 50 145. Leape, Quantitative Risk Assessment in Regulation of Environmental Carcinogens, 4 HARV. ENVTL. L. REV. 86, 103 (1980). 146. 1975 NAS REPORT, supra note 140, at 41. 147. S. REP. No. 284, supra note 6, at 86. 148. Id. 149. 2 NATIONAL ACADEMY OF SCIENCES, DECISION MAKING IN THE ENVIRONMENTAL PRO- TECTION AGENCY 230 (1977). 150. Id. at 231. OMB and Cost-Benefit Analysis To overcome these inherent valuation problems, other analysts have looked to damage awards from juries and to the results of questionnaires which poll individuals as to how much they would be willing to pay to live longer.' 5 ' The results indicate that, not surprisingly, the rich are willing to pay more than the poor. In addition, one difficulty with questionnaires is that respondents may not fully comprehend the significance of the questions, and may therefore provide superficial responses. As the National Academy of Sciences has properly concluded after reviewing all of these methodologies: "[T]here is at present no single reliable and generally accepted way of placing a dollar value on the bene- fits that will be realized under a public program that reduces mor- tality or morbidity."1 52 Since costs and benefits are likely to occur not simply now but in the future, technical cost-benefit analysis also requires that fu- ture costs and benefits be discounted to present values. To begin with, "[i]n some sense, then, discounting future benefits and costs, represents discrimination against future generations."'' 53 Moreover, it poses inherent technical problems in choosing an appropriate discount factor both for costs (e.g., lost future jobs) and benefits (e.g., saved future lives). In many cases, these dis- count factors are largely arbitrary and they can seriously distort the cost-benefit balance.' 54 All of these technical problems are exacerbated in practice be- cause agencies find it difficult, if not impossible, to develop and apply consistent standards for gathering the needed cost-benefit information. A report of the General Accounting Office con- cluded, for example, that inconsistent agency estimates of recrea- tion benefits for water projects could "cause a project's final benefit-cost ratio to range from a high of 1.58 to a low of .89, depending on the method used to compute the number of visitors using the facility."' 55 Thus, inconsistent techniques can artifi- cially change the regulatory decision from "no" to "yes." The 151. Id. at 232. 152. Id. at 234. 153. Id. at 236. 154. Discount rates for Federal Water projects have varied from 2-1/2% to 6-7/8%. See Rodgers, supra note 16, at 198 and 198 n.42. 155. COMPTROLLER GENERAL OF THE UNITED STATES, REPORT To THE CONGRESS: AN OVERVIEW OF BENEFIT-COST ANALYSIS FOR WATER RESOURCES PROJECTS - IMPROVEMENTS STILL NEEDED 8 (1978). 1991] 174 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 National Academy of Sciences reached the inescapable conclusion that: [h]ighly formalized methods of benefit-cost analysis can seldom be used for making decisions about regulating chemicals in the environment. Thus the development of such methods should not have high priority. However, the benefit-cost and decision frameworks . . . can be useful in organizing and summarizing relevant data on regulatory alternatives which the decision maker must review. 156 In an effort to confront some of the technical problems with cost-benefit analysis, in December 1983, EPA published "Guide- lines for Performing Regulatory Impact Analysis" to assist Agency personnel in complying with the requirements of Execu- tive Order 12,291.' 57 These Guidelines follow the outline of OMB's guidance documents, and broadly address the analytical techniques that may be used and the information that may be de- veloped in analyzing identification, quantification, valuation, present value and distributional factors relating to EPA regulations. Although the Guidelines recognize that health and environ- mental effects may be difficult to identify, quantify and value, they encourage agency analysts to perform this function in the major- ity of cases. Thus, the Guidelines provide that: [t]he major objective of economic valuation is to transform esti- mates of changes in physical or biological effects into monetary estimates of benefits. This is done by using the amount indi- viduals would pay for such changes as a measure of their value, i.e., benefits should be measured in terms of willingness to pay. 158 The Guidelines suggest that in valuing human morbidity (illness), EPA should look first to the estimate of the direct cost of medical treatment, unless it is feasible to use willingness to pay as an indi- cator. For human mortality (death), however, the Guidelines state that valuation can follow either approach. If mortality is val- ued by the willingness-to-pay method, however, the Guidelines 156. 1975 NAS REPORT, supra note 140, at 44. See also COMPTROLLER GENERAL OF THE UNITED STATES, REPORT TO THE CONGRESS: COST-BENEFIT ANALYSIS CAN BE USEFUL IN ASSESSING ENVIRONMENTAL REGULATIONS, DESPITE LIMITATIONS (1984) [hereinafter GAO REPORT]; CONSERVATION FOUNDATION REPORT, supra note 15. 