Copyright 1986 by the Georgetown Law Journal Association; Henry H. Perritt, Jr
August, 1986


NEGOTIATED RULEMAKING BEFORE FEDERAL AGENCIES: EVALUATION OF
RECOMMENDATIONS BY THE ADMINISTRATIVE CONFERENCE OF THE UNITED STATES
74 GEO1. L.J. 1625

Henry H. Perritt, Jr. [FNa1]


 

INTRODUCTION AND SUMMARY

  In 1982, the Administrative Conference of the United States (ACUS)   [FN1] issued Recommendation 82-4, [FN2] encouraging the use of negotiated rulemaking procedures under federal agencies. The concept of negotiated rulemaking arose from dissatisfaction with notice and comment and hybrid rulemaking pursuant to the Administrative Procedure Act (APA) [FN3] due to their increasingly adjudicatory and adversarial character. In 1985 ACUS issued further recommendations after several agencies had tried negotiated rulemaking.

  Negotiated rulemaking is a realistic alternative to adversarial administrative procedures. The technique permits affected interests to have greater control over the content of agency rules while ensuring fairness and balanced participation. It also permits agencies to obtain a more accurate perception of the costs and benefits of policy alternatives than agencies can obtain by digesting voluminous records of testimonial and documentary evidence presented in adversarial hearings.

  Between the time ACUS Recommendation 82-4 was issued and the end of 1985, federal agencies used negotiated rulemaking four times. [FN4] The Federal Aviation Administration (FAA) used negotiated rulemaking to develop a new flight and duty time regulation for airline flight crews. The Environmental Protection Agency (EPA) used negotiated rulemaking to develop proposed rules on non-conformance penalties for vehicle emissions and on emergency exemptions from *1628 pesticide regulations. The Occupational Safety and Health Administration (OSHA) encouraged labor, public interest, and industry representatives to negotiate a standard for occupational exposure to benzene. The benzene negotiations did not result in agreement among the parties on a proposed rule, but the other three negotiations did result in at least partial agreement, producing proposed rules based on the negotiations. Three of the rules have been promulgated in final form.

  Other agencies also have shown an interest in negotiated rulemaking, and several other major negotiated rulemaking efforts are under way. To guide the use of negotiated rulemaking, the ACUS published an additional set of recommendations in December, 1985. [FN5]

  The 1982 ACUS recommendations address negotiated rulemaking within the legal framework of the APA and judicially developed administrative law concepts. The 1985 ACUS recommendations take the next step, addressing the dynamics of the negotiation process in the rulemaking setting. Together, the two ACUS recommendations present a framework for planning a rule negotiation. This article summarizes the political and legal developments that led to the use of negotiated rulemaking, briefly describes the four completed experiences with the negotiation process, and explains the justification for the 1985 ACUS recommendations. The article is largely drawn from my report to the ACUS based on an investigation of the four completed rule negotiations. [FN6]

  The article is divided into nine parts. Part II reviews the evolution of the negotiated rulemaking concept, emphasizing the proposal Philip J. Harter presented in 1982 on negotiated rulemaking. [FN7] Part II also presents conceptual models drawn from political science and dispute resolution literature that lead to hypotheses as to the preconditions for successful negotiation. Part III reviews the benzene negotiations. [FN8] Part IV reviews the negotiations of the FAA. Part V *1629 reviews the two negotiations of the EPA. Part VI summarizes other past experiences and future agency plans with respect to negotiated rulemaking. Part VII compares the four negotiations. Part VIII discusses major legal issues affecting future use of negotiated rulemaking. Part IX explains the basis for the 1985 ACUS recommendations.

  It is important to view both the 1982 and the 1985 recommendations of the ACUS as a guide to issues to be considered rather than a formula to be followed. Negotiation is intrinsically a process that cannot be specified entirely in advance. Accordingly, what will 'work' in a particular case depends on a number of factors: substantive issues, perception of the agency's position by affected parties, relationships among the parties, authority of party representatives in the negotiations, negotiating style of the representatives, divergence of views within each constituency represented, and skill of agency personnel and mediators. [FN9] Some of these variables almost certainly will change several times during the negotiations. An agency cannot expect that the pattern followed successfully by another agency, or even by itself on another issue, can be transplanted without modification to another negotiation.

  The article explains that the APA need not be amended to provide for negotiated rulemaking. Amending the APA would risk destroying the flexibility it provides to adapt the negotiation process to the needs of different regulatory situations. Moreover, the four agency experiences show that the Federal Advisory Committee Act (FACA), [FN10] as interpreted by the sponsoring agencies and participants, was not a serious impediment to effective negotiations. The purpose of the FACA is satisfied when a properly balanced rulemaking negotiation is conducted, and the statute should not be interpreted to impose additional requirements that may jeopardize the success of negotiation. Under current judicial and agency interpretation of the FACA, caucuses and other working group meetings may be held in private when necessary to promote an effective exchange of views. Agencies should not be deterred from using negotiated rulemaking by a perception that the negotiating group will be an 'advisory committee' under the FACA. Some uncertainty can be reduced, however, if the General Services Administration amends its regulations to make it clear that meetings of caucuses and subgroups may be closed.

  Perhaps the most important insight to be gained from an assessment of the four completed negotiated rulemaking efforts is that an agency sponsoring a negotiated rulemaking should take part in negotiations. Negotiations are unlikely to succeed unless all parties are continually motivated by a perception that a negotiated rule would be preferable to a rule developed under traditional processes. As an incentive to negotiate an agreement, the agency should help *1630 create realistic expectations of the consequences of not reaching a consensus. Agencies must be mindful throughout the negotiations of the impact that agency conduct and statements have on party expectations. The agency may need to communicate with participants, perhaps with the assistance of a mediator or facilitator, to ensure that each has realistic expectations about the outcome of agency action in the absence of a negotiated agreement.

  Negotiated rulemaking is only one of several alternative dispute resolution techniques that can be used by administrative agencies. Negotiated rulemaking is designed to facilitate resolution of interest disputes. [FN11] Other techniques are more suitable for rights disputes, such as arbitration, fact-finding, and mediation. [FN12]

II. LEGAL AND POLITICAL BACKGROUND

A. HISTORY OF THE NEGOTIATED RULEMAKING IDEA

  Negotiated rulemaking [FN13] is a process for resolving interest disputes, similar in many ways to the legislative process. Negotiated rulemaking emerged as a distinct administrative law concept in the late 1970's in reaction to the unsuitability of notice and comment and hybrid rulemaking [FN14] processes for making quasi-legislative administrative agency decisions.

  New administrative processes arose from increased use of administrative litigation to address problems previously dealt with in markets, through private contractual negotiation, [FN15] or by elected representatives in legislative assemblies. *1631 The increased use of administrative regulation to address these problems arose from a combination of increased government intervention to correct market imperfections [FN16] and the impracticability of legislative assemblies taking on the resulting burden of decisionmaking.

  When rules for societal conduct are made in private markets or in representative assemblies, negotiation is relied upon. [FN17] Negotiation occurs as part of the legislative process in a representative assembly at two levels: first, during the process of electing representatives and second, during the interaction among the representatives in making the compromises necessary to pass laws. At both levels, no single interest or constituency gets all it wants; each must rank its demands and trade off one for another.  [FN18] There is an incentive to find common ground to form a majority to elect a favored candidate, or to pass desired legislation.

  The migration of problem solving to administrative agencies from markets and legislatures made negotiation more difficult. Agencies make rules because the legislature has left unresolved certain disputes among conflicting interests. In exercising rulemaking responsibility, agencies lack many attributes that facilitate accommodation among conflicting interests. For example, the accommodation and compromise that result from the election of representatives is missing, [FN19] which forces agencies to deal with a broad range of atomistic interests [FN20] and prevents them from benefiting from some subordinate aggregation. In addition, there is no assurance that even a multimember agency represents the major conflicting interests on any given issue.

  Clearly, the legislative delegation of details to specialized agencies arose out of necessity. However, because an administrative agency lacks the institutional attributes that make private accommodation an integral part of a legislative process, it must find other ways to strike a balance between competing views in developing a rule that will be reasonably satisfactory to those bound by it. In addition, the agency is obligated by law to pursue statutory objectives regardless of the views of affected interests.

  The conceptual basis of administrative law has shifted in recent years in search of ways to facilitate accommodation of conflicting interests.  [FN21] Original administrative law models emphasized judicial review to keep agencies within statutory bounds and decisionmaking procedures designed to promote the accuracy, rationality, *1632 and reviewability of agency decisions. [FN22] Increasingly, administrative law has served to provide a surrogate political process to ensure the fair representation of a wide range of affected interests in the administrative decisionmaking process. [FN23] The conceptual shift was motivated by a growing recognition that administrative procedures modeled on judicial processes do not fit the needs of administrative rulemaking. The main problem is that the procedures available to agencies largely have remained static. They were designed around the original conceptual models and are ill suited to a surrogate political process.

  A major part of the mismatch between administrative procedure and the decisionmaking requirements of delegated legislative power arose from the failure to distinguish rights disputes from interest disputes. [FN24] Adjudication is designed only to deal with rights disputes. [FN25] Rights disputes involve application of preexisting legal standards or rules of decision to facts determined by the adjudicator. Interest disputes, in contrast, are characterized by the absence of preexisting rules for decision. Resolution of interest disputes requires parties to work out the rules according to an accommodation of their interests. While the political process and private contractual negotiation are well suited to resolving interest disputes, the adjudicatory process is ill suited to resolve these disputes.  [FN26] Adjudicatory procedures nevertheless were superimposed on the rulemaking process. [FN27] As agencies were delegated more responsibility for legislating, the need for a process suited to interest disputes increased.  [FN28]

  Conceptual development of new administrative processes accelerated in the 1970's. Professor Stewart, writing in 1975, considered some of the alternative procedures available to agencies given responsibility for a surrogate political process. [FN29] He concluded that all of the obvious possibilities for application within traditional procedural frameworks were seriously flawed.  [FN30] He observed that agencies themselves are likely to afford certain interests disproportionate influence, and that the courts are ill suited to force an alteration in the balance of interests. [FN31]

  This led Stewart to explore more 'explicitly political' mechanisms for interest representation: [FN32]

    Since, in the absence of authoritative rules of decision, the resolution *1633 of the conflicting claims of a large number of competing interests is essentially a political process, a solution to the problems raised by the transformation of administrative law into a system of interest representation might better be achieved by a more direct and explicitly political scheme for securing the representation of all relevant interests affected by administrative decisionmaking. . . . Policy would result from a process of bargaining among the representatives of affected interests. [FN33]

  Stewart considered two ways to achieve better representation of interests: popular election of agency members, or their appointment for specific terms by private organizations designated by Congress. [FN34] He concluded that neither alternative had any real prospect of adoption in the foreseeable future. [FN35]

  Near the time that Stewart's article appeared, academic and policy currents converged, identifying negotiation as a process that should be examined as an alternative to adjudicatory litigation in making administrative rules. The policy currents were much influenced by John T. Dunlop, Secretary of Labor from 1974-1975and before that head of President Nixon's wage and price controls program. As much as anyone in the United States at the time, he was expert on the process of negotiation as a means of resolving workplace disputes. As director of wage and price controls, he administered an extremely broad delegation of rulemaking authority from Congress, [FN36] which struggled to conform to restrictions traditionally imposed by administrative law.

  Secretary Dunlop set a high priority on finding new ways to address hotly disputed issues within the jurisdiction of these regulatory agencies. He authored a paper identifying eleven problems with the existing approach to regulation and then suggested that 'the parties who will be affected by a set of regulations should be involved to a greater extent in developing those regulations.' [FN37] The paper offered as a useful example the method Dunlop used to implement section 13(c) of the Urban Mass Transportation Act, [FN38] which precluded distribution of certain mass transit subsidy funds until the Secretary of Labor certified that employees would not be affected adversely by the federally funded activities:

    Rather than prepare regulations, the Department brought together union and transit representatives and got them to prepare a three year agreement as to what protection employees should receive as a consequence of the federally funded activities. The Department mediated and provided technical assistance, helping to create the standards to apply to individual cases presented to it. [FN39]

Dunlop also sought to mediate an agreement between the steelworkers union and the steel industry on a health standard for coke oven emissions. [FN40]

  *1634 Dunlop's ideas required more conceptual development before they were practical. Professor Stewart in 1976 commented on the ideas advanced in Dunlop's The Limits of Legal Compulsion:

    I am very skeptical as to whether the collective bargaining type of alternative suggested will really work in very many areas. Most types of governmental regulation definitely do not involve the situation which characterizes collective bargaining--two more or less well-recognized groups with a mutual interest in a negotiated compromise (as well as objectives that conflict in part). Much governmental regulation is designed to protect loosely scattered, disorganized interests held by individuals--such as those of consumers or persons concerned with environmental quality. Normally these interests do not have authoritative spokesmen who are in a position to reach binding compromise agreements with the regulated Industry on matters of policy. The only leverage which such disorganized interests often have (apart from seeking champions and publicity in the legislature) is to attempt to goad the agency into action on their behalf and, failing administrative protection, recourse to the judiciary. The 'legal game-playing' [a phrase used in the Dunlop paper] may be an important and perhaps necessary tool enabling interests to protect themselves, and I am very unclear as to how the collective bargaining model could be applied in such situations. [FN41]

  Dunlop's idea of using negotiation as a regulatory procedure responded to Stewart's exploration of explicitly political processes. It was another means to permit political bargaining not considered explicitly in Stewart's article. The negotiation idea nevertheless presented difficulties, especially concerning the representation process.

  These difficulties were addressed by Philip J. Harter, under sponsorship of the ACUS. Harter was motivated to define the regulatory negotiation idea further because he thought discussions of the negotiation process at the Department of Labor and in congressional hearings [FN42] failed to give sufficient attention to the relationship between negotiation and the APA.

  Harter wrote a law review article [FN43] in which he carried the Dunlop concept further, addressing some of the concerns raised by Stewart's letter. Harter reviewed developments leading to Stewart's conclusion that the interest representation model had supplanted the traditional model of administrative law. [FN44] Harter concentrated on procedural issues, noting that under the hybrid rulemaking *1635 concept, [FN45] agencies had become umpires bound by the record. Under this concept, the rulemaking record was the vehicle of ultimate interest group control over agency decision. [FN46] The record- generating procedures focus on facts, however, while theoretically reserving to the agency the unilateral right to make policy. [FN47] Adjudicatory procedures for developing a record in hybrid rulemaking do not permit direct interest group interaction on policy issues.

  The procedural inadequacy of hybrid rulemaking led Harter to focus on practical alternatives to the fact oriented adjudicatory procedure. He found examples of such alternatives in the Consumer Product Safety Commission 'offeror process,' the National Institute of Building Sciences, innovations by Secretary Dunlop, ad hoc efforts involving dam building and uranium mine siting controversies, sequential negotiations frequently involved in formulating notices of proposed rulemaking safety and health consensus standards developed by private groups, the National Coal Policy Project, site specific environmental negotiation, and negotiated settlements of lawsuits challenging EPA regulatory decisions. [FN48] He sought to synthesize from these examples criteria for processes Stewart presumably would acknowledge as 'explicitly political.' [FN49] These criteria could be used in generalizing the use of explicitly political processes to improve administrative agency decisionmaking.

  Formulation of ACUS Recommendation 82-4 proceeded contemporaneously with Harter's article. The ACUS committee charged with reviewing the proposed recommendation was well balanced, representing a broad range of interests. Agency interest in the subject matter of the recommendation was intensified by the introduction of S. 1601 [FN50] at the time the recommendation was being considered. S. 1601 provided for the establishment of regulatory negotiation committees to allow direct participation of affected interests in federal rulemaking. The bill was circulated to federal agencies for their comment through the Office of Management and Budget (OMB) legislative clearance process. *1636 The necessity of formulating a position on S. 1601 undoubtedly heightened agency interest in the negotiated rulemaking concept.

  Recommendation 82-4 was adopted by the ACUS with little controversy. The recommendation suggests that:

    Agencies should consider using regulatory negotiation, as described in this recommedation, as a means of drafting for agency consideration the text of a proposed regulation. A proposal to establish a regulatory negotiating group could be made either by the agency (for example, in an advance notice of proposed rulemaking) or by the suggestion of any interested person. [FN51]

  The four efforts reviewed in parts III through VI of this article show that negotiated rulemaking is a practical alternative to notice and comment or hybrid rulemaking, and that Recommendation 82-4 is basically sound. ACUS Recommendation 85-5, developed in light of the four efforts, is discussed in part IX.

B. THEORETICAL FOUNDATION

  Implementation of the negotiated rulemaking concept in practice requires consideration of the characteristics of the negotiation process, interest groups, and regulatory programs.

1. Dynamics of Regulatory Negotiation

  Notwithstanding efforts to impose adjudicatory procedures on the formulation of regulatory policy, regulatory program decisions are made in a political environment. [FN52] The statutory framework of any program is as much a function of the political process that shaped it as it is a function of objective economic or scientific processes. In developing and pursuing a regulatory strategy within this statutory framework, an administrator is unlikely to pursue a strategy that he knows will enrage those groups most influential in the Congress, with the press, or with his superiors in the executive branch, regardless of what 'rational' analysis says about the merits of various options.

  Regulatory negotiation is but one mechanism to accomplish political accommodation. Negotiation will succeed only when persons able to use other processes have an incentive to participate in negotiations and to reach negotiated agreement. Incentives operate at several different levels: at the level of the negotiation itself, and at lower levels, where negotiations within constituencies are necessary to produce party positions. The best way to understand incentives to negotiate is first to consider the viewpoint of a hypothetical, monolithic party. Having described the incentives for this hypothetical party, one then can overlay complications that influence real world regulatory negotiations, especially complications inherent in intragroup interest aggregation.

  A useful conceptual structure for understanding incentives to negotiate is the one offered by Professors Fisher and Ury in their book on the negotiation process. [FN53] They explain that the participation of any party to a negotiation will be guided by that party's 'Best Alternative to Negotiated Agreement' (BATNA). If a party's BATNA is superior to what can be obtained in negotiation, the party will not participate. [FN54]

  For potential participants in a regulatory negotiation, BATNAs are determined by perceptions of what the agency will do in the absence of a negotiation.  [FN55] A rational, monolithic party will participate in regulatory negotiation only if it perceives the probable negotiation outcome to be superior to its BATNA, determined by the party's estimate of probable unilateral agency action. Different parties are likely to have different BATNAs because they predict the unilateral agency outcome differently, or because they have different predictions of the cost impact and benefit of agency action.

  The BATNA determined participation incentive is not invariant; it likely will change over time for each party, as the party gets additional information about the agency's intentions. Even more important, other negotiators, neutral mediators or convenors, and the agency itself can influence party BATNAs, and hence party incentives to negotiate and to agree.

  The most appropriate analogy to a regulatory negotiation is not a traditional labor-management negotiation, where BATNAs are determined by each party's assessment of the opponent's ability to inflict injury or to offer rewards. Instead, the appropriate analogy is to civil litigation settlement negotiation in which party predictions of what a nonparty, the judge or jury, will do determine BATNAs. In regulatory negotiations, as in settlement negotiations, the third party decisionmaker can influence party perception of likely outcome in the absence of a negotiated settlement. In other words, the agency or judge changes BATNAs by what she says about her intentions.

  This model suggests that regulatory negotiations are most likely to be successful when the agency (or some other credible source) persuades each potential participant that unilateral agency action has undesirable consequences for that participant. Lower BATNAs mean greater incentives to negotiate a solution.

  This analysis yields the first hypothesis for effective regulatory negotiation: parties will negotiate only if they perceive the outcome of unilateral agency action to be worse for them than what is attainable in the negotiation. The hypothesis has three corollaries. Regulatory negotiations are more likely to be successful if: (1) the parties agree on what the outcome will be in the absence of negotiations; [FN56] (2) the parties disagree on what the outcome will be in the absence of negotiations and are all pessimistic rather than optimistic; and (3) the *1638 agency actively influences party perceptions of BATNAs, emphasizing to each party the undesirable consequences of unilateral agency action in terms relevant to each party. [FN57]

  The preceding analysis of BATNA driven choices about participation has assumed that parties are monolithic: that parties affected by regulatory decisionmaking behave like rational individuals dealing with a single estimate of risk. Real parties do not behave this way. Real parties are represented by individuals, expressing the views of a group of constituents who usually have divergent views. The relations within constituency groups and between constituency groups and individual negotiators complicate the regulatory negotiation dynamics.

  Experienced mediators know that three agreements are necessary to any successful two-party negotiation: (1) an agreement between negotiator A and his constituents; (2) an agreement between negotiator B and her constituents; and (3) an agreement between negotiators A and B. [FN58] Agreements (1) and (2) can be called intraparty agreements. Frequently the most difficult mediation job involves achieving the intraparty agreements rather than achieving the negotiator-negotiator agreement.

  The intraparty problem is more difficult in regulatory negotiations than in labor negotiations, because the parties to regulatory negotiation are likely to be ad hoc groups or coalitions without formal processes worked out for internal decisionmaking. Ordinarily in regulatory negotiation, there is no equivalent of the 'exclusive representation' principle from the law of collective bargaining [FN59] available to bind constituents to the position taken by a group representative. At any point a constituency may disavow its putative representative in the negotiations, take its own position on matters under negotiation, present formal positions to the agency inconsistent with positions taken by its representative, or sue to have the negotiated rule set aside by the courts. Some risk of this occurring probably is inevitable, but regulatory negotiation cannot be successful unless a way can be found to resolve as many of these potential intraparty disagreements through a representation process in the negotiations instead of outside the negotiations. When representatives must make compromises at the bargaining table, the intraparty problem becomes worse.

  It is important for someone involved in the negotiation process, either the representatives themselves, the mediator or convenors, or the agency personnel, to be adroit at diagnosing intraparty problems and working creatively to facilitate intraparty agreement. Someone familiar with the internal structure and decision processes of institutions such as labor unions, corporations, and public interest groups involved in a particular negotiation can be invaluable. Even more valuable would be a mediator who already knows the key decisionmakers within a particular constituency, and thus is trusted to some degree by them.

  *1639 Individual group representatives in regulatory negotiations incur personal risk by participating and greater personal risk by reaching agreement. Any party representative who sits down with an adversary and agrees on a regulation is responsible in some measure for the regulation. If, on the other hand, the same party representative presents positions in the traditional administrative and judicial processes, the representative is not responsible for the outcome to the same degree; the decisionmaker, agency, or judge can be blamed, and the representative can point to the purity of his or her advocacy as proof of aggressive pursuit of constituency objectives.

  Personal risk is a difficult problem to deal with, and it always will remain in the background as a continuing bias against meaningful regulatory negotiation. Optimally, other participants, mediators or convenors, and the agency will recognize the source of the risk of and be alert to opportunities to reduce the risk of participation--compared with the risk of nonparticipation--on an ad hoc basis for each individual representative. As the undesirable consequences of unilateral agency action are communicated to constituents of representatives, the relative risk of participation is reduced. Periodic constituency meetings may be desirable, [FN60] and full reporting to constituents as negotiations proceed is essential. In addition, selecting negotiators with strong power bases can serve to permit freer exploration of compromises.

  Moreover, it is important to shield the give and take among negotiation participants from publicity. Fragmentary or inaccurate reports of positions taken or compromises considered by their representatives may alarm constituents, endangering the minimum level of constituent support necessary to make the participation of any representative meaningful.

  The choice between open and closed meetings among negotiators, however, also bears on the relative power of different kinds of interest representatives. Public interest groups frequently compensate for small financial and staff resources by mobilizing the press to influence public and congressional opinion. It is easier to accomplish this mobilization when the deliberations of the negotiating group are visible to the press. It is harder to do when the meetings not only are closed to the public, but also when the negotiators pledge nondisclosure. Thus the legal requirement of open meetings may result in a subtle enhancement of public interest group power, even when no outsider comes to the meetings or tries to publicize committee activities. More generally, open meetings benefit groups possessing less influence with agency decisionmakers. The decision whether to open negotiation to public scrutiny bears on the question of who gets to be included in formulating agency policy.

2. Characteristics of Interest Groups

  Political decisions in a complex society are made primarily through the interaction of interest groups. [FN61] Legislative action occurs when the balance of political *1640 power favors change. [FN62] The balance of political power generally is determined by the strength and intensity of feeling of groups within society perceiving that they have similar interests on a particular subject. [FN63]

  Interest groups arise to provide economies of scale in the exercise of power, and to permit individual concerns about particular issues to be focused within representative institutions. [FN64] The proposition that policy decisions are made by group interaction raises both macro and micro issues. The macro issues relate to the interaction among groups. The micro issues relate to interaction within groups.

  The macro issues raise questions of group power, issue maturity, and intensity of feeling. One commentatorhas drawn the following general conclusion about issue maturity: 'The objective must have been the subject of sufficient political debate so that the groups interested in it and opposed to it can be identified, their positions and relative strengths evaluated, and potential sources of support have time to develop.' [FN65]

  Issue maturity plays an important role in the development of how intensely different groups feel about a particular issue and how strongly they prefer different alternatives. Issue maturity also ensures that a range of alternatives has been formulated for consideration.

  Intensity of feeling is an important variable in the calculus of public opinion. [FN66] The influence of a particular group or faction, which combined with the influence of other groups or factions determines popular will, is a function of numerosity and intensity of feeling. [FN67] Transaction costs reduce the desire of interest groups to have intense involvement in the full range of political decisions. [FN68] In other words, the cost-benefit ratio for participation in policy formulation is unfavorable when a potential participant is indifferent among outcomes. [FN69]

  Microanalysis of interest groups focuses on the representation, or interest aggregating function, that interest groups perform. [FN70] It also permits consideration of the ways in which interest groups function to intensify member interest in particular issues, to formulate concrete alternatives, and to articulate member positions. [FN71]

  Interest groups have overcome the fragmentation, transaction-cost, and free- rider difficulties impeding group action through 'organizational entrepreneurship.' [FN72] Especially when the issues involved are complicated, group members may defer almost entirely to the decisions of group representatives. This deference can make negotiations involving such representatives more fruitful. Trade unions or trade union federations dealing with complex technical regulatory disputes present an example of this phenomenon. Rank and file members have a low level of awareness of technical issues, and they tend to defer to a handful of staff experts.

  On the other hand, some public interest groups may prefer short term litigation victories to negotiation. Greater publicity associated with a dramatic victory and extreme statements made in litigation tend to facilitate fund raising and other facets of membership support. These groups tend to be less risk averse than business groups, and they therefore may prefer the all- or-nothing characteristic of litigation rather than accommodation. [FN73] In addition to lower risk aversion, public interest groups and trade unions may perceive that adversarial procedures better compensate for scarce technical resources than negotiation. [FN74] Careful selection of test cases in court permits targeting of scarce resources on key legal issues rather than dissipating them on complex factual disputes. [FN75]

  These are only tendencies of public interest groups. Obviously the most effective consumer groups do more than litigate; many participate regularly in the legislative process, which always involves negotiation and accommodation. For example, the consumer groups involved in the negotiated rulemaking efforts addressed in this article participated actively and in good faith. [FN76]

  Because large firms are more risk averse than small firms or organizational entrepreneurs, a negotiated resolution of a regulatory dispute is likely to be more attractive to interest groups dominated by a few large firms than to public interest groups or other groups with more fragmented membership.  [FN77] Negotiation, however, while attractive, may be more difficult because the large constituents have the resources to develop idiosyncratic positions and priorities, making intragroup compromise more difficult than it is in groups in which constituents defer to their representatives.

3. Characteristics of Regulatory Programs

  Interest groups interact with regulatory programs. Indeed, interest groups arose in part because of increased government regulation. [FN78] The nature of the interaction is determined in part by the nature of the regulatory program, because the nature of the program affects the intensity of interest group feeling on *1642 regulatory issues. [FN79]

  Professor James Q. Wilson offers a classification of regulatory programs based on the incidence of costs and benefits likely to influence intensity of group feeling. [FN80] He suggests that programs fall into three categories. Programs in Wilson's first category concentrate their benefits on a small group and distribute their costs over wide sectors of the population. Economic regulation of railroads and airlines, milk prices and taxicabs fall into this category. Programs in Wilson's second category concentrate both benefits and costs on a small group. Regulatory programs relating to labor- management relations fall into this category; labor or management benefits at the expense of the other. Wilson's third category encompasses most recent consumer and environmental protection and health and safety regulation. Here, benefits are diffused over large parts of the population and costs are concentrated on relatively narrow sectors. As Wilson points out, the development of policy is particularly difficult with respect to the third category of regulation, because the number of transactions subject to the regulation is likely to be far greater than in the first or second categories: a few hundred license applications before the Federal Communications Commission (FCC) compared with safety and health practices in nine million workplaces. One can add to Wilson's list a fourth category. This category includes programs whose costs and benefits both are diffused. An example would be automobile emissions device inspection programs.

  Programs in Wilson's second category are better candidates for negotiation than programs in the fourth category, because it is easier to mobilize interest representatives for the bargaining process when the interest groups are few in number and narrow in scope. Moreover, programs with narrow impact are less likely to attract intervention by Congress, the press, the White House, and the public. [FN81] Between the extremes represented by the second and fourth categories, Wilson's first and third categories present intermediate levels of difficulty in organizing interest representatives for regulatory negotiation.

C. CRITERIA FOR NEGOTIATED RULEMAKING

  ACUS Recommendation 82-4 offers criteria to select subjects for negotiated rulemaking. The recommendation was drafted for the conference by Philip J. Harter. Harter's contemporaneous law review article addresses the same criteria in somewhat greater detail than the recommendation. Harter dissected the negotiation process into six distinct facets:

    (1) assembling the negotiators; [FN82]

    (2) the negotiations themselves; [FN83]

    (3) achieving consensus; [FN84]

    (4) the means of reporting consensus to the agency; [FN85]

    *1643 (5) agency action; [FN86] and

    (6) judicial review. [FN87]

Finally, he articulated a set of nine principles that are embodied almost verbatim in ACUS Recommendation 82-4. [FN88]

  Harter recognized that people will negotiate only as long as they believe negotiations will resolve their dispute in a manner more favorable to each of them than other dispute resolution techniques. [FN89] The acceptability of negotiation as a dispute resolution process is determined by the relative power of the interested parties. [FN90] Harter identified four sources of power:  [FN91]

    (A) the ability to use a set of preexisting criteria to structure decisionmaking; [FN92]

    (B) availability of some other formal process for decisionmaking; [FN93]

    (C) the ability to gain access to these alternative decisional processes in a way that a party could use its own resources most effectively; [FN94] and

    (D) the power to delay a decision by any of these means.

