HUNTON & WILLIAMS RESPONSES TO QUESTIONS RESULTING FROM TESTIMONY BY TURNER T. SMITH, JR. HUNTON & WILLIAMS Questions By Senator Heflin 1. In testimony provided to the Committee there is a statement which claims that the bill does not require a "cookbook" approach to cost-benefit analysis and risk assessment. Yet, it seems to me that the language of the bill is quite clear in the exact ingredients which each draft analysis must contain. Can you explain to this Committee how agencies as varied as the Administrative Office of Courts and the Environmental Protection Agency can promulgate regulations using the same specific ingredients for cost-benefit analysis as stated in this bill. Answer: Section 622(c)(2) of S. 343 contains essentially five requirements. Each draft cost-benefit analysis must contain: An analysis of the "benefits" and the "costs" of the proposed rule (including an explanation of how each will result from the proposed rule) and an analysis of whether those benefits "are likely to exceed" the identified costs. An identification of "reasonable" alternatives for achieving the identified benefits of the proposed rule. Included among these must be, where they qualify as "reasonable" alternatives, the "no action" alternative, alternatives that will accommodate any differential impact of costs and benefits as between geographic regions and as between persons with differing levels of resources with which to comply, and any alternatives that employ "performance or other marketbased standards" and that permit the greatest flexibility in achieving the benefits through marketbased mechanisms. The agency must analyze the cost and the benefits of each reasonable alternative that it identifies, presumably (by analogy to practice under NEPA) in less detail than for the chosen alternative and must provide an analysis of whether the proposed rule will provide greater net benefits to society than any of the other reasonable alternatives that it has considered. Third, the agency must ask itself whether it is feasible to establish a regulatory program that operates through the application of market-based mechanisms. Fourth, where the proposed rule is based on scientific evaluations or information or is subject to the risk assessment requirements of Subchapter III, the agency must describe the actions it has undertaken to verify the quality, reliability, and relevance of those scientific evaluations or information in accordance with the risk assessment requirements of Subchapter III. Finally, the agency must assess the aggregate effect of the rule on small businesses with fewer than 100 employees, including an assessment of the net employment effect of the rule. It is important at the outset to recognize that the draft cost-benefit analysis required is not primarily a numerical analysis. As the Senate Judiciary Committee said in 1981 in its Report No. 97-284 (hereinafter "S. Rep.") in discussing the equivalent provisions of S. 1080, "the regulatory analysis required by [the cost-benefit provisions of S. 1080] establishes an explicit process to rationally structure the way an agency resolves the trade-offs of regulatory decision-making." S. Rep. at 74; see also S. Rep. at 151-54. Further, as that Committee recognized then, "regulatory analysis as a method for making decisions, for choosing alternative courses of action, is not fundamentally different from the billions of health and safety decisions individuals make each day." S. Rep. at 71. These principles are reflected in the language of S. 343. For example, the term "benefit" means the "reasonably identifiable significant benefits, including "social and economic benefits" that are expected to result "directly or indirectly" from implementation of a rule or an alternative to a rule. Section 621(5). The term "costs" means the "reasonably identifiable significant" costs and "adverse effects, including social and economic costs, reduced consumer choice, substitution effects, and impeded technological advancement" that are expected to result "directly or indirectly" from a rule or an alternative. Section 621(6). These definitions are quite similar to the definitions of benefit and cost included in Sections 621(4) and (5) of S. 1080, and S. Rep. 97-284 recognizes expressly how flexible they are. S. Rep. at 141-42. The definition of benefits and costs is broad enough to encompass the pros and cons in reasoned decision-making that each such agency finds relevant to its purposes. Thus, the particular costs, benefits, and alternatives to be considered by the agency taking action will be shaped by the nature of the mandate given it by Congress and the types of costs and benefits relevant to the action it contemplates taking. In short, as this Committee said in S. Rep. 97-284, "To the extent [cost-benefit analysis] is actually used to drive a decision process (and is not merely a cosmetic), and to the extent that the users act objectively, in good faith, and with requisite analytical rigor, rational decision-making is promoted." S. Rep. 97-284 at 74 (quoting Barum Report to the Administrative Conference) (emphasis in Barum original). As for the