View Header

THE WHITE HOUSE

Office of the Vice President


For Immediate Release September 1, 1993
                        Accompanying Report of 
                   the National Performance Review
 
                    OFFICE OF THE VICE PRESIDENT
 
                           Washington, DC
                           September 1993

REG05:
Streamline Agency Rulemaking Procedures

Background

In an effort to coordinate policies within each agency and within the executive branch, numerous layers of review and analyses have been added on top of the relatively simple notice-and-comment requirements of the Administrative Procedure Act (APA). Clearance is required by numerous offices within the agency (and again within the department if the agency is part of a department) and by the Office of Management and Budget (OMB). As has been previously noted, the entire process is repeated twice--first at the proposal stage, then again when the rule is finalized. After a several-month-long study by a contractor, one agency found that it took an 18-foot flow chart with 373 boxes to describe its then-current rulemaking process.[Endnote 1]

Another reason for the extensive internal review is the fear of litigation challenging the rule. Agencies compile extensive records to defend themselves if rules are challenged in court--even for issues no one has raised during the public comment period. An agency can spend considerable resources developing a rule only to have it nullified in court because the rule was based on an impermissible statutory interpretation. If a successful challenge is brought after the rule is issued, an entire program may be thrown into a state of disarray.[Endnote 2]

Some agencies have statutory provisions that help reduce the amount of litigation preparation needed during the rulemaking process. For example, some statutes prohibit litigants from challenging a rule on any basis other than ones raised during the public comment period and require challenges to be brought within 60 days of the day the rule is issued.[Endnote 3]

The straightforward APA notice-and-comment rulemaking process has now become so formalized that its name--informal rulemaking--seems a misnomer. Some agencies, however, have additional statutory requirements that make their rulemaking process even more cumbersome. For example, the Federal Trade Commission is required to conduct trial-type hearings that allow cross-examination of witnesses.[Endnote 4]

NEED FOR CHANGE

Although different agencies have different internal rulemaking processes, within an agency too often the same review process is used and the same number of clearances are required regardless of the significance or complexity of a rule.[Endnote 5] A process with numerous checks makes sense for significant rules (e.g., ones that announce a major policy, initiate a new regulatory program, have a significant effect on the economy, or are particularly controversial). However, a large number of the thousands of rules issued each year are not significant and do not warrant extensive review by numerous offices. Anecdotal evidence indicates that insignificant rules (which may be a clarification of an existing rule or a small change needed to undo an unintended consequence of an existing rule) often languish on reviewers' desks because the rules are not high priority and the reviewers are busy with other demands.[Endnote 6] Even relatively minor changes to an existing program can easily take nine months or more.

The failure to establish a process that differentiates between significant and insignificant rules is costly. The current lengthy process precludes quick, minor adjustments to existing programs, wastes the time of numerous reviewers, and frustrates staff responsible for getting the rules out and keeping programs functioning. Evidence that the rulemaking process has become too cumbersome is provided by the frequency with which agency staff turn to other methods of establishing agency policy. Instead of issuing rules, agencies issue policy statements, guidance documents, and memos to agency personnel that are often not required to go through public comment, extensive internal agency clearance or OMB review.[Endnote 7]

SOME EFFORTS AT STREAMLINING. The Clinton administration's regulatory review executive order is a major step in streamling the regulatory process. It will streamline the review process by requiring the Office of Information and Regulatory Affairs (OIRA) to review only significant rules.

The Environmental Protection Agency (EPA) also has made efforts to streamline its process and avoid the "double" review process (i.e., all clearances required both at proposal and final rule stage) by issuing "direct final" rules for certain rules on which it believes the public will not comment. Although EPA issues many rules that are controversial and that benefit from public comment, a significant number of rules receive no comments. These rules include:

--minor modifications of existing rules (such as directions about where to submit information or technical changes to testing standards that have been adopted by a professional association);

--federal approval of changes to a state plan implementing federal law affecting only one or a small group of companies (especially where the federal agency has very little discretion); and

--"significant new use" rules limited to specific chemicals.

In this direct final rulemaking approach, EPA publishes a notice in the Federal Register saying that a rule will become effective in 60 days unless, within 30 days, someone submits written notice of an intent to file an adverse or negative comment. If no such notice of intent is filed, the rule becomes final without going through a second round of intra- and inter-agency review. If even one person files such a notice, EPA withdraws the rule, republishes it as a proposed rule, and proceeds with normal noticeand -comment rulemaking and the second round of review. (Republishing the rule as a proposal serves two functions. First, it gives an opportunity to comment to people who may have kept silent because they wanted the rule to go into effect immediately. Second, it self-imposes a penalty--additional steps and further delay--which serves to prevent the agency from overreaching in the use of this technique.)

This approach avoids the second round of clearances and review, which otherwise delays rules, wastes time, and should be superfluous (especially if all reviewers are warned the first time that the rule might go final without further review). Theoretically, the second review ought to be very quick, but clearing any document through numerous government offices takes time. The paper shuffling also wastes the reviewers' time by requiring them to look at something twice when once would have sufficed.

EPA uses this process routinely in two areas--approval of changes to state implementation plans and "significant new use" rules for chemicals.

--In a pilot project testing this procedure for EPA approval of changes to state implementation plans under the Clean Air Act, 90 rules were processed in an average of 181 days, which was considerably better than the 419 days it took to process 81 comparable rules prior to the use of the direct final process.[Endnote 8] For the last 10 years, EPA has used the process nationwide to approve changes to state plans.