  The relative attractiveness of negotiations will be influenced by these aspects of power. [FN95] Harter offered the following criteria to define situations where negotiation would be most effective. [FN96] He stressed that he did not envision mechanical application of the criteria or satisfaction of every criterion. [FN97]

1. Countervailing Power [FN98]

  Each party must have power to affect the decisional outcome. This can flow from the capacity to influence the legislature, the ability to run an effective public relations campaign, substantial litigation resources, or any other way of obtaining an outcome favorable to the party, or from inflicting costs on opponents, in another forum. Negotiation will be effective as a decisional process only if no *1644 one party has power sufficient to overwhelm the others. [FN99] Increased power on one side, however, strengthens incentives for opposing sides to seek a negotiated solution.

2. Limited Number of Parties [FN100]

  It is difficult to negotiate when a large number of people and demands must be accommodated. 'Pure' negotiation as a method for producing a new tax bill, for example, would not work. [FN101] The number of people whose interests are involved would probably be so large so as to preclude effective communication. [FN102]

3. Mature Issues [FN103]

  The issues must be readily apparent and the parties must be ready to decide them. If information is lacking or the parties are still establishing their positions, [FN104] negotiation cannot be utilized because the parties do not know what their positions are or what compromises they are prepared to make. The importance of issue maturity also has been recognized by other students of political behavior. [FN105] Issue maturity is significant even when disputants are individuals. Each disputant must have some time to think about her position on a new question or proposal. Issue maturity is far more important when disputants are groups. Constituency positions must be determined, and this requires aggregation and trading off within the group.  [FN106] The more complex the issue, and the more novel the possible solution, the longer it will take for this process to occur.

4. Inevitability of Decision [FN107]

  There must be pressure for resolution of the matter. [FN108] Effective negotiation requires compromise that involves making concessions. Most people are reluctant *1645 to make concessions until they are forced to do so by the prospect of something unpleasant. Deadline pressure can result from fear that the decision will be taken away from the negotiators. [FN109]

5. Opportunity for Gain [FN110]

  Negotiation must have the potential to produce gain for all parties. Negotiated solution to 'zero sum games' are difficult to achieve. [FN111] Effective mediation helps negotiating parties discover alternative formulations and to perceive the true value of their BATNAs. [FN112] The heightened perception of loss in alternative forums may be great enough to overcome any fear about loss through negotiations. [FN113]

6. Absence of Fundamental Value Conflict [FN114]

  The regulation to be developed cannot involve compromise of deeply held beliefs or values. This would involve costs higher than most parties could tolerate. [FN115] This criterion does not mean that negotiation cannot be used to resolve minor issues involved in controversies in which fundamental values are at stake. This criterion also does not mean that the implementation of regulations developed on a fundamental matter cannot be negotiated.  [FN116]

  Moreover, the fundamental value criterion frequently is misunderstood. Merely because substantial costs are involved does not mean that fundamental values are involved. Instead, fundamental values are those having ideological, rather than merely economic significance, or, perhaps, economic risks of such magnitude that they seriously threaten a party's very survival.

7. Permitting Trade Offs [FN117]

  The negotiation process is an evaluative one. The parties determine what is most important to them and direct their behavior toward accomplishing it. If there is only one issue to resolve, one position to assume, negotiation is unlikely to result in agreement. A situation where there are two or more issues to resolve, however, with the attendant possibilities of gain on one issue offsetting loss on another, is more amenable to negotiation. [FN118]

*1646 8. Research Not Determinative of Outcome [FN119]

  Resolution of a dispute should not depend on research results. The parties may be unwilling to formulate or compromise positions in the face of scientific uncertainty. Research results might produce a clear victory for one interest.  [FN120] Even if a clear win or loss might result from the research findings, however, the parties still might choose to negotiate parallel or peripheral issues. [FN121] The parties might negotiate issues to be researched and the way in which the research should be conducted.

9. Agreement Implementation [FN122]

  Some kind of effective implementation process must be present. Lack of an implementation process would destroy the core prerequisite for negotiation: the parties' belief that their own interests will be furthered by negotiating.

  Harter recognized implicitly two forms of negotiated rulemaking, one in which the agency participates in the negotiations and another in which it does not.  [FN123] The same two variants had been addressed more explicitly in a Harvard Law Review note published about the same time. [FN124]

  The objective of negotiated rulemaking is to reach 'consensus' among the participants as to the content of the proposed rule. Harter characterized the definition of 'consensus' as 'one of the most difficult and complex questions in regulatory negotiation.' [FN125] He concluded that experience was necessary before anyone could develop more concrete ideas on what consensus should entail, pointing out, however, that the existence of a consensus is more a matter of feel than of mathematical calculation. [FN126] The Harter formulation necessarily omitted detailed formulation of the conditions conducive to closure on an agreement. [FN127]

  The Harter guidelines, reflected in ACUS Recommendation 82-4, proved sound in practice application. The following parts of the article explore application of the guidelines in four actual rule negotiations. The article begins with the benzene negotiation, because this negotiation did not produce consensus and is therefore a useful basis for scrutiny to determine what can go wrong, and because *1647 the author personally interviewed most of the major participants in the benzene negotiation. [FN128]

III. THE BENZENE NEGOTIATIONS

A. INTRODUCTION

  In the summer of 1983, the Occupational Safety and Health Administration  (OSHA) began to develop a revised standard for workplace exposure to benzene through negotiated rulemaking. An earlier benzene standard had been invalidated in litigation that ultimately reached the Supreme Court. [FN129] Negotiations proceeded for more than a year, producing agreement in principle. The parties did not report to the OSHA, however, because no agreement could be reached on some details and because the political climate had changed, and there was dissension among the various interests (and others similarly situated). Philip J. Harter and Gerald Cormick facilitated the negotiations.

  Although one reason the negotiations adjourned without agreement was that the negotiators expected the OSHA to issue a standard unilaterally, more than a year elapsed before the OSHA published a proposed rule. [FN130]

B. HISTORY OF THE REGULATION OF BENZENE BY THE OSHA

  Benzene is a clear, colorless, highly flammable liquid with a strong odor. It evaporates rapidly under ordinary atmospheric conditions, giving off vapors nearly three times heavier than air. The petrochemical and petroleum refining industries produce nearly all the benzene used in the United States. Benzene is used as a solvent or a reactant alone or with another liquid. Some 274,000 workers are exposed to benzene in seven major industries: petrochemicals, petroleum refining, coke and coal manufacturing, rubber tire manufacturing, bulk storage terminals, bulk plants, and transportation. [FN131]

  The OSHA adopted the original standard for benzene exposure in 1971. This was a 'national consensus standard,' [FN132] setting a time-weighted-average exposure level (TWA) [FN133] of 10 parts per million (ppm); ceiling concentrations of 25 ppm; and permitted excursions above the ceiling not to exceed 50 ppm for more than ten minutes in any eight hour work period.  [FN134]

  Revision of this standard was triggered by a 1974 report prepared by the National *1648 Institute of Occupational Safety and Health (NIOSH),  [FN135] acknowledging that benzene might cause leukemia but recommending retention of the national consensus standard for the time being. In 1976, the United Rubber Workers (URW) petitioned unsuccessfully to lower the standard on an emergency basis. Later that same year, NIOSH submitted a revised report concluding that benzene could cause leukemia and that no safe level for benzene exposure could be established. [FN136] It recommended that the OSHA establish a new standard prohibiting exposure in excess of 1 ppm. [FN137]

  In response, the OSHA issued voluntary guidelines in January, 1977, recommending that worker exposure to benzene not exceed an eight hour TWA of 1 ppm in any eight hour shift of a forty hour week. Four months later the OSHA promulgated an emergency temporary standard for occupational exposure to benzene [FN138] setting an eight hour TWA limit of 1 ppm with a ceiling level of 5 ppm for any fifteen minute period during an eight hour work period.  [FN139] Challenges to the standard were filed in Courts of Appeals for both the District of Columbia Circuit and the Fifth Circuit. [FN140] The Fifth Circuit issued a temporary restraining order against the OSHA, [FN141] and the emergency standard never went into effect. [FN142]

  One week later, the OSHA published notice of a proposed final benzene standard. [FN143] Hearings on the proposed standard were held from July 19 through August 10, 1977. Ninety-five persons testified. [FN144]

  The final standard of the OSHA, issued in February, 1978, was identical to the emergency standard, limiting exposure to an eight hour TWA of 1 ppm, with a ceiling of 5 ppm for any fifteen minute period in an eight hour work period. It also prohibited eye and skin contact and required measurement of employee exposure, engineering controls, work practices, personal protective clothing and equipment, signs and labels, employee training, medical surveillance, and record keeping.

  Producers and users appealed to the Fifth Circuit seeking preenforcement review of the final standard. [FN145] The Fifth Circuit invalidated the standard because it was based on findings that were unsupported by the administrative record. [FN146] The court concluded that the OSHA had exceeded its authority because it had not been shown that the eight hour TWA exposure limit was 'reasonably necessary or appropriate to provide safe and healthful employment' as required by section 3(8) of the Act. [FN147] The court further concluded that section 6(b)(5) [FN148] did not give the OSHA the unbridled discretion to set standards designed to create absolutely risk free workplaces without regard for cost. [FN149]

  The Supreme Court affirmed, [FN150] with a plurality concluding that the OSHA's rationale for lowering the permissible exposure limit from 10 ppm of benzene to 1 ppm was based on a series of unsupported assumptions indicating that leukemia might result from exposure to 10 ppm of benzene. [FN151] There had been no showing that leukemia was caused by exposure to 10 ppm of benzene and that it would not be caused by exposure to 1 ppm. [FN152] The plurality reasoned that section 3(8) of the Act implies that before the Secretary of Labor can set a permanent health or safety standard, he must first find that the workplace is unsafe in the sense that significant risks are present and can be eliminated or lessened by a change in practices. [FN153] Therefore, the Secretary exceeded his power by avoiding this threshold responsibility when he relied on a special policy for carcinogens that placed the burden of proving the existence of a safe level of exposure to benzene on the industry.  [FN154] The plurality also noted that the Act's legislative history supports the conclusion that Congress was concerned not with absolute safety but with the elimination of significant risk of harm. [FN155]

  Justice Stevens, joined by Chief Justice Burger and Justice Stewart, also concluded that the burden was on the OSHA to show, on the basis of substantial evidence, that it is at least more likely than not that long-term exposure to 10 ppm of benzene presents a significant risk of material health impairment. [FN156] The OSHA, they concluded, did not satisfy this burden.  [FN157]

  Chief Justice Burger, concurring, emphasized that the requirement for the OSHA to 'retrace its steps' with greater care does not displace its ability to make policy judgments and noted that the Act also required the Secretary to determine that the economic effects of the new standard bore a reasonable relationship to the expected benefits and that the OSHA had not done this.  [FN158] Justice Rehnquist, concurring in the judgment, expressed the view that section 6(b)(5) of the Act, relied on by the Secretary in promulgating the benzene standard, was an unconstitutional delegation of legislative power to the executive branch. [FN159]

  Justices Marshall, Brennan, White, and Blackmun dissented. They stressed that the Act required a reviewing court to uphold the Secretary's determination *1650 if supported by 'substantial evidence in the record considered as a whole.' [FN160] Therefore, the Secretary's standard was fully in accord with the statutory mandate of section 6(b)(5) of the Act that standards for toxic materials or harmful physical agents most adequately assure that no employee will suffer material impairment of health or functional capacity. [FN161]

  Following the Supreme Court's decision, the OSHA considered various quantitative risk assessments to determine whether it should try again to adopt a standard below 10 ppm. Of the 274,000 workers exposed to benzene, the OSHA estimates that some 262,000 are exposed to concentrations at or below 1 ppm as an eight hour TWA, while approximately 10,000 are exposed to levels between 1 and 5 ppm. The OSHA estimates that fewer than 1,500 workers are presently exposed to benzene levels averaging 5 ppm or more. [FN162]

  On April 14, 1983, Dr. Sidney M. Wolfe, Director of the Public Citizen Health Research Group, wrote Thorne G. Auchter, Assistant Secretary of Labor for Occupational Safety and Health, requesting that the OSHA issue an emergency temporary standard under section 6(c)(1) of the OSHA statute. Auchter denied the petition, noting that only eight to nine percent of workers actually are exposed to benzene in excess of 1 ppm on an eight hour TWA basis. He committed the agency, however, to issue a new permanent standard on an expedited basis, not later than June, 1984. Specifically, he committed to a timetable, pursuant to which the OSHA would add a benzene standard to its 'regulatory agenda' by June, 1983; submit a proposed standard to the OMB by November, 1983; publish a proposed standard in the Federal Register by December, 1983; hold a fact- finding hearing by February, 1984; and publish the final standard by June, 1984. [FN163]

C. DECISION TO NEGOTIATE

  At about the same time that Auchter and Wolfe were corresponding over the plans of the OSHA for a revised benzene standard, Philip Harter and Gerald Cormick held discussions with the OSHA staff about the potential for regulatory negotiation to facilitate the standard setting activities of the OSHA. As a result of these discussions, Assistant Secretary Auchter decided to explore negotiated rulemaking as a means of resolving the benzene dispute. The OSHA's interest in regulatory negotiation was heightened by a perception that negotiations had been useful in connection with development of the coke over standard. [FN164]

  Preliminary discussions with the parties and actual negotiations proceeded from July, 1983, until October, 1984. Initially the convenor/facilitators  [FN165] determined that the parties were skeptical that negotiations could produce a consensus on OSHA action, but they thought meetings might be helpful. After the meetings began, the participants decided to attempt to narrow their differences *1651 and to reach agreement on a standard under a protocol that required agreement on a 'total package' before anything substantive was reported to the OSHA. Representation of the petroleum industry, one of the key industry stakeholders, proved difficult because of dissension among firms within the industry about the wisdom of negotiating. Nonparticipant trade unions were concerned that options being considered in the negotiation might set an adverse precedent for other health standards of the OSHA. Nevertheless, the participants nearly agreed on a standard before finally adjourning in the fall of 1984.

D. ISSUES IN NEGOTIATION

  The central issue in the benzene negotiations was the Permissible Exposure Limit (PEL). The existing standard was set at 10 ppm, and the effort of the OSHA to set a standard at 1 ppm had been invalidated in the Supreme Court. The steel industry, facing difficulty in meeting the 10 ppm standard and questioning evidence of health hazards below that level, opposed reducing it. The rubber industry had been meeting a 1 ppm level since the late 1970's.

  Between these extremes were the petroleum and chemical industries, which thought they could accept a level of 2 ppm, although they were not persuaded that it was justified on a scientific or medical basis. A 1 ppm standard was a problem for the petroleum and chemical industries because of variability in measurements. Eighty-five to ninety percent of monitoring results would show levels lower than 1 ppm, but measurements above that level also would occur, involving different locations and circumstances hard to predict or evaluate. Thus frequent measurements in a single facility would show noncompliance with a 1 ppm limit at least some of the time. [FN166]

  Labor vigorously opposed anything higher than 1 ppm, and the number '1' became an article of faith within the labor movement. New scientific data that had become available since the promulgation of the 1978 standard reinforced the labor view that a 1 ppm standard was justified based on health risk.

  Industry hoped to deal with its own concerns, while satisfying labor's desire for a standard of 1 ppm, primarily through the concept of averaging: requiring an employer to show compliance with a 1 ppm standard by an average of measurements taken over some time period. Many different ideas were explored with respect to averaging: a standard expressed in terms of 'ppm-hours,' such as a limit of 40 ppm-hours; providing that a single 'exceedence' of the eight hour TWA would not result in a violation as long as the average eight hour TWA for the previous five readings was within the standard; providing that a single 'exceedence' of the eight hour TWA would not result in a violation as long as the average eight hour TWA for the next five readings was within the standard.

  Labor was concerned that a standard that allowed averaging would set a precedent for other health standards. Some labor participants preferred setting a simple PEL of 1 ppm, and dealing with the problems caused by exceeding the standard through enforcement guidelines.

  The impact of a new benzene standard on tort liability also was a concern for industry, particularly petroleum refiners. Many of those familiar with the negotiations *1652 perceived the tort liability issue as more important to some industry participants than the PEL. Gasoline has benzene in it. Benzene causes leukemia and other blood disorders, and may cause tumors. Everyone who drives a car is exposed to low levels of benzene. Employees exposed to benzene in enterprises outside the petroleum industry may sue manufacturers of benzene instead of, or in addition to, their own employers because of workers' compensation limitations on suit, or for other reasons. Thus the class of potential plaintiffs against petroleum industry defendants is extremely large. Asbestos litigation, and litigation over benzene-caused disease in Gulf Coast shipyard industry made industry executives sensitive to the potential for benzene litigation in which their companies might be defendants.

  The relationship between the ability of petroleum companies to defend such lawsuits and a finding by the OSHA of significant health risk at levels of 10 ppm or below is speculative. Section 4(b)(4) of the Occupational Safety and Health Act [FN167] says that the OSHA standards shall not be outcome determinative in tort litigation. Nevertheless, it was widely perceived that a finding of significant health risk could affect adversely the industry's ability to protect itself against large tort liability. Labor was basically uninterested in tort liability, but it was willing to try to develop language satisfactory to the petroleum refiners as long as the language did not jeopardize the ability of the standard to survive judicial review.

  The negotiators were faced with a dilemma, however, in formulating a compromise. Industry wanted to avoid a strong finding of health risk below 10 ppm to mitigate tort liability problems. If the finding was weakened to reduce tort liability problems, however, the chance that a negotiated standard would be invalidated in the courts increased, leaving the parties in the position of having compromised on some issues without the benefits of the total negotiated package.

  The rubber industry had been concerned previously only with that part of a standard that might deal with skin contact. The industry hoped for an exemption for uses of liquids with a content of less than 0.3% benzene by volume but labor was opposed to an outright exemption. The negotiators agreed upon an exemption for employers who could show both that liquids contacting employees' skin contained less than 0.3% benzene by volume, and absorption rates low enough that a 1 ppm level would not be exceeded. [FN168]

  Labor understood that variability of benzene levels presented legitimate problems for standard definition. Labor participants sought to determine actual industry practice regarding temporary excursions and to write those practices into a standard; for example, requiring action and frequent monitoring after a spill or a ventilation system failure. They perceived, however, that industry participants were never willing to be pinned down about what should be required when the standard was exceeded, that is, what action should be taken when the 'action level' was exceeded.

  A separate goal for labor was to use the benzene negotiations to induce the OSHA to accept the proposition that medical surveillance was most needed for *1653 older workers, some of whom had left active service. Industry participants were reluctant to embrace this idea, apparently out of fear that it would stir up tort claims by the older workers. In addition, industry noted that screening of older workers does not permit prophylactic action to the same degree as screening of younger workers.

  In general, the labor participants were unaccustomed to the idea that different parts of a benzene standard could be traded off against each other, based on economic impact. Labor was accustomed to thinking of each piece of a standard--the PEL, engineering controls, monitoring, medical surveillance, action requirements--as independent components, each with its own rationale. Participation by labor in discussing tradeoffs resulted only from the realization that industry viewed the standard as a total package and was concerned with the overall economic and health impacts.

E. BENZENE AS A CANDIDATE FOR NEGOTIATED RULEMAKING

  In retrospect, benzene was both a good and a bad candidate for negotiated rulemaking. Benzene was a good candidate because formulation of a rule required several different subjects to be addressed, creating the potential for tradeoffs; [FN169] because party positions had crystallized during seven years of rulemaking proceedings and litigation reaching the Supreme Court of the United States; [FN170] and because a limited number of parties were involved. [FN171]

  The history of the litigation also made benzene a good candidate for another reason: a standard generally supported by organized labor had been invalidated in the courts because of aggressive opposition by industry, and promulgation of a revised standard had been delayed for seven years. This demonstrated that a measure of acceptability to industry could speed attainment of labor's health and safety objectives. [FN172]

  Benzene was a bad candidate because the years of administrative and court litigation had solidified the positions of the parties and the agency and generated a voluminous record lending support to each of those positions. In addition, benzene was a bad candidate because of the perception that benzene exposure creates significant health risks, including the risk of fatal disease. This increased the likelihood that basic values would be at issue in negotiations. [FN173] Finally, though benzene is a single substance, the issues confronting the different industries that would be affected by a benzene rule were diverse, resulting in a complex, multipolar negotiation and making it difficult for the participants to assimilate all the data relevant to the issues under discussion. [FN174]

  *1654 Some participants thought benzene was a bad candidate because the issues were essentially technical, reinforcing their view that a 'correct' standard could be determined objectively. [FN175] The technical nature of the issues, in the view of these participants, meant that the most useful process for developing a standard would be probing examination of research data--an activity for which negotiations are less well suited than formal adjudicatory hearings. They thought that labor's disparity of resources and technical expertise would be a more prominent handicap in negotiations than in traditional administrative litigation. They also feared that the negotiations would delay the issuance of a standard, despite the commitment of the OSHA to issue a final standard by June, 1984. [FN176]

  Another reason benzene was a bad candidate was the parties' differing perceptions of the implications of the Supreme Court's decision. Labor believed that a 1 ppm standard could pass court review if the OSHA did a better job of marshaling scientific evidence in support of the standard. Taking this view, there was little to be accomplished in negotiations, since the only remaining problem was a technical data analysis. Industry, on the other hand, viewed the Court's decision as a repudiation of attempts by the OSHA to revise the 10 ppm national consensus standard. This perception suggested that negotiations would be concerned with the full range of issues related to benzene exposure in the workplace. [FN177]

  Benzene also was a bad candidate because the statute authorizing a standard makes a finding of 'significant health risk' a prerequisite for agency power to regulate. The importance of this finding to a valid standard had been underscored by the Supreme Court in Industrial Union Department v. American Petroleum Institute, [FN178] and the participants in the benzene negotiation were acutely aware of it. The centrality of the significant health risk finding presented a dilemma that utlimately proved insurmountable. If a negotiated standard was to survive judicial review, [FN179] the OSHA must find that significant health risks resulted from the existing standard of 10 ppm. A finding of health risks at PELs lower than 10 ppm was perceived, however, by the petroleum industry at least, as increasing potential tort liability. Industry thought the OSHA was inclined to read the requirement of the Supreme Court decision as saying, 'Quantify the risk, making it seem as high as possible at the lowest possible exposure levels.' They hoped to frame a risk finding that would recognize a risk at a PEL of 10 ppm, but would not say that a risk existed at the new PEL. The participants, however, were unable to develop language that satisfied both the tort and statutory criteria.

F. INCENTIVES TO PARTICIPATE--IN GENERAL

  Both labor and industry believed that uncertainty could be reduced by negotiations. The traditional adversary process usually results in opposing parties taking extreme positions, and consequently building a record that supports agency *1655 action anywhere in between. One participant described the result as a 'crap shoot.' In contrast, if the parties negotiated an NPRM, or at least narrowed the issues through negotiations, they could make the ultimate standard more predictable. All of the parties perceived themselves as risk averse--industry more than labor. Thus reducing uncertainty was a benefit.

  Virtually all participants lacked confidence in the OSHA; there was considerable concern that the OSHA would 'screw up' a new standard through the regular process, either by promulgating a standard that could not be defended in the courts or by imposing collateral requirements with high costs for relatively little health benefit.

  The participants recognized that an individual employer, even a member of one of the participating trade associations, or a union entity or public interest group, could challenge a negotiated standard. Most of the participants were nevertheless reasonably confident that any such challenger would be hard pressed to convince a court to invalidate a standard that unions and trade associations most directly affected had not only agreed upon but also actively defended. Others, however, thought that a challenge to a negotiated standard might be strengthened by the negotiation process, because of their belief that a court would find patent violations of the FACA, or a violation of the APA prohibitions on ex parte contact, or prohibitions on delegation of governmental authority to private citizens.

G. INCENTIVES FOR THE OSHA TO PARTICIPATE

  The OSHA did not participate in the benzene negotiations directly. It did not take part in the discussions among industry, labor, and public interest representatives. It did, however, pay the convenor/mediators [FN180] and indicated its willingness to use a negotiated standard as the basis for its rulemaking.

  Enthusiasm for the process was higher at policy levels within the agency than at 'working levels.' Many health standards personnel of the OSHA and their legal staffs had major doubts whether benzene was a good candidate for negotiated rulemaking, and further doubts about the way the negotiations were handled, believing that the convenor/mediators consistently were too optimistic about the prospects for agreement. The commitment of the OSHA to the process, low from the beginning, was retarded further by its lack of participation in the negotiations themselves.

  Nevertheless, sharp adversarial conflict during rulemaking and in court challenges make the agency's job more difficult, and in the words of one official of the OSHA, negotiated rulemaking was 'worth a try.'

H. INCENTIVES FOR LABOR TO PARTICIPATE

  Organized labor had two reasons to believe that negotiations would be preferable to the traditional rulemaking process. First, a negotiated standard accompanied by an agreement not to litigate the legality of the standard could save organized labor significant litigation expenses. Second, the perception that the *1656 OSHA was in unfriendly hands increased labor's fear that an agency promulgated standard would be delayed and might not satisfy labor's desires as well as a negotiated standard.

  Labor's incentive to negotiate was reinforced by industry's power to delay. Labor did not believe a 1 ppm standard was vulnerable to judicial attack. It thought a court would likely find the OSHA entitled to opt for greater health protection, working from uncertain scientific evidence, as long as it supported a lower standard with a finding of significant health risks at benzene levels higher than the standard. On the other hand, industry had already delayed promulgation of a lower standard for eight years, and could delay it further by arguments presented to the OMB [FN181] and by litigating aggressively.

  Some labor participants perceived the OMB as a bigger threat than the courts to a standard from the OSHA. To them, industry support for a standard before the OMB was more important than an agreement not to challenge the standard in court. Labor's motivation to negotiate was strengthened by an informal assurance by the OMB that any proposal by the OSHA based on negotiated agreement would be cleared by the OMB within twenty-four hours.

  Labor's incentives to participate were mitigated by a perception that the OSHA had committed itself to issuing a standard within a short time and a belief that this standard would adopt a PEL of 1 ppm. In addition, some labor participants were more comfortable with the traditional hybrid rulemaking process than with negotiation, believing that adversary administrative litigation, especially cross-examination of industry witnesses, is the best way to develop scientific data and a factual record to support an adequately protective standard. They were pessimistic, however, that a standard acceptable to them could be promulgated within the eight months promised by the OSHA, given what labor perceived as the unsympathetic attitude of the OSHA and the OMB.

  Other labor participants, and some nonparticipants allied in interest with the labor participants, were suspicious of the negotiation process. As noted above in the discussion of benzene as a candidate for negotiated rulemaking, some participants feared that labor's resource limitations would be magnified in negotiations, and that negotiations might delay issuance of a standard. Nevertheless, because they thought the OSHA wanted negotiations to occur, they reluctantly agreed to participate.

I. INCENTIVES FOR INDUSTRY TO PARTICIPATE

  Before identifying positive incentives for industry to participate, it is useful to note a change in industry position since the 1977 standard was issued. During the three years between adoption of the emergency standard and invalidation of the final 1 ppm standard by the Supreme Court, a number of firms had reduced benzene levels in their workplaces. Thus the marginal cost for the industry to reach the 1 ppm standard in 1983-1984 was less, in real dollars, that it would have been in 1977-1978.

  *1657 Industry's incentives to participate in negotiations depended on its perception of probable action by the OSHA in the absence of a negotiated standard. After the Supreme Court litigation, it appeared virtually certain that the OSHA would find a way to justify a 1 ppm standard, and might accompany this standard with findings, not only that exposures at the 10 ppm level posed health hazards, but also that a 1 ppm standard presented health hazards. In addition, a variety of ancillary issues would be addressed in a standard, such as health monitoring, averaging, and the action level, that might be influenced in a negotiation. There was a perception, at least among some in the industry, that some alternative needed to be found to adversarial rulemaking, and that regulatory negotiation was worth a try. Some other industry participants were pessimistic that anything useful could come from the negotiation project but elected to participate for defensive reasons. They were afraid a process excluding them might produce an undesirable outcome.

J. DIFFERENCES AMONG THE INDUSTRY PARTICPANTS AND WITHIN INDUSTRY

CONSTITUENCIES

  The possibility of a 1 ppm PEL presented different concerns to the different industries involved. The petroleum, chemical, and rubber industries were able to tolerate a 1 ppm standard better than was the steel industry. On the other hand, the petroleum and chemical industries were more concerned about tort liability than was the steel industry. This concern with tort liability made petroleum and chemical participants more willing to accept a 1 ppm standard in exchange for the absence of an agency finding that benzene exposure created significant health risks below levels of 10 ppm.

  The rubber industry had much narrower concerns than the other industries, limited to that part of the standard than would address dermal contact. The rubber industry expressed little concern about the PEL during the negotiations. Representatives of the different industries did not undercut each other's positions in formal negotiated sessions, but neither did they offer aggressive support for positions other than their own.

  The American Petroleum Institute (API) played the leading role on the industry side. API participation was viewed as essential by labor. API also had more internal difficulties than the other industry groups participating. These facts combined to make API difficult in plenary sessions and in industry caucuses. Some of the other industry participants resented what they perceived as API dominance, and occasionally maneuvered to have the chemical industry participants, rather than API, speak for industry in sessions of the full negotiating group.

  Most of the industry participants in the benzene negotiations worked in the health and safety functions of regulated enterprises or trade associations. Industrial relations officers in the same organizations were anxious about direct dealings between labor and industry in the benzene negotiations. One concern related to the possible content of a negotiated standard--the possibility that it might prejudice positions in future labor-industry negotiations. Another concern related the possible precedent setting effect of industry-level negotiations; both the petroleum and chemical industries had resisted industrywide bargaining *1658 with trade unions, and opposed conduct that might make such industrywide bargaining more likely in the future.