judgment concerning the relationship between costs and benefits that each of the various Federal agencies must make, it is important to recognize, as the Senate Judiciary Committee did in Report No. 97-284, that: "[A]l intelligent decision-making has to be based on some sort of S. Rep. 97-284 at 73 (citations omitted). Here again, requirements of S. 343 are remarkably similar to those the Senate approved unanimously in 1981 in S. 1080. In short, as I indicated in my testimony, each type of agency is simply required to ask, in the context of its own mission and its own facts, "is this action "worth it'?". As for the second major requirement, the consideration of reasonable alternatives, the use of the word "reasonable" allows each differently situated agency to determine what alternatives, if any, are reasonable in the context of the particular type of action it is taking. The word "reasonable" also qualifies the requirement to consider the no-action alternative, the different regional and income level differences as to impact, and the possible use of market-based alternatives. Where differential regional and income impacts are irrelevant to a particular agency's mission, or where market-based approaches are irrelevant, there will be no "reasonable" alterna tives that the agency must consider. The legislation simply requires that the agency ask itself the question whether these factors are relevant, and then explain why they are not, if they are not. As for the requirement that the agency analyze whether a proposed rule will provide "greater net benefits to society" than any of the alternatives, this is simply a requirement that the agency choose the course of action which it deems "best" in the context of its statutory mandate and factual circumstances, after considering the relevant quantified and unquantifiable benefits and costs. As the Judiciary Committee said in Report No. 97-284: S. Rep. at 82. [T]he Committee does not view regulatory analysis as, "black box" for making regulatory choices. It is a process to The requirement that benefits and costs be analyzed does not imply that the analysis must always be done by a particular numerical methodology, nor that all benefits and costs must be quantified. The definitions of benefits and costs do not mention quantification. The final costbenefit analysis adopted with the final rule: shall include, to the extent feasible, a quantification or numerical Section 622(e)(1)(A) (emphasis added); see S. Rep. at 153-54 (section-by-section analysis of the similar language from Section 622(e) of S. 1080). In short, non-quantifiable benefits and costs must not be ignored, and may frequently govern the result. Quantification, which must be reasonably precise and explained when used, is only required "to the extent feasible." Feasibility will be a function of the difficulty of quantification, the importance of the decision, the time available for decision-making, and a number of other factors within the discretion of each type of agency taking action. The third major requirement is simply that each agency, whatever the nature of its regulatory program and responsibilities, must ask itself whether that program can be operated through the application of market-based mechanisms. In many cases, that determination may be made on a generic basis by an agency, where market-based mechanisms are not usefully employed in the type of program it is dealing with. In other cases, the question will be more complicated and more specific to the particular regulatory action being taken. In any case, the important requirement is that the agency consider the matter and articulate its views with regard to it. Fourth, in cases where relevant, each agency must explain what actions it is taking to ensure the quality of the scientific evaluations or information that it has relied upon in making its decisions. It is to be guided by the risk assessment requirements of Subchapter III in this regard. In many cases, Federal agencies may make decisions on major rules that do not depend 35-770 96-9 upon scientific evaluation or information, in which case this provision is not relevant. Where they do use scientific evaluations or information, this provision requires only that in the particular context in which they take action, they must explain what steps they have taken to ensure the reliability of the evaluations or information on which they rely. In different types of regulatory programs, and considering different types of scientific evaluations or information, this requirement may be satisfied by different explanations. In short, there is great flexibility built into this requirement. The bottom line is that agencies must attempt to use good science where they use science, and must articulate how they have done this. Finally, an agency must assess the "aggregate effect" of the rule on small businesses with fewer than 100 employees, including an assessment of the net employment effect of the rule. This presumably requires that the agency assess the total impact of the regulatory burden on such small businesses, an important factor if the job creation and entrepreneurial