--In 1989, as a result of an informal reg neg process with industry and environmental groups, EPA adopted the direct final rulemaking process for chemical-specific significant new use rules (SNURs) issued under the Toxic Substances Control Act.[Endnote 9] SNURs for at least a hundred chemicals have been issued as direct final rules.[Endnote 10]

EPA also uses direct final rules occasionally for other types of rules. The agency has rarely had to withdraw a rule because public comments were filed, and no one has ever challenged these rules for failure to comply with the APA.[Endnote 11]

Some agencies have streamlined the process by issuing broad, generic rules to resolve common issues or to establish standards applicable to multiple industries instead of conducting separate rulemakings for each issue or industry. For example, the Social Security Administration and the Nuclear Regulatory Commission each have successfully issued rules concerning common problems arising in disability claims and applications for power plants, respectively. On the other hand, the Occupational Safety and Health Administration was not allowed by a court to use a generic, instead of a hazard-by-hazard, approach for permissible exposure levels for air contaminants.[Endnote 12]

The National Marine Fisheries Service has established categories of rulemaking notices that appear in final form in the Federal Register after reduced review. For example, when a quota of fish previously specified by regulation has been taken in a particular fishery, a notice closing that fishery is published in the final rule section of the Federal Register. Violators of these closure notices may be subject to a variety of civil enforcement proceedings. These notices are not reviewed by either the Department of Commerce or OIRA.[Endnote 13]

Agencies also conserve agency resources by adopting voluntary national consensus standards rather than developing criteria
in-house.[Endnote 14]

ACTIONS

  1. Streamline internal agency rulemaking procedures. (1)

The President should direct heads of regulatory agencies to review and streamline their internal rulemaking processes. The Regulatory Coordinating Group (RCG), in consultation with the Administrative Conference of the United States (ACUS), should assist the process by providing a forum for agencies to exchange information and coordinate hiring of contractors (if any are necessary) to avoid duplicative efforts. A variety of principles should be adopted to streamline the process and reduce the cost of issuing rules, including:

--Agencies should differentiate between significant and insignificant rules and use a shorter process for insignificant rules. The new regulatory review executive order requires OIRA review only for significant rules. Agencies also should strive to group or tier rules and to adjust the process so that the number of steps in the clearance process varies with the significance or importance of the rule.

--Executive departments (such as the Department of Health and Human Services) should review their internal delegations to determine whether agencies headed by Presidential appointees (such as the Food and Drug Administration) can issue some insignificant rules without departmental review. It might be appropriate to develop a triage system similar to the one in the new regulatory review executive order.

--Agencies should determine whether there are continuing, unnecessary roadblocks in the process and then determine how to remove them.

--Agencies should get early input from top policymakers on the choice of regulatory approach to be taken so that agency resources are not wasted drafting regulations that are inconsistent with the desired policies.

2. Use "direct final" rulemaking to reduce needless double review of noncontroversial rules. (1)

The head of each regulatory agency should use "direct final" rulemaking or a similar approach at least once in the next year (or explain why it cannot be done).[Endnote 16] The direct final process should only be used for rules that the agency believes are so noncontroversial that no one would file adverse or negative comments on the proposal. Under this process, the agency would publish a direct final rule in the Federal Register. The Federal Register notice should explain that the rule would become effective in 60 (or some other appropriate number) days if, within 30 (or some other appropriate number) days, the agency did not receive notice from any person of the intention to file adverse or negative comments.[Endnote 17] If the agency is notified that adverse comments would be filed, the agency would be required to withdraw the direct final rule, republish it as a proposed rule, and go through the usual notice-and-comment procedures.

3. Develop specifications for rulemaking petitions. (1)

The heads of regulatory agencies should issue regulations specifying what must be in petitions for rulemaking and describing the process for filing them.[Endnote 18] Such regulations should explain that detailed petitions, complete with a text of a proposed rule and underlying analyses, may help the agency respond to the petition more quickly. Sufficiently detailed petitions may form a basis for a notice of proposed rulemaking. After a public comment period, the agency could then either withdraw the proposal or modify it as necessary and adopt it as a final rule.[Endnote 19]

4. Evaluate and draft proposed legislative changes to speed the rulemaking process. (4)

Heads of regulatory agencies, in coordination with the Regulatory Coordinating Group (RCG) and ACUS should:

--identify existing statutes that unnecessarily require cross-examination and other adjudicative fact-finding procedures in rulemakings and should recommend whether the administration should seek legislative changes;

--identify which, if any, statutes should be amended to limit the amount of time parties have to challenge a rule, or to limit the issues on judicial review to those that were raised during the public comment period, and should recommend whether the administration should seek legislative changes; and

--recommend whether the administration should seek amendments to ensure that regulatory statutes provide sufficient flexibility for agencies to issue "generic rules" that would settle, in one proceeding, issues that would otherwise recur in numerous separate rulemakings or enforcement proceedings.

The Chair of the RCG, in coordination with ACUS, should study whether:

--The administration should seek an amendment to the Federal Advisory Committee Act to exempt advisory committees that meet only once from its requirements.[Endnote 20]

--It would be feasible, useful and constitutional for agencies to be able to seek judicial resolution (similar to a declaratory judgment) by a court of difficult statutory interpretations that are important to resolve before an agency finalizes a rule and expends significant resources.[Endnote 21] If the administration seeks legislation, the procedure to be established should ensure that affected interests are represented and that issues are not presented prematurely (i.e., before the agency has definitively determined that there is an actual problem of statutory interpretation).

CROSS-REFERENCES TO OTHER NPR ACCOMPANYING REPORTS

Department of Labor, DOL03: Expand Negotiated Rulemaking and Improve Up-front Teamwork on Regulations.

Department of Education, ED05: Streamline and Improve the Department of Education's Grants Process.

Department of Health and Human Services, HHS02: Reengineer the HHS Process for Issuing Regulations.

Streamlining Management Control, SMC08: Expand the Use of Waivers to Encourage Innovation.