  One of the most serious difficulties experienced during the negotiation involved internal dissension within API. Much of the difficulty was structural; hundreds of petroleum companies would be adversely affected by a regulation requiring reduced benzene exposures, many more than in any of the other industries. Moreover, the large size of the industry leaders meant that large numbers of people within each company felt entitled to be included in discussion of negotiating positions taken by industry representations.

  The petroleum industry participated through a three-tiered committee system. At the highest level was a committee of thirty to forty company vice presidents with responsibility for environmental and health matters. At least one member of this group was intransigent in opposing industry participation in the benzene negotiations. Chief executive officers of petroleum companies were briefed from time to time on the negotiations and were supportive, but they never effectively restrained their vice president who had more direct control. As a result, the petroleum industry representatives were regularly subjected to criticism by others from within their own industry. Fear of adverse public reaction prevented a decision by the petroleum industry to withdraw from the negotiations. Concerns within the industry, however, made meaningful compromises at the bargaining table difficult.

  Despite these internal difficulties, however, the petroleum industry finally supported a compromise benzene standard with considerable unity. This compromise was the basis of the near-agreement in the negotiations.

K. INTRALABOR MOVEMENT DIFFERENCES

  Labor participants perceived intraconstituency differences as less of a problem for labor than for industry. Union health and safety representatives regularly worked together on matters involving the OSHA, and thus were accustomed to the process of compromise in formulating strategic administrative and judicial litigation stategies. Moreover, labor participants thought they had greater authority to speak for their constituents than industry representatives.

  In part, labor unity was due to general agreement that the 1978 standard was the minimum that would be acceptable. It was further enhanced by respect for the lead role the AFL-CIO traditionally had taken on health and safety issues, and by personal respect for the AFL-CIO spokesperson.

  Opposing these unifying forces were some differences among the participating unions and some strong feelings held by unions that did not participate. The United Rubber Workers (URW) had greater trust in its industry counterparts than did the other unions. It was more willing to accept various compromises proposed during the negotiations. The Oil, Chemical & Atomic Workers (OCAW) historically had invested heavily in health and safety issues, although in the opinion of internal union critics had little to show for the investment. Faced with declining membership and failed merger explorations, OCAW was perceived as having an acute need for some tangible evidence of success on the health and safety front. A negotiated benzene rule would be such evidence, on an issue of particular concern to rank-and-file refinery workers. There also were *1659 some differences between those accustomed to the give and take of collective bargaining and those more accustomed to the formal adversary process of traditional rulemaking.

  Labor's limited resources discouraged additional conferences to iron out differences and arrive at a common position outside the full negotiations with industry representatives. As a result, labor differences tended to surface at the table for everyone to see.

  Labor constituency problems arose, not so much within represented unions as within the labor movement as a whole. Most approached the OSHA standard setting from a global perspective rather than an industry-specific perspective. Accordingly, the precedent setting effects of particular elements of a benzene standard were a concern. After agreement on a standard seemed a real possibility, the United Auto Workers (UAW) and Amalgamated Clothing and Textile Workers Union (ACTWU) were quite critical of the process, and this criticism within the labor movement apparently discouraged participating union representatives from taking major risks to reach final agreement.

L. NEGOTIATION PROTOCOLS

  It is important to recall that, when the benzene negotiations began, no one expected an agreement on a total package but, largely due to the enthusiasm of Assistant Secretary Auchter, the effort was transformed into an attempt actually to negotiate a standard. Some resentment that the process had been transformed lingered and was amplified when Auchter resigned while negotiation was under way.

  The parties agreed that none would promote, as products of the negotiation process, anything that was not agreed to by all participants. This ground rule did not, however, preclude consideration of an agreement that would leave certain issues to be decided by the OSHA. In this regard early agreement of the URW and the rubber industry trade association [FN182] was something of an irritant to industry participants.

  The parties also agreed, at least tacitly, that none of the participants would challenge any aspects of a standard that had been agreed to in the negotiation. In this way, both labor and industry could buy themselves a measure of certainty. On the other hand, the union participants understood that the trade associations could not guarantee that none of their corporate members would litigate the legality of an ultimate standard, even if the participants reached agreement. Nevertheless, it was perceived that an agreement would be a powerful psychological motivation for a court to sustain a standard from the OSHA, which is entitled to considerable deference under both the APA and the OSHA Act.

M. THE NEGOTIATION PROCESS

  Perhaps the most salient feature of the benzene negotiation process is that the agency with the statutory responsibility for establishing a standard did not actually *1660 participate in the negotiations. Instead, the OSHA said that it would use any negotiated standard as a basis for its rule, encouraged the parties to work out their differences in good faith, attempted to increase incentives for affected interests to participate by threatening to proceed with its own standard regardless of the pace of negotiations, and kept itself informed, through the mediators, on negotiating progress.

  This kind of agency nonparticipation had been recognized as one of the two basic forms of negotiated rulemaking. [FN183] One advantage when the agency does not participate is that deliberations may be more candid in the absence of the ultimate decisionmaker.

  In addition to this theoretical justification for the low profile of the OSHA, however, there is significant evidence that the OSHA limited its participation because it did not want the FACA [FN184] to apply to the benzene negotiations, and it perceived that if the agency did not participate, the FACA would not apply. [FN185] The OSHA subsequently was encouraged to limit its participation by the perception that substantial agreement was reached in the cotton dust rulemaking without agency involvement. Indeed the cotton dust agreement [FN186] was used by some personnel of the OSHA and their lawyers as a post hoc model for the structure of the benzene negotiations. [FN187]

  Many of the private participants did not want the OSHA to take part in the negotiations, fearing that the OSHA would adopt compromises in bits and pieces without understanding the tradeoffs involved.

  Among the private participants, there was some initial discomfort with the give and take of negotiations. Industry representatives were more comfortable than labor representatives with the idea that elements of a total package could be traded off against each other based on cost and health benefits. Over the course of discussions, the participants gradually accepted the idea that they should attempt to negotiate a 'total package,' largely because industry constituents were unwilling to proceed unless they could be assured that certain aspects of a standard undesirable to them could be made up by labor concessions on other issues.

  Everyone recognized the political problems within the petroleum industry. At one point, the API participants were forbidden to discuss a standard of 1 ppm, a restriction several participants from both industry and labor characterized as 'silly.' During the period this restriction was in effect, others continued to negotiate, *1661 with the API representatives as observers. One objective of proceeding in this manner was to present the API constituents with the possibility that everyone else would negotiate a standard without their participation unless unrealistic constraints on their representative were relaxed. [FN188]

  The problem of finding health risk initially was handled in a discussion of general principles to be included in a preamble to a negotiated standard. The facilitators urged dealing with the preamble only in terms of general principles until the remainder of the standard was resolved. Participant attorneys urged, however, that preamble language be addressed in detail. One industry attorney produced a draft of actual preamble language that, rather than reflecting a compromise position based on the discussion, 'tilted' substantially toward industry's position. Labor lawyers who were not participating in the negotiations suggested only modest changes. When the draft preamble was presented to the labor negotiators, however, they took it as an act of bad faith in the negotiations.

  Virtually all the participants, however, thought that the quality of most of the discussions in the negotiation was principled rather than being merely power based, and that the interest representatives genuinely sought to understand one another and to narrow their differences as much as possible. Indeed the negotiators came much closer to agreeing than most participants thought was possible when they began discussions.

N. REASONS NO RULE WAS AGREED UPON

  Several conditions contributed to the failure of the benzene negotiations, despite the agreement in principle.

  Two aspects of the OSHA's role were harmful: nonparticipation by the OSHA and distribution by the OSHA of a draft standard midway through the negotiations.

  Nonparticipation by the OSHA reinforced the consensus ground rule that the contents of the negotiation would not be used as the basis for a rule unless the parties formally could agree on a total package. If the OSHA had been present during the negotiating sessions, it would have gained its own contemporaneous impressions of what the parties could accept in a final standard. If present, even if the parties could not agree formally on a total package, or indeed on anything, the OSHA nevertheless would have the benefit of deliberations in promulgating its own standard. [FN189]

  Nonparticipation by the OSHA forced the parties to communicate the 'results' of the negotiations to the OSHA, which in turn made adoption of a rigid consensus definition more likely. To the extent that results were to be communicated in a formal document, the likelihood of failure because of unwillingness or political inability to commit in writing to a compromise position wasincreased. Communications about the status of negotiations occurred via the mediators, but divergent information also was communicated indirectly by parties to agency personnel. Because the agency could not reach its own conclusions about negotiating *1662 progress, mediator credibility with the agency and the parties was reduced. In addition, nonparticipation by the OSHA gave it less of a stake in successful negotiations and therefore less motivation to use its ultimate power to create incentives for parties to negotiate meaningfully.

  Even without active participation in the negotiations, the OSHA could have been more supportive of the negotiations. Staff level personnel of the OSHA gave conflicting signals to participants as to how committed the agency was to the negotiation process. As a result, some participants were never sure that the OSHA really wanted the negotiations to succeed.

  Distribution by the OSHA of a draft standard midway through the negotiations hardened party positions and reduced the likelihood of reaching agreement. Labor thought the draft of the OSHA standard looked better than what the negotiators were about to agree upon. [FN190]

  The convergence of election campaigns for the presidency of the United States and of the United Steelworkers of America (USW) impeded negotiations. The first made organized labor reluctant to appear to be cooperating with the Reagan Administration; the second diverted the attention of the USW participants and probably made the USW participants more reluctant to be identified with a negotiated standard that might be unpopular within the union or with its new administration.

  The departure of Assistant Secretary Auchter from the OSHA before the negotiations were complete also was a problem. Auchter was enthusiastic about the negotiation process, and he also was perceived as having sufficient influence to ensure that a standard would be issued by the OSHA unilaterally if the negotiations did not succeed. After he resigned, the prospect of speedy clearance by the OMB and issuance of a unilateral standard by the OSHA diminished. This changed party BATNAs, [FN191] lessening the incentive to agree in the negotiations.

  Superimposed on these difficulties were others: the controversial nature of the averaging concept for union constituents; and a residual belief among some industry constituents that the negotiation process was a mistake that should be avoided in favor of aggressive litigation--which had, after all, already delayed a tighter standard for seven years.

  At least some of the labor participants had a continuing conviction that hybrid rulemaking would have been a superior process to negotiation in developing a standard. This conviction was a reinforced by some difficulty in getting specific data from industry representatives to support the need for averaging or from the OSHA as the negotiations proceeded, though some data was provided by both.

  In addition, some participants thought the negotiations dragged on too long and the meetings were too infrequent. This broke momentum and added to constituency pressures as more and more constituents found out about the negotiations or became convinced that negotiations might actually produce a standard. *1663 Increased constituent awareness put more pressure on participants to stiffen their positions. Infrequent meetings increased the likelihood that memories about positions or concessions would become fuzzy, leading to perceptions that participants were changing their positions between sessions. On the other hand, time was required to work on constituency problems and the negotiations probably could not have moved faster.

  Serious intraconstituency problems on both the industry and labor sides-- particularly within the petroleum industry--caused participants to seem to embrace a solution to a problem at one session, only to back away at the next session. This presented opposing participants with a moving target.

  There was some sentiment that the negotiators moved too abruptly from considerations of general principles in outline form to the drafting of detailed language, particularly preamble language, which dramatized the difficulty of dealing with the issue of finding a health risk. One labor participant characterized the preamble draft as an 'affront,' revealing for the first time that industry was using the negotiations 'to take us for a ride.'

  Two substantive issues blocked agreement, and might have been fatal to consensus on a standard even in a more favorable political environment. The first was the difficulty in reconciling industry desire for a weaker finding of health risk against the need for an unequivocal finding to sustain a standard of less than 10 ppm in litigation. The most obvious reason for the failure to reach agreement is that the parties simply ran out of ideas for ways to resolve their differences. Labor was wedded to the idea of a 1 ppm standard. The only way industry could accept a 1 ppm standard was to have a rule that allowed some flexibility when the standard was exceeded. The only idea for reconciling these positions that occurred to anyone was some type of averaging, but averaging proved unacceptable in principle to the labor movement as a whole.

  Serious consideration by labor participants of solutions to the exceedence problem was retarded to some degree by the precedent set by the coke oven negotiations. [FN192] The coke oven controversy was addressed by adopting a relatively simple and strict standard and dealing with implementation problems through separate compliance directives. Some labor participants wanted to take the same approach in the benzene negotiations--an idea the industry participants strongly opposed.

  The second substantive reason for failure was fear by labor representatives that the averaging concept, if adopted by the benzene negotiators, would become a precedent for averaging in other standards. This concern was made more acute when one industry attorney, who had participated in some of the industry negotiations, made a presentation to the OSHA in the presence of officials from other unions that the averaging concepts discussed in the benzene negotiations should be used in another health standard.

  Moreover, labor feared that enforcement of certain averaging concepts would be more difficult. Enforcement of a forty hour TWA standard would require more than one visit by a compliance officer, and labor was concerned that inspections already were difficult to arrange. Some petroleum industry companies *1664 were initially opposed to averaging but became convinced that averaging was necessary because of the variability of samples taken at a particular point in time.

  In the end, labor, after a period of consultation with nonparticipant unions, took a broad look at what was emerging from the negotiations and concluded that it would prefer what the OSHA would do unilaterally. The expectation that the OSHA was committed to moving quickly, without waiting for the negotiations, combined with the perception that a unilateral standard from the OSHA would suit labor better than what could be negotiated removed any incentive for labor to continue.

  At the same time, some industry constituents were becoming convinced that the OMB would block or delay a standard from the OSHA unacceptable to industry, which made it difficult to achieve unity behind a position.

  In the jargon of negotiating theory, the BATNAs of both industry and labor shifted to make a negotiated solution less attractive.

O. FEDERAL ADVISORY COMMITTEE ACT PROBLEMS

  The Department of Labor's regulations relating to the administration of advisory committees restrict flexibility. [FN193] The OSHA Act and the regulations require preparation of agendas, [FN194] advance notice of meetings, [FN195] meetings open to the public, [FN196] and verbatim transcripts and minutes. [FN197] The regulations provide less flexibility to accommodate regulatory negotiation than the FACA itself, by failing to authorize closure of a meeting to the public and by requiring a transcript in all cases.

  Earlier sections of this part of the article explained that the limited role of the OSHA in the benzene negotiations resulted in part from agency and participant desire not to conduct the negotiations under the FACA. To the extent that the limited role of the OSHA was responsible in part for the failure to reach a consensus on a standard, the FACA was indirectly responsible.

  Personnel of the OSHA and their lawyers perceived two impediments to successful negotiation flowing from the FACA. First, the statute and implementing regulations require a charter, approval from the General Services Administration (GSA), advance notice of meetings, and minutes; these requirements slow the negotiation process, and speed was thought to be necessary. [FN198] Second, and more fundamentally, the FACA and implementing regulations were thought to require that meetings, not only of the full negotiating committee, but also of subgroups and caucuses, be open to the public. This was thought to be inconsistent with good faith negotiation.

  Open meetings would hamper good faith negotiations, under this view, because risks to individual participants of making concessions would be increased. In fact, on one occasion in the benzene negotiations, a trade publication reported that an industry representative had said that 'we can live with' a PEL of 1 ppm. *1665 As a result of the story, that representative's constituents became alarmed and caused the representative to feel that his job might be threatened. This is an example of constituency problems that some participants thought would be created by open meetings. If representatives feared adverse constituency reaction to concessions, they would make no concessions and there would be no movement from initial postures, making fruitful negotiations impossible.

  One participant viewed the open meeting problem, not in terms of constituency relations, but in terms of estoppel or prejudice to future positions if negotiations were unsuccessful. Closed meetings were necessary, under this view, to prevent participants from being embarrassed by the OSHA's knowing what they would be willing to consider to reach agreement. Participation by the OSHA in the negotiations, desirable for other reasons, would have exacerbated this concern. Other participants found the concern with open meetings completely unpersuasive. These participants believed that changes in position would be communicated to the OSHA by other participants, and that the principle of open meetings by advisory committees is valuable.

  Some participants suggested that an adequate framework for negotiating occupational safety and health standards exists under the advisory committee provisions of the OSHA Act. Section 7 of the Act [FN199] authorizes the Secretary to establish two types of advisory committees: National Advisory Committee on Occupational Safety and Health, and advisory committees to assist the Secretary in setting standards. The OSHA, by regulation, had divided the class of standards advisory committees further into 'continuing committees' and ad hoc committees. [FN200] Continuing committees assist with the development of standards in areas where there is frequent rulemaking. Ad hoc committees assist with particular rulemaking proceedings.

  The basic authority to establish occupational safety and health standards contained in section 6 of the Act [FN201]contemplates a role for section 7 advisory committees at the option of the Secretary. If the Secretary requests recommendations from an advisory committee on a contemplated standard, the advisory committee is required to submit its recommendations to the Secretary within ninety days of its appointment. The Secretary may extend or reduce the period for developing recommendations, but it may not exceed 270 days.  [FN202]

  When an advisory committee is appointed and the Secretary determines that a standard should be issued, the rule must be published for comment within sixty days after receipt of the committee's recommendations, or the expiration of the deadline for the committee's recommendations. [FN203] Thereafter, the statutory standard setting process proceeds in the same manner regardless of whether an advisory committee has been involved, with an opportunity for the public to submit written comments and hearings if they are requested in the comments. [FN204]

  The makeup of section 7 advisory committees was addressed by the Seventh *1666 Circuit in National Roofing Contractors Association v. Brennan.  [FN205] In that case, a trade association of roofing contractors challenged their exclusion from an advisory committee established to advise the Secretary of Labor on safety standards for sloping roofs. The court found that the interests of roofing contractors were represented adequately by general contractors on the committee. [FN206]

  Some personnel of the OSHA and their lawyers thought that coke oven health standards had been negotiated effectively within the section 7 advisory committee framework under the active leadership of Assistant Secretary Eula Bingham. This, in their view, is evidence that successful negotiations can occur under procedures mandated by the FACA and by section 7 of the OSHA Act.

  Other participants thought the section 7 process too rigid to accommodate negotiated rulemaking.

P. UNILATERAL OSHA-PROSPOSED RULE

  On December 10, 1985, the OSHA published its own notice of proposed rulemaking (NPRM) to limit benzene exposure in the workplace. [FN207] The preamble to the NPRM noted that the negotiations had not produced a 'joint document,' [FN208] but otherwise made no substantive reference to the negotiation attempt. As expected, the OSHA proposal included a PEL of 1 ppm, determined over an eight hour TWA. [FN209] In addition, the NPRM addressed several of the issues raised in the negotiations. Variability in the petroleum industry may, in the final rule, be handled by providing that a single exceedence of the PEL would not necessarily result in a citation, if the employer can show that exposures are below the PEL a large majority of the time by showing the results of at least five eight hour TWA measurements in the same area within a reasonable period of time. [FN210] The NPRM also granted an exemption to the rubber industry for solvents containing less than 0.5% of benzene, as agreed by the rubber industry and union representatives. [FN211] Significantly, the OSHA found a significant risk of occupational disease to exist at exposure levels of 10 ppm but not at 1 ppm under the proposed standard. [FN212]

  On December 10, 1984, shortly after the negotiations had adjourned, the Steelworkers had filed in the United States Court of Appeals for the District of Columbia Circuit a petition for a writ of mandamus, seeking to require OSHA to issue an NPRM for a benzene standard within 30 days of the court's decision, and to issue a permanent standard within seven months after the NPRM.  [FN213] After oral argument, the court ordered the OSHA to submit a timetable for post-NPRM activities. [FN214] The timetable submitted in response called for OSHA to complete its staff analysis by November 10, 1986.  [FN215] The court denied the petition *1667 for faster action on the proposed rule and declined to retain jurisdiction on February 25, 1986.  [FN216]

  The OSHA told the author by telephone on October 30, 1986, that public hearings on the benzene standard were planned for February, 1987.

IV. THE FAA FLIGHT AND DUTY TIME NEGOTIATIONS

A. INTRODUCTION

  The Federal Aviation Administration (FAA) utilized negotiated rulemaking to develop revisions to its flight and duty time regulations after two failures to revise the regulations through traditional rulemaking. [FN217] Negotiations began in the summer of 1983 and resulted in agreement on a proposed rule issued by the FAA in March, 1984. [FN218] A final rule based on the negotiated agreement was promulgated in July, 1985. [FN219]

B. HISTORY OF FLIGHT AND DUTY TIME REGULATIONS

  Section 601(a)(5) of the Federal Aviation Act [FN220] requires that the Administrator of the FAA prescribe reasonable rules and regulations governing in the interest of safety the maximum hours or periods of service for airmen and other employees of air carriers. The FAA promulgated restrictions on flight and duty time in the mid-1950's that remained essentially unchanged, despite substantial changes in aircraft technology and industry structure. [FN221]

  The existing regulations applicable to major scheduled airlines limited annual flight time to 1,000 hours, monthly flight time to 100 hours, and limited flight time in consecutive seven day periods to thirty hours. In addition, the regulation prescribed minimum rest periods. [FN222] Special rules applied to flag carriers, [FN223] and supplemental air carriers and commercial operators. [FN224]

  These unchanged regulations, poorly suited for the reality of modern commercial air transportation, generated more requests for interpretations than any other provision of the Federal Aviation Regulations. The agency had issued more than 1,000 pages of interpretations by 1983.

  On several occasions in the ten years preceding the regulatory negotiation the FAA made proposals for changing the flight and duty time regulations,  [FN225] but *1668 'because of the complexity of the flight time rules and the economic interests affected, none of the past proposals succeeded in resolving the problems to the satisfaction of the affected parties.' [FN226] The 1980 proposal was opposed in part by 'virtually all affected segments of the air transportation community' [FN227] and was withdrawn after it was opposed by the Airline Pilots Association, Alaskan operators, rotorcraft operators, and others. [FN228]

  After these failures to revise the flight and duty time regulations through the traditional rulemaking processes the FAA turned to negotiated rulemaking in early 1983.

C. PRENEGOTIATION ACTIVITIES

  On May 12, 1983, the agency published a notice in the Federal Register   [FN229] of its intent to hold a regulatory negotiation through an advisory committee. It solicited comments concerning the issues that it should consider, the interests affected, the membership of the committee, the procedures that should be followed, and other pertinent matters. [FN230] The notice reported that the FAA already had appointed Nicholas Fidandis, Director, Mediation Services, of the Federal Mediation and Conciliation Service, to act as convenor/mediator. The notice indicated that there had been preliminary inquiry among representatives of affected interests to explore the feasibility of negotiated rulemaking. It concluded that regulatory negotiation could be successful in this instance.

  The notice expressed concern, however, that someone might use the regulatory negotiation process 'simply to delay the development of an NPRM.'  [FN231] It declared that the FAA would proceed to develop an NPRM on its own if the negotiation process failed to produce a consensus, and that the committee would be dissolved by mid-August, 1983, if it was unable to reach agreement by that time. [FN232] The notice identified eighteen specific issues to be explored in the negotiations and listed the interests and participants.

  If a person or interest requested inclusion in the negotiations in response to the notice, the FAA said it would determine (1) 'whether that interest would be substantially affected by the rule, (2) if so, whether it would be adequately represented by an individual already in the negotiating group, and (3) whether in any event the requester should be added to the group or whether interests can be consolidated and still provide adequate representation.'  [FN233]

  The notice declared the intention of the FAA to 'issue the negotiated proposal in a notice of proposed rulemaking unless it is inconsistent with the statutory authority of the agency or other statutory requirements, or it is not appropriately justified.' [FN234] The notice said, 'For the process to be successful, the interests represented should be willing to accept the final product of the advisory *1669 committee.' [FN235]

D. ESTABLISHMENT OF NEGOTIATING COMMITTEE

  On June 28, 1983, the FAA published a final notice [FN236] establishing an advisory committee for the negotiations. It rejected some requests for membership on the negotiating committee but added representatives of six groups.

  It noted that nonmembers of the committee would be given an opportunity to present information to the committee and that all interested individuals or organizations would be given full opportunity to comment on the NPRM.

E. SUITABILITY OF SUBJECT MATTER

  The flight and duty time rulemaking was well suited for negotiated rulemaking. The number of interests were manageable; ultimately eighteen participants were involved. [FN237] The issues were mature, having been subject to discussion and unsuccessful notice and comment rulemaking over a period of more than twenty years. [FN238] Furthermore, fundamental values were not perceived as involved. Flight and duty time restrictions have obvious safety implications, but participants agreed that tradeoffs were appropriate among requirements aimed at increasing safety. [FN239] Also, the correct form of an ultimate rule was not determinable as a matter of objective scientific evidence. [FN240] The principal disputes related to accommodating the details of a rule to operations requirements in a variety of carrier environments.

  In addition to satisfying the Harter criteria, other factors indicated the possibility of successful negotiation. Chief among these was the lack of success of the FAA with two NPRMs on the subject within the preceding seven years. The first was withdrawn because of opposition from industry. The second was withdrawn because of opposition from the Airline Pilots Association.  [FN241] Recent experience therefore reinforced party perceptions that the FAA might promulgate something unacceptable to its interests, a perception that is a prerequisite to negotiated agreement. [FN242]

F. INCENTIVES TO PARTICIPATE

  The labor interests had an incentive to participate because they wanted an enforceable rule, applicable to both major (part 121) and minor (part 135) carriers. [FN243] They believed negotiation was the only foreseeable way to get this rule. *1670 Traditional rulemaking had not produced any changes in more than twenty years.

  Part 135 carriers were not unhappy with the status quo, but they were convinced that some new rule covering their operations was inevitable, and they wanted to influence the content of the rule.

  Part 121 carriers, represented by the Air Transport Association (ATA), initially were reluctant to participate, presumably because they were satisfied with the status quo. Ultimately, however, the FAA convinced them and others that it would issue a rule and would proceed with negotiations in the absence of the ATA, possibly inducing individual part 121 carriers to participate or proceeding without part 121 carrier representation in the negotiations at all.

  The FAA was inclined to participate because of frustration with the traditional process, and a perception that a unilaterally promulgated rule would be easier to defend if the parties tried and failed to develop their own rule. Moreover, the FAA thought that it could blunt criticism of the agency's efforts if the critics were themselves unable to develop a regulation.

G. INTRACONSTITUENCY DIFFERENCES

  Few intraparty problems were manifest in the negotiations. One would expect that the trunk carriers would have disagreed among themselves, but ATA, once it decided to participate seriously in the negotiations, convened a meeting of sixteen carriers and hammered out a proposal all the carriers could accept.

  The FAA, however, experienced some internal problems. Program staff were inclined to prefer a rule requiring more rest than that preferred by the legal staff. Midway through the negotiations a new FAA Administrator was appointed who had been involved as a member of the National Transportation Safety Board (NTSB) in developing flight and duty time recommendations. After the initial negotiations were concluded and the NPRM published, the internal FAA critics of the proposed rule were able to induce the new Administrator to make changes in the proposed rule.

H. THE NEGOTIATIONS

  The advisory committee held sixteen days of formal meetings between June 29 and September 26, 1983. More than thirty informal meetings of subgroups also occurred. The results of the negotiation are aptly summarized by the FAA in the resulting NPRM:

    The committee . . . thoroughly discussed the major issues involved in the regulation of flight time limits. Numerous proposals and justifications were drafted by participants and submitted to the committee for review. . . . Although the committee did not reach consensus on any particular proposal, its deliberations were successful because committee members gave serious consideration to and entered into candid discussion of thevarious proposals and justifications submitted to them. Thus the committee succeeded in narrowing the differences among parties and in reaching substantial agreement on some issues. In addition, the committee identified major areas of concern and all parties obtained significant, new information on a subject which has been discussed, *1671 without resolution, for years. [FN244]

  After the September 26, 1983, committee meeting, the FAA took responsibility for the production of an NPRM, which the committee reviewed on February 14, 1984. The NPRM memorialized consensus resolution of some issues and offered the FAA's independent judgment on other issues about which consensus had not been reached, informed by the FAA's presence in the negotiating sessions. In the February 14 meeting, a majority of the committee members recommended publishing the NPRM in the Federal Register as submitted. Several members of the committee dissented from this recommendation because they wanted to address certain issues that had not been resolved to their satisfaction before the NPRM was published. [FN245]

  On March 28, 1984, the FAA published the NPRM as submitted to the committee, with certain modifications in the preamble to address issues not resolved by the negotiators. [FN246] Interested parties had forty-five days to comment on the NPRM. [FN247] The FAA synthesized the comments and presented them to the committee on September 11, 1984. [FN248] The agency incorporated input received at this meeting, made additional changes, and submitted a draft final rule to the OMB.

I. FINAL RULE

  The FAA promulgated the final flight and duty time rule on July 18, 1985.   [FN249] The rule included certain changes from the NPRM, and was accompanied by slightly more than 350 column inches of preamble.

  The FAA discussed the use of negotiated rulemaking in a background section at the beginning of the preamble, terming it 'essential in achieving the highly successful result that is apparent in these amendments.' [FN250] The remainder of the preamble, however, primarily addressed the substance of the rule, explaining the rationale of the FAA and evaluating the merits of the 140 comments received during the forty-five day comment period. [FN251] Written comments submitted by interests represented on the negotiating committee were, according to the FAA, largely restatements of their negotiating positions.  [FN252] In some instances, the FAA justified aspects of the rule by referring to the activities of the negotiating committee, but only fifty column inches--less than fifteen percent of the total preamble--were devoted to discussion of the negotiation process in support of the rule.