157. OFFICE OF POLICY ANALYSIS, EPA, GUIDELINES FOR PERFORMING REGULATORY IM- PACT ANALYSIS (1983). 158. Id. at 10. OMB and Cost-Benefit Analysis state that recent studies indicate that workers are generally paid $4 to $70 more for working in an environment with a 1 in 100,000 greater than average risk. The Guidelines go on to conclude that "[t]his translates into a value for a statistical life of roughly $400,000 to $7,000,000 (in 1982 dollars)."' 159 V. REFINEMENT OF THE CURRENT REGULATORY PROCESS Most observers of the regulatory process would probably agree that by coordinating the process, reducing the number of un- resolved conflicts among agencies and providing a structured means for evaluating costs and benefits, Executive Order 12,291 has been successful. Indeed, both the Government Accounting Office and the American Bar Association have generally endorsed the cost-benefit analysis regulatory reform effort.' 60 However, the regulatory review process requires refinement to: (1) assure a more accurate statement of the rulemaking agency's basis and purpose and adherence to the statutory standards; (2) enhance the accuracy of the administrative record; and (3) avoid the problems of an unduly strict cost-benefit analysis. There have been recent legislative and agency attempts to effect some of these changes. A. Assuring an Accurate Statement of Basis and Purpose and Adherence to the Statutory Standards The administrators of the third century Chinese Empire appar- ently understood clearly that "[p]olitics is the hidden face of the law. The latter concerns everyone; the former is the exclusive property of the prince. Politics is secret and impenetrable, whereas the law is public and universal."'' Executive Order 12,291 is the modern means by which the "prince" controls his administrators and assures himself that the "hidden face of the law" - politics - plays a prominent role in shaping the adminis- tration's regulatory programs. Like the Chinese prince, the Presi- 159. Id. at 11. 160. See GAO REPORT, supra note 156; AMERICAN BAR ASSOCIATION, ADMINISTRATIVE LAW SECTION, RECOMMENDATION, reprinted as Appendix to Strauss and Sunstein, The Role of the President and OMB in Informal Rulemaking, 38 ADMIN. L. REV. 181, 206 (1986). 161. J. LEVI, THE CHINESE EMPEROR 151 (1987). The Chinese Emperor also under- stood that "[a] prince must provide himself with the means of controlling the administra- tion, for his subordinates' one idea is to usurp his place. The art of politics lies in detecting the thoughts of one's executives and finding out all about them - to be able to terrorize them." Id. 1991] 176 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 dent has a strong, clear and legitimate interest in accomplishing these dual objectives. And yet, as the Chinese administrators un- derstood 1,600 years ago, "politics is secret... whereas the law is public and universal."' 16 2 This is the unresolved tension which is inherent in the system created by Executive Order 12,291. The exercise of control and the opportunity to interject political considerations may be best served by a "secret and impenetrable" process. As some exper- ienced OMB officers have noted, "like any other deliberative pro- cess, it can flourish only if the agency head or his delegate, and OMB as the president's delegate, are free to discuss frankly the merits of a regulatory proposal."'' 63 However, in order for the public to have continued confidence in the regulatory system, and the courts to exercise meaningful judicial review, the agency must "say what it means, and mean what it says." If the agency has a hidden reason for choosing a particular regulatory course, or ba- ses its decision on hidden, non-public factual material, both pub- lic confidence and judicial review will be seriously undermined. How can these seemingly irreconcilable considerations be rec- onciled? Can politics continue to be the "hidden face of the law" and the law continue to be "public and universal"? The answer is yes, to a degree. OMB should be allowed to confer in private with the agency administrators both to exercise the President's control over his executives and to monitor their consistency with the President's political agenda. In this limited respect, it may be helpful to view OMB as part of EPA. There is no administrative law constraint on open, confidential discussion among the various groups within EPA itself. However, OMB may not interject into the agency's decisional process factors which the governing statute does not permit. Many environmental statutes do not permit the consideration of cost.' 64 Yet, there are reports that OMB personnel have, in prac- tice, applied cost-benefit principles, and urged the agency to do the same, even where the governing statute prohibits such analysis. 1 65 162. Id. 163. DeMuth and Ginsburg, supra note 9, at 1085. 