ENDNOTES

  1. Although that agency has since modified its process, the process depicted on the chart is not atypical of rulemaking procedures throughout the government.
  2. See Corrosion Proof Fittings v. EPA, 947 F.2d 1201 (5th cir. 1991) (EPA's asbestos rule, 10 years in the making, remanded to agency for lack of substantial evidence).
  3. See the Clean Air Act, 42 U.S.C. 7607(b)(1) and (d)(7)(B) (1991 Supp.).
  4. 15 U.S.C. 57a (1988). The Securities and Exchange Commission has similar requirements in the Securities Act Amendments, 15 U.S.C. 78f(e) (1988), as does the Environmental Protection Agency (EPA) in the Toxic Substances Control Act, 15 U.S.C. 2605(c) (1988).
  5. Some very minor or repetitive rules have an expedited process, but most agency processes fail to adequately match the thoroughness of the process with the complexity of the rule. The American Waterways Operators reports that a staff member from the Office of Information and Regulatory Affairs noted that "the Coast Guard's rigorous internal review process applies equally to rules of sweeping consequence and minor technical guidelines." The American Waterways Operators, "The Federal Regulatory Process: Opportunities for and Obstacles to Government/Industry Cooperation," May 1992, p. 17.
  6. Interview with Richard Wilson, Director, Office of Mobile Sources, Environmental Protection Agency, May 14, 1993.
  7. To a certain extent, the use of such other methods allows agencies to differentiate between significant and insignificant rules and is a useful development. However, there are drawbacks to these methods-- although technically these policies cannot bind people outside the agency, as a practical matter, they may have that effect. Unlike rules (which are published in the Federal Register and codified in the Code of Federal Regulations), less formal statements of agency policy are not necessarily compiled in one place and may not be accessible to the public. There is a general sense that agencies are turning to policy statements and guidance too frequently. See Emery, Fred, Rulemaking as an Organizational Process (1982), p. 46 (unpublished report to ACUS); Anthony, Robert, "Agency Efforts to Make Nonlegislative Documents Bind the Public," Admin. Law Review, Vol. 44 (1992), p. 31. an expanded version of the latter article will shortly appear as Interpretive Rules, Policy Statements, Guidance, Manuals and the Like--Should Agencies Use Them to Bind the Public?, 1992 Administrative Conference of the U.S. Recommendations and Reports.
  8. Environmental Protection Agency, Requirements for Preparation, Adoption, and Submittal of State Implementation Plans; New SIP Processing Procedures to Save Time and Resources, 47 FR 27073, 27074 (June 23, 1982) (in this notice EPA uses the term "immediately final" rather than "direct final").
  9. Interview with James Nelson, Associate General Counsel, Environmental Protection Agency (May 11, 1993). See 40 C.F.R. 721.170(d)(4)(1).
  10. See note 9 above.
  11. Interviews with Alan Eckert, Associate General Counsel, Environmental Protection Agency, May 6, 1993; James Nelson, Associate General Counsel, EPA, May 11, 1993; and Howard Hoffman, Senior Attorney, EPA, May 10, 1993. In response to an informal request to all attorneys in Environmental Protection Agency's Office of the General Counsel, no direct final rule was identified as having been challenged in court for failure to comply with the Administrative Procedure Act (APA).
  12. AFL-CIO v. OSHA, 965 F.2d 962 (11th Cir. 1992).
  13. Memorandum to Jeffrey Lubbers, Team Leader, NPR Improving Regulatory Systems Team, from Gloria Gutierrez, Acting Chief Financial Officer and Assistant Secretary for Administration, Department of Commerce, Aug. 5, 1993.
  14. Administrative Conference of the United States, Recommendation 78-4, "Federal Agency Interaction with Private Standard-Setting Organizations in Health and Safety Regulations," 1 C.F.R. 305.78-4.
  15. These principles should apply to all agencies even though agency processes vary significantly from agency to agency, depending, in part, on the agency's mission and culture.
  16. Agencies can use direct final rules without new legislation, although the process does pose some litigation risk. Arguably, direct final rules meet the requirements of the APA. However, an argument could be made to the contrary. Agencies should determine when it is appropriate to take that risk after consultation with their general counsels. One factor that may make it easier for EPA to use the process is the special provisions in the Clean Air Act and the Toxic Substances Control Act that require anyone challenging a rule to bring the judicial challenge within 60 days after the rule is published in the Federal Register. 42 U.S.C. 7607(b)(1) (1991 Supp.); 15 U.S.C. 2618 (1988). Agencies that do not want to take the litigation risk associated with direct final rules should structure review procedures to accomplish nearly the same thing. Instead of a direct final rule automatically going into effect unless the agency withdraws it, the agency could publish in the Federal Register a notice of no adverse comment that makes the rule final. Internal and OMB review procedures should not require review of such a notice if everyone is notified that this process might be used.
  17. The rule cannot become effective until at least 30 days after the close of the public comment period. See 5 U.S.C. 553.
  18. Administrative Conference of the United States, Recommendation 86-6, "Petitions for Rulemaking," 1 C.F.R. 305.86-6.
  19. In setting up such procedures, care must be taken not to turn over to private parties the governmental functions of setting the agency's agenda and analyzing rules. The regulations should ensure that the agency's decision to withdraw a proposal would not be subject to additional judicial review merely because the agency has invited rulemaking petitions.
  20. 5 U.S.C. App. (1988).
  21. The Carnegie Commission recommended this procedural innovation. Carnegie Commission on Science, Technology, and Government, Risk and the Environment: Improving Regulatory Decision Making (Washington, D.C., June 1993), p. 111.

REG06:
Encourage Alternative Dispute Resolution When Enforcing Regulations

BACKGROUND

Disputes constantly arise when agencies enforce regulations. Enforcing regulations and encouraging compliance with regulatory requirements in the least costly, least adversarial, and most effective way are crucial to ensuring that the policies underlying regulations are implemented.

Enforcement of regulations often requires substantial litigation. Adjudication--the resolution of disputes before agency administrative law judges or in federal court--is often lengthy, expensive, and highly contentious.[Endnote 1] Ultimate outcomes are frequently unsatisfactory and may undercut future efforts to ensure effective compliance by creating an atmosphere of antagonism and distrust. In the long run, encouraging compliance is likely to be more effective than litigating over enforcement.

Current enforcement dispute resolution systems tend to emphasize litigation, and there are times when this emphasis is proper. Regulated entities must understand that they must comply with regulations or face sanctions. Successful litigation by the government may serve as a deterrent to future wrongdoing. Ensuring regulatory compliance is, after all, a federal law enforcement role.

However, the fact is that a high percentage of enforcement disputes are ultimately resolved not through litigation but through settlement negotiations.[Endnote 2] There are a variety of reasons for this, including inadequate resources to pursue all cases through the full litigation process, and insufficient or questionable evidence in particular cases. Moreover, in some situations, the remedies available through the adjudicatory process (be it at the administrative level or in court) may not provide the optimal practical solutions. Negotiated settlements often occur, however, only after substantial adversarial maneuvering, or at the proverbial "courthouse steps."

NEED FOR CHANGE

Various alternatives to litigation and administrative adjudication can be useful in resolving a significant number of enforcement disputes earlier, with fewer resources, and with better results than might otherwise occur. Alternative means of dispute resolution (ADR) can be an alternative to litigation, and it can also serve as an improved way to handle negotiated resolutions. Because ADR processes are designed to develop consensual solutions to disputes, they should reduce contentiousness and may result in relationships between the regulators and the regulated that are more cooperative and more conducive to future voluntary compliance.

These alternatives encompass a collection of techniques designed to help disputing parties resolve their conflicts in mutually acceptable ways.[Endnote 3] Although the scope of what constitutes ADR varies depending on who is using the term, it is generally considered to include dispute resolution systems where a neutral third party, or "neutral," helps parties resolve their dispute.