  The agency did meet objections to the short comment period, however, by pointing to the fact that the meetings of the negotiating committee were open to the public and to the fact that the negotiators heard oral presentations by nonmembers of the committee at its September 11, 1984, meeting. [FN253]

  *1672 The agency made the following major changes in the NPRM:

    (1) Based on comments from operators, it increased the weekly flight time limitation of 32 hours imposed on part 135 operators to 34 hours. The FAA noted that 34 hours was 'within the range of weekly flight hour limitations discussed' during the negotiation process. [FN254]

    (2) It eliminated proposed section 121.471(h), relaxing limitations for part 121 operators flying propeller driven aircraft with seating capacities between 31-60. The proposed paragraph (h) was not the product of consensus in the negotiations, but was included at the request of the Regional Airlines Association, which had proposed it in negotiations. Comments on paragraph (h) were split 27-27, with regional carriers favoring the provisions and regional pilots and their organizations opposing it. The agency concluded that safety considerations militated in favor of deleting the provision and requiring regional operations involving small aircraft larger than 30 passengers to conform to part 121 limitations, even though this would subject some regional operators to different limitations for their part 121 and 135 operations.  [FN255]

    (3) The proposed nine hour limitation on scheduled flight time between rest periods was changed to eight hours. The nine hour proposal was not the result of consensus among the negotiators, but was a compromise figure arrived at by the FAA. [FN256] Based on comments submitted by pilots objecting to the nine hour limitation, and on the absence of information on the benefits of a nine hour cap invited by the FAA but not submitted by carriers, the agency reverted to an eight hour limitation in the final rule. [FN257]

    (4) The proposed floor of 7.5 hours on the duration of rest periods was increased to 8 hours, based on comments from pilots and consumer groups opposing the shortened rest period. [FN258]

    (5) Language was added requiring that 'compensatory rest' [FN259] begin within twenty-four hours after commencement of the reduced rest period.  [FN260] This change represented a clarification of an issue raised at the last negotiating committee meeting, and was, in the FAA's opinion, consistent with the intent reflected in the negotiated proposal. [FN261]

    (6) Part 135 operators conducting both scheduled and unscheduled operations are permitted under the final rule to conduct all operations under the rules for scheduled operations, after obtaining an appropriate operations specification amendment. [FN262] This represented an addition to the proposed rule in response to comments from part 135 operators objecting to the administrative burden of conducting their operations under different rules.

*1673 J. DYNAMICS OF NEGOTIATION AND REASONS FOR SUCCESS

  The negotiations moved slowly until the FAA submitted a draft rule to the participants. This reinforced the view that the FAA would move unilaterally. It also reminded the parties that there would be things in a unilaterally promulgated rule that they would not like--thus reminding them that their BATNAs were worse than what was being considered at the negotiating table.

  Participation by the Vice President's Office, the Office of the Secretary of Transportation, and the OMB at the initial session discouraged participants from thinking they could influence the contents of the rule outside the negotiation process. One attempt to communicate with the Administrator while the negotiations were underway was rebuffed. [FN263]

  The participants tacitly agreed that it would not be feasible to develop a  'total package' to which the participants formally could agree. Instead, their objectives were to narrow differences, explore alternative ways of achieving objectives at less disruption to operational exigencies, and educate the FAA on practical issues. The mediator had an acute sense that the negotiation process should stop before agreement began to erode. Accordingly, he forbore to force explicit agreement on difficult issues, took few votes, and adjourned the negotiations when things began to unravel. In addition, the FAA, the mediator, and participants were tolerant of the political need of participants to adhere to positions formally, even though signals were given that participants could live with something else.

  Agency participation in the negotiating sessions was crucial to the usefulness of this type of process. Because the agency was there, it could form its own impressions of what a party's real position was, despite adherence to formal positions. In addition, it was easy for the agency to proceed with a consensus standard because it had an evolving sense of the consensus. Without agency participation, a more formal step would have been necessary to communicate negotiating group views to the agency. Taking this formal step could have proven difficult or impossible because it would have necessitated more formal participant agreement. In addition, the presence of an outside contractor who served as drafter was of some assistance. The drafter, a former FAA employee, assisted informally in resolving internal FAA disagreements over the proposed rule after negotiations were adjourned.

K. THE FACA ISSUES

  The negotiation group was chartered as an advisory committee, held public meetings, published minutes, and otherwise complied with the FACA. [FN264] Sensitive matters were handled in caucus or through other informal means without serious objection. In the view of some participants, the public nature of the negotiations was an advantage. It permitted the agency to respond to later critics of the rule by noting they had a chance to present their views, not only by responding to the original Federal Register notice and through the section 553 notice and comment process, but also by coming to the meetings. In fact, the NTSB and a *1674 group of commuter pilots made formal presentations to the participants at one of the later meetings.

L. THE APA ISSUES

  The FAA did not believe that it had delegated its authority to the negotiators, despite its commitment to publish a consensus rule. Because it was a participant in the negotiations, it could forestall consensus, and thus relieve itself of the obligation to publish a rule that in its view was inappropriate.

  Ex parte contact after the negotiations had adjourned, to check out the acceptability of changes being made in response to comments and to the views of the new FAA Administrator, was considered to be well within the limitations of Home Box Office v. FCC [FN265] and Sierra Club v. Costle. [FN266] Factual information was not being sought or received; the only information sought was reaction to agency exercise of policy discretion. Moreover, any alternatives presented in these conversations were supportable by the record and could satisfy the arbitrary and capricious standard of review that courts use in reviewing agency action. Therefore the alternatives were not dependent on the contents of the ex parte communications.

V. ENVIRONMENTAL PROTECTION AGENCY NEGOTIATIONS

A. INTRODUCTION

  The Environmental Protection Agency (EPA) has shown particular enthusiasm for regulatory negotiation. It has used negotiated rulemaking successfully to develop two proposed rules, one dealing with nonconformance penalties (NCPs) for vehicle emissions and another dealing with emergency exemptions from pesticide regulations. [FN267]

B. THE NEGOTIATION PROJECT

  In February, 1983, the EPA announced a 'Regulatory Negotiation Project.' [FN268] The announced purpose of the project was to test (1) the utility of developing regulations by negotiation; (2) the types of regulations most appropriate for negotiated rulemaking; and (3) the procedures and circumstances that foster the most effective negotiations. [FN269] The EPA published a notice in the Federal Register to solicit suggestions for regulations that might be candidates for negotiated rulemaking, requesting reference to summary criteria derived from ACUS Recommendation 82-4. [FN270]

  *1675 The EPA considered and rejected negotiation of a number of rules suggested as candidates. [FN271] Candidates were rejected for timing reasons, because they involved too many different interests, because they involved complex, unsettled questions of science, or because they involved 'generic' issues too complicated to be broken down into negotiable components.  [FN272] One environmental group suggested NCPs as one of several candidates for regulatory negotiation. [FN273]

C. NONCONFORMANCE PENALTIES

  In April, 1984, the EPA announced its intention to use negotiated rulemaking to develop an NPRM on nonconformance penalties (NCPs) for heavy duty vehicles under section 206(g) of the Clean Air Act. [FN274] Negotiations resulted in agreement on an NPRM issued eleven months later. [FN275]

1. Background of Nonconformance Penalties

  The Clean Air Act authorizes the EPA to set emissions limits for motor vehicles. Section 206(g) of the Act requires the EPA to issue certificates of conformity to any class or category of heavy duty vehicles or engines exceeding an emission standard but within an upper limit associated with the standard if the manufacturer pays an NCP. The penalty approach was intended to mitigate the adverse effects of 'technology forcing standards,' by permitting technological laggards to sell their engines or vehicles by payment of a penalty, while avoiding competitive disadvantage to technological leaders. [FN276]

  The EPA proposed NCPs in two traditional rulemakings. In 1979, the agency proposed NCPs as part of its regulation for 1983 and Later Model Year Heavy- Duty Engines, [FN277] and as part of its Gaseous Emission Regulations for 1983 and Later Model Year Light-Duty Trucks. [FN278] The proposals avoided establishing NCPs for hydrocarbon (HC) and carbon monoxide (CO) because of the EPA's belief that manufacturers could meet emission standards. In 1983, as part of the Reagan Administration's regulatory relief initiatives, the agency relaxed the HC and CO emission standards, obviating, in the view of the EPA, the need for NCPs in the final regulations. [FN279] The EPA nevertheless believed NCPs might be necessary for oxides of nitrogen (NOx) and particulate standards scheduled to be effective in 1978, and for previously promulgated standards, if standards for other pollutants subsequently made compliance more difficult. [FN280]

  In Natural Resources Defense Council v. Ruckelshaus, [FN281] the district court set *1676 a timetable forcing the EPA to promulgate Phase I NCP rules by August 31, 1985, and Phase II rules by December 31, 1985. Phase I rules were to address when NCPs would be made available, how upper limits would be chosen, the general formula for calculating the penalties, and procedures for testing the degree of emissions nonconformity. [FN282] Phase II would apply the Phase I concepts to determine particular emissions standards for which NCPs would be available, specific upper limits, and numerical values for the variables in the penalty rate formula for particular subclasses of engines.  [FN283]

2. Prenegotiation Activities

  A Nuclear Resources Defense Council representative suggested NCPs as a candidate for negotiated rulemaking. The convenor retained by the EPA, ERM- McGlennon Associates, and the EPA conducted prenegotiation efforts aimed at determining whether NCP was a suitable candidate, and at identifying appropriate participants for the negotiation. The process began with the program office of the EPA. Once program office support was assured, the convenor called forty-seven representatives of manufacturing, consumer, environmental, state, agency, and industry association interests. [FN284] These representatives generally expressed support for the negotiation project. The convenor also met personally with a smaller number of representatives to assess issues, interests, and positions.

  On the basis of these contacts the EPA decided to proceed with negotiated rulemaking for the NCP rule. The EPA began the process by sponsoring an organizational meeting at which participants met each other and discussed procedural and logistical issues such as frequency and location of negotiations, agenda formulation, and minutes. Potential participants had shown a willingness to negotiate before the organizational meeting; the purpose of the meeting was to prevent surprises once substantive negotiations got underway.

  The convenor obtained final participation commitments from participants and organized a six hour training session before the first actual negotiating session. [FN285] The goals of the training program were to:

    (1) educate participants on the fundamentals of environmental negotiations;

    (2) improve participant awareness of the dynamics of negotiations;

    (3) develop negotiating skills and techniques; and

    (4) assist participants in understanding the different interests that would be represented at the negotiating table. [FN286]

  Over twenty negotiators attended the organizational meeting. At the meeting, ERM-McGlennon proposed a simple set of operating protocols: (1) decision-making by consensus; (2) no contact with the press during negotiations; and (3) agreement on a deadline for concluding negotiations. The proposal was intended to be broad enough to allow participants to create specific protocols and thus 'own' the first agreement of the process. The negotiators were content to modify the general guidelines drafted by ERM-McGlennon.

  *1677 The group discussion over resource pool protocols was far more extensive. The resource pool protocols were designed for the administration of a $50,000 fund [FN287] designed to defray the extraordinary expenses of participating in the negotiations and to support any necessary research or technical analysis required during the negotiation process. After extensive discussions, the committee reached a consensus on how to administer the pool, how to manage contracts, and how to distinguish between federal and private funds in the pool. [FN288] The committee agreed that individual stakeholder requests would be reviewed by the entire group and managed by an independent organization. In the NCP negotiations, the pool was administered by the American Arbitration Association under the direction of an NCP work group.

3. Establishment of the Negotiating Committee

  In its April, 1984, Federal Register notice the EPA articulated its reasons for selecting the NCP rule for negotiation, [FN289] identified key issues for the negotiation, and proposed participants. [FN290] The notice solicited requests by other persons wishing to participate, requiring such requests to be accompanied by an explanation of the requester's interests and a showing of why representation of that interest was not adequately provided for in the proposed composition of the negotiating committee. It said that such requests would be evaluated according to the following criteria:

    (1) whether the requester would be substantially affected by the rule;

    (2) whether the requester was already adequately represented in the negotiating group; or

    (3) whether the requester should be added to the group for other reasons.   [FN291]

  Issues to be addressed in the negotiations were identified. [FN292] The notice listed various manufacturers, industry associations, and environmental groups as proposed parties to the negotiations. [FN293] Only one of the proposed participants did not participate in the negotiations. Three parties were later added to the committee. The total membership of the committee included twenty-three participants.

4. Incentives to Participate

  Environmental groups participated in the negotiations because they favored NCPs and thought negotiations would result in NCPs. NCPs would make it easier for the EPA to promulgate emission standards based on 'leader technology.'  [FN294] In the absence of NCPs, these groups believed, the EPA would be inclined *1678 to set lower standards, so a greater portion of the industry could meet them. Thus, NCPs represented an escape valve that would permit tougher standards. In addition, the litigation strategy of these groups was enhanced by regulations providing for NCPs. NCPs create an alternative for laggards other than going out of business, thereby mitigating impracticability arguments in industry challenges to standards. Laggard manufacturers had an incentive to participate to ensure that NCPs were as low as possible. Leader manufacturers had an incentive to participate to ensure that NCPs were as high as possible to reduce the competitive advantage that otherwise might accrue to laggard manufacturers.

5. Intraconstituency Problems

  Intraparty disagreements were minimal in the NCP negotiations because the relatively small number of heavy duty vehicle manufacturers made it possible to have virtually all of them as participants.

  Intraconstituency disagreements were also minimal within the environmentalist community. A relatively stable coalition of environmental groups had developed over the years, essentially delegating representational authority to one or two staff members who handled certain regulatory issues on behalf of the environmental movement as a whole. Such delegation was necessitated by resource limitations.

6. The Negotiations

  Negotiations began on June 14, 1984, and concluded on October 12, 1984.   [FN295] They were conducted under a federal advisory committee charter. Committee meetings were open to the public and noticed in the Federal Register. [FN296]

  The EPA hired a contractor to study the NCP negotiation process.   [FN297] The contractor surveyed participants at the beginning and end of the negotiations. [FN298] The results of the surveys can be summarized as follows:

    (1) Many participants were concerned at the outset that the negotiating process would be more trouble than it was worth, but they were generally satisfied with the result.

    (2) Participants were particularly concerned about what role the EPA would play as a participant, and expressed some criticism at the end of the negotiations that the EPA was too passive, especially regarding technical data.

    (3) There was a range of enthusiasm for the process. Vehicle manufacturers had more enthusiasm than other participants. Support for the process seemed to correlate with a perception that it decreased adversarial postures and promoted deeper understanding of the issues and trade offs involved.

    (4) Multiple representatives of an interest, or of similar interests, were perceived as reducing power, because of the need to promote and retain consensus among the representatives. The single environmental group representative was perceived as having the most power.

    *1679 (5) Some participants thought that litigation might be a more resource efficient way for public interest groups to influence rulemaking than participation in negotiations. Almost everyone agreed that participation in the negotiations required substantial resources, and that environmental groups have fewer resources than other institutional interest representatives.

    (6) Some participants expressed the view at the outset that a benzene type consensus guideline was desirable. [FN299] No concern about the more flexible consensus rule actually used was expressed at the end of the negotiations.

    (7) In general, support was expressed for participation by the OMB.

    (8) Some participants suggested that it would have been useful to have appointed a neutral technical expert who could have assisted in evaluating data while shielding proprietary data from disclosure to competitors or the public.

  Toward the end of the negotiations, the EPA made available to the negotiators a microcomputer with software that permitted negotiators to evaluate the impact of different regulatory options. Most of the participants thought this helpful.

7. Postconsensus Process

  The committee reached consensus on the core conceptual issues on October 12, 1984. The proposal of the EPA, based on the group's consensus, appeared in the Federal Register on March 6, 1985. [FN300] Only thirteen comments were received. All were from participants and all supported the consensus proposal. Six specifically requested additional negotiations.

8. Final Rule

  The Phase I NCP rules were promulgated in final form on August 30, 1985.   [FN301] The EPA provided independent justification for the rule, referring to the negotiated rulemaking, but did not rely on the consensus reached in negotiations to justify the contents.

D. PESTICIDE EXEMPTIONS

  The EPA used negotiated rulemaking successfully a second time in revising regulations to implement exemptions to pesticide regulations. Negotiations proceeded from late 1984 to early 1985, resulting in agreement on a proposed rule published in the spring of 1985 and a final rule published in January, 1986.

1. Bankground

  Section 18 of the Federal Insecticide, Fungicide and Rodenticide Act (FIFRA)   [FN302] gives the Administrator of the EPA the discretionary authority to exempt any federal or state agency from any provision of the FIFRA if the Administration determines that emergency conditions require it. Regulations implementing section 18 were promulgated in 1973 [FN303] and were not changed. [FN304]

  *1680 In the fall of 1982, the EPA Office of Pesticide Programs conducted a review of the regulations and programs to grant state and federal agency exemptions. The review and audit raised four concerns about section 18 regulations. [FN305] In addition, the House Subcommittee on Department Operations, Research and Foreign Agriculture expressed concern that states or industry might be using section 18 to circumvent the stringent data and risk control requirements of the Act. [FN306] In January, 1984, the EPA began to consider revisions in the regulations and held three public hearings to solicit the public's views in preparation for issuing proposed revised regulations.  [FN307] After the hearings, the EPA decided to explore negotiated rulemaking as the procedure for developing revised regulations. [FN308]

2. Formation of the Negotiating Committee

  In August, 1984, the EPA published a Federal Register notice declaring its intent to form an advisory committee to develop new section 18 regulations.  [FN309] The notice reported that the EPA had contacted affected interests and that they were interested in participating in negotiations. The notice identified a list of issues to be considered in the negotiations. [FN310] The notice also identified environmental groups, state organizations and officials, federal agencies, user groups, and manufacturers and processors as tentative parties. [FN311]

  The notice invited requests for participation, to be accompanied by an explanation of the interest to be represented and a showing of why the interest would not be represented adequately by the proposed participants.

  The notice reported the intent of the EPA to use any consensus reached in the negotiation as the basis for an NPRM, 'unless it is inconsistent with our statutory requirements, or it is otherwise unjustified.' [FN312] It also limited the negotiation process to four months. If the committee did not reach consensus by January 28, 1985, the committee would be terminated and the EPA would proceed on its own to develop a revised regulation. [FN313] A number of groups were added to the negotiating committee after the August, 1984, notice. [FN314]

3. Prenegotiation Preparations

  ERM-McGlennon Associates served as the convenor for the pesticide negotiations, though an in-house facilitator was used after the negotiations began. In a manner similar to the NCP negotiations, a training session was held and convenor-*1681 drafted protocols were discussed. Because of the utility of the organizational meeting in the NCP negotiations, the program office for the section 18 rule specifically requested an organizational meeting. Protocols were less controversial in the pesticide negotiations than in the NCP negotiations. The pesticide exemption committee's resource pool was managed by the National Institute for Dispute Resolution. [FN315]

4. Incentives to Participate

  The issues involved in section 18 exemptions involved conflict among interests. Agricultural interests generally favored a more permissive exemption policy, though there were different concerns among agricultural groups depending on crop and region of the country. States generally favored a more permissive exemption policy, though sharp differences existed between health and agricultural agencies within a particular state. Environmental groups tended to favor a more restrictive exemption policy.

  The inevitability of change created an incentive for all groups to participate in shaping the change. The existing program appeared to be collapsing. Between 1976 and 1982 there had been a 750% increase in self- administered 'crisis exemptions.' The congressional staff study had concluded that changes in the regulation were necessary. [FN316]

5. The Negotiation

  The full committee met four times. The committee divided into three working groups that met separately from the full committee. Working group results were considered at full committee meetings. The full committee developed the consensus on the preamble and the proposed rule.

  An early ground rule, insisted upon within the committee, was that a consensus be reflected by commitment to a formal document. Otherwise, participants were unwilling to invest the time to participate in the negotiation. The ultimate commitment to support the consensus document before the agency emerged in the last negotiating session.

  The facilitator and from the Office of General Counsel of the EPA. Thus the pesticide exemption negotiation is the only completed negotiated rulemaking in which an internal facilitator was used.

6. Postconsensus Process

  On January 16, 1985, the committee reached full consensus on the exact wording of text and preamble of the NPRM. The EPA published the NPRM on April 8, 1985. Nineteen comments were received. Three were from participants and supported the proposal. The other comments raised relatively minor points.

*1682 7. Final Rule

  The final rule was published on January 15, 1986. [FN317] Comments on the proposed rule were received from nineteen sources. [FN318] The EPA's discussion of the comments in the preamble to the final rule referred to the negotiations with respect to two of the comments, but otherwise offered EPA's own rationale in support of the final rule.

VI. OTHER NEGOTIATIONS--PAST AND PRESENT

A. INTRODUCTION

  After the author's report to the ACUS was completed and ACUS Recommendation 85-5 was adopted, agency interest in negotiated rulemaking increased. This part identifies the major rule negotiations not specifically addressed in parts III- V. It also comments on early Labor Department efforts to prompt consensus among interests affected by proposed rules.

  The Department of Labor has had more experience with negotiated rulemaking besides the benzene effort discussed in part III and the experiences considered in this part of the article, under the advisory committee process authorized by section 7 of the OSHA Act. [FN319] The negotiating efforts addressed in this part form a useful context within which to evaluate the benzene effort.

B. COKE OVEN NEGOTIATIONS

  The OSHA regulations controlling employee exposure to coke over emissions [FN320] were adopted in accordance with agreements reached through negotiations among the OSHA, industry, and organized labor. The standard sets a Permissible Exposure Level (PEL) of 150 micrograms per cubic meter of air, averaged over any eight hour period. [FN321] Employers are permitted to show that engineering and work practice controls to reduce exposures to the PEL are not feasible. [FN322] In cases where a showing of nonfeasibility is made, the regulation directs employers to undertake other protective measures, such as requiring the use of respirators. [FN323]

C. THE COTTON DUST EXPERIENCE

  Negotiated rulemaking was used to a limited extent, without agency involvement, to resolve disagreements between industry and labor over revisions in the occupational health standard for cotton dust. Cotton dust can cause byssinosis. On June 23, 1978, the OSHA promulgated a regulation [FN324] setting permissible exposure limits for cotton dust in the yarn manufacturing, slashing and weaving, and knitting and nontextile industries. [FN325] The standard was challenged in litigation *1683 reaching the Supreme Court.  [FN326]

  In reaction to the litigation, the OSHA published an Advance Notice of Proposed Rulemaking on February 9, 1982, [FN327] followed by an NPRM on June 10, 1983. [FN328] The NPRM promised hearings.

  After three weeks of hybrid rulemaking hearings in the fall of 1983, counsel for the American Textile Manufacturers Institute and Amalgamated Clothing and Textile Workers Union met informally in an attempt to resolve about fifteen issues in dispute between them. They reached agreement on about twelve of the issues, and discussed how they would handle the other three--who would take positions and who might remain silent. They communicated their agreement to the OSHA by submitting posthearing briefs with identical pages 2-76.

  Their negotiations were facilitated by the existence of a completed hearing record and the parties' knowledge of the major issues. As a result, both sides had a better sense of possible outcomes. As part II.B.1 of this article explains, negotiated resolution of a dispute is more likely when party estimates of outcome in the absence of negotiations are similar. That theoretical proposition is substantiated by the cotton dust negotiations.

  Success in the cotton dust negotiations was important for three reasons. First, it encouraged the OSHA to believe that negotiated rulemaking could be successful in some circumstances, and strengthened the desire of the OSHA to continue the benzene negotiations, which had begun some six months before the cotton dust agreement. Second, it illustrates the possibility of negotiated agreement on controversial rules [FN329] without agency participation, when the incentives of the private parties are strong. Third, it illustrates the potential usefulness of negotiated rulemaking at late stages of the rulemaking process, even when negotiations have not been used to develop a proposed rule.

  The success of the cotton dust negotiations was jeopardized, however, when the OMB held up final release of the negotiated standard out of concern that the parties to the negotiation did not represent all affected interests.

D. MDA NEGOTIATIONS

  The Department of Labor decided to use negotiated rulemaking to facilitate setting a standard for 4,4'--Methylendedianiline (MDA). On October 22, 1985, the Department published a notice in the Federal Register [FN330] chartering a negotiating committee and soliciting requests for additional participants in the negotiations.

  MDA is an adhesive. It has been identified by the National Toxicology Program as an animal carcinogen. Studies indicate that MDA is a potential human carcinogen. [FN331] Several thousand workers experience dermal and respiratory exposure to MDA in the chemical, rubber, and adhesives industries.

  *1684 The OSHA announced its intention to control workplace exposure to MDA on September 20, 1983, in an Advance Notice of Proposed Rulemaking.  [FN332] The EPA also has regulatory responsibility for regulating MDA under the Toxic Substances Control Act (TSCA). [FN333] The EPA decided to defer to the regulatory action of the OSHA under section 9 of the TSCA, [FN334] obligating the OSHA to notify the EPA by early January, 1986, concerning what regulatory action the OSHA intended to take. [FN335]

1. Suitability of MDA for Negotiated Rulemaking

  Personnel of the OSHA believe MDA is a better candidate than benzene   [FN336] for negotiated rulemaking for the following reasons:

    (1) The possibility of EPA regulation creates special incentives for industry and labor groups to assist the OSHA in developing a standard promptly. Labor has less influence at the EPA than at the OSHA, and industry may fear that the EPA is less sympathetic to the economic impact of health and safety standards than the OSHA.

    (2) The absence of human carcinogenicity studies makes fundamental values less an issue for MDA than for benzene because there is more room for argument as to the human health effects of MDA.

    (3) Fewer industries and labor organizations are concerned with MDA than with benzene.

    (4) The absence of large scale health effects already manifesting themselves removes the tort liability issue which affected the benzene negotiations.

    (5) Because there is no existing standard for MDA and consequently no litigation history, party positions are less concrete and less polarized.

    (6) The process the OSHA intends to use for MDA standard negotiations is expected to work better than the process used for benzene negotiations.

2. Negotiating Process

  The OSHA utilized the FAA and EPA negotiations as medels for its MDA negotiations. It chartered an advisory committee as the forum within which negotiations would occur, expecting this to produce certain advantages. The OSHA would not be inhibited from participating in the negotiations. [FN337] Administrative resources are available from the Labor Department's advisory committee management office to support the negotiation.

  The OSHA is participating actively in the MDA negotiations, and expects that its participation will increase the usefulness of the negotiations. [FN338] Because the OSHA contemplates that consensus will be defined more broadly than in the benzene negotiations, it expects to gain a sense of the feasibility of different types *1685 of standards and the rationales for parties' positions, and hopes to achieve some narrowing of issues even if the negotiators are unable to agree on a PEL. [FN339]

  The initial Federal Register notice, following the example of the FAA and EPA notices, identified thirteen specific issues for the negotiations to address, listed potential participants, and invited comment on the makeup of the negotiating group. [FN340] The notice committed the OSHA to use any consensus reached through negotiation as the basis for its NPRM for MDA,  [FN341] but declared that the OSHA will terminate the negotiations and proceed on its own if no consensus is reached within six months. [FN342] It is, of course, too early to evaluate the effectiveness of MDA negotiations. The nature of the process envisioned, however, as well as the decision by the OSHA to try negotiated rulemaking for another health standard, shows that the benzene effort, though not successful in producing an agreed upon standard, was successful as a pilot effort in exploring a new process.

  Negotiations on the MDA standard began in July, 1986, and have proceeded since then, producing hope that a consensus might be reached by early 1987.  [FN343]

  The OSHA's MDA negotiations had not begun at the time the ACUS adopted Recommendation 85-5. Since the adoption of ACUS Recommendation 85-5, however, the author has attended MDA negotiating sessions and has met informally with most of the participants. No one knows as of this writing whether the MDA negotiation will produce consensus. Nevertheless, it is possible to offer some observations about the process, comparingit with the benzene negotiation process.

  Active participation by the OSHA is helpful. In fact, in the MDA negotiations, the OSHA participant regularly prepares draft NPRM language synthesizing the OSHA tentative views and those expressed by participants in the *1686 negotiations. This relieves participants form having to agree or disagree with the language prepared by an opposing interest. In addition, OSHA preparation of postdiscussion drafts provides concrete material for negotiators to consider, and reinforces the view that the OSHA will develop an NPRM on its own if the negotiations do not produce a consensus. On the other hand, the process imposes significant administrative burdens on the OSHA participant, and it poses an intellectual challenge to capture adequately the views expressed in the negotiation sessions.

  Compliance with the FACA does not appear to be presenting any difficulties. The negotiation sessions regularly are open to the public and participants report that this poses no particular difficulty.

  Midway through the negotiations all participants were enthusiastic about the process, reporting that whether or not consensus is achieved, all parties have achieved a greater understanding of the legitimate difficulties that opponents have with particular approaches to setting standards. The greatest concern at the midpoint of the negotiations is the procedure that will be used by the OSHA to process any consensus recommendation.

E. ADDITIONAL EPA RULE NEGOTIATIONS

  After ACUS Recommendation 85-5 was published, the EPA continued to use the negotiated rulemaking process. Negotiations over a farmworker protection standard continued through the spring of 1986. [FN344]

  The committee began by developing the issues the regulation should address and the factors that would enter into their resolution. Its workgroups brought to bear first hand experience from the field and from the operations of the farms. Consultants to the EPA wrote a working draft of a potential regulation that the committee used to focus its discussions. One of the major interests involved decided, however, that it would be better off not participating further. The remaining interests decided that they wanted to continue to review the draft regulation and to attempt to write a sound, workable standard. It was clear that with one of the major interests absent, such discussions could no longer be regarded as working out a consensus rule. The group was converted therefore into a standard advisory committee and it was made clear that the EPA alone would take responsibility for the proposed rule. The truncated group met twice more and reviewed the evolving draft each time. The EPA's Office of Pesticide Programs built on the earlier discussions. The EPA has not yet published its notice of proposed rulemaking, but it is expected shortly.