benefits of small business are not to be inadvertently overwhelmed by government regulation. Should this Subcommittee believe that, notwithstanding the provisions I have just cited, further clarification of the flexible nature of the rational decision-making framework provided by the cost-benefit and alternatives analysis requirements of S. 343 be provided, I suggest that they insert in Section 622 an adaptation of the language found also in Section 635(b)(3) with regard to risk assessment: A risk assessment [in the case of Section 622, a "cost-benefit Further, cost-benefit analysis could be defined so that it is clear that it requires an assessment of costs and benefits rather than some numerically precise calculation. The word "outweigh" in Sections 622(d)(2)(B)(i) and 623(a)(1) could be changed to "justify," and Section 622(e)(2)(B) could be amended to add "An agency is not required to make such evaluation primarily on a mathematical or numerical basis." 2. Section 624, Judicial Review, subjects the determination or designation by the President or an agency that a rule is or is not a major rule to judicial review. I believe that this interlocutory judicial review may raise a serious seperation [sic] of powers constitutional question. Do you see such a constitutional question? Answer: The requirement for judicial review by the President may present a constitutional issue. I do not believe that review on an interlocutory basis of the determination of an "agency" that a rule is or is not a major rule raises a constitutional separation of powers question. 3. A standard of proof, in the bill, for granting of a petition to withdraw an existing rule is the showing that there is a reasonable likelihood that the rule outweights [sic] the benefi or a reasonable question exists as to a reasonable alternative. The agency must make a derermination under those standards to grant or deny the petition in 180 days from the date of filing of the petition. Would this require the agency to make a complete review of all "reasonable" alternatives in that time period? Answer: I do not believe that this provision would require a "complete review" of all "reasonable" alternatives within 180 days. Rather, it would require the agency to review those alternatives presented to it by the petitioner, to determine on a preliminary basis whether any of them are "reasonable," and then to decide whether or not there is a "reasonable likelihood" that any of those alternatives judged to be reasonable would yield more net social benefits than the agency's current practice. "Complete review" of the alternatives would not be required until after the agency has determined to grant the petition under the "reasonable likelihood" test. Questions From Senator Thurmond 3. Could you both please address the concern that S. 343 would move discretion from Federal agencies to individual Federal judges? In your view, would this result from the legislation and, if so, would it be desirable? Answer: I do not believe that S. 343 would move discretion from Federal agencies to individual Federal judges. Rather, S. 343 provides instructions from Congress, in Section 623, as to how agencies must exercise their discretion within the existing statutory frameworks and parameters already provided by Congress. The Federal courts will then, on the basis of the normal standards of review, such as substantial evidence on the rulemaking record as a whole, and as spelled out in Section 628 of S. 343, engage in normal judicial review. If there is any question whether the provisions of Section 628(c)(1)(C) promote the intrusion of Federal courts into the exercise of agency discretion more than is normally the case under the traditional "arbitrary and capricious" and "contrary to law" standards of review through its requirement that the court determine whether the agency has determined that the interpretation selected rather than other permissible constructions of the statute is the one that "maximizes net benefits to society," this Subcommittee could change that provision so that it reads as follows: 4. (C) the agency has explained in a reasoned analysis why it Mr. Smith, I think there is a broad consensus that regulations should be based on so-called 'good science". However, it seems that experts can be found that will swear to both sides of almost any issue, as demonstrated in courts every day. My question is whether you believe there will be much disagreement and litigation over exactly what "good science" is, in practice? Answer: I believe that there may be some disagreement about what "good science" is, but that the new requirement will promote a more even-handed approach by agencies in determining what scientific conclusions to reach. It will likely encourage agencies to evaluate with some care ideas outside the mainstream of scientific consensus before taking major public decisions, and imposing major costs on society, in reliance on those views. As for litigation, given that scientific factual matters are at the heart of many rulemaking appeals today, I suspect that the new requirement will not contribute greatly to increased litigation. |