ADR has become increasingly popular as a way to help resolve the wide variety of disputes involving individuals, businesses, and governments. Although some variations of ADR have been around for centuries, in recent time, court backlogs have served as an impetus for wider consideration of ADR techniques. Courts (both federal and state) provide ADR programs to litigants.[Endnote 4] For decades, labor and management have used ADR to resolve disputes. Private industry has begun to look to ADR as a way to prevent disputes from escalating and absorbing large amounts of resources.[Endnote 5] Community-based programs and family mediation programs have used ADR processes to address interpersonal disputes. An increasing number of states have incorporated the use of ADR into courts and agencies. As President Clinton said (while Governor of Arkansas), "[One] thing we need, in every state, is an alternative dispute resolution mechanism that would encourage people to look for a way, short of litigation, to resolve their disputes."6

ADR comes in a variety of forms. Some, like mediation, are fairly well-known. Others, such as minitrials, are less familiar. The more common types of ADR include:

--mediation,

--early neutral evaluation,

--factfinding,

--settlement judges,

--minitrials, and

--arbitration.[Endnote 7]

ADR is generally an appropriate dispute resolution mechanism when the dispute is one in which the parties might be willing to accept a negotiated settlement.[Endnote 8] Because ADR processes are inherently flexible, and can be designed to meet the needs of the disputing parties, they should be considered in a wide variety of cases.

THE ADMINISTRATIVE DISPUTE RESOLUTION ACT. The Administrative Dispute Resolution Act of 1990 (ADR Act) governs agency use of ADR in non-rulemaking disputes.[Endnote 9] It reflects Congress' interest in and recognition of the potential benefits of ADR, and it authorizes and encourages the innovative use of ADR by providing a flexible framework that agencies can adapt to their own needs.

The Act requires relatively little, but authorizes much. It requires each federal agency to develop a policy for using ADR. It directs agencies to review their types of disputes, how they are currently resolved, and whether ADR might be useful. The Act also requires that each agency appoint a senior official as a "dispute resolution specialist" to serve as the coordinator and point person for agency ADR activities. These requirements are aimed at ensuring that agencies make educated and thoughtful decisions about ADR.

The Act amended the Administrative Procedure Act to provide express authority to use ADR where the disputants agree to use it. It provides new authority for federal agencies to enter into binding arbitration, with some limitations. It provides statutory protection for confidentiality of ADR processes, and contains provisions relating to the acquisition of dispute resolution "neutrals."

Unfortunately, however, agency compliance with the ADR Act has been inconsistent. Despite clear success stories (see below), some agencies have vacancies in the dispute resolution specialist positions mandated by the Act or have not appointed to the post an official with sufficient rank.[Endnote 10]

FEDERAL AGENCY USE OF ADR. Over the last decade, some federal agencies have begun to give serious attention to the use of ADR. ADR has been used by several agencies in enforcement cases and other regulatory proceedings. Agencies have also used ADR in resolving contract, personnel and Equal Employment Opportunity disputes. A wide range of agencies--from the Environmental Protection Agency (EPA) and the Departments of Labor and Health and Human Services to the Federal Deposit Insurance Corporation (FDIC) and the Resolution Trust Corporation--have used ADR. Others--like the U.S. Air Force and the Internal Revenue Service--are in the planning stages for significant ADR programs.[Endnote 11]

Some of the results of the use of ADR to date have been impressive. The Department of Labor last year ran a pilot program in one of its regional offices using mediation to resolve enforcement cases involving the workplace, including cases involving the Occupational Safety and Health Administration (OSHA) and wage and hour regulations. The Department's evaluation of the program demonstrated that both agency and industry parties were quite satisfied with the use of ADR, and that in many cases, the mediated resolution was quicker and better than either face-to-face negotiation or litigation.[Endnote 12] The Department has recently decided to expand the program to all of its regions.[Endnote 13] The Farmers Home Administration found that it saved millions of dollars using mediation to avoid foreclosures in farmer-lender disputes.[Endnote 14] The FDIC saved over $400,000 in litigation costs from a small pilot mediation program involving creditor claims arising from failed banks.[Endnote 15] EPA has had significant successes using mediation to resolve Superfund hazardous waste cleanup proceedings.[Endnote 16]

To maintain momentum, a new Presidential directive on ADR is needed. The only existing executive order on ADR, Executive Order 12778, is not specific enough to provide useful guidance to the agencies.[Endnote 17] The directive should not only be linked to the ADR Act's provisions, but it also should address other approaches to encouraging agency use of ADR, including increased budgetary incentives, personnel appraisal incentives, and streamlined procurement of outside neutrals.

BUDGETARY INCENTIVES. Although the goals of ADR include the conservation of resources in the long run, it is important that the Office of Management and Budget (OMB) provide agencies with budgetary incentives to use ADR. This has at least two aspects: ensuring that the resources necessary to develop ADR systems are available, and ensuring that budget incentives over the long term do not undercut ADR use, but rather reward the consensual resolution of disputes and discourage the creation of large adjudicatory backlogs.

First, some resources are necessary to develop and implement an ADR program. While over the long run, ADR should provide savings to both the government and the private sector, seed money and personnel resources are often needed to get agency systems up and running. Thus, at least in the early implementation stages, some specific budget for ADR implementation may be necessary.

Second, the criteria for increasing agency budgets often provide skewed incentives. For example, providing additional resources to an agency with a large adjudicatory backlog might remove the incentive to reduce that backlog. Moreover, the savings from using ADR may not be felt by the same component of the agency to which the costs accrue.

PERSONNEL INCENTIVES. The ADR Act requires that the dispute resolution specialist and other agency personnel be provided training in ADR.[Endnote 18] This is crucial. Although ADR concepts are not extremely complicated, those who would use ADR, as well as those deciding whether and when to use it, need to understand how to participate effectively in these processes, understand their benefits, and know when it is appropriate to use them.

Performance evaluation criteria also must provide the appropriate incentives to encourage agency personnel to use ADR and resolve disputes consensually. For example, some agencies give credit in individual performance appraisals for the number of cases opened, but not for the number of cases resolved or the efficiency and effectiveness of resolution. This should be altered so that both types of numbers are important.