  Thus, the farmworker protection negotiations, like those in benzene, did not result in consensus. Nevertheless, at least some of the participants, including senior agency officials, have viewed the negotiations as significantly productive, *1687 and expect them to have a major effect in shaping the proposed standard. [FN345]

  In addition, the EPA has proposed negotiations to develop proposed rules on air pollutant emissions from wood burning stoves, [FN346] on hazardous waste injection wells, [FN347] and on RCRA permits. [FN348]

F. FEDERAL TRADE COMMISSION NEGOTIATION INITIVATIVE

  On February 12, 1986, the Federal Trade Commission (FTC) announced that it was considering formation of an advisory committee to negotiate a consensus on recommended revisions to the FTC's rule on informal dispute settlement procedures, rule 703. [FN349] The informal dispute settlement procedure rule implemented section 110(a)(1) of the Magnuson-Moss Warranty Act. [FN350] That section provides that warrantors may incorporate into their written warranties a requirement that consumers resort to an informal dispute settlement procedure before pursuing judicial remedies available under the Act for warranty claims. The FTC's rule has been criticized as being both unduly burdensome and insufficiently stringent. [FN351] The FTC believed that negotiation of possible changes was particularly appropriate because of the voluntary nature of rule 703. [FN352] Based on preliminary inquiry by two professional mediators retained by the FTC to act as 'convenors,' John McGlennon of ERM-McGlennon Associates and Gail Bingham of The Conservation Foundation, the FTC believed that negotiation could succeed with respect to the rule. [FN353] The February 12 notice proposed procedures for negotiation and listed 26 potential parties to be members of a negotiating party. [FN354]

  On August 20, 1986, the FTC established a 'Rule 703 Advisory Committee' to develop proposed revisions to rule 703. It also announced that its first meeting would be held on September 23, 1986. [FN355] The Commission received 22 comments in response to its proposal to use the negotiated rulemaking in connection with rule 703 revisions. Only one commentor opposed use of the process, arguing that consumer groups and consumer protection agencies could not represent the interests of consumers adequately.  [FN356] The FTC responded that a wide variety of consumer interest representatives would participate on the committee and opined that consumer views would be represented adequately. In addition, the *1688 Commission noted that subsequent procedural steps, after the advisory committee completed its work, would provide ample opportunity for individual consumers and others to comment on any committee consensus. [FN357]

  The FTC identified from the comments three groups whose interests would not be represented adequately by the parties appearing on the February 12 meeting list: organizations that assist consumers in presenting claims in auto industry dispute resolution procedures, state legislatures, and the recreational vehicle industry. [FN358] It added representatives of those three groups to the committee and reported that some representatives on the February 12 meeting list had agreed to withdraw to prevent the committee from becoming too large.  [FN359] The representatives on the original list but not on the final list were the Conference of Consumer Organizations, U.S. Office of Consumer Affairs, Association of Home Appliance Manufacturers, and Residential Warranty Corporation. [FN360]

G. NEGOTIATED RULEMAKING BY STATE AND FEDERAL AGENCIES

  At the American Bar Association annual meeting in New York in August, 1986, the Dispute Resolution Committee of the Section on Administrative Law presented a program on negotiated rulemaking as an alternative dispute resolution technique. The author was the program organizer and moderator. [FN361] The panel reviewed the several federal agency initiatives with negotiation of rules analyzed in this article.

  Experience with negotiated rulemaking as a separately defined decisionmaking process has been limited at the state level. It is common, however, for state agencies to meet informally with affected groups to develop rules. Such informal interaction more often occurs through bilateral contact between the agency and a single representative in preparing a proposal, than with all parties simultaneously. Some states, particularly New Jersey, use a preproposal notice to affected interests. In Minnesota, agencies have an incentive to obtain agreement on proposed rules with affected interests to avoid the necessity of a formal hearing held by the independent office of administrative judges. The office charges the agency an hourly fee for administrative law judge (ALJ) services, and the ALJ has veto power over the agency's proposed rule.

  The participants agreed that further use of negotiated rulemaking at the state level depends, as it does at the federal level, on all parties (including the agency) having a clear incentive to negotiate a consensus rather than following the usual formal process. [FN362]

*1689 VII. COMPARATE ANALYSIS

  The four negotiations studied were different in subject matter, power of parties, cohesiveness of interests represented, procedure followed, and commitment and competence of the sponsoring agencies.

  Only the EPA formally solicited suggestions for candidates for negotiated rulemaking. This canvass shifted the burden to major interest groups to propose subjects on which they would be willing to negotiate, even as they rejected other subjects. The agencywide initiative may have some benefits in terms of communicating clear agency support for the concept. It also raised fears among some affected interests, however, that the EPA was overly committed to a process and might force its use without sufficient consideration for its utility in a particular proceeding.

  The EPA picked subjects that had not already failed in traditional rulemaking processes. The OSHA and the FAA picked subjects for negotiation that had proved intractable in traditional rulemaking and judicial review. [FN363] Other differences between the history of the benzene rulemaking effort and the history of the flight and duty time effort, however, made benzene a riskier candidate than flight and duty time. No one involved in the flight and duty time, NCP, or pesticide controversies really believed that a rule could be determined by objective examination of scientific data. At least some of the benzene participants believed that the matter was simply one of analyzing scientific data. This inhibited realistic exchange of concessions in the benzene negotiations because of an underlying perception that the correct standard could be determined objectively rather than through bargaining.

  BATNAs [FN364] in the benzene negotiation were relatively attractive, reducing incentives to negotiate. The benzene litigation had exaggerated points of disagreement and led to divergent perceptions of what the OSHA was expected to do in response to the Supreme Court's decision. Labor expected action by the OSHA similar to the rule invalidated by the Supreme Court; industry expected it could block this standard in litigation or in the OMB clearance process. Thus each side perceived a relatively favorable BATNA, reducing the incentives to negotiate.

  BATNAs in the flight and duty time negotiation, by comparison, were relatively unattractive. The FAA's unsuccessful attempts at flight and duty time rulemaking had convinced the participants that unilateral FAA action could be harmful to their interests.

  Somewhat surprisingly, there was little controversy over who should participate in the negotiations. In all of the initiatives except the one involving benzene, the agencies published lists of proposed participants, permitting excluded interests to make a showing of why they should be added to the negotiating group. [FN365] Agencies experienced no difficulties with this approach. The benzene effort mighthave benefitted from a similar approach to identify participants. *1690 The benzene negotiations would have proven less difficult if some labor unions not participating could have been involved to a greater degree instead of objecting to aspects of a potential agreement from the sidelines. In the benzene negotiation, one problem was failing to include everyone who had the incentive and the power to block a negotiated resolution.

  The process of identifying participants in the flight and duty time negotiation focused on potential intraconstituency disagreements within the major airline industry association, creating an incentive for the association to address such disagreements. Less attention was paid to intraconstituency disagreements at the threshold of the OSHA negotiation.

  Some participants in the NCP negotiations, particularly environmental groups, may have found the negotiation process relatively more attractive than participants in the benzene or flight and duty time negotiations because some NCP participants perceived themselves as having relatively little influence with the EPA through notice-and-comment or hybrid rulemaking and more influence through the negotiation process. In contrast, organized labor has long enjoyed substantial informal influence with Labor Department agencies. Therefore negotiations threatened to lessen labor's relative power advantage. The airline and pilot organizations similarly have considerable power to affect FAA policy, but each side perceived its opponent as having equivalent power.

  It is difficult to conceive how any of the four negotiations would have worked at all without the help of the facilitator/mediators. All three agencies used both convenors and mediators. [FN366] In some cases the convenor became the facilitator or mediator; in other cases, the facilitator or mediator was a different person from the convenor. The facilitator/mediators played different roles in the four negotiations. Those involved in the benzene negotiation were less active in the formal benzene meetings and more active outside the formal negotiating sessions than the other neutrals. The mediator in the flight and duty time negotiations concentrated his effort on formal negotiating sessions and caucuses. He was characterized as being 'phenomenally effective' in caucuses. Despite the differences in styles, involvement of the facilitator/mediators almost certainly was essential to keep all of the negotiations moving. [FN367] Few participants or observers thought that either external or internal facilitator/mediators have advantages over the other. Instead, mediation experience and competence was considered important. Mediator skills involve more than getting along well with people; they involve an instinctive awareness of group functioning and how to move toward closure.

  The most striking difference between the benzene negotiations and the other negotiations is that the OSHA did not participate in the benzene negotiations, while agencies did participate in the other negotiations. [FN368] Agency participation *1691 militated against harmful divergence in expectations about what the agency would do unilaterally, and therefore conditioned BATNAs to maintain incentives for negotiated agreement. In the benzene negotiation, the participants came to believe they would do better outside the negotiations, working directly with the OSHA or OMB, and therefore their BATNAs shifted so that a negotiated agreement was unattractive.

  The FAA and EPA were committed to the negotiating process as the primary means of developing a regulation; the OSHA was not. At least some of the difference in agency commitment can be attributed to the lower level of participation by the OSHA. The FAA and its parent agency, the Department of Transportation, and the EPA had key personnel who were enthusiastic about the regulatory negotiation and who had the skills to mediate effectively, supplementing the efforts of the facilitator/mediator; the OSHA and Department of Labor did not. In particular, both the Department of Transportation and the EPA have policy level officials responsible for promoting regulatory negotiation; the Department of Labor did not. [FN369]

  Intraconstituency differences were manifest in both the benzene and flight and duty time negotiations. Both the ATA, representing airlines, and the API, representing petroleum companies, struggled to forge a consensus that could be communicated by their negotiators. The ATA had more success than the API. Notably, the FAA proposed bringing individual company representatives into the negotiations. This may have galvanized the ATA to ensure that it remained an effective participant on behalf of the entire constituency.

  In three of the four experiments (benzene, flight and duty time, and NCP) the participants fell short of formal agreement on some of the major issues;  [FN370] indeed one could argue that fewer fundamental matters separated the benzene negotiators when negotiations were adjourned than separated the flight and duty time negotiators. The benzene negotiations, however, are perceived widely as having failed while the flight and duty time negotiations are perceived as having succeeded. The different perceptions, despite the similarity of results, are attributable to the difference in what the agency did when the negotiations were adjourned. The FAA, having participated in the negotiating sessions, had an understanding of what the parties could accept. The OSHA, not having participated, had no such understanding. Communicating the points of agreement on benzene to the OSHA required transmission of a document. The benzene negotiators nevertheless had agreed on a ground rule that no document would be transmitted in the absence of a formal agreement on a total package. [FN371] No such ground rule restricted the flight and duty time or NCP participants or agencies.

  Only the EPA used the 'resource pool' concept, which permits supporting *1692 travel expenses and possible expert, neutral technical assistance.  [FN372] Participants thought the resource pool was useful and worth utilizing in other negotiations. On the other hand, as the NCP experience showed, negotiation of a resource pool protocol can be controversial and may get the negotiations off to a bad start.

  In the three cases in which negotiators reached full or partial agreement, the agencies published the negotiated rule for comment, [FN373] clearly indicating changes from the negotiated text. In addition, both the FAA and EPA allowed the negotiators to review comments submitted in response to the Federal Register notice. [FN374] This approach satisfied participants.

  Compliance with the requirements of the Federal Advisory Committee Act worried everyone. Compliance with the open meeting requirement for meetings of the full negotiating group, however, created few problems in the FAA and EPA negotiations. Agencies including the OSHA generally followed the spirit of paragraph 12 of ACUS Recommendation 82-4, closing sessions only when necessary to protect confidential data and caucuses only when necessary to protect the deliberative process. Some participants and agency personnel believed the open meetings permitted challenges to the negotiating process to be deflected more successfully than if the meetings had been closed. [FN375] In the benzene negotiations, efforts to avoid the FACA requirements distorted the negotiation process, primarily by foreclosing active agency participation. [FN376]

  No court challenges to the negotiated rules have materialized to date. Thus, there is no experience by which to judge the need for legislation dealing with authority of agencies to use negotiated rulemaking [FN377] or the need for changes in the standards of judicial review. [FN378]

VIII. ANALYSIS OF SPECIFIC LEGAL ISSUES

A. INTRODUCTION

  Initially, it was thought that negotiated rulemaking could be successful only if authorized specifically by Congress. In 1982, the ACUS recommended that Congress amend the APA to authorize explicitly the use of negotiated rulemaking. [FN379] The commentators whose reassessment of administrative law gave rise to the regulatory negotiation concept thought judicial review might need to be altered to accommodate negotiated rulemaking. [FN380] Some discussion of the APA *1693 problems associated with negotiated rulemaking appears in the literature. [FN381]

  Since 1982, however, it is has become clear that negotiated rulemaking can be accomplished within the flexible framework provided by section 553 of the APA. Through the negotiated rulemaking process, agencies provide affected parties with a greater opportunity to influence the substance of agency rules than existed under the minimum requirements of section 553 merely by having notice of an agency-developed proposal and providing comments. Apparent problems with delegation of governmental authority to private parties, ex parte communication in violation of the spirit of section 553, and open meeting requirements under the FACA have turned out to be largely illusory.

  There are some caveats, however. If a negotiated rulemaking is to withstand challenge under the standards for judicial review set forth in section 706 of the APA, the agency must provide a reasoned justification for the final rule. It would not be sufficient for the agency merely to adopt without comment the work of the parties to negotiation. Moreover, the requirements of the FACA continue to trouble participants in negotiated rulemaking. Because the FACA has been interpreted in a practical way and no affected parties were motivated to challenge the negotiation process, the FACA has not yet been a problem.

  The analysis of administrative law issues is necessarily somewhat academic because no judicial challenges to negotiated rulemaking have materialized so far. That no challenges have materialized is, of course, perhaps the strongest endorsement of negotiation as a process that satisfies the needs of the affected interests.

B. DELEGATION DOCTRINE

  Superficially, negotiated rulemaking seems to pose problems with the delegation doctrine. A central precept of democratic political theory is that governmental decisions ought to be made only by politically accountable officials. [FN382] The delegation doctrine prohibits such officials from delegating their policymaking authority to persons or institutions that are not politically accountable.

  In the early years of administrative law, members of the legal profession were preoccupied with the delegation doctrine as they sought to rationalize the exercise of political power by nonelected administrative agency officials.  [FN383] It is well *1694 accepted now, however, that delegation of quasi-legislative authority to administrative agency officials is permissible when the statute delegating the power adequately circumscribes the exercise of the power to permit judicial scrutiny of whether the agency has stayed within its delegated authority. [FN384] Implicit in this doctrine is the idea that agency officials are themselves politically accountable to some degree. They are government officers, appointed pursuant to law. [FN385] When cabinet level officers make decisions, for example, political accountability is assured by the political forces operating on the President in connection with high level appointments, including the constitutional requirement for Senate confirmation. Accountability is less evident when lower level officials make decisions, but it presumably is ensured to some degree by appointment procedures set forth in the civil service laws and by various restrictions placed on the officials' conduct.

  If, however, the agency delegates its authority to a group of private citizens, further delegation problems arise. One problem is whether such delegation is within the agency's authority from the Congress. Another problem is that the private delegates are even less politically accountable than agency officials. [FN386]

  Compliance with the delegation doctrine concerns proponents of negotiated rulemaking. [FN387] While it is appropriate to think about delegation issues associated with negotiated rulemaking, it is important not to exaggerate the magnitude of the problem. The early proponents of regulatory negotiation offered a number of reasons why the delegation doctrine is not violated by negotiated rulemaking.

  First, under all the current conceptions of negotiated rulemaking, negotiators play only an advisory role to the agency; the agency retains the final decisionmaking authority. [FN388] This advisory role has been approved by the courts in a variety of other circumstances. [FN389] Impermissible delegation obviously does not occur, for example, when private parties to an agency proceeding get together *1695 privately and compromise their differences. [FN390]

  Second, the nature of negotiated rulemaking, if it is pursued under the ACUS recommendations, ensures adequacy of representation of affected groups. Thus, it provides its own form of political accountability, which probably is greater than when the agency makes rules unilaterally. Negotiated rulemaking pursuant to the ACUS recommendations thus avoids the problem of unaccountable decisionmaking that the delegation doctrine is intended to avoid. [FN391]

  It bears emphasizing that the delegation doctrine overlaps other requirements imposed on agency decisionmaking under the APA and substantive statutes.  [FN392] Delegation problems are avoided when rules are subject to judicial review under APA requirements and need not be addressed separately. [FN393]

  There are thus two entirely independent ways for negotiated rulemaking to satisfy the delegation doctrine. First, if the rule ultimately resulting from negotiated rulemaking passes judicial scrutiny under the arbitrary and capricious, within statutory authority, and in accord with statutory procedures standards, it has passed muster under the delegation doctrine. Second, if the affected interests have been represented fairly in the negotiation, political accountability exists and there is no need to be wary of delegation because the harm it seeks to avoid has been avoided from the outset.

C. CONFLICT BETWEEN GOALS OF INFORMALITY AND JUDICIAL REVIEWABILITY

  Negotiated rulemaking is a reaction, in some respects, to the schizophrenic nature of informal rulemaking under the APA. APA requirements for rulemaking are driven by conflicting goals of informality and judicial reviewability.  [FN394]

  As originally conceived, informal rulemaking under the APA was not adjudicatory in character. Kenneth Culp Davis, one of the deans of American administrative law, had this to say about rulemaking in 1972:

    The rulemaking procedure described in [section 553 of the APA] is one of the greatest inventions of modern government. It can be, when the agency so desires, a virtual duplicate of legislative committee procedure.

. . . .

    [R]ulemaking procedure is superior to adjudicative procedure in many ways, including the following: . . . An administrator who is formulating a set of rules is free to consult informally anyone in a position to help, such as the business executive, the trade association representative, the labor leader. An administrator who determines policy in an *1696 adjudication is usually inhibited from going outside the record for informal consultation with people who have interests that may be affected. In policymaking through adjudication, either the quest for understanding is likely to be impaired or the tribunal's judicial image is likely to be damaged. [FN395]

  He admitted, however, that adjudicatory procedure might be 'indispensable' when 'facts about parties are in dispute.' [FN396]

  The participation norm, and all of the administrative law doctrines reviewed here, militate against negotiation of a final rule without some opportunity for public comment. [FN397] Using negotiations to prepare a proposed rule, and then allowing notice and comment rulemaking, as occurred in all three of the successful negotiated rulemaking experiments, is a sound approach with few apparent disadvantages. Moreover, such publication and comment mitigates the effect of complaints by nonparticipants in the negotiations that they were denied a fair opportunity to influence the content of the rule. [FN398]

  The federal courts have been concerned with more than participation by affected interests; the courts also have insisted on meaningful judicial review of agency decisionmaking. Increasingly, they required that informal rulemaking be more formal. Prominent among the techniques employed to increase formality was hybrid rulemaking.

D. HYBRID RULEMAKING

  'Hybrid rulemaking' refers to a set of procedures imposed by Congress or the courts onto the requirements explicitly established by the APA. [FN399] Under hybrid rulemaking, agencies must develop an evidentiary base for rules under procedures less formal than full trial-type hearings but more elaborate than section 553 notice and comment procedures as they originally were envisioned. [FN400] Examples of statutory requirements for hybrid rulemaking are found in the Occupational Safety and Health Act [FN401] and the Magnuson-Moss Act. [FN402]

  Beginning with United States v. Nova Scotia Food Products Corp. [FN403] the idea that informal rulemaking ought to be accomplished 'on the record' gained support. [FN404] The paradigm case imposing hybrid rulemaking procedures is Mobil Oil Corp. v. FPC, [FN405] in which the court interpreted a statute requiring that a rule be *1697 supported by substantial evidence to obligate the agency to use at least certain elements of adversary, adjudicative procedures to develop the rule. Other courts have followed similar reasoning. [FN406]

  In Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, [FN407] the Supreme Court rejected the idea that courts routinely may require more than minimum procedures imposed by section 553 of the APA.  [FN408] Vermont Yankee apparently has made courts more reluctant to require trial-type procedures in notice and comment rulemaking, [FN409] but they continue to require that the 'record' in notice and comment rulemaking support the agency's decision. [FN410]

  ACUS Recommendation 76-3 identifies, for agency consideration, additions to notice and comment procedures that may be warranted in some rulemaking proceedings, while recommending that adversarial procedures not be superimposed on informal rulemaking as a general rule. [FN411] Negotiated rulemaking is yet another way to develop a rule and to marshal facts supporting the rule without the expense, delay, and rigidity associated with adversarial adjudicatory procedures.

E. EX PARTE COMMUNICATIONS

  Negotiated rulemaking does not contravene a policy against ex parte communication, derived from the hybrid rulemaking concept. [FN412] This section explains the ex parte communication prohibition and considers why it should not impede negotiated rulemaking.

  Notions of fairness and due process preclude ex parte communication between parties and the decisionmaker in an adjudicatory process. Fundamental concepts of adjudicatory decisionmaking contemplate decisions based on the formal record, and adversaries having the right to know and to counter information being put before the decisionmakers. [FN413] Ex parte communication jeopardizes *1698 these two principles in two respects: because the opposing parties do not know the content of the ex parte communication, they are deprived of an opportunity to respond to it; and the ex parte communication usually is not made a part of the record, and thus should not influence the decision.

  Remedies for violating a ban on ex parte communication include setting aside the agency's decision, and remanding for reconsideration of the decision after all interested parties are given as opportunity to participate in proceedings where all the information available to the agency is subject to challenge. [FN414]

  The 'on the record' requirements resulting from the hybrid rulemaking cases naturally raised questions about the appropriateness of ex parte communication in informal rulemaking.

  In Home Box Office v. FCC, [FN415] the D.C. Circuit articulated stringent limitations on ex parte contact in conjunction with informal rulemaking:

    (1) Once an NPRM is issued, agency officials or employees likely to be involved in the decisionmaking process should refuse to discuss matters relating to the disposition of a rulemaking proceeding with any interested private party, or an attorney or agent for any such party prior to the agency's decision; [FN416] and

    (2) If ex parte contacts nonetheless occur, written documents or summaries of oral communications must be placed in the rulemaking file established for public review and comment. [FN417]

  The court was motivated by three considerations: the possibility that the APA notice and comment process and the official record would be a sham, with the real decisional process based on secret communications; the need for adversarial comment on matters communicated to the agency; and the inconsistency between secrecy and 'fundamental notions of fairness implicit in due process and with the ideal of reasoned decisionmaking on the merits.'  [FN418] These considerations implicate both the judicial reviewability and the participatory goals of the APA.

  The Home Box Office restrictions have been applied unenthusiastically by the courts of appeals [FN419] and have been much criticized by commentators  [FN420] and by *1699 the Administrative Conference [FN421] as inconsistent with the realities of agency decisionmaking and the concept of informal rulemaking.

  Major questions about the correctness of the Home Box Office limitations were raised by the D.C. Circuit in Sierra Club v. Costle. [FN422] Sierra Club involved ex parte contacts among the EPA and coal industry representatives, members of Congress, and White House staff after the end of the comment period on a proposed limitation on sulfur dioxide emissions. The court declined to invalidate the rule because of the contacts:

    Where agency action resembles judicial action, where it involves formal rulemaking, adjudication, or quasi-adjudication among 'conflicting private claims to a valuable privilege,' the insulation of the decisionmaker from ex parte contacts is justified by basic notions of due process to the parties involved. But where agency action involves informal rulemaking of a policymaking sort, the concept of ex parte contacts is of more questionable utility. . . . Later decisions of this court . . . have declined to apply Home Box Office to informal rulemaking of the general policymaking sort involved here, and there is no precedent for applying it to the procedures found in the Clean Air Act Amendments of 1977. [FN423]

In addition to meetings of a policymaking nature in Sierra Club, documents also were submitted to the EPA ex parte. Important to the court's conclusion that these ex parte communications were permissible was that the rule was supported by information in the record and there was no indication that the documents submitted after the comment period played any significant role in the agency's support for the rule, or that the challenger was deprived of an opportunity to respond to anything that was outcome determinative.

  Moreover, ex parte contact is permitted during the development of NPRMs.   [FN424] The courts also have permitted agencies to use outside assistance during the deliberative phases or rulemaking, after the record is closed. In United Steelworkers of America v. Marshall, [FN425] for example, the union challenged reliance by the OSHA on outside consultants to analyze the rulemaking record and to help prepare the preamble of the final rule. The consultants participated *1700 in the rulemaking proceeding and submitted material for the record. As part of their deliberative assistance, the consultants prepared additional reports, which were submitted to the OSHA but not made part of the record. The union argued that these were impermissible ex parte communications.

  The D.C. Circuit rejected this argument: '[T]he communications between the agency and the consultants were simply part of the deliberative process of drawing conclusions from the public record. The consultants acted after the record was closed as the functional equivalent of agency staff . . ..'  [FN426]

  Relying on an opinion by Judge Friendly in a Freedom of Information Act suit involving the same proceeding, [FN427] the court concluded that

    while the reports might contain some factual matter . . . such information was necessarily incident to and not severable from the process of summary and analysis. He suggested, moreover, that to the extent the reports drew inferences from and weighed the evidence they were more truly 'deliberative' and thus better candidates for exemption than mere summaries of the record.  [FN428]

The reports were no less deliberative merely because they reponded to criticisms of on the record evidence by the same consultants. The response contained no new evidentiary material, only analysis and evaluation of the record. 'Thus, the earlier participation of these consultants as expert witnesses in no way disqualifies them as aides in the final decision.'  [FN429]

  It is reasonably safe to offer the following conclusions about ex parte contact in the context of negotiated rulemaking: [FN430]

    (1) The rule ultimately adopted by the agency must be supported by factual information contained in the official record. There is no reason that the negotiators cannot discuss and even agree upon factual information to be put in the record.

    (2) Persons with an interest in the content of the rule must be afforded an opportunity to know the factual basis for the rule and to challenge facts submitted by opponents in an appropriate adversarial context. This opportunity may be provided by an opportunity to participate in the negotiations or by an opportunity to comment afterward as part of the notice and comment process.

    (3) If the negotiation takes place among appropriately balanced interest representatives, the opportunity for adversarial exploration of policy and factual issues is preserved in the negotiation itself. [FN431]

    (4) Consultation between the agency and the negotiation participants after the record is closed is permissible if the consultation focuses on policy rather than new factual matters.

    (5) Placing summaries of discussions in the record so nonparties to the discussions *1701 can know of their substance and have an opportunity to respond, while not necessary in every case, enhances the likelihood that the ex parte contact will be found permissible by a court. [FN432]

F. JUDICIAL REVIEW UNDER ARBITRARY AND CAPRICIOUS STANDARD

  Section 706 of the APA [FN433] allows a court to overturn a rule if the agency's action was 'arbitrary, capricious, and an abuse of discretion, or otherwise not in accordance with law.' [FN434] It is common for substantive statutes authorizing agencies to make rules to include similar standards for judicial review, or to incorporate the APA standard by reference. [FN435]

  The Supreme Court recently reiterated what 'arbitrary and capricious' means:

    Normally an agency rule would be arbitrary and capricious if the agency has [1] relied on factors which Congress has not intended it to consider, [2] entirely failed to consider an important aspect of the problem, [3] offered an explanation for its decision that runs counter to the evidence before the agency, or [4] is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. [FN436]

  While the agency is given considerable discretion, the Court has 'frequently reiterated that an agency must cogently explain why it has exercised its discretion in a given manner.' [FN437] Agency explanation takes on added importance because 'an agency's action must be upheld, if at all, on the basis articulated by the agency itself.' [FN438] The obligation to explain is not trivial. The explanation of the OSHA of its original benzene standard, for example, filled 184 pages of an appendix to the standard in the Federal Register, [FN439] but the Supreme Court nevertheless invalidated the standard. [FN440]

  Uncertainty about facts may permit the agency to take one course of action instead of another based on its policy judgment, but the agency must explain how it moved 'from the facts and probabilities on the record to a policy conclusion.' [FN441] It is not sufficient for an agency merely to recite the terms 'substantial *1702 uncertainty' as a justification for its actions.  [FN442]

  The Supreme Court's Vermont Yankee decision [FN443] does not dilute the arbitrary and capricious review; Vermont Yankee merely requires courts to apply the arbitrary and capricious standard in scrutinizing the record for support of an agency rule instead of imposing particular procedures. Negotiated rulemaking, as a procedural innovation, is entirely consistent with the spirit of Vermont Yankee, but the rule resulting from the negotiations nevertheless must pass muster under the arbitrary and capricious standard of judicial review.

  Consequently a rule promulgated merely because it is agreed upon in negotiations among affected parties might be vulnerable to attack as being arbitrary and capricious, [FN444] unless the agency offers its own rationale and support for the content of the rule. [FN445] While part of the rationale can be that affected interests reached agreement in negotiations, rationality also requires a nexus between the rule and factual record support for it.  [FN446] A detailed record of the negotiations could chill interchange among negotiators, but there is no reason negotiators should not discuss what needs to be put in the record to support their consensus proposal, and what the agency should say in its rationale.

  The same action taken by negotiators and sponsoring agencies to avoid delegation and ex parte communication problems will also reduce the likelihood that a court would find a negotiated rule to be arbitrary and capricious.

  Even if there are flaws in the agency's articulated rationale, [FN447] it is appropriate for a reviewing court to consider whether the negotiating procedure used nevertheless sufficiently guaranteed that the agency considered all the relevant factors. [FN448] 'That the rule reflects a consensus of the affected parties therefore goes a long way towards meeting the goals of Overton Park and ensures that the rule is neither arbitrary nor capricious.' [FN449]

  Nevertheless, if the negotiated rulemaking procedure is to supply support for an agency decision that otherwise would be vulnerable to attack under the arbitrary and capricious standard, the reviewing court must assure itself that the negotiation procedure adequately ensured the participation of all the affected interests. Review of the procedure would include the following:

    (1) Reviewing the notice of intent to establish a rule negotiation to ensure *1703 that it adequately informed the public what issues would be considered so that all interests affected could know that their interests would be involved in the negotiation. [FN450]

    (2) Reviewing the merits of the consensus developed in the negotiations only if the challenger can demonstrate that its interests were not adequately represented in the negotiation and that it made a reasonable effort to participate in response to the original notice of negotiation. [FN451]

    (3) Reviewing the rule itself to determine if it is within the agency's authority, [FN452] and if it reflects all the factors required by statute to be considered. [FN453] In this part of the review it is appropriate to permit agency action within a wide ambit of reasonableness, [FN454] since an adequate negotiation process ensures that the agency has taken a 'hard look' at relevant factors and considered alternatives. [FN455]

G. FEDERAL ADVISORY COMMITTEE ACT PROBLEM

  The FACA [FN456] presents a potentially serious threat to the effectiveness of negotiated rulemaking. A strict interpretation of the FACA requires that regulatory negotiation sessions be open to the public. This interpretation, if applied by agencies or insisted upon by the courts, could make candid exploration of compromise by interested parties difficult.