164. See GAO REPORT, supra note 156, at 15-17. See also supra notes 61-70 and accompa- nying text. 165. See Olson, supra note 14, at 52 nn.261-62. 1991] OMB and Cost-Benefit Analysis 177 One way to address this situation would be for the President to amend the Executive Order to include some of the provisions which would have been in Senate Bill 1080, had it been adopted. After all, S. 1080 was the Administration Bill, which had been in- troduced by Senator Laxalt at President Reagan's request. S. 1080 contained express requirements that the agency include in its final rule "a memorandum of law supporting the determina- tion of the agency that the final rule is within the authority dele- gated by law and consistent with congressional intent."' 166 In addition, the Bill expressly provided that cost-benefit analysis would not be appropriate "where the enabling statute pursuant to which the agency is acting directs otherWise .... *"167 Perhaps OMB should also be required to communicate all of its comments in writing, so that legitimate disputes about what factors were considered could be resolved by in camera inspection of the docu- ments by the court. These requirements would go a long way to assure that agency deliberations pursuant to the Executive Order do not rely on nonstatutory factors. To be sure, there remains the chance that a decisionmaker will base his decision on factors which Congress has deliberately excluded from the regulatory calculus. This risk is certainly not new and is not created solely by the mandates of Executive Order 12,291. In the final analysis, it is for the courts in judicial review and for Congress in its oversight of the agencies to police this type of deceit. However, it is not enough to say, as did then-Judge Scalia dur- ing a D.C. Judicial Conference, that an agency may merely give "a plausible justification that would make it reasonable to adopt the rule if that were the agency's genuine motivation." ' 68 If judicial review were diminished to the type of cynical exercise Judge Scalia seemed to posit, then federal judges would be transformed into theater critics judging the quality of the performance, rather than the quality of the rulemaking. This is obviously not what our system deserves, nor what it-bargained for in adopting the Ad- ministrative Procedure Act.169 166. S. REP. No. 284, supra note 6, at 184. 167. Id. at 199. 168. See D.C. Circuit Judicial Conference, supra note 70. 169. The House Report accompanying the 1946 adoption of the Administrative Proce- dure Act said that: "The required statement of the basis and purpose of rules issued should not only relate to the data so presented [in the agency record] but with reasonable 178 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 16:143 B. Enhancing the Accuracy of the Administrative Record Based on the appellate court rules, EPA's regulations, and the practice that has evolved over the years, EPA has undoubtedly created both a judicial and public expectation that all of the docu- ments which it relied upon in adopting the regulation are in the public record. If interested parties are able to submit documents to EPA outside of this public record, public confidence will be eroded and the process of judicial review will be substantially un- dermined. The practical implementation of Executive Order 12,291 has reportedly involved just such submissions through the offices of OMB. 170 As a result, the Executive Order should be amended to make it clear that such "backdoor" access to the decisionmaker is improper. S. 1080 provided that: [i]n promulgating a final rule, the agency may not substantially rely on any factual or methodological material that was not identified in the notice of proposed rule making unless such material was placed in the rule making file in time for interested persons to comment thereon during the period for public par- ticipation in the rule making.17' The incorporation of this provision into the Executive Order would have a positive effect in enhancing the accuracy of the ad- ministrative record. C. Enhancing the Quality of Rulemaking by Using the Cost-Benefit Organizational Approach Finally, the President should expressly recognize the benefits of the organizational approach to cost-benefit analysis and the limi- tations of the technical cost-benefit balancing method even in cir- cumstances in which its use is permitted by the governing statute. S. 1080 would have made it clearer that the technical approach has serious limitations, particularly when applied to environmen- tal, health and safety regulation. It would have provided that "[a]n agency shall describe the nature and extent of the non- quantifiable benefits and costs of a proposed and a final rule pur- suant to this section in as precise and succinct a manner as possi- fullness explain the actual basis and objectives of the rule." S. Doc. No. 248, 79th Cong., 2d Sess. 259 (1946), quoted in 1 K. DAVIS, ADMINISTRATIVE LAW TREATISE § 6:5, at 462 (1978). 