HIRING OF NEUTRALS. Current contracting processes for obtaining a neutral can be extremely cumbersome. Neutrals are potentially available from a variety of sources. ACUS, for example, has a roster of neutrals from which agencies can select.[Endnote 19]

However, getting a candidate from the roster does not obviate the need for a system to hire and arrange to pay (if necessary) for the neutral's services. Depending on the amount of money involved, a more or less complicated competitive bidding system may be required. This process is frequently incompatible with the short time frames that may be necessary to use ADR effectively.

It is important that agency personnel interested in using ADR have access to neutrals in a timely fashion. Depending on whether the neutral is someone from the same agency, a different agency, or outside the government, mechanisms (including ones for sharing of neutrals among agencies) should be established to ensure ready availability. Agencies also have been and should be encouraged to continue training their personnel in mediation and related ADR techniques, so that they can make use of neutrals from their own and other agencies to help resolve disputes more quickly and less expensively.

ACTION

  1. Increase use of alternative means of dispute resolution. (2)

The President, through directive, should strongly encourage agencies to use ADR.[Endnote 20] Governor Clinton recognized the need for every state to use ADR processes to avoid needless litigation; President Clinton should take the opportunity to apply that dictum to every federal agency.[Endnote 21]

Presidential support of ADR is necessary to avoid budget incentives that undercut ADR use, to train agency staff in ADR, to structure performance appraisals to reward proper use of ADR, and to make it easier to hire neutrals in a timely fashion. Such support would let agency personnel know that they should take the dictates of the ADR Act of 1990 seriously, without infringing on agencies' discretion to decide when to use ADR. Presidential leadership on this issue is a low-cost but highly effective way to engender long-term and widespread acceptance of the view that resolving disputes through ADR is not only appropriate, it is, in many cases, preferable.

CROSS-REFERENCES TO OTHER NPR ACCOMPANYING REPORTS

Department of Labor, DOL04: Expand the Use of Alternative Dispute Resolution by the Department of Labor.

Reinventing Human Resource Management, HRM08: Improve Processes and Procedures Established to Provide Workplace Due Process for Employees.

ENDNOTES

  1. Agencies have varying enforcement authorities. Many enforcement statutes permit agencies to bring enforcement proceedings before agency "administrative law judges" (with judicial review on the administrative record in federal court). Other agencies must seek enforcement (with the aid of the Department of Justice) in federal district court. In either case, litigation costs and delays can be a concern.
  2. See, for example, Edles and Nelson, Federal Regulatory Process: Agency Practices and Procedures, 2d ed. (Prentice Hall Law and Business looseleaf), p.
  3. "The administrative process at most agencies could not operate if all cases had to be pursued to final, formal decisions."
  4. See Administrative Conference of the U.S., Recommendation 86-3, "Agency Use of Alternative Means of Dispute Resolution," 1 C.F.R. 305.86-3 (1993); Harter, Philip J., "Points on a Continuum: Dispute Resolution Procedures and the Administrative Process," Administrative Law Journal, Vol. 1, No. 1 (1986) p. 141.
  5. For example, the United States District Court for the District of Columbia has an alternative means of dispute resolution (ADR) program at both the trial and appellate levels. Finkelstein and Stanley, "The Federal Angle," Washington Lawyer (May/June 1992) p.
  6. State and local courts in many states, including California, New Jersey, and Florida have set up ADR programs. Telephone interview with Diana FarthingCapowich, Program Manager, State Justice Institute, August 12, 1993.
  7. Companies in the insurance, automotive, and electronics industries are among those using ADR processes to resolve disputes. See Center for Public Resources, Mainstreaming: Corporate Strategies For Systematic ADR Use, Practical Guide Series, looseleaf (New York, undated), Section E.
  8. "Bush vs. Clinton: The Candidates Debate," Reader's Digest (October 1992), p. 61.
  9. In mediation, a trained neutral third party helps disputants negotiate a mutually agreeable settlement. A mediator, by talking with each side, together and separately, can often help parties come to agreement by helping them develop options and explore acceptable settlements. The mediator does not have the power to decide a dispute; rather, the parties must agree to any resolution themselves. Mediation is often a component of other ADR processes. In early neutral evaluation, a neutral factfinder, often one with substantive expertise, evaluates the relative merits of the parties' cases. After hearing an informal presentation of the highlights of parties' cases or positions, the neutral provides a nonbinding evaluation. This should give parties a more objective perspective on the strengths and weaknesses of their cases, thereby making further negotiations more likely to be productive. In factfinding, neutrals determine disputed facts. This can be particularly useful where disagreements about the need for or meaning of data are impeding resolution of a dispute, or where the disputed facts are highly technical and would be better resolved by experts. Factfinding usually involves an informal presentation by each side of its case, whereupon the neutral provides an advisory opinion on the disputed facts, which can be used by the parties as a basis for further negotiation. A settlement judge serves essentially as a mediator or neutral evaluator in cases pending before a tribunal. The settlement judge is usually a second judge from the same body as the presiding judge. In some cases, a settlement judge may give an informal advisory opinion. See Administrative Conference of the United States, Recommendation 88-5, "Agency Use of Settlement Judges," 1 C.F.R. 305.88-5; Joseph, Daniel and Michelle Gilbert, "Breaking the Settlement Ice: The Use of Settlement Judges in Administrative Proceedings," Administrative Law Journal, vol. 3, no. 3 (1989), p. 571. A minitrial is not really a trial, but a structured settlement process in which the disputants agree on a procedure for presenting their cases in a highly abbreviated form (usually no more than a few hours or a few days) to senior officials for each side who possess settlement authority. This process allows those officials to see first hand how their case, and that of the other parties, play out; this can serve as a basis for more fruitful negotiations. Often a neutral presides over the hearing, and may subsequently mediate the dispute or help parties evaluate their cases. Arbitration is a relatively formal process, in which parties jointly select the decisionmaker(s), to whom they turn over the decisionmaking. The arbitrator(s), after hearing each side's case using procedures agreed on in advance by the parties, issues a decision. The standards for decision may also have been agreed on in advance by the parties. The arbitrator's decision may be nonbinding (in which case the process resembles neutral evaluation), or if the parties agree, it may be binding. See ACUS Recommendation 87-5, "Arbitration in Federal Programs," 1 C.F.R. 305.87-5 (1993); Administrative Conference of the United States, Agency Arbitration, Studies in Administrative Law and Procedure 88-1 (1988) (discussing constitutional and statutory issues). Current law contains limitations on the use of binding arbitration by federal agencies. See section 4 of ADR Act, codified at 5 U.S.C.A. 571- 583 (1993 Supp.).
  10. For example, litigation is necessary if a party needs to establish a judicial precedent. If there is a fundamental principle at stake that a party is absolutely unwilling to negotiate, ADR is likely to be unsuccessful. However, since most cases do settle eventually, it is important for parties to assess carefully whether an issue really is non-negotiable. If the costs of ADR would exceed the costs of resolution through other means, then other means should generally be used. Where a public record of the proceeding is needed, ADR may be inappropriate (although this concern may be accommodated by agreement of the parties). If a program requires complete uniformity of outcome, then individualized resolutions by the parties might not satisfy that need.
  11. 5 U.S.C.A. 571-583 (1993 Supp.).
  12. Conversations with staff of the Administrative Conference of the United States. See also, letter from Leonard L. Riskin, Professor of Law, University of Missouri-Columbia, to Senator Christopher Bond, July 5, 1993, commenting that "notwithstanding the passage of the [ADR Act] in 1990, some agencies have show[n] great reluctance to employ alternative methods of dispute resolution," p. 2.
  13. See U.S. Air Force, Secretarial Memorandum re Implementation of the Administrative Dispute Resolution Act of 1990, January 12, 1993; letter to Brian C. Griffin, Chairman of ACUS, from David A. Mader, Assistant Commissioner, Internal Revenue Service, July 15, 1993.
  14. U.S. Department of Labor, Report to the Secretary of Labor on the Philadelphia ADR Pilot Project (October 14, 1992), p. 52.
  15. Conversation with Department of Labor officials.
  16. Riskin, Leonard L., The Farmer-Lender Mediation Program: Implementation by the Farmers Home Administration, Report to ACUS (November 1991), p.
  17. (FmHA reports net savings to states in fiscal year 1989 as between $18 and $37 million and net savings to the federal government as between $6.6 and $14.8 million.)
  18. Federal Deposit Insurance Corporation (FDIC) staff paper, "Summary of FDIC ADR Creditor Claims Pilot Project," April 1993. ("$410,475 in legal fees and expenses avoided.")
  19. Environmental Protection Agency, Superfund Enforcement Mediation, Regional Pilot Project Results (October 1991); and EPA, "Superfund Enforcement Mediation Case Studies" (April 1992).
  20. Executive Order No. 12778, Civil Justice Reform (October 23, 1991), Federal Register 55195 (October 25, 1991). Although the Order supports ADR training and notes that ADR can be appropriate in litigation involving the United States and its agencies, it lacks guidance or specifics on how ADR techniques can and should be employed. Moreover, EO 12778's provisions relating to ADR primarily address ADR use in court litigation. The Order's mention of administrative proceedings (section 3) is limited to a citation to an Administrative Conference of the United States (ACUS) Recommendation on case management that includes a short section encouraging ADR use. ACUS, Recommendation 86-7, "Case Management as a Tool for Improving Agency Adjudication," 1 C.F.R. 305.86-7. This is especially weak given the opportunities presented by the Administrative Dispute Resolution Act, which is designed to facilitate ADR use at the agency level.
  21. Pub. L. 101-552, 3(c); codified at 5 U.S.C.A. 571 (1993 Supp.).
  22. 5 U.S.C.A. 573(c) (1993 Supp.).
  23. The executive order should address the use of ADR broadly, rather than limiting it to the regulatory context.
  24. See note 6 above and accompanying text.