  The FACA was enacted in 1972 [FN457] after 15 years of intermittent congressional concern over the utilization of advisory committees by administrative agencies. [FN458] The FACA regulates the creation, composition, and functioning of advisory committees. It requires that committee meetings be open to the public. Exceptions to the open meetings requirement originally were available only when the subjects to be discussed qualified for an exemption under the Freedom of Information Act. [FN459] In 1976, however, the FACA was amended to apply the Sunshine Act's [FN460] open meeting exceptions to advisory committees. [FN461] As *1704 amended, the FACA permits advisory committee meetings to be closed to the public only when the meeting will involve (1) matters covered by exemptions 1-7 of the Freedom of Information Act, (2) criminal accusations directed at a person or other formal censure of a person, (3) frustration of proposed agency action if prematurely known, or (4) agency participation in formal rulemaking or litigation.  [FN462]

  Remedies for violation of the FACA are limited and probably do not include invalidation of agency decisions made in reliance on advisory committee proceedings that violate the FACA. [FN463]

  General Services Administration (GSA) regulations implementing the FACA   [FN464] permit certain meetings to occur without compliance with the FACA,  [FN465] including meetings for the purpose of exhanging facts or information; [FN466] meetings initiated by a private group (rather than by the agency) for the 'purpose of expressing the group's view,' providing the agency does not use the group as a 'preferred source of advice or recommendations'; [FN467] and meetings for the purpose of 'obtaining the advice of individual attendees and not for the purpose of utilizing the group to obtain consensus advice or recommendation.' [FN468]

  In addition, the preamble to the GSA regulations [FN469] suggests that subcommittees of advisory committees are covered by the FACA only when they report directly to the agency rather than through the main advisory committee, and that informal meetings of two or more members of advisory committees are not covered when the purpose of the meeting is:

    (1) to gather information;

    (2) to conduct research; or

    (3) to draft option papers for the full advisory committee. [FN470]

  The FACA produces impediments to successful regulatory negotiation. First, the statutory and regulatory requirements for a charter, GSA approval, advance notice of meetings, and minutes slow the negotiation process. Second, and more fundamentally, the belief that the FACA and the implementing regulations require that meetings of the full negotiating committee as well as subgroups and caucuses be open to the public is likely to prevent effective negotiation.

  Open meetings increase risks to individual participants and may deter them *1705 from making concessions. On one occasion in the benzene negotiations, trade publication reports led to strong constituency reaction, impairing the authority of an industry representative. If representatives fear constituency reaction to tentative concessions, they will make concessions grudgingly. Such reluctance to move from initial postures makes fruitful negotiations difficult. [FN471]

  Some interests prefer keeping the threat of open negotiating sessions alive, however. As explained earlier, [FN472] closing meetings may have an adverse effect on the power of public interest representatives.

  The concerns about the FACA warrant an attempt to characterize the magnitude of the constraints imposed by the FACA. Three questions frame the characterization: (1) Is a regulatory negotiation group an 'advisory committee' within the meaning of the Act? (2) If it is, does the Act require that meetings of the group and of any subordinate bodies be open to the public? (3) If the meetings must be open to the public, does this represent a real threat to the negotiation?

  Answering the first question, regulatory negotiation groups probably are advisory committees under a literal interpretation of the FACA. It is difficult to argue that a regulatory negotiation group is not established by an agency 'in the interest of obtaining advice or recommendation.' [FN473] The agency, not having delegated its final authority to the negotiators, [FN474] must be seeking advice. This reasoning leads to the conclusion that the FACA is applicable. The FACA, however, is concerned with communication of advice to the agency and not how the advice is developed among private interests. Under this view, the FACA applies only to sessions at which the agency is present, not to negotiating sessions or caucuses from which the agency is absent. [FN475] There is some support for the position that an agency can structure a negotiated rulemaking in a manner that would permit a court to find that no advisory committee is involved, [FN476] but an agency could not be confident that a court would sustain the agency's position.

  Assuming a regulatory negotiation group is an advisory committee, the second question is whether the meetings must be open to the public. Here also, the answer is probably 'yes,' although arguments exist that meetings might be closed to protect effective deliberations, especially where the committee membership is balanced. [FN477]

  *1706 As previously discussed, GSA regulations permit meetings to be closed under certain conditions. Conceivably, a regulatory negotiation group could be structured to allow closed meetings by meeting with the agency only for the purpose of exchanging facts or information; initiating meetings rather than having the agency initiate meetings; and meeting with the agency only for the purpose of communicating 'the advice of individual attendees and not for the purpose of using the group to obtain consensus advice or recommendations.' Proceeding within these limitations, however, would distort the negotiation process, and would not entirely eliminate the possibility that a court would find that the spirit of the GSA regulations were violated, or that the regulations contravened the FACA.

  Even if full meetings of the negotiation group must be open, however, strong arguments exist that caucuses and subgroup meetings can be closed. [FN478] Support for this argument is found in National Anti-Hunger Coalition v. Executive Committee of the President's Private Sector Survey on Cost Control.  [FN479] The court found that subgroups formed by an advisory committee to provide information and recommendations for consideration to the committee were not themselves advisory committees. [FN480] It reached this conclusion because the subgroups were not established or utilized by the agency creating the full advisory committee. Instead, the subgroups were established and utilized by the advisory committee, thus taking them out of the literal definition of advisory committee in the FACA. [FN481] Moreover, the court reasoned that Congress could not have intended that

    interested parties . . . should have access to every paper through which recommendations are evolved, have a hearing at every step of the information-gathering and preliminary decision-making process, and interject themselves into the necessary underlying staff work so essential to the formulation of ultimate policy recommendations. [FN482]

  Additional support for the right to close meetings of subgroups of a negotiating committee is found in the Supreme Court's decision in FCC v. ITT World Communications. [FN483] There, the Court decided that the Sunshine Act [FN484] did not require that meetings between a panel of the FCC and foreign officials be open to the public. It concluded that Congress recognized that "informal background discussions that clarify issues and expose varying views' are a necessary part of an agency's work,' [FN485] and found that applying the requirements of the FACA to *1707 such discussions would 'impair normal agency operations without achieving significant public benefit.' [FN486]

  Assuming that at least some of the meetings of a regulatory negotiation group must be open, the final question is presented: does it matter? The conventional wisdom is that public meetings will chill the negotiation process. The experience of the benzene, flight and duty time, NCP, and pesticide exemption participants, however, suggests that the nominally open meetings made little difference to the negotiations. It is important to remember, however, that the FACA was interpreted in those negotiations as permitting closed meetings of subgroups and caucuses.

  There are two reasons for the difference between expectations and experience. The first is that more sensitive exploration of concessions almost always takes place away from the formal negotiating table. Secondly, no one from the press or the nonparticipating public made an issue of caucuses or subgroup meetings being closed, nor did the negotiations result in much publicity. If regulatory negotiation becomes more common, and is applied to highly controversial issues, press or public interest representatives probably will become more aggressive in reporting on negotiations and in insisting that the FACA be observed rigorously. This trend could result in greater impediments of the FACA to the negotiations process. On the other hand, practical ways of conducting business away from public meetings will almost certainly continue to exist.

H. REPRESENTATION ISSUES

  The APA does not address representation problems that arise in selecting participants for negotiated rulemaking. There are, of course, models that could be borrowed from the National Labor Relations Act [FN487] or the Railway Labor Act [FN488] or from the class certification procedures under the Federal Rules of Civil Procedure. [FN489] These models suggest that the following issues may need to be addressed when large numbers of persons or entities are to be represented in negotiations: (1) Whether the representative shall be the exclusive representative of the constituency. (2) If representation is to be exclusive, how the unit to be represented shall be defined. (3) How frequently the unit can be redefined and an election held for a new representative. (4) What duties the representative has in representing its constituents fairly.

  Despite the analytical usefulness of the labor law or class action models, important differences exist between representation problems in negotiated rulemaking and in collective bargaining or class action litigation.  [FN490] First, interests affected by agency rules usually already have some representation arrangements through trade associations, trade unions, or public interest groups. In the majority of negotiated rulemaking proceedings, these arrangements can be expected to *1708 work well, and there is no need to superimpose another process to designate representatives.

  Second, rule negotiations do not envision long term relations among the parties like collective bargaining. In this respect, the similarity is greater between negotiated rulemaking and class action litigation than between negotiated rulemaking and collective bargaining.

  Third, erecting a formal statutory mechanism for selecting representatives for rule negotiations could cause more harm than benefit by creating additional opportunities and incentives to litigate compliance with the representation procedures.

  Therefore, the most practical approach appears to be that adopted by the ACUS: encouraging agencies to be mindful of representation issues and to be creative in finding ways to solve anticipated representation problems.  [FN491] Agencies may of course wish to use the labor law or class action procedures as rough models to define options for dealing with negotiated rulemaking representation disputes, possibly even holding 'elections' through the notice and comment procedure.

IX. CONCLUSIONS AND RECOMMENDATIONS

  This part presents conclusions and recommendations derived from the analysis of negotiated rulemaking contained in the preceding eight parts. The structure is the same as that of ACUS Recommendation 85-5, which was based on the author's report to the conference.

A. SUMMARY

  Negotiations have proven effective in some circumstances in developing widely acceptable proposals for agency rules. Experience has shown that 'negotiated rulemaking' is a practical alternative to notice and comment or hybrid rulemaking, and that ACUS Recommendation 82-4 is basically sound. Some parts of the recommendation, however, need amplification. ACUS Recommendation 85-5 was warranted because interest in negotiated rulemaking was growing, and it is natural for interested persons to wonder how Recommendation 82-4 has worked in practice.

  It is important to view the original recommendations and Recommendation 85-5 as a guide to issues to be considered rather than as a formula to be followed. Negotiation is intrinsically a process that cannot be specified entirely in advance. [FN492] Accordingly, what will 'work' in a particular case depends on substantive issues, perception of the position of an agency by affected parties, past relationships among the parties, authority of party representatives in the negotiations, negotiating style of the representatives, number and divergence of views among individual units within each constituency represented, and skill of agency personnel and mediators.

  *1709 An agency cannot expect to transplant automatically and without modification the pattern followed sucessfully by another agency, or even by itself on another issue, to another negotiation. The ACUS properly urged that elements of Recommendations 82-4 and 85-5 be taken as a conceptual framework with which to design actual negotiations in a particular proceeding, and not as a formal model. [FN493]

  The provisions of Recommendation 82-4 relating to defining participation in negotiations proved sound. Who needs to be included at the outset is determined by the same criterion as who needs to agree with the negotiating result: whoever's opposition to the negotiating result is likely to be fatal--or at least to pose a risk unacceptable to the participants. Thus if X is unlikely to be able to affect the result, by litigating or inducing the OMB or the Congress to intervene, there is no need to include X in the negotiations; nor is it important whether X, having been included, agrees with the negotiation result. On the other hand, if X is likely to be able to affect the result significantly, there are stronger incentives to include X and to ensure that X agrees with the result of the negotiation. In other words, the need for X's agreement is a continuous positive function of X's power to litigate successfully in the rulemaking proceeding and in the courts. This, in turn is a function of how much X cares, X's resources, and the probable merits of X's position. [FN494]

  Representation should be addressed through an initial Federal Register notice, providing an opportunity for additional parties to request participation. This approach, used by the FAA and EPA, worked well. [FN495] The agency that sponsors negotiated rulemaking should specify an initial list of participating interests and representatives, allowing comment and requests for additional participation. The FAA experience shows the utility of this step. It assists in deflecting subsequent criticisms of the process.

  Recommendation 82-4 also proved basically sound with respect to defining consensus. As it suggests, agencies and negotiators should take a flexible view toward defining 'consensus.' Insisting on formal subscription to a total package recommendation may make negotiating success impossible; it may be infeasible for a participant to acknowledge agreement with a compromise.  [FN496] Especially if the agency participates in the negotiations, formal, total package agreement is not necessary. A sophisticated agency participant will have a sense, when negotiations are concluded, concerning what a given party can live with and what will be so obnoxious to that party that it will be motivated to litigate. The agency will also have a sense of what concessions are linked to gains on other issues. Moreover, a flexible attitude toward consensus may permit substantial agreement on a framework for a rule with certain difficult issues reserved for adversarial comment and decision by the agency. In other words, a consensus can *1710 include an agreement to dispute certain issues before the agency or in the courts. This outcome still provides the benefit of narrowing the issues. The FAA experience is a good example. On the other hand, a total package ground rule forces the parties to make more compromises, instead of leaving tough issues for the agency to resolve.

  A possible rule of thumb is to say that a consensus has been obtained when all participants agree informally that they will not actively oppose a particular resolution of issues, though certain of their constituents might register formal opposition.

  After the text of a rule is negotiated, the agency should publish it as an NPRM, allow public comment, and articulate its own rationale for the rule finally adopted. [FN497] APA review of a negotiated rule can be protected by negotiated agreement about what the parties will put in the rulemaking record and what the agency will say in support of the rule.

  The ACUS concluded that the four agencies' experiences did not show that the FACA, as interpreted by the sponsoring agencies and participants, was a serious impediment to effective negotiations. The purpose of the FACA is satisfied when a properly balanced rulemaking negotiation is conducted, and the statute should not impose additional requirements that jeopardize the success of negotiation. Under current judicial and agency interpretation of the FACA, caucuses and other working group meetings may be held in private where this is necessary to promoted an effective exchange of views. Agencies should not be deterred from considering negotiated rulemaking by a perception that the negotiating group will be an 'advisory committee' under the FACA. Accordingly, the ACUS made no further recommendation pertaining to the FACA in 1985. Uncertainty can be reduced, however, if the GSA amends its regulations to make it clear that meetings of caucuses and subgroups can be closed. [FN498] The recommended amendment is consistent with judicial interpretations of the FACA. [FN499]

  It is premature to consider amendments to the APA. The APA itself was a product of several decades of experimentation and evolution of concepts at the agency level, in the context of specific decisions. More experimentation and evolution of negotiated rulemaking concepts is necessary before a conclusion is appropriate whether amendments to the APA are desirable and, if they are, what form they should take. The risk of amending the APA now is that flexibility to adapt the negotiation process to the needs of different regulatory situations might be destroyed.

  An example of reduced flexibility is S. 451. [FN500] The bill contains a number of desirable provisions, especially those relating to coverage of the FACA and the procedure for closing negotiation group meetings. In other respects, however, it confines the negotiation process too narrowly, especially with respect to definition of consensus, [FN501] selection and role of mediators, [FN502] and continued activities of a negotiating group after it has reached consensus and made a recommendation *1711 to an agency.  [FN503]

B. RECOMMENDATION 85-5

    (1) An agency sponsoring a negotiated rulemaking should take part in the negotiations. Agency participation can occur in various ways. The range of possibilities extends from full participation as a negotiator to acting as an observer and commenting on possible agency reactions and concerns. Agency representatives participating in negotiations should be sufficiently senior in rank to be able to express agency views with credibility.

  This recommendation says that agencies should use negotiations, when they are undertaken, as the agencies' primary channel for communicating with the parties. This is the major lesson learned from the unsuccessful effort to negotiate a benzene health standard. The utility of negotiating is an inverse function of group influence outside the negotiations, through direct dealings with the agency, contact with the OMB, litigating, and lobbying with the Congress. The willingness of a group to participate meaningfully in negotiated rulemaking is likely to be affected by the group's perception of its ability to influence the content of the rule through other channels. A group that has a low estimate of its influence with the agency may embrace negotiations as a way of participating equally with other interests, an improvement of what it perceives its effect would be in more traditional notice and comment rulemaking. Conversely, a group that believes it has high influence may be more reluctant to participate meaningfully. Similarly, a group that has been successful in judicial challenges to agency decisions is more likely to prefer the litigation alternative to negotiations. In addition, a group perceiving itself as having substantial influence at the OMB or in the Congress is less likely to be enthusiastic about negotiating. The willingness to negotiate is of course not dichotomous; the factors that make a group less willing to participate may permit it to participate but not to be willing to make many changes in its position.

  Participation by the agency--and by the OMB [FN504]--reduces the real or perceived potential for parties to undermine the negotiating process by making 'end runs' to the agency or to the OMB. Agency participation also affords an opportunity for greater access to the agency than some parties might expect in hybrid rulemaking. Participation also increases the likelihood that the agency and the OMB will support, and understand the basis of, negotiated recommendations. *1712 Participation increases the range of options available regarding consensus. The same logic might militate in favor of congressional staff participation in some negotiations. If, for some reason, participation by an agency or the OMB is not acceptable, the facilitator/mediators should serve as a channel between the negotiations and the OMB and congressional staff.

  There are of course various forms that agency participation can take. In most cases, the agency should be represented on the committee and should ask questions of participants to gain substantive knowledge and also to probe for a party's 'soft spots' in its position. Experience suggests, however, that the agency should avoid pushing the participants toward one result or another. That kind of participation can stifle debate. There is a fine line between an agency's showing its hand too early and offering a proposal in an effort to stimulate movement in the negotiations and to remind participants what the prospects of deadlock may be.

  The early proponents of negotiated rulemaking recognized two forms of negotiated rulemaking, one in which the agency participates in the negotiations, the 'Agency Participation Model,' and another in which it does not, the 'Agency Oversight Model.' [FN505] The agency oversight model might be useful in some circumstances. The parties might be motivated so strongly to negotiate a resolution of their disagreements that agency participation is unnecessary. [FN506]

    (2) Negotiations are unlikely to succeed unless all participants (including the agency) are motivated throughout the process by the view that a negotiated agreement will provide a better alternative than a rule developed under traditional processes. The agency, accordingly, should be sensitive to each participant's need to have a reasonably clear expectation of the consequences of not reaching a consensus. Agencies must be mindful, from the beginning to the end of negotiations, of the impact that agency conduct and statements have on party expectations. The agency, and others involved in the negotiations, may need to communicate with other participants--perhaps with the assistance of a mediator or facilitator--to ensure that each one has realistic expectations about the outcome of agency action in the absence of a negotiated agreement. Communications of this character always should consist of an honest expression of agency actions that are realistically possible.

  This recommendation stresses that agencies should understand the major influence they have on party expectations and therefore on party BATNAs.  [FN507] This is a major lesson learned from the unsuccessful benzene negotiations. The utility of a party's participation is properly understood in terms of the likelihood of a negotiated settlement that is equal to or better than every party's BATNA; otherwise no consensus settlement is likely. This means that the agency must be willing to communicate selectively possible agency action that the particular party receiving the communication will find unfavorable. Just like a judge seeking to induce private litigants to settle a lawsuit, the agency, as decisionmaker, must point out the flaws in each party's argument and position, and it must *1713 ensure that each party understands why the outcome of agency action in the absence of negotiated agreement may not be as favorable as the party thinks.

  Agency participation in negotiations, addressed in recommendation number 1, promotes realistic party expectations. Agencies also can promote realistic party expectations by preparing draft rules or outlines and making them available to other parties. Instead of fearing premature disclosure of tentative staff views, the agency should welcome disclosure of such views, because they can educate affected interests on what the agency might do in the absence of a negotiated rule. In addition, such drafts can form a useful framework or agenda for negotiations.

  Such conditioning of party expectations by the agency is delicate; it can reinforce party fears that the agency will not participate in the negotiations in good faith, or can chill meaningful party contributions by signaling that the agency has its mind made up. Much conditioning of party expectation can be done by the facilitators, but facilitator representations regarding agency intentions will not be credible for long unless the agency acts to reinforce the representations.

    (3) The agency should recognize that negotiations can be useful at several stages of rulemaking proceedings. For example, negotiating the terms of a final rule could be a useful procedure even after publication of a proposed rule. Usually, however, negotiations should be used to help develop a notice of proposed rulemaking, with negotiations to resume after comments on the notice are received, as contemplated by paragraphs 13 and 14 of Recommendation 82-4.

  This recommendation encourages agencies considering negotiated rulemaking to be creative in exploring where negotiations best fit. The Harter model, and all of the completed negotiated rulemaking efforts, envision negotiations over the content of an NPRM. This is probably the best time for negotiations.

  Negotiations also can begin, however, after an NPRM is published. [FN508] In post-NPRM negotiations, the NPRM could sharpen the issues to be dealt with by the negotiators and could make certain options more concrete and less theoretical. [FN509] Post-NPRM negotiations are more vulnerable to APA ex parte communication criticisms, [FN510] but the agency can mitigate these concerns by putting minutes of post-NPRM negotiating sessions in the rulemaking record. When negotiations begin after a record is developed in an NPRM, through hearings or otherwise, the agency could sit down with participants to decide, 'what conclusions should we draw from this record?' In other words, the 'negotiation' would be a series of informal meetings with the parties to the hybrid rulemaking hearings. The relevant statutes do not preclude this procedure; the open meeting requirements of the FACA might be handled under the litigation exception, [FN511] and representation problems already might have been resolved in connection with the rulemaking hearings. [FN512]

    *1714 (4) The agency should consider providing the parties an opportunity to participate in a training session in negotiations skills just prior to the beginning of the negotiations.

  Training provided in connection with the EPA negotiations was well received, and parties to the negotiation of farmworker protection standards asked for similar training.

    (5) The agency should select a person skilled in techniques of dispute resolution to assist the negotiating group in reaching an agreement. In some cases, that person may need prior knowledge of the subject matter of the negotiations. The person chosen may be styled 'mediator' or 'facilitator,' and may, but need not, be the same person as the 'convenor' identified in Recommendation 82-4. There may be specific proceedings, however, where party incentives to reach voluntary agreement are so strong that a mediator or facilitator is not necessary.

  Virtually all of the participants in the four completed rule negotiations agreed that the mediators made essential contributions to the process. This recommendation strengthens the import of paragraph 10 in Recommendation 82-4,  [FN513] suggesting that mediators be an integral part of most rule negotiations.

  There are advantages and disadvantages of outside and inside facilitators. 'Inside' facilitators may be inhibited in dealing with intraconstituency problems and in intervening with other governmental agencies, such as the OMB. In addition, private parties may be reluctant to accept the neutrality of a facilitator from within the agency. On the other hand, the use of an inside facilitator in the pesticide negotiations worked well, and in appropriate cases inside facilitators may be effective.

  Mediation skills are highly personal, and it would be unwise to establish any exclusive institutional source of rulemaking mediation services. It may be desirable, however, to organize training and educational programs and to procure mediation services competitively, to ensure that opportunities for potential neutrals are distributed equitably. If competitive procurement occurs, however, it is essential to ensure speed and to have a process for ensuring that bidders are qualified by mediation experience.

    (6) In some circumstances, federal agencies such as the Federal Mediation and Conciliation Service or the Community Relations Service of the Department of Justice may be appropriate sources of mediators or facilitators. These agencies should consider making available a small number of staff members with mediation experience to assist in the conduct of negotiated rulemaking proceedings.

  Some agencies considering negotiated rulemaking have expressed concern that the process might cost more than notice and comment rulemaking. Recommendation 85-5 encourages agencies considering the cost of negotiated rulemaking to explore resources available from government agencies, such as mediation services from the Federal Mediation and Conciliation Service, the Community Relations Service within the Justice Department, and negotiation training available from the Department of Justice.

    (7) The agency, the mediator or facilitator, and, where appropriate, other participants *1715 in negotiated rulemaking should be prepared to address internal disagreements within a particular constituency. In some cases, it may be helpful to retain a special mediator or facilitator to assist in mediating issues internal to a constituency. Agencies should consider the potential for internal constituency disagreements in choosing representatives, in planning negotiations, and in selecting persons as mediators or facilitators. The agency should also recognize the possibility that a group viewed as a single constituency at the outset of negotiations may later become so divided as to suggest modification of the membership of the negotiating group.

  Constituency disagreements threatened both the FAA and the OSHA negotiations. The success of the FAA negotiation resulted, in part, from more timely resolution of such disagreements. It has long been recognized that the most difficult challenge to a negotiated agreement involves not the process at the negotiating table but the process of resolving intraconstituency disagreements away from the table. This recommendation encourages agencies to anticipate intraconstituency problems. This does not mean that the agency or the private interests know in advance exactly how to 'resolve' such disagreements--only that the convenors assess the likelihood that such disagreements will occur and understand the adequacy of existing institutional mechanisms for resolving them.

  Intraconstituency problems may be more difficult under certain types of interest representation arrangements than others. For example, trade unions exist for the purpose of aggregating employee interests and therefore are experienced in resolving differing positions within the constituency. Similarly, public interest groups have a certain authoirity in speaking for their otherwise diffuse constituencies. Business interests, in contrast, usually have fewer established mechanisms for resolving internal differences.

  If such internal differences are expected, and existing institutions are not well suited for resolving them, negotiation should not proceed in the absence of a mediator who has the experience, skill, and acquaintance with the constituency and its major personalities. This mediator may be able to reduce intraconstituency differences, thereby validating the authority of the spokesperson for the affected group. In some cases, a special mediator might be retained to assist the main mediator in dealing with an internal constituency problem.

  The potential for internal constituency disagreements also should influence selection of interest representatives. If a trade association may be unable to represent its constituents effectively, it may be desirable to include individual company representatives. It also may be desirable to convene subgroups of major subinterests to provide the mediator or facilitator with a forum to adjust internal constituency disagreements.

  Agencies and mediators or facilitators should be wary, however, of fragmenting representation too much. Permitting subinterests to be represented separately merely moves disagreements among these subinterests from other resolution forums to the bargaining table itself.

    (8) Where appropriate, the agency, the mediator or facilitator, or the negotiating group should consider appointing a neutral outside individual who could receive confidential data, evaluate, it, and report to the negotiators. The parties would need to agree upon the protection to be given confidential data. A similar *1716 procedure may also be desirable to permit neutral technical advice to be given in connection with complex data.

  Use of this neutral individual could alleviate concerns with disclosure of proprietary information to competitors or to the public that otherwise might frustrate negotiating progress. Similarly, microcomputers and appropriate software may be helpful to the negotiators in evaluating the implications of different approaches, as they were in the NCP negotiations.

    (9) Use of a 'resource pool' may be desirable to support travel, training, or other appropriate costs incurred by participants or expended on behalf of the negotiating group. The feasibility of creating this pool from contributions by private sources and the agency should be considered in the prenegotiation stages.

  The resource pool concept iscontroversial for several reasons. Public interest organizations have expressed to the author the view that their representatives should be compensated for their time when they participate in rule negotiations. Other interest representatives strongly oppose this idea. Agency personnel are wary of delegating to private parties authority for deciding how public funds should be spent, in part because of the need to comply with legal restrictions on expenditure of public funds. On the other hand, costs are incurred by participants in negotiated rulemaking, and it seems fair to reimburse participants at least for out-of-pocket travel expenditures and to provide a flexible mechanism for paying for necessary analytical efforts undertaken by the negotiating committee. The ACUS recommendation on resource pools was deliberately general, to raise the issue without taking a position on the controversial aspects.

C. FUTURE ACUS ACTIVITIES

  The foregoing recommendations were adopted by the ACUS in response to the author's report. The report also made the following recommendations to the ACUS regarding its own activities.

    (1) The ACUS should investigate means of resolving intraconstituency disagreements, with particular attention to difficulties experienced by the API in the benzene negotiations, and by the ATA in the flight and duty time negotiations.

  Negotiations are unlikely to be successful if serious internal disagreements exist within a represented interest and cannot be resolved. Such disagreements were the focus of trade association and mediator effort in the OSHA and FAA negotiations, with differing success. This subject is worth further investigation to identify mediation or representation approaches that facilitate resolution of intraconstituency differences.

    (2) As the OSHA pursues its initiative to negotiate an MDA standard, the ACUS should monitor the activity to permit detailed comparison with benzene negotiations.

  Many of the characteristics of the benzene negotiations appear to be unique to health standards controversies involving the AFL-CIO and major industry groups. The OSHA has involved these groups under a different negotiating procedure than was used in the benzene negotiations, to develop a standard for 4,4'-Methylendedianiline (MDA). [FN514] When the MDA effort has run its course, it can *1717 be compared with the benzene effort.

    (3) ACUS should investigate use of microcomputers for modeling and telecommunications in connection with rule negotiations.

  A microcomputer was used by the NCP negotiators to model alternatives considered in the negotiations. The negotiators found this computer aid helpful. Negotiators in another negotiation of the EPA involving farmworker protection have arranged to use microcomputers for telecommunicating documents. This technology appears to afford significant benefits for a variety of rule negotiations. The ACUS should develop information on useful applications of microcomputers for future efforts at negotiated rulemaking.

D. CONCLUSION

  Negotiated rulemaking is a credible alternative to traditional adversarial procedures before administrative agencies and thus has become a respectable part of the search for 'regulatory reform.' A growing list of agencies are considering its use in connection with a variety of rule changes. [FN515] Even when negotiation fails to produce consensus among participating interests, it narrows disputes and informs the sponsoring agency better than development of a record through litigation before the agency. [FN516] While no informed commentator proposes rule negotiation as the exclusive means of quasi- legislative administrative decisionmaking, the process is a promising innovation recognizing that policymaking is a political process rather than a quasi-judicial trial. This recognition is long overdue in American administrative law.

  In using negotiated rulemaking effectively, it is important for agencies and students of the administrative process to understand the dynamics of negotiation in the regulatory context, to understand how negotiation meets the constraints of administrative law, to think hard about the ideas embodied in the ACUS recommendations and to be sophisticated about creating incentives for interest groups to resolve their own differences rather than advocating rigid positions for agencies and courts to sort out.

  The analysis of negotiation dynamics, interest group behavior, and regulatory program characteristics in this article is meant to offer a conceptual framework within which this understanding can be sought. Evaluation of regulatory improvement alternatives such as negotiated rulemaking can be enriched further by synthesizing from state and local administrative experiences as well as federal agency experiences. [FN517]

[FNa1]. Professor of Law, Villanova University. S.B. 1966; S.M. 1970, Massachusetts Institute of Technology; J.D. 1975, Georgetown University Law Center.