170. See D.C. Circuit Judicial Conference, supra note 70 (Comments of Mr. Fitzpatrick), reprinted in 105 F.R.D. 251, 334-335 (1984); Olson, supra note 14, at 55-64. 171. S. REP. No. 284, supra note 6, at 190. OMB and Cost-Benefit Analysis ble.' 1 72 The addition of similar language in Executive Order 12,291, along with a narrative description of the importance of giving credit to such non-quantifiable factors, would provide a valuable extra measure of assurance that these types of factors are not overlooked in the cost-benefit process. At the same time, the President should also re-emphasize the importance of employing the less rigid cost-benefit organizational method as a way of assuring that society's resources are being marshalled in an efficient manner. In the Superfund program in particular, tough choices must be made about the use of limited resources and the cost-benefit organizational method can contrib- ute to the decisionmaking process. V. RECENT REFORM EFFORTS In 1989, OMB and Rep. Conyers agreed upon OMB disclosure requirements. 73 In early 1990, however, the Administration withdrew from this informal sidebar agreement. According to correspondence between White House Counsel Gray and Rep. Conyers, the Administration withdrew from the agreement be- cause parts of it would "fundamentally impede the President's conduct of his constitutional responsibilities."' 74 A replacement sidebar has been proposed by OMB. 75 The new agreement would be implemented by either an OMB circular or an executive order, and would modify existing OMB disclosure requirements. 76 Rep. Conyers has also introduced legislation that would in- crease disclosure of OMB's review of agency activities. 77 The language of this proposed legislation is also included in the reauthorization bill for OMB's Office of Information and Regula- tory Affairs. The bill, The Paperwork Reduction and Federal In- formation Resources Management Act of 1989, requires OMB to make public correspondence received concerning agency activi- ties under review and requires the appropriate agency to be noti- fied of all meetings involving OMB's Office of Information and Regulatory Affairs staff and anyone who is not an employee of the 172. Id. at 199. 173. 21 Env't Rep. (BNA) 334 (June 15, 1990). 174. Id. 175. 21 Env't. Rep. (BNA) 358 (June 22, 1990). 176. Id. 177. H.R. 3695, 101st Cong., 1st Sess. (1990). 1991] 180 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 16:143 federal government. It also provides the agency an opportunity to send a representative to the meeting. In October 1990, the bill was passed by the House.' 7 8 Its Senate counterpart, the Federal Information Resources Management Act of 1989 (S. 1742), was introduced by Sen. Bin- gaman. S. 1742 is similar to the House legislation, but also re- quires OMB to advise the appropriate agency of any oral communication concerning an agency rulemaking and sets dead- lines for OMB to complete reviews. In addition, the bill requires OMB to maintain a public file which includes reasons for any changes to a rule and the source of the authority for the change. The Senate bill was reported out of committee, but did not re- ceive full Senate action in the 101st Congress. 179 These bills would strengthen the accuracy of the administrative record, and therefore would improve the current OMB regulatory review practice. On the other hand, the bills have also been criticized for potentially limiting beneficial confidential discussions between agencies and the OMB. VI. CONCLUSION Growth in the volume of environmental regulations and the magnitude of their economic impact is likely to continue. This growth will inevitably place increasing emphasis on cost-benefit analysis. EPA is faced with ever more complex environmental challenges and fewer resources to do the job. Moreover, as U.S. economic competitiveness continues to be challenged by foreign producers who may be free of these types of controls, the eco- nomic impacts of environmental regulation take on even greater significance. As a result, it is inevitable that OMB will have a continuing role in policing the development of regulations regardless of the polit- ical party which controls the White House. Moreover, Executive Order 12,291 has made substantial positive contributions toward improving the cost efficiency of agency rulemaking. Yet, the cur- rent process under Executive Order 12,291 requires modification to reduce the potential for the erosion of public confidence and to preserve the value of independent judicial review. In addition, the more technical cost-benefit balancing which the Executive Or- 178. 136 CONG. REC. H 11895 (daily ed. Oct. 23, 1990). 179. 136 CONG. REC. S16827 (daily ed. Oct. 23, 1990). 1991] OMB and Cost-Benefit Analysis 181 der has engendered at EPA is unduly rigid, too costly to conduct, and often deceiving in its results. Instead, a method more like the cost-benefit organizational approach, which has proven useful in organizing information and focusing a decisionmaker's attention on the relevant factors, should be employed.