REG07:
Rank Risks and Engage in "Anticipatory" Regulatory Planning

BACKGROUND

Agency planning tends to be short-range rather than long-range. There is almost always a new crisis demanding immediate attention. The budget cycle is annual. Top agency policymakers often are at the agency for only two years. All of this gives agencies incentives to develop short-term rather than longterm plans.

Nonetheless, long-term strategic planning is needed. Regulatory agencies have limited resources and limited demands they can make on the public. They share responsibility with Congress for ensuring that the nation gets the most for the money it spends on drug safety, the environment, and other matters. This requires agencies not only to set strategic goals, but also to determine whether those goals are being met.

This, in part, is the impetus behind the Government Performance and Results Act of 1993.[Endnote 1] The Act's implementation should do much to advance strategic, performance-based planning by the agencies. But regulatory agencies should take additional steps to plan their future regulatory agendas (especially those relating to regulation of environmental, health, and safety risks). The potential targets of regulations are so many, the implications so large, and regulatory resources so comparatively limited, that prioritization and long-range planning is imperative.

Regulatory agencies have begun to take some steps in this direction. Since 1985, executive agencies have been required to submit their planned significant regulatory actions to the Office of Management and Budget (OMB), which has published compilations annually as the Regulatory Program of the United States.[Endnote 2] Unfortunately, however, this process was not used as a significant planning document within the agencies or within the administration as a whole.[Endnote 3] Moreover, the Program was not published in a timely fashion; for example, the Program for 1992 was not distributed until February of 1993.[Endnote 4] The new regulatory review executive order will place greater emphasis on planning and will make the annual regulatory planning process more meaningful and timely.

The Environmental Protection Agency (EPA) has begun a process that, although still in its early stages, can provide a model for the other 20 or so agencies that regulate environmental, health, or safety risks, including the Food and Drug Administration, the Occupational Safety and Health Administration (OSHA), the Consumer Product Safety Commission, and the Nuclear Regulatory Commission.[Endnote 5] Its 1987 report, Unfinished Business, was based on an effort by 75 senior career managers to compare and rank the relative risks posed by 31 environmental problems regulated by the agency.[Endnote 6] These rankings, while based on professional judgments, were limited by incomplete data and the infancy of risk comparison methodologies. But the effort was a landmark in that it crossed programmatic lines and also explicitly pointed out the disparity between the rankings and the resource allocation at EPA.[Endnote 7] Interestingly, the EPA experts' rankings were almost the reverse of prevailing public opinion about risk.[Endnote 8]

In 1990, EPA turned to its Science Advisory Board to evaluate Unfinished Business and the agency's response to it. The Board's special 39-member Relative Risk Reduction Strategies Committee produced a report, Reducing Risk, that contained a series of recommendations to the administrator on how to undertake risk prioritization so that EPA could "target its environmental protection efforts on the basis of opportunities for the greatest risk reduction."[Endnote 9]

The Carnegie Commission has applauded Unfinished Business and Reducing Risk as a "groundbreaking enterprise."[Endnote 10] It has specifically recommended that "other risk reduction agencies should attempt to incorporate into their own decisionmaking and institutional culture the global perspective on risk that these reports and their implementation embody."[Endnote 11]