[FN1]. The Administrative Conference of the United States (ACUS) was established to improve the administrative procedure of federal agencies. The President appoints a full-time chairman to the ACUS for a five year term. The council, which is the executive board, consists of the chairman and 10 other members appointed by the President for three year terms. Only half of those 10 members may be drawn from federal agencies. The total membership of the ACUS may not exceed 91 members or be fewer than 75 members. In addition to the council, the membership comprises 44 government members (heads of agencies or their designees) from 36 agencies and approximately 36 nongovernmental members appointed by the chairman, with the council's approval, for two year terms. The council must call at least one plenary session annually. The membership meeting in plenary session is called the Assembly of the Administrative Conference. Only the chairman is entitled to compensation for his services. Administrative Conference Act, 5 U.S.C. § §  571-576 (1982).

[FN2]. 47 Fed. Reg. 30,708 (1982) (codified at 1 C.F.R. §  305.82-4  (1986)).

[FN3]. 5 U.S.C. § §  551-559, 701-706 (1982).

[FN4]. The following table summarizes the four completed negotiations:

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE 

TABULAR OR GRAPHIC MATERIAL SET FORTH AT THIS POINT IS NOT DISPLAYABLE 

When this article went to press other rule negotiations were underway. See parts VI.D & F.

[FN5]. ACUS Recommendation 85-5, 50 Fed. Reg. 52,893, 52,894  (1985) (codified at 1 C.F.R. §  305.85-5 (1986)).

[FN6]. H. PERRITT, ANALYSIS OF FOUR NEGOTIATED RULEMAKING EFFORTS  (1985) (available from Administrative Conference of the United States). Abbreviated summaries of the report appeared in Perritt, Negotiated Rulemaking in Practice, 5 J. POL'Y ANALYSIS AND MGMT. 482 (1986); Perritt, Negotiated Rulemaking and Administrative Law, 38 ADMIN. L. REV. 471 (1986).

[FN7]. Harter, Negotiating Regulations: A Cure for Malaise, 71 GEO. L.J. 1 (1982).

[FN8]. The benzene negotiations are reviewed in greater detail than the other negotiations because the author interviewed most of the participants and because these negotiations have not been reported elsewhere.

[FN9]. Convenors and mediators play different roles in negotiated rulemaking. Convenors assist agencies in deciding whether a particular subject is suitable for negotiated rulemaking. They determine who are affected interests and informally contact those interests to explore their desire to participate in a negotiation. 1 C.F.R. §  305.82-4 para. 4 (1986). Mediators become involved in the actual negotiation, helping the parties to agree on a procedure and to narrow their substantive differences. 1 C.F.R. §  305.84-4 para. 10; 1 C.F.R. §  305.85-5 para. 5 (1986). 'Facilitator' is just another name for a mediator. A facilitator or mediator may help to reduce anxiety experienced by those who object to the idea of using a labor-management negotiation model for administrative rulemaking.

[FN10]. 5 U.S.C. app. § §  1-15 (1982). The FACA is designed to assure public access to the deliberations of advisory committees to federal agencies and is discussed in greater detail in part VIII.G of this article.

[FN11]. The difference between interests disputes, which are legislative in nature, and rights disputes, which are adjudicatory in nature, is discussed infra notes 21-28 and accompanying text. An additional technique for resolving interest disputes deserves brief comment. When a class action lawsuit against agency action or inaction is filed, the Federal Rules of Civil Procedure provide the opportunity to accomplish some of what can be accomplished in negotiated rulemaking. There are, however, several reasons why negotiations within the class action framework are much less satisfactory than negotiated rulemaking under the ACUS recommendations. The class action approach requires that a lawsuit be filed, and that the plaintiffs or intervenors represent a sufficiently broad range of interests for a negotiated settlement to be fair and successful. Moreover, while federal judges presumably are permitted by recent amendments to rule 16 of the Federal Rules of Civil Procedure to consider negotiations within a class action framework, there is no guarantee that a particular federal judge would.

[FN12]. See ACUS Recommendation 86-3, 51 Fed. Reg. 25641, 25643  (1986) (proposing greater agency use of arbitration, mediation, fact finding, and mini-trials).

[FN13]. 'Regulatory negotiation' refers to use of negotiation in any decisionmaking process by an administrative agency. 'Negotiated rulemaking' is a specific application of regulatory negotiation, the use of negotiation in the rulemaking process. This article concentrates on negotiated rulemaking. The Administrative Conference also sponsors work on other forms of regulatory negotiation. See A. ADMAS & J. FIGUEROA, EXPEDITING SETTLEMENT OF EMPLOYEE GRIEVANCES IN THE FEDERAL SECTOR (1985) (Report to the Administrative Conference of the United States) (evaluation of arbitration to settle federal employee appeals).

[FN14]. Terminology is a problem in comparing negotiated rulemaking to more traditional processes. This article uses the term 'adjudication' to refer to an adversarial decisionmaking process in which parties present their views to the agency by examining witnesses and qualifying documents through something resembling formal rules of evidence. 'Hybrid rulemaking' is a procedure imposed by statute or by the courts under the APA in which adjudicatory procedures are used in the rulemaking process, though the type of decisionmaking involved would be termed 'rulemaking' and not 'adjudication' under the APA. Hybrid rulemaking is discussed more fully in part VIII.D of this article.

[FN15]. See L. FULLER, THE PRINCIPLES OF SOCIAL ORDER 172-87  (1981) (contrasting private negotiation with governmental decisionmaking). See generally, Harter & Eads, Policy Instruments, Institutions, and Objectives: An Analytical Framework for Assessing 'Alternatives' to Regulation, 37 ADMIN. L. REV. 221 (1985) (many different mechanisms, including governmental standards, criminal law, civil liability, labor relations, can change regulated behavior).

[FN16]. See S. BREYER, REGULATION AND ITS REFORM 15-35 (1982) (discussing rationales for regulation).

[FN17]. See J. DUNLOP, DISPUTE RESOLUTION: NEGOTIATION AND CONSENSUS BUILDING 1-22 (1984) (describing and comparing private and political negotiation); J. HURST, DEALING WITH STATUTES 19 (1982) (legislature is basically arena for bargaining); V.O. KEY, POLITICS, PARTIES AND PRESSURE GROUPS 7 (1942) (politics is struggle among groups or interests); Dunlop, The Negotiations Alternative in Dispute Resolution, 29 VILL. L. REV. 1421 (1984) (describing negotiation process).

[FN18]. See generally Jaffe, The Illusion of the Ideal Administration, 86 HARV. L. REV. 1183, 1188 (1973) (describing pervasiveness of political compromise in context of administrative regulation).

[FN19]. Some degree of accommodation occurs with appointment of administrative agency policymaking personnel and in legislative oversight of agencies. Nevertheless, agencies are accountable to the public only indirectly, and are subject to political forces only when the press or the public takes an interest in an issue before an agency, or through occasional intervention by the legislature or the executive.

[FN20]. The breadth of the interests an agency must deal with is a function of the type of program within the agency's responsibility. Part II.B.3 of the article discusses the characteristics of regulatory programs.

[FN21]. See generally Stewart, The Reformation of American Administrative Law, 88 HARV. L. REV. 1667 (1975).

[FN22]. Id. at 1670.

[FN23]. Id. at 1671; see also Jaffe, supra note 18, at 1184, 1188  (1973) (criticizing model based on assumption that Congress establishes an objective that 'is capable of disinterested and nonpolitical administration' and suggesting political model).

[FN24]. See Perritt, 'And the Whole Earth Was of One Language': A Broad View of Dispute Resolution, 29 VILL. L. REV. 1221, 1221-29 (1984) (distinguishing rights and interest disputes).

[FN25]. See Fuller, The Forms and Limits of Adjudication, 92 HARV. L. REV. 353, 369 (1978) (discussing adjudication in terms of claim of rights).

[FN26]. Id. at 368.

[FN27]. Part VIII.D. of the article discusses hybrid rulemaking. The point is not that 'adjudication' under § §  556 and 557 of the APA is used in rulemaking. Instead, the point is that certain trial-type procedures associated with adjudication have been superimposed on informal rulemaking conducted under §  553 of the APA.

[FN28]. See Harter, Dispute Resolution and Administrative Law: The History, Needs, and Future of a Complex Relationship, 29 VILL. L. REV. 1393 (1984) (discussing development of administrative agencies).

[FN29]. Stewart, supra note 21, at 1760-89.

[FN30]. Id.

[FN31]. Id. at 1789-90.

[FN32]. Id. at 1790.

[FN33]. Id.

[FN34]. Id.

[FN35]. Id. at 1791.

[FN36]. See Amalgamated Meatcutters v. Connally, 337 F. Supp. 737 (D.D.C. 1971) (finding broad delegation to Wage and Price Control program constitutional).

[FN37]. Dunlop, The Limits of Legal Compulsion, 27 LAB. L.J. 67, 72  (1976).

[FN38]. 49 U.S.C. § §  1601-1618 (1984).

[FN39]. Dunlop, supra note 37, at 72.

[FN40]. Part VI.B of the article discusses the coke oven standard in greater detail.

[FN41]. Letter from Richard B. Stewart to Henry H. Perritt, Jr., March 10, 1976. This excerpt refers to the absence of institutional mechanisms for interest aggregation. Part II.B.2 of the article discusses the characteristics of interest groups. The author of this article was Deputy Under Secretary of Labor during Dunlop's tenure as Secretary.

[FN42]. See HEARING ON H.R. 746, REGULATORY PROCEDURE ACT OF 1981, BEFORE THE SUBCOMM. ON ADMINISTRATIVE LAW AND GOVERNMENTAL RELATIONS OF THE HOUSE COMM. ON THE JUDICIARY, 97th Cong., 1st Sess. 683-84 (1982) (testimony of C. Boyden Gray, Counsel to the Vice President, advocating meetings between agencies and affected interests as part of rulemaking process).

[FN43]. Harter, supra note 7.

[FN44]. Id. at 1-11.

[FN45]. See generally S. BREYER & R. STEWART, ADMINISTRATIVE LAW AND REGULATORY POLICY 592-607 (2d ed. 1985) (discussing development of hybrid rulemaking).

[FN46]. Harter, supra note 7, at 14.

[FN47]. Id. at 16.

[FN48]. Id. at 25-42. Federal environmental statutes have been written in many cases to facilitate the use of federal court litigation to promote negotiated settlement of controversies between administrative agencies and private interests. 'By enacting the sixty-day notice provision, Congress intended to encourage resolution of disputes outside the courts, and to secure the participation of the EPA in certain suits.' Garcia v. Cecos Int'l, Inc. 761 F.2d 76, 98 (1st Cir. 1985) (quoting ADA-Cascade Watch Co. v. Cascade Resource Recovery, Inc., 720 F.2d 897, 908 (6th Cir. 1983) (Merritt, J., dissenting)). See also United States v. Olin Corp., 606 F. Supp. 1301, 1303 (N.D. Ala. 1985) (describing consent decree in suit over DDT pollution). Harter might also have noted negotiations within 'rate bureaus' under the Interstate Commerce Act, and the experience of the Federal Railroad Administration with air brake regulations. In 1982, the Federal Railroad Administration amended rules pertaining to railroad air brakes. The specific changes adopted resulted from a notice of proposed rulemaking (NPRM) adopted at the joint request of the Association of American Railroads--the main industry trade association--and the Railroad Labor Executives Association--a confederation of labor organizations representing rail employees. 47 Fed. Reg. 36,792 (1982). Comments submitted in response to the NPRM generally were supportive, resulting in a final rule reflecting the labor-industry agreement. Id.

[FN49]. Harter, supra note 7, at 42-61.

[FN50]. S. 1601, 97th Cong., 1st Sess., 127 CONG. REC. 19,771 (1981). Harter drafted the bill pursuant to a request from Sen. Roth's staff. Sen. Roth apparently introduced S. 1601 to illustrate an alternative to the regulatory reform proposals then under consideration in Congress. Most of the other proposals concentrated on making rulemaking more adjudicatory in character.

[FN51]. 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82-4 (1986)).

[FN52]. See generally Jaffe, supra note 17, at 1183 (describing pervasiveness of political compromise in context of administrative regulation).

[FN53]. R. FISHER & W. URY, GETTING TO YES (1981).

[FN54]. Some simple examples are useful. If I am considering buying a car from you and you know I can get the identical car from your competitor across the street for $10,000, I will negotiate with you only if I think our negotiations may produce agreement on a price less than $10,000. The $10,000 is my 'Best Alternative to Negotiated Settlement' (BATNA). Alternatively, if I am a plaintiff in a personal injury lawsuit and the expected net value of a jury verdict to me is $25,000, I will negotiate a settlement with you, the defendant, only if the settlement is worth more to me than $25,000. My BATNA in these negotiations is $25,000.

[FN55]. What the agency will do will be affected by what courts, the President, and Congress will do. Consequently, perceptions of these influences are also important.

[FN56]. See Perritt, supra note 24, at 1256 (settlement via dispute resolution more likely when parties agree on probable trial result).

[FN57]. Agency persuasion of the type suggested in the text tends to promote similar party perceptions of unilateral agency decision, or to make parties more pessimistic. Both similar perceptions and pessimistic perceptions make BATNAs less attractive and negotiation relatively more attractive.

[FN58]. Dunlop, supra note 17, at 1433 (referring to the need for three agreements to conclude a negotiation).

[FN59]. See National Labor Relations Act §  9(a), 29 U.S.C. §  159(a) (1982) (exclusive representative); Railway Labor Act §  2, Ninth, 45 U.S.C. §  152, Ninth (1982) (same). See generally T. SULLIVAN, RESOLVING DEVELOPMENT DISPUTES THROUGH NEGOTIATIONS 39-62 (1984) (contrasting structure for labor representation with lack of structure in public dispute negotiation).

[FN60]. This occurred with the American Petroleum Institute and and the Air Transport Association in the benzene and flight and duty time negotiations, respectively.

[FN61]. V.O. KEY, supra note 17, at 7 (politics is struggle among groups or interests). See generally Sunstein, Interest Groups in American Public Law, 38 STAN. L. REV. 29 (1985); Cass, Models of Administrative Action, 72 VA. L. REV. 363 (1985).

[FN62]. V.O. KEY, supra note 17, at 7-8.

[FN63]. See generally J. HURST, supra note 17, at 1-29. Prof. Hurst discusses the limitations that resources, diffusion of interest, and other factors may present in obtaining legislative response. He suggests that legislatures are fundamentally institutions that provide broad arenas for bargaining among diverse interests. Id. at 19. Most people interest themselves in the legislative process, however, only when a matter under consideration has material importance for them. Id. at 23.

[FN64]. V.O. KEY, supra note 17, at 203 (referring to representation function performed by interest groups).

[FN65]. J. LOGSDON, THE DECISION TO GO TO THE MOON: PROJECT APOLLO AND THE NATIONAL INTEREST 181 (1970).

[FN66]. R. DAHL, A PREFACE TO DEMOCRATIC THEORY 48-50 (1956).

[FN67]. Id. at 38.

[FN68]. Id. at 44.

[FN69]. Id. at 42-43.

[FN70]. V.O. KEY, supra note 17, at 203.

[FN71]. Id. at 203-04.

[FN72]. Stewart, The Discontents of Legalism: Interest Group Relations in Administrative Regulation, 1985 WIS. L. REV. 655, 674.

[FN73]. Id. at 674. Organizational entrepreneurship refers to the tendency of organizers of public groups to seek publicity and political support necessary to build and maintain their organizations. Id.

[FN74]. This perception existed in the benzene negotiations. Part III.N of the article discusses how this perception contributed to lack of agreement on a rule.

[FN75]. The costs of trying a complex factual issue on remand to the agency may be high after a victory in the appellate courts.

[FN76]. For example, in the case of the FAA negotiations, the Aviation Consumer Action Project (ACAP) dismissed a lawsuit against the FAA once it appeared to the ACAP that the FAA was serious about negotiating and that negotiating would lead to a rule faster than litigation, which would have resulted at most in an order to the agency to promulgate a rule by a certain date.

[FN77]. Stewart, supra note 72, at 671-72.

[FN78]. V.O. KEY, supra note 17, at 201-02.

[FN79]. The role of intensity in political interaction is summarized in the text accompanying notes 66-69.

[FN80]. See Wilson, The Politics of Regulation, in THE POLITICS OF REGULATION 357, 366-67 (J. Wilson ed. 1980) (classifications based on group perception of costs and benefits).

[FN81]. See Stewart, supra note 72, at 670.

[FN82]. Harter, supra note 7, at 67-82.

[FN83]. Id. at 82-92.

[FN84]. Id. at 92-97.

[FN85]. Id. at 97-99.

[FN86]. Id. at 99-102.

[FN87]. Id. at 102-07.

[FN88]. The congruence between the Harter criteria and the ACUS recommendations is not surprising: Harter developed the recommendation under contract to the ACUS.

[FN89]. Harter, supra note 7, at 42-43.

[FN90]. Harter uses the term 'relative power.' The idea is the same as the BATNA concept discussed in part II.B.1 of the article.

[FN91]. Harter, supra note 7, at 43-44. In the referenced discussion, Harter talked about the acceptability of any form of dispute resolution. The textual discussion here applies his ideas to the acceptability of negotiation in particular.

[FN92]. See Eisenberg, Private Ordering Through Negotiation: Dispute Settlement and Rulemaking, 89 HARV. L. REV. 637, 639 (1976) (use of precedent in negotiations).

[FN93]. The use of norms does not automatically establish the use of a formal process. Examples are given of 'relationships, such as parent-child, union-management, . . . [which] exhibit features of dependency and/or intimacy that inhibit the parties from seeking resolution through the intervention of a third party.' Harter, supra note 7, at 44 n.245.

[FN94]. A party that has broad based public appeal may wish for the decision to be made by public opinion. Another party with access to skilled lawyers may prefer complex judicial proceedings. Access to such alternative forums or processes enhances relative power in a negotiating forum.

[FN95]. Id. at 44. Part II.B.1 of the article considers in more detail the determinants of whether negotiation is an attractive alternative to a party.

[FN96]. These criteria stress the use of negotiation as a tool to 'produce a sound regulation or to facilitate the regulatory process.' Harter, supra note 7, at 44.

[FN97]. Id. at 44-45 & n.250.

[FN98]. Id. at 45-46. This corresponds to p 4(e) of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82- 4 (1986)).

[FN99]. For example, negotiation would not effectively resolve a conflict between a power company that seeks to build a dam, town residents who support the power company, and a small group of environmentalists who oppose the dam. The environmentalists would bring little power to the table. If the dispute resolution was carried out in a courtroom or a congressional committee, however, the environmentalists would have the advantage of procedural safeguards. The presence of these safeguards would ensure the protection of their interests--provided they have sufficient persuasive power, whether legal analysis or political clout.

[FN100]. Harter, supra note 7, at 46. This corresponds to p 4(c) of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82-4 (1986)).

[FN101]. Pure negotiation means nonrepresentative negotiation. Each individual represents his or her own interests and all the individuals gather together to negotiate. In fact, however, tax bills are produced by a representative negotiation, in which a few individuals (legislators) represent many interests. Nonrepresentative negotiation is distinguished by the absence of interest aggregation mechanisms.

[FN102]. Fifteen parties is suggested as the 'rough practical limit,' if for no other reason than 'it seems difficult to get more people than that around a table in reasonable comfort.' Harter, supra note 7, at 46 & n.257.

[FN103]. Id. at 47. This corresponds to p 4(a) of ACUS Recommendation 82- 4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82-4 (1986)).

[FN104]. Examples of this would be lobbying for public support, media support, etc. Harter, supra note 7, at 47 & n.259.

[FN105]. See J. LOGSDON, supra note 65, at 81 (issue maturity determining factor in success of political decision).

[FN106]. Harter, supra note 7, at 47.

[FN107]. Id. at 48. This is addressed in p 4(a) of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82- 4 (1986)).

[FN108]. See Dunlop, supra note 17, at 1436 (deadline serves vital function in negotiations).

[FN109]. Harter, supra note 7, at 47-48 & n.263.

[FN110]. Id. at 48-49. This corresponds to p 4(e) of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82- 4 (1986)).

[FN111]. Harter, supra note 7, at 48 n.284.

[FN112]. Perritt, supra note 24, at 1232-35. Part II.B.1 of the article explains the BATNA concept.

[FN113]. Harter, supra note 7, at 49.

[FN114]. Id. at 49-50. This corresponds to p 4(b) of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82- 4 (1986)).

[FN115]. For example, the establishment of a public school curriculum through negotiation would be difficult if the only participants were a fundamentalist Baptist preacher and an equally committed evolutionist.

[FN116]. Harter, supra note 7, at 49-50.

[FN117]. Id. at 50. This corresponds to p 4(d) of ACUS Recommendation 82- 4, 47 Fed. Reg. 30,708, 30,709 (1982) (codified at 1 C.F.R. §  305.82-4 (1986)).

[FN118]. Harter, supra note 7, at 50 n.274.

[FN119]. Id. at 50-51. This criterion is not addressed explicitly in ACUS Recommendation 82-4.

[FN120]. See Perritt, supra note 24, at 1253-56 (explaining how similar perceptions of probable outcomes make negotiated resolution more likely). Uncertainty can improve the attractiveness of negotiated resolution to risk averse parties, but similar predictions of outcome probably are more important. Id.

[FN121]. Harter, supra note 7, at 51 n.276.

[FN122]. Id. at 51. This criterion is addressed in part by pp 13-14 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,710 (1982), (codified at 1 C.F.R. §  305.82-4 (1986)), proposing publication of a negotiated rule in the Federal Register and giving the negotiators an opportunity to review comments submitted by nonparticipants.

[FN123]. Harter, supra note 7, at 57-58.

[FN124]. See Note, Rethinking Regulation: Negotiation as an Alternative to Traditional Rulemaking, 94 HARV. L. REV. 1871, 1875-76 (1981) (distinguishing between 'Agency Oversight Model,' in which agency does not participate, and 'Agency Participation Model,' in which it does).

[FN125]. Harter, supra note 7, at 92.

[FN126]. Id. at 93.

[FN127]. Paragraph 11 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,710 (1982) (codified at 1 C.F.R. §  385.82-4 (1986)), addressed consensus in very general terms: 'Consensus . . . means that each interest represented in the negotiating groups concurs in the result, unless all members of the group agree at the outset on another definition.' That paragraph contemplated, however, that the negotiators might issue a report indicating agreement on some issues and disagreement on other issues.

[FN128]. The other three negotiations had been studied by other investigators when the author began work for the Administrative Conference. This article, drawing on the research for the ACUS, relies to a greater extent on the completed work of the other investigators to explore the FAA and EPA negotiations.

[FN129]. American Petroleum Inst. v. OSHA, 581 F.2d 493 (5th Cir. 1978), aff'd sub nom. Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607 (1980).

[FN130]. 50 Fed. Reg. 50,512 (1985).

[FN131]. Brief for the Secretary of Labor at 7, In Re United Steelworkers of Am., 783 F.2d 1117 (D.C. Cir. 1986).

[FN132]. This standard was established under §  6(a) of the Occupational Safety and Health Act (the Act), 29 U.S.C. §  655(a) (1982).

[FN133]. A time weighted average is an average calculated by multiplying values of a variable by the length of time a particular value existed, and then dividing by the total time period. For example, if the atmospheric concentration of a substance was 5 parts per million for 10 minutes and 2 parts per million for 20 minutes, the time weighted average would be [(5 x 10) + (2 x 20)]/30, or 3 parts per million.

[FN134]. 43 Fed. Reg. 5,919 (1978).

[FN135]. The report was prepared under §  22(d) of the Act, 29 U.S.C. §  669(d) (1982).

[FN136]. Id.

[FN137]. Emergency Temporary Standard for Occupational Exposure to Benzene; Notice of Hearing, 42 Fed. Reg. 22,516 (1977).

[FN138]. Id.

[FN139]. Emergency Temporary Standard, 42 Fed. Reg. at 27,465.

[FN140]. Industrial Union Dep't v. Bingham, 570 F.2d 965, 966 (D.C. Cir. 1977); American Petroleum Inst. v. OSHA, 581 F.2d 493 (5th Cir. 1978), aff'd sub nom. Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607 (1980).

[FN141]. American Petroleum Inst. v. OSHA, 581 F.2d at 500.

[FN142]. Id. at 510.

[FN143]. 42 Fed. Reg. 27,452 (1977).

[FN144]. Occupational Exposure to Benzene, 42 Fed. Reg. 32,263 (1977).

[FN145]. There were two groups of petitioners: a 'producers group' led by the American Petroleum Institute and a 'users group' led by the Rubber Manufacturers Association (RMA). Some member companies of the RMA filed in the Sixth Circuit, while others filed in the Second Circuit. These cases were transferred to the Fifth Circuit in March, 1978. American Petroleum Inst. v. OSHA, 581 F.2d 493, 499 (5th Cir. 1978), aff'd sub nom. Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607 (1980). Some of the positions taken in the multidistrict litigation were reflected subsequently in negotiating positions.

[FN146]. American Petroleum Inst. v. OSHA, 581 F.2d at 504-05.

[FN147]. Id. at 502.

[FN148]. Id.

[FN149]. Id.

[FN150]. Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607, 662 (1980).

[FN151]. Id. at 630-38.

[FN152]. Id.

[FN153]. Id. at 639-46.

[FN154]. Id. at 658-59.

[FN155]. Id. at 646-52.

[FN156]. Id. at 653.

[FN157]. Id.

[FN158]. Id. at 663 (Burger, C.J., concurring).

[FN159]. Id. at 672-88 (Rehnquist, J., concurring).

[FN160]. Id. at 695 (Marshall, J., dissenting) (citing 29 U.S.C. §  655(f) (1982)).

[FN161]. Id. at 688-94.

[FN162]. Brief for the Secretary of Labor at 7-8, In Re United Steelworkers of Am., 783 F.2d 1117 (D.C. Cir. 1985).

[FN163]. Id. at 7 (quoting from Thomas G. Auchter letter to Sidney M. Wolfe). A proposed standard was not published until December, 1985. 50 Fed. Reg. 50,512 (1985).

[FN164]. Part VI.B. of the article discusses the coke oven negotiations.

[FN165]. Philip Harter and Gerald Cormick served as convenors and facilitators.

[FN166]. This is known as the 'exceedence problem.'

[FN167]. 29 U.S.C. §  653(b)(4) (1982).

[FN168]. The Rubber Manufacturers Association and the United Rubber Workers reported their agreement to the OSHA in a letter on Feb. 16, 1984.

[FN169]. This satisfied Harter's seventh criterion (permitting tradeoffs). See Harter, supra note 7, at 50, and p 4(d) of ACUS Recommendation 82-4. 47 Fed. Reg. at 30,709.

[FN170]. This satisfied Harter's third criterion (issue maturity). See Harter, supra note 7, at 47, and p 4(a) of ACUS Recommendation 82-4. 47 Fed. Reg. at 30,709.

[FN171]. This satisfied Harter's second criterion (limited number of parties). See Harter, supra note 7, at 46, and p 4(c) of ACUS Recommendation 82-4. 47 Fed. Reg. at 30,709.

[FN172]. This satisfies Harter's first criterion (countervailing power). See Harter, supra note 7, at 45, and p 4(e) of ACUS Recommendation 82-4. 47 Fed. Reg. at 30,709.

[FN173]. This violates Harter's sixth criterion (fundamental values). See Harter, supra note 7, at 49-50, and p 4(b) of ACUS Recommendation 82-4. 47 Fed. Reg. at 30,709.

[FN174]. This made it difficult for Harter's seventh criterion (tradeoff), see Harter, supra note 7, at 50, and p 4(d) of ACUS Recommendation 82-4, to be realized in fact. 47 Fed. Reg. at 30,709.

[FN175]. This violates Harter's eighth criterion (research not determinative of outcome). See Harter, supra note 7, at 50-51.

[FN176]. No standard was issued until more than a year after the negotiations adjourned. 50 Fed. Reg. 50,512 (1985).

[FN177]. This muddied Harter's fifth criterion (opportunity for gain). See Harter, supra note 7, at 48-49.

[FN178]. 448 U.S. 607 (1980).

[FN179]. This implicates Harter's ninth criterion (implementation). See Harter, supra note 7, at 51.

[FN180]. Actually the OSHA covered less than 50% of the mediators' costs. The remainder was covered by neutral nongovernment sources and by the mediators themselves.

[FN181]. Exec. Order No. 12,291, 46 Fed. Reg. 13,193 (1981), reprinted as note to 5 U.S.C.A. §  601 (West 1985 Supp.), requires that agencies submit 'major rules' to the Office of Management and Budget (OMB) and withhold final action until they have received and responded to the views of the OMB on the rule. The executive order is intended to improve the cost/benefit ratio of agency rules.

[FN182]. The Rubber Manufacturers Association and the URW reported their agreement to the OSHA in a letter on Feb. 16, 1984.

[FN183]. See Harter, supra note 7, at 57-67 (advantages and disadvantages of agency participation or nonparticipation).

[FN184]. The FACA is addressed more fully in part VIII.G of the article.

[FN185]. The theory is that the FACA applies only to the communication of advice by a group to an agency. If negotiations took place only among private citizens, and a negotiated recommendation was communicated to the agency by a single person--the mediator--the FACA would not be applicable. This theory finds some support in the GSA guidelines for the FACA. 48 Fed. Reg. 19,325 (1983) (preamble) (only groups reporting directly to agency, rather than through intermediary, are advisory committees under the Act). Substantial disagreement exists concerning the merits of this interpretation of the FACA.

[FN186]. Part VI.C of the article discusses the cotton dust agreement.

[FN187]. Legal advisors of the OSHA developed this interpretation of the FACA only in reaction to a decision at the policy level to proceed with negotiated rulemaking, combined with participants' opposition to negotiating under the strictures of the FACA. The statutory interpretation derived from the observation that purely private negotiation, as occurred with the cotton dust standard, would not be subject to the FACA. Benzene negotiations without participation by the OSHA would be only a small step from the cotton dust procedure.

[FN188]. The facts asserted in the text were reported to the author, independently, by several participants in the negotiations, who were promised anonymity by the author.