As a by-product of its recent focus on risk prioritization, EPA has devoted increasing attention to the identification of future risk. In 1988, the EPA Science Advisory Board produced a report, Future Risk: Research Strategies for the 1990s, that catalogued risk reduction strategies for the future.[Endnote 12] Its first recommendation was that EPA should shift its regulatory strategy "from end-of-pipe controls to preventing the generation of pollution."[Endnote 13] Administrator Carol Browner recently announced her intention to make "pollution prevention the guiding principle for all our programs at the [EPA]."[Endnote 14]

EPA has also employed its Office of Policy Planning and Evaluation (OPPE) to plan for the future. OPPE's Office of Strategic Planning and Environmental Data has a Futures Staff for this purpose.[Endnote 15] The Office also has recently joined with the Science Advisory Board to sponsor a new Environmental Futures Committee at the behest of Administrator Browner.[Endnote 16]

Another agency now engaged in risk prioritization is the Research and Special Programs Administration in the Department of Transportation, which regulates the safety of gas, offshore, and hazardous liquid pipelines. About 20 to 25 pipeline-related risks will be ranked as high, medium, or low as part of this initiative.[Endnote 17]

NEED FOR CHANGE

As the Carnegie Commission report declared, "the fundamental problem in regulatory decision making at the agency level, as at the presidential and interagency levels, is how to set priorities. It is a great challenge for science-based regulatory agencies to compare and rank individual risks and families of risks within the universe they regulate."[Endnote 18]

The scope of agencies' regulatory mandates far exceeds their rulemaking resources. For example, OSHA, which must issue health and safety standards for thousands of chemicals, carcinogens, and other dangers in most of our nation's workplaces, only has sufficient resources to pursue 15 to 20 rulemaking efforts at any given time.[Endnote 19] Hence the dire need for prioritization. Prioritization by agencies will also help inform Congress about agency resource needs and may help Congress reorder its own thinking while also informing regulated parties about the agency's likely future agenda.

Risk prioritization can help agencies distinguish between regulatory targets that demand immediate attention and those that allow for mid- or long-range corrections. But in this fast-changing technological world, regulatory agencies must also begin to anticipate future problems before it is too late or before an expensive regulatory scheme must be imposed.

For example, EPA is finding solutions now to the potential problem posed by the lithium in batteries used to power electric motor vehicles.[Endnote 21] OSHA, for another example, might have better anticipated the workplace stresses and injuries caused by the computerized office (e.g., non-ergonomic chairs, video display terminal exposure, and carpal tunnel syndrome).

It is a truism to say that tomorrow's regulatory problems will likely differ markedly from today's. Changes in risk prioritizing must be anticipated from numerous sources. New knowledge about the hazards of existing technologies (e.g., chlorofluorocarbons or CFCs) and exposures (e.g., radon) is always being developed. Solutions to one problem may cause another (e.g., increased indoor pollution in offices with improperly designed energy efficient programs). Changes in behavior (e.g., increased use of lawn-care pesticides), introduction of new technologies (e.g., biotechnology), side effects of government intervention (e.g., floodplain management), or changes in societal concerns (e.g., greater concern about costs or fairness of solutions) all bear on future risk regulation. These sorts of changes must be anticipated to the greatest feasible extent.[Endnote 22]

In short, regulatory agencies cannot simply be reactive. They must develop ways of, and devote resources to, anticipating the need for new research or regulatory initiatives--and possibly for new or adjusted statutory mandates to forestall or remedy those problems.

ACTIONS

  1. Rank the seriousness of environmental, health, or safety risks. (1)

Heads of regulatory agencies involved in the regulation of environmental, health, or safety risks should direct their agencies to rank the seriousness of those risks to permit better prioritization of their regulatory agendas. This is consistent with the new regulatory review executive order, which will require agencies to consider relative risks of different substances or activities within their jurisdictions. How that prioritization can best be accomplished should be left to individual agencies.

EPA's initial use of its program managers with a review by outside experts assembled by its Science Advisory Board has worked well. An alternate approach was espoused by a recent study by the Administrative Conference of the U.S. The study, done at OSHA's request, urged that agency to establish a formal prioritization committee staffed by politically sensitive technical experts who would serve staggered terms and be eligible for reappointment.[Endnote 23] Agencies should compare their own resource allocations with their internal risk rankings. Agencies should coordinate their development of risk prioritization processes and standards through the Regulatory Coordinating Group. Once risk lists are better developed and tested, they should be compared and possibly even coordinated across agency lines.

No one can pretend, however, that this will be an easy task. Analytical methodologies and basic data must be improved, and public perceptions of risk (which often differ markedly from that of scientists) must be factored into the equation.[Endnote 24] Comparing various types of health risks with disparate impacts on different populations (or human health concerns with degradation of ecosystems) requires value judgments that may be impossible to quantify. Thus, precision in rankings may be unattainable. But ranking of risks into high, medium and low categories is realistic. The payoffs from such prioritization should far exceed the cost of the undertaking. It would inform the legislative process (since agency priorities are largely statute-driven), the budget process, the interagency coordination process, and ultimately, the public's understanding of risk.[Endnote 25]

2. Develop long-range future plans and anticipatory approaches to regulatory problems. (1)

The heads of regulatory agencies should develop longrange future plans and anticipatory approaches to regulatory problems within their jurisdiction. Because of the differences in organizational structures among agencies, it is not possible to dictate just how such anticipatory planning can best be built into agency decision-making. Agencies should consider creating a staff to address long-range planning. Agency officials who serve in such offices should also meet under the sponsorship of the Regulatory Coordinating Group to discuss approaches to the problem.

CROSS-REFERENCES TO OTHER NPR ACCOMPANYING REPORTS

Department of Transportation, DOT01: Measure Transportation Safety; and DOT11: Improve Intermodal Transportation Policy Coordination and Management.

Environmental Protection Agency, EPA03: Shift EPA's Emphasis Toward Pollution Prevention and Away from Pollution Control; and EPA07: Establish Measurable Goals, Performance Standards and Strategic Planning within EPA.

Federal Emergency Management Agency, FEMA02: Develop a More Anticipatory Customer-Driven Response to Catastrophic Disasters.

Mission-Driven, Results-Oriented Budgeting, BGT02: Effectively Implement the Government Performance and Results Act of 1993.