[FN189]. Some of the participants feared OSHA participation because they thought the OSHA would use the deliberations without an understanding of its subtleties.

[FN190]. At least one participant believes that the release by the OSHA of the draft standard was helpful. The draft standard made no difference in party perceptions by adopting a 1 ppm standard; all participants expected that outcome. Release of the draft may have caused some helpful movement on the industry side, because it made recalcitrant constitutents take seriously the threat of independent adoption by the OSHA of a 1 ppm standard. This observation is based on interviews with both industry and labor negotiators.

[FN191]. Part II.B.1 of the article explains the BATNA concept.

[FN192]. Part VI.B of the article discusses the coke oven negotiations.

[FN193]. The requirements of the FACA are reviewed in part VIII.G of the article.

[FN194]. 29 C.F.R. §  1912.26 (1985).

[FN195]. Id. §  1912.27.

[FN196]. Id. §  1912.28.

[FN197]. Id. §  1912.33.

[FN198]. Subsequently, the EPA and the FAA were able to satisfy these steps in three to four weeks, with the cooperation of the OMB and the GSA.

[FN199]. 29 U.S.C. §  656 (1982).

[FN200]. 29 C.F.R. §  1912.2 (1985).

[FN201]. 29 U.S.C. §  655 (1982).

[FN202]. Id. §  655(b)(1).

[FN203]. Id. §  655(b)(2).

[FN204]. Id. §  655(b)(3)-(4).

[FN205]. 495 F.2d 1294 (7th Cir. 1974), cert. denied, 419 U.S. 1105  (1975).

[FN206]. Id. at 1296.

[FN207]. 50 Fed. Reg. 50,512 (1985).

[FN208]. Id. at 50,515.

[FN209]. Id. at 50,553.

[FN210]. Id. at 50,555.

[FN211]. Id. at 50,560.

[FN212]. Id. at 50,538-39.

[FN213]. United Steelworkers of Am. v. Rubber Mfrs. Ass'n, 783 F.2d 1117, 1119 (D.C. Cir. 1986).

[FN214]. Id.

[FN215]. Id.

[FN216]. Id. at 1120.

[FN217]. This was not the first Department of Transportation experience with negotiated rulemaking. In 1982, the Federal Railroad Administration amended rules pertaining to railroad air brakes. The changes resulted from an NPRM adopted at the joint request of the Association of American Railroads, the main industry trade association, and the Railway Labor Executives Association, a confederation of labor organizations representing rail employees. 47 Fed. Reg. 36,792 (1982). Comments submitted in response to the NPRM generally were supportive, resulting in a final rule reflecting the labor-industry agreement. Id.

[FN218]. The FAA negotiations are the subject of a study by John N. Nay and John D. Waller of the Performance Development Institute, 600 Maryland Avenue, S.W., Suite 302, Washington 20024, under Transportation Systems Center Contract No. DTRS-57-84-C-00144.

[FN219]. 50 Fed. Reg. 29,306 (1985).

[FN220]. 49 U.S.C. §  1421(a)(5) (1982).

[FN221]. 14 C.F.R. § §  121.470-.525, 135.261 (1984).

[FN222]. Id. §  121.471.

[FN223]. Id. §  121.480-.493.

[FN224]. Id. §  121.500-.525.

[FN225]. Id. §  135.261.

[FN226]. Notice of Proposed Rulemaking, 49 Fed. Reg. 12,136, 12,137  (1984).

[FN227]. 48 Fed. Reg. 21,339, 21,340 (1983).

[FN228]. Id. at 21,340.

[FN229]. Id. at 21,339.

[FN230]. Id.

[FN231]. Id. at 21,343.

[FN232]. Id.

[FN233]. Id. at 21,341.

[FN234]. Id.

[FN235]. Id.

[FN236]. 48 Fed. Reg. 29,771 (1983).

[FN237]. This satisfied Harter's second criterion and p 4(c) of ACUS Recommendation 82-4. Part II.C of the article discusses Harter's criteria for negotiated rulemaking and ACUS Recommendation 82-4.

[FN238]. This satisfied Harter's third criterion and p 4(a) of ACUS Recommendation 82-4.

[FN239]. This satisfied Harter's sixth criterion and p 4(b) of ACUS Recommendation 82-4.

[FN240]. This satisfied Harter's eighth criterion.

[FN241]. This implicates Harter's first criterion and p 4(e) of ACUS Recommendation 82-4, relating to countervailing power.

[FN242]. This implicates Harter's fifth criterion, opportunity for gain, and is central to the BATNA concept.

[FN243]. See 14 C.F.R. pt. 121 & 135 (1986) (part 121 applies to trunk carriers; part 135 applies to commuter carriers).

[FN244]. 49 Fed. Reg. 12,136, 12,137 (1984).

[FN245]. Id. at 12,137-38.

[FN246]. Id. at 12,136.

[FN247]. This was a significantly shorter comment period that is usually afforded in traditional rulemaking.

[FN248]. Flight Time Limitations and Rest Requirements, 50 Fed. Reg. 29,306, 29,306 (1985).

[FN249]. Id. at 29,306.

[FN250]. Id.

[FN251]. Id. at 29,307-08.

[FN252]. Id.

[FN253]. Id. at 29,307.

[FN254]. Id. at 29,309.

[FN255]. Id. at 29,310.

[FN256]. Id.

[FN257]. Id. at 29,311.

[FN258]. Id. at 29,312.

[FN259]. 'Compensatory rest' is additional rest to make up for reduced rest, below the normal minimum of nine hours, down to the revised floor of eight hours, given at the next proceding rest period.

[FN260]. 50 Fed. Reg. 29,306, 29,312 (1985).

[FN261]. Id. at 29,312.

[FN262]. Id. at 29,315.

[FN263]. This conclusion is based on interviews with Department of Transporation personnel who participated in the negotiations.

[FN264]. Part VIII.G of the article discusses more generally the FACA.

[FN265]. 567 F.2d 9 (D.C. Cir.) (per curiam), cert. denied, 434 U.S. 829 (1977). This case is discussed in parts VIII.D, hybrid rulemaking, and VIII.E, ex parte communication.

[FN266]. 657 F.2d 298 (D.C. Cir. 1981).

[FN267]. The EPA has also used negotiated rulemaking for agricultural pesticides. See Intent To Form an Advisory Committee To Negotiate Proposed Farmworker Protection Standards for Agricultural Pesticides, 50 Fed. Reg. 38,030 (1985) (notice of intent to use negotiated rulemaking to develop an NPRM for Farmworker Protection Standards). This initiative and other EPA negotiation proposals are discussed in part VI.E.

[FN268]. 48 Fed. Reg. 7,494 (1983).

[FN269]. Id.

[FN270]. Id. at 7,495.

[FN271]. 49 Fed. Reg. 17,576, 17,579 (1984).

[FN272]. The process leading to the rejection of the rule governing disposal of low level radioactive waste as a candidate has been investigated in detail. See D. Fish & L. Suskind, The Prenegotiation Phase of Regulatory Negotiation: The Case of the Low-Level Radioactive Waste Rule (unpublished manuscript).

[FN273]. McGlennon, Negotiating Regulations: Success at EPA, in A.L.I.- A.B.A. COURSE OF STUDY: ENVIRONMENTAL LAW 489 (Feb. 14-16, 1985).

[FN274]. 42 U.S.C. § §  7401-7642 (1982 & Supp. III 1985); 49 Fed. Reg. 17,576 (1984).

[FN275]. 50 Fed. Reg. 9,204 (1985).

[FN276]. Id.

[FN277]. Id. at 9,205 (citing 44 Fed. Reg. 9,464 (1979)).

[FN278]. Id. (citing 44 Fed. Reg. 40,784 (1979)).

[FN279]. Id. (citing 48 Fed. Reg. 1,413, 1,424 (1983)).

[FN280]. Id.

[FN281]. No. 84-748 (D.D.C. Sept. 14, 1984).

[FN282]. 50 Fed. Reg. 9,204, 9,205 (1985).

[FN283]. Id.

[FN284]. McGlennon, supra note 273. McGlennon was the convenor for the NCP negotiation.

[FN285]. Id. at 492.

[FN286]. Id.

[FN287]. Half the money came from the EPA. The remainder that the EPA helped arrange came from private sources. The National Institute for Dispute Resolution assisted in designing and administering the resource pool.

[FN288]. The discussion of the prenegotiation phase is heavily borrowed from McGlennon, supra note 273, at 492-94.

[FN289]. The Federal Register notice discusses how NCPs meet Harter's criteria for negotiated rulemaking.

[FN290]. Intent To Form Advisory Committee To Negotiate Nonconformance Penalty Regulations, 49 Fed. Reg. 17,576, 17,577 (1984).

[FN291]. Id. at 17,577.

[FN292]. Id.

[FN293]. Id. at 17,578.

[FN294]. 'Leader technology' is that technology possessed by the most advanced member of the industry.

[FN295]. 50 Fed. Reg. 9,204, 9,205 (1985).

[FN296]. Id. at 9,205.

[FN297]. The EPA hired Prof. Lawrence Susskind of the Massachusetts Institute of Technology.

[FN298]. The results of Prof. Susskind's study are available from Chris Kirtz, Director of the Regulatory Negotiation Project, U.S. Environmental Protection Agency.

[FN299]. See part III.L of the article. This guideline prohibits written communication from the negotiation group to the agency unless a consensus emerges on all matters of concern.

[FN300]. 50 Fed. Reg. 9,204 (1985).

[FN301]. 50 Fed. Reg. 35,374 (1985).

[FN302]. 7 U.S.C. §  136w (1982).

[FN303]. 40 C.F.R. pt. 166.

[FN304]. 50 Fed. Reg. 13,944 (1985).

[FN305]. Id.

[FN306]. 50 Reg. 13,944 (1985) (discussing House Subcommittee on Department Operations, Research and Foreign Agriculture, Regulatory Procedures and Public Health Issues in the EPA's Office of Pesticide Programs (Dec. 17, 1982)).

[FN307]. 50 Fed. Reg. 13,944 (1985).

[FN308]. Id.

[FN309]. 49 Fed. Reg. 31,145 (1984). The Federal Register notice explained why the pesticide exemption rules met Harter's criteria for negotiated rulemaking.

[FN310]. Id. at 31,147.

[FN311]. Id. at 31,148.

[FN312]. Id. at 31,146.

[FN313]. Id. at 31,147-48.

[FN314]. The original notice listed fewer participants, id., than were eventually members. 50 Fed. Reg. 13,944, 13,945 (1985).

[FN315]. The EPA contributed half the total amount, and the National Institute for Dispute Resolution contributed the remainder.

[FN316]. 50 Fed. Reg. 13,944 (1985).

[FN317]. Id.; 51 Fed. Reg. 1896 (1986).

[FN318]. Id.

[FN319]. Part III.O of the article discusses problems with the OSHA Act §  7 committee process.

[FN320]. 29 C.F.R. §  1910.1029 (1984).

[FN321]. 29 C.F.R. §  1910.1029(b) (1984).

[FN322]. 29 C.F.R. §  1910.1029(f) (1984).

[FN323]. Id.

[FN324]. 29 C.F.R. §  1910.1043 (1979).

[FN325]. 43 Fed. Reg. 27,350 (1978).

[FN326]. American Textile Mfg. Inst. v. Donovan, 452 U.S. 490 (1981). The rather complicated procedural history of the litigation is described in the NPRM of the OSHA. 48 Fed. Reg. 26,962 (1983).

[FN327]. 47 Fed. Reg. 5,906 (1982).

[FN328]. 48 Fed. Reg. 26,962 (1983).

[FN329]. As noted, the cotton dust health standard, like the benzene standard, had been the subject of litigation reaching the Supreme Court. American Textile Mfg. Inst., 452 U.S. 490.

[FN330]. 50 Fed. Reg. 42,789 (1985).

[FN331]. Id. at 42,791.

[FN332]. 48 Fed. Reg. 42,836 (1983).

[FN333]. 15 U.S.C. § §  2601-2629 (1982).

[FN334]. Id. §  2608(a).

[FN335]. 50 Fed. Reg. 27,674 (1985).

[FN336]. Benzene was selected by the OSHA for negotiated rulemaking, but now the OSHA believes that benzene was, in some respects, unsuitable for negotiation.

[FN337]. 50 Fed. Reg. 42,789, 42,792 (1985).

[FN338]. Id. at 42,792.

[FN339]. Id. at 42,791 (referring to 'rank[ing] of priorities' and 'identif  [ication] of acceptable solutions' as well as agreement on regulatory text and supporting rationale).

[FN340]. Id. at 42,791-92.

[FN341]. Id. at 42,790.

[FN342]. Id. at 42,792.

[FN343]. See 50 Fed. Reg. 42,787 (1985) (establishing Negotiated Rulemaking Advisory Committee on MDA to 'advise the Secretary of Labor regarding the building of consensus by affected interests on issues associated with a proposed OSHA standard on MDA'); 50 Fed. Reg. 42,789 (1985) (plans for MDA rule negotiation, proposing committee procedures, and identifying 13 issues to be addressed); 50 Fed. Reg. 48,655 (1985) (extension of time for submitting comments and requests for membership on negotiation committee); 50 Fed. Reg. 44,337 (1985) (regulatory agenda proposing schedule for negotiated rulemaking to result in NPRM by March, 1986); 51 Fed. Reg. 6,748 (1986) (responding to EPA notice of July 5, 1985, 50 Fed. Reg. 27,674, and announcing consideration of negotiated rulemaking to develop standard); 51 Fed. Reg. 14,202 (1986) (regulatory agenda announcing plans for mediated rulemaking, with NPRM scheduled for Oct. 1986 and final action by Dec. 1987); 51 Fed. Reg. 24,452 (1986) (appointing members of negotiating committee and announcing first meeting for July 22-24, 1986; stating that oral presentations by members of public will be permitted under control ofmediator/chairman; relaxing requirement for verbatim transcript and providing for detailed minutes instead, and modifying quorum requirements so that two members representing employers and two representing management must be present to conduct business); 51 Fed. Reg. 27,919 (1986) (announcing meetings and topics: Aug. 5, 1986, scope and application, definitions, recordkeeping, emergencies, hygiene facilities, and housekeeping; Sept. 9, 1986, personal protective equipment, exposure monitoring, and site visitation; Oct. 7, 1986, health effects, risk assessment, medical surveillance, medical appendices, biological monitoring, and removal and rate retention; Nov. 16, 1986, technological and economic feasibility; Dec. 9, 1986, permissible exposure limits and action level discussions; and Jan. 13, 1987, review of committee recommendations for Federal Register publication); 51 Fed. Reg. 35,571 (1986) (announcing location changes for meetings on Oct. 7, 1986, Nov. 18, 1986, Dec. 9, 1986, and Jan. 13, 1987); 51 Fed. Reg. 36,876 (1986) (renewing negotiation committee and changing name to 'Mediated Rulemaking Advisory Committee').

[FN344]. 50 Fed. Reg. 42,223 (1986) (announcing establishment of advisory committee to negotiate Farmworker Protection Standards for Agricultural Pesticides, and initial meeting on Nov. 4, 1985, to complete outstanding procedural matters, to determine how to address substantive issues and to begin addressing them); 50 Fed. Reg. 48,130 (1985) (announcing meetings of negotiating committee for Dec. 6, 1985, to continue work on substantive issues identified in notice); 51 Fed. Reg. 6,595 (1986) (announcing two day meeting of negotiating committee for March 6-7, 1986, to address procedural issues arising since last meeting and to continue work on substantive issues); 51 Fed. Reg. 13,091 (1986) (announcing two day meeting of negotiating committee for May 5-6, 1986, to review latest version of draft proposal).

[FN345]. The breakdown in the farmworker protection negotiations and subsequent developments was reported to the author by mediator Philip J. Harter in an MCI letter dated Nov. 3, 1986.

[FN346]. 51 Fed. Reg. 4,800 (1986) (requesting comments on possible establishment of advisory committee to negotiate issues leading to NPRM on New Source Performance Standards for Residential Wood Combustion Units, under §  111 of Clean Air Act).

[FN347]. 51 Fed. Reg. 25,401 (1986) (requesting comments on possible establishment of advisory committee to negotiate issues leading to NPRM implementing restrictions on injection of hazardous waste mandated by § §  3004(f) and 3004(g) of the Resource Conservation and Recovery Act).

[FN348]. 51 Fed. Reg. 25,739 (1986) (requesting comments on possible advisory committee to negotiate issues leading to NPRM amending current regulations governing major and minor modifications to Resource Conservation and Recovery Act permits).

[FN349]. 51 Fed. Reg. 5,205 (1986) (citing 16 C.F.R. pt. 703).

[FN350]. 15 U.S.C. §  2310(a)(1) (1982).

[FN351]. 51 Fed. Reg. at 5206.

[FN352]. Id.

[FN353]. Id.

[FN354]. Id. at 5207-08.

[FN355]. 51 Fed. Reg. 29,666 (1986).

[FN356]. Id. at 29,667.

[FN357]. Id.

[FN358]. Id.

[FN359]. Id.

[FN360]. Compare 51 Fed. Reg. 5,207-08 (1986) with 51 Fed. Reg. 29,668 (1986).

[FN361]. Panel members were Philip J. Harter, Chairman of the Dispute Resolution Committee; Chris Kirtz, Office of Policy for the United States Environmental Protection Agency; Duane Harves, Chief Administrative Law Judge of the state of Minnesota; Steven LeFelt, Deputy Director of the New Jersey Office of Administrative Law; William Penny, Assistant Commissioner and General Counsel of the Tennessee Department of Environment; and Warren Kaplan, law student liason.

[FN362]. The panel also gave attention to the use of alternative dispute resolution techniques in adjudication (resolution of 'contested' cases). Significant resistance to governmental bodies submitting to arbitration continues to be common, but the represented states have been energetic in trying to expand the use of settlement conferences and other forms of informal resolution.

[FN363]. Both approaches are consistent with ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982).

[FN364]. Part II.B.1 of the Article explains the BATNA concept.

[FN365]. This approach is recommended by p 7 of ACUS Recommendation 82-4,  47 Fed. Reg. 30,708, 30,709-10 (1982).

[FN366]. Paragraphs 3 and 4 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), recommend use of convenors in every regulatory negotiation. Paragraph 10 recommends the use of mediators in some cases.

[FN367]. This experience raised questions about the optional character of  p 10 of ACUS Recommendation 82-4. Paragraph 5 of ACUS Recommendation 85-5 urges that mediators be used in virtally all rule negotiations. 50 Fed. Reg. at 52,895 (1985).

[FN368]. Paragraphs (g) and 8 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), recommend agency participation.

[FN369]. When this article was written the author was working with the Office of the Secretary of Labor to develop options for institutionalizing departmental support for negotiated rulemaking and other regulatory improvements.

[FN370]. Paragraph 11 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), addresses consensus and contemplates the possibility of less than total agreement on all issues.

[FN371]. Establishment of such a ground rule probably was essential for the negotiations to proceed. Paragraph 11 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), contemplates that negotiators report on less than total agreement, identifying the issues on which they agree and those on which they disagree.

[FN372]. The resource pool concept is a way--without legislation--of implementing p 9 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,710 (1982) (certain affected interests may require direct reimbursement to participate and lead to negotiation; Congress should clarify agency authority to provide this reimbursement).

[FN373]. Paragraph 13 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), recommends such publication.

[FN374]. Paragraph 14 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), recommends such review.

[FN375]. This information is based on interviews with FAA and DOT personnel involved in the negotiations.

[FN376]. This information is based on interviews with OSHA and Solicitor of Labor personnel.

[FN377]. Paragraph 14 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709-10 (1982), recommends such legislation.

[FN378]. Harter recommended such changes. See Harter, supra note 7, at 102.

[FN379]. Paragraph 2 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982).

[FN380]. See Harter, supra note 7, at 102-07 (asserting rule should be sustained if within agency jurisdiction and reflects consensus; if not, traditional standards of review should apply); Note, supra note 124, at 1883- 90. See generally Stewart, supra note 21.

[FN381]. See Wald, Negotiation of Environmental Disputes: A New Role for the Court?, 10 COLUM. J. ENVTL. L. 1, 17-25 (1985) (examining judicial review of negotiated rulemaking in light of APA); Harter, The Role of Courts in Regulatory Negotiation--A Response to Judge Wald, 11 COLUM. J. ENVTL. L. 51 (1986) (same).

[FN382]. See J. LOCKE, SECOND TREATISE OF GOVERNMENT §  141, at 74-75 (C. Macpherson ed. 1980) (legislature cannot transfer power of making laws to any other hands).

[FN383]. See Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607, 671, 672 (1980) (Rehnquist, J., concurring in the judgment) (Congress improperly delegated legislative authority to Secretary of Labor); Panama Refining Co. v. Ryan, 293 U.S. 388, 421 (1935) (congressional authority to delegate functions to administrative agencies is not unlimited), Schecter Poultry Corp. v. United States, 295 U.S. 495 (1935) (invalidating statute delegating authority to private industry group); Stewart, supra note 21, at 1672-73 (traditional model of administrative law premised on belief that imposition of administratively determined sanctions on private individuals must be authorized by legislature through rules that control agency action); see generally J. DICKINSON, ADMINISTRATIVE JUSTICE AND THE SUPREMACY OF LAW IN THE UNITED STATES (1927).

[FN384]. Industrial Union Dep't, 448 U.S. at 686 (Rehnquist, J., concurring in the judgment); Yakus v. United States, 321 U.S. 414, 426 (1944).

[FN385]. Political accountability is obvious, albeit indirect, when the agency officials are appointed by the President. See Buckley v. Valeo, 424 U.S. 1, 135 (1976) (invalidating regulatory scheme depriving President of power to appoint agency members by vesting it in legislative branch officers).

[FN386]. See Carter v. Carter Coal Co., 298 U.S. 238 (1936) (invalidating statute delegating wage and hour standard setting to representatives of coal producers and coal miners); Aqua Slide 'N' Dive Corp. v. Consumer Product Safety Comm'n, 569 F.2d 831, 843-44 (5th Cir. 1978) (courts should not defer to opinions of governmental consultants as much as to agency personnel); see generally Jaffe, Law Making by Private Groups, 51 HARV. L. REV. 201 (1937); Liebmann, Delegation to Private Parties in American Constitutional Law, 50 IND. L.J. 650 (1975).

[FN387]. See Harter, supra note 7, at 107-09 (concluding regulatory negotiation is not an impermissible delegation); Note, supra note 124, at 1880- 83 (asserting 'nondelegation doctrine' problem nonexistent if agency replicates process of pluralistic decisionmaking thereby having adequate representatives).

[FN388]. Harter, supra note 7, at 109 (because agency makes final decision its authority is not delegated); Note, supra note 124, at 1882-83 (if agency has final authority no delegation problem exists). In all of the negotiated rulemaking efforts reviewed in this article, agencies unequivocally reserved final decisionmaking authority.

[FN389]. See Sunshine Anthracite Coal Co. v. Adkins, 310 U.S. 381, 399  (1940) (approving against delegation challenge statute permitting coal producers to propose minimum prices and other sales conditions to public commission that could approve, disapprove, or modify them); First Jersey Sec., Inc. v. Bergen, 605 F.2d 690, 697-700 (3d Cir. 1979) (approving self- regulation in securities markets), cert. denied, 444 U.S. 1074 (1980); Association of Am. Physicians & Surgeons v. Weinberger, 395 F. Supp. 125, 140 (N.D. Ill. 1975) (private establishment of standards to govern health care not invalid delegation because agency provides hearing on standards). See generally Note, supra note 124, at 1882 n.63 (citing other cases approving delegation to private organizations).

[FN390]. Part VI.C of the article discusses the cotton dust negotiation, in which this occurred.

[FN391]. Harter, supra note 7, at 109; Note, supra note 124, at 1883.

[FN392]. These requirements are considered in part VIII.F of the article. See Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607, 686 (Rehnquist, J. concurring) (delegation doctrine requires that courts be able to review agency decisions under ascertainable standards).

[FN393]. See id. at 646 (Stevens, J., majority opinion) (construing statutory requirement narrowly and invalidating administrative regulation to avoid constitutional delegation problems).

[FN394]. See Levin, Scope-of-Review Doctrine Restated: An Administrative Law Section Report, 38 ADMIN. L. REV. 239, 260 (1986) (controversy); id. at 274-77 (difficulty in deciding what 'record' is in informal rulemaking).

[FN395]. K. DAVIS, ADMINISTRATIVE LAW TEXT §  6.03, at 142-43 (3d ed. 1972).

[FN396]. Id. at 143.

[FN397]. Paragraph 13 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,710 (1982), provides for publication of a negotiated rule for comment.

[FN398]. Part IV.I of the article discusses the FAA response to these complaints.

[FN399]. See generally Pedersen, Formal Records and Informal Rulemaking, 85 YALE L.J. 38 (1975); Williams, 'Hybrid Rulemaking' under the Administrative Procedure Act: A Legal and Empirical Analysis, 42 U. CHI. L. REV. 401 (1975).

[FN400]. See generally ACUS Recommendation 76-3, 1 C.F.R. §  305.76-3  (1983) ('Procedures in Addition to Notice and the Opportunity for Comment in Informal Rulemaking').

[FN401]. 29 U.S.C. §  655(b)(3) (requiring public hearing on written objections to proposed rules).

[FN402]. 15 U.S.C. §  57(a)(c)(2) (1982) (requiring opportunity for rebuttal submissions and for cross-examination when FTC determines that regulations involve disputed issues of material fact).

[FN403]. 563 F.2d 240 (2d Cir. 1977).

[FN404]. See generally Pedersen, supra note 399.

[FN405]. 483 F.2d 1238 (D.C. Cir. 1973). But see Wisconsin Gas Co. v. FERC, 770 F.2d 1144, 1167 (D.C. Cir. 1985) (Mobil Oil 'is no longer good law' after Vermont Yankee).

[FN406]. See International Harvester Co. v. Ruckelshaus, 478 F.2d 615, 629-31 (D.C. Cir. 1973) (requiring cross-examination in some circumstances); Appalachian Power Co. v. EPA, 477 F.2d 495, 503 (4th Cir. 1973) (same); Kennecott Copper Corp. v. EPA, 462 F.2d 846, 850 (D.C. Cir. 1972) (requiring EPA to articulate basis for rule in detail).

[FN407]. 435 U.S. 519 (1978).

[FN408]. See Byse, Vermont Yankee and the Evolution of Administrative Procedure: A Somewhat Different View, 91 HARV. L. REV. 1823, 1823 (1978) (with rare exceptions courts cannot require more procedures than specified in APA); Stewart, Vermont Yankee and the Evolution of Administrative Procedure, 91 HARV. L. REV. 1805, 1808-09 (1978) (same).

[FN409]. See Lead Indus. Ass'n v. EPA, 647 F.2d 1130, 1169-71 (D.C. Cir.) (industry failed to show 'compelling circumstances' requiring cross- examination), cert. denied, 449 U.S. 1042 (1980) (United Steelworkers of Am. v. Marshall, 647 F.2d 1189, 1203 n.6, 1227-28 (D.C. Cir. 1980) (denying right to cost-benefit analysis and cross-examination), cert. denied, 453 U.S. 913 (1981).

[FN410]. See Sholly v. NRC, 651 F.2d 792 (D.C. Cir. 1981) (Tamm, MacKinnon, Robb, and Wilkey, JJ., dissenting from denial of rehearing en banc), vacated on other grounds, 459 U.S. 1194 (1983); PPG Indus. v. Costle, 630 F.2d 462 (6th Cir. 1980); Cotter v. Harris, 642 F.2d 700 (3d Cir. 1981); Levin, supra note 394, at 283 (questioning whether Vermont Yankee allows procedural review separately from 'substantive' arbitrary and capricious reivew). See generally Rodgers, A Hard Look at Vermont Yankee: Environmental Law Under Close Scrutiny, 67 GEO. L.J. 699 (1979).

[FN411]. 1 C.F.R. §  305.76-3 (1983).

[FN412]. See Note, supra note 124, at 1887-89 (1981) (suggesting that ex parte communication problem could be avoided in negotiated rulemaking by judicial review limited to question of whether party claiming harm was represented adequately in negotiations).

[FN413]. See Friendly, 'Some Kind of Hearing,' 123 U. PA. L. REV. 1267, 1279-94 (1975) (identifying 'right to know opposing evidence' and 'right to have the decision based only on the evidence presented' as basic elements of adjudicatory process).

[FN414]. See Home Box Office v. FCC, 567 F.2d 9 (D.C. Cir.) (per curiam), cert. denied, 434 U.S. 829 (1977).

[FN415]. 567 F.2d 9 (D.C. Cir.) (per curiam), cert. denied, 434 U.S. 829 (1977).

[FN416]. Id. at 57.

[FN417]. Id.

[FN418]. Id. at 56. Another formulation of the concerns raised by ex parte communication in notice and comment rulemaking was articulated in ACUS Recommendation 77-3, 1 C.F.R. §  305.77-3 (1983) ((1) decisionmakers may be influenced by communications made privately, thus creating a situation seemingly at odds with the widespread demand for open government; (2) significant information may be unavailable to reviewing courts; and (3) interested persons may be unable to reply effectively to information, proposals, or arguments presented in an ex parte communication).

[FN419]. Compare Iowa State Commerce Comm'n v. Office of Fed. Inspector, 730 F.2d 1566, 1576 (D.C. Cir. 1984) (rate determinations for Alaska pipeline not invalid because ex parte contacts occurred); Katherine Gibbs School v. FTC, 612 F.2d 658, 670 (2d Cir. 1979) (complaints about contacts between decisionmaker and agency advocates should be addressed to Congress); Hercules, Inc. v. EPA, 598 F.2d 91, 127 (D.C. Cir. 1978) (contacts between decisionmaker and agency staff advocates for getting assistance in understanding record reluctantly approved because of volume of record) and  Action for Children's Television v. FCC, 564 F.2d 458 (D.C. Cir. 1977) (rejecting challenge to ex parte discussions leading to acceptance of industry self-regulation in lieu of Commission regulation of children's programming) with United States Lines, Inc