Creating Quality Leadership and Management, QUAL02: Improve Government Performance Through Strategic and Quality Management.

ENDNOTES

  1. Pub. L. 103-62 (1993). The Act authorizes 10 three-year pilot projects to test program performance measures and requires Congress to vote in 1997 on mandatory government-wide implementation of five-year strategic planning, annual program goal-setting, and annual program performance reporting beginning in fiscal year 1998.
  2. E.O. No. 12498, Federal Register 1036 (1985).
  3. Discussions with regulatory officials from various agencies convened at Department of Transportation, May 13, 1993, Elliott, Donald, "TQMing OMB: Or Why Regulatory Review under Executive Order 12291 Works So Poorly and What President Clinton Can Do About It," draft, August 17, 1993, p. 24, Law and Contemporary Problems (Durham, North Carolina: forthcoming 1994).
  4. Regulatory Program of the United States Government (April 1, 1992-March 31, 1993), officially transmitted to Congress on January 15, 1993, but not available to agencies or to the public until February.
  5. Other risk regulatory agencies include the Animal and Plant Health Inspection Service, the Food Safety and Inspection Service, and the Forest Service (Department of Agriculture); National Oceanic and Atmospheric Administration (Department of Commerce); Army Corps of Engineers (Department of Defense), Office of Surface Mining, and the Fish and Wildlife Service (Department of Interior); Mine Safety and Health Administration (Department of Labor); and Coast Guard, the Federal Aviation Administration, the National Highway Traffic Safety Administration, and the Research and Special Programs Administration (DOT).
  6. Environmental Protection Agency, Unfinished Business: A Comparative Assessment of Environmental Problems (Washington, D.C., 1987).
  7. Of course, to a large extent, EPA's budgetary priorities are set by legislation. See Environmental Protection Agency, Reducing Risks (Appendix B.4), SAB-EC-90-021B (Washington, D.C., 1990).
  8. For example, while the public ranked nuclear accident radiation and industrial pollution of waterways as among the highest risks, EPA experts ranked them among the lowest. Conversely, the public was not very concerned about radon and indoor air pollution, which were ranked as high risks by the EPA experts. See Breyer, Stephen, Breaking the Vicious Circle: Toward Effective Risk Regulation (Harvard University Press 1993), Chart 7, "Rating Risks --How the Public and EPA Rate Health Risks Associated with Environmental Problems," derived from Russell, "What, Me Worry?" American Health 47 (June 1990). This problem may be affected by the media's perception of risk as well. See "Stories on Cancer's Causes Are Said to Be Misfocused: Media Overplay Minor Environmental Threats to Health, Experts Contend," Washington Post (July 27, 1993), p. A6.
  9. Environmental Protection Agency, Reducing Risks: Setting Priorities and Strategies for Environmental Protection, SAB-EC-90-021 (Washington, DC, 1990), p. 6.
  10. Carnegie Commission on Science, Technology and Government (Carnegie Commission), Risk and the Environment: Improving Regulatory Decision Making (Washington, D.C., 1993), p. 81.
  11. See also the favorable review "Developing a Mindset for Risk Prioritization," American Industrial Health Council Quarterly (Fall 1990), p. 3.
  12. Environmental Protection Agency, Future Risk: Research Strategies for the 1990s, SAB-EC-88-040 (Washington, D. C., 1988).
  13. Ibid., at pp. 5, 8-9.
  14. Carol M. Browner, Administrator, Environmental Protection Agency, Memorandum to EPA employees, "Pollution Prevention Policy Statement: New Directions for Environmental Protections," June 15, 1993, p. 1.
  15. Interview with David Rejelski, Futures Staff Chief, Office of Strategic Planning and Environmental Data, Office of Policy, Planning and Evaluation, EPA, June 24, 1993.
  16. The Charter of the new committee will be to conduct a short-term (five-year horizon) and longterm (20-year horizon) "scan of future developments that will affect environmental quality and the Nation's ability to protect the environment." The final report is scheduled for September 1994. Interview with Donald G. Barnes, Director, Science Advisory Board, July 7, 1993; Environmental Protection Agency, Draft Charter, Science Advisory Board, Environmental Futures Project (undated).
  17. Telephone interview with Judith S. Kaleta, Chief Counsel, Research and Special Programs Administration (RSPA); RSPA FY 1994 Budget Submission, pp. 145-149 (detailing its "Risk-Based Prioritization Plan").
  18. Carnegie Commission, p. 73.
  19. Shapiro, Sidney A., and Thomas McGarity, "Reorienting OSHA: Regulatory Alternatives and Legislative Reforms," Yale Journal on Regulation, vol. 6, no. 1 (Winter 1989), p. 14, n. 85.
  20. Interview with David Rejelski.
  21. Because of increasingly strict air quality rules in California, electric vehicles are expected to become more prevalent there in the next decade.
  22. The foregoing paragraph is adapted from Andrews, Richard, "Long-Range Planning in Environmental, Health, and Safety Regulatory Agencies," (draft background paper for Carnegie Commission, supra note 10, prepared May 4, 1991), pp. 6-7.
  23. Administrative Conference of the U.S., Recommendation 87-1, "Priority Setting and Management of Rulemaking by the Occupational Safety and Health Administration," 1 C.F.R. 305.87-1, 52 Federal Register 23,629 (June 24, 1987); derived from Shapiro and McGarity, supra note 19, pp. 18-23.
  24. See note 8 above.
  25. Congress and agencies often complain that longrange plans cannot be followed because the crisis of the day dictates congressional policy, which in turn establishes regulatory mandates. Political responsiveness to mood swings in the American Public, coupled with paralysis in the institutions of government, makes it very difficult to address problems requiring a steady, sustained, and unphotogenic response. A panel of The National Academy of Public Administration, Beyond Distrust: Building Bridges Between Congress and the Executive (Washington, D.C., January 1992), p. 76. To move away from a situation where the crisis of the minute dictates long-term policy, the public needs to be better educated about the problems facing the country. Reports of a study done by the Public Agenda Foundation, founded by Cyrus Vance, demonstrated that when members of the public were given short, balanced presentations on solid waste and global warming issues, they were able to grasp the scientific issues. Exposure to these presentations changed their views on how the issues should be handled to parallel the experts' views more closely. Carnegie Commission on Science, Technology, and Government, Risk and the Environment: Improving Regulatory Decision Making (Washington, D.C., June 1993), p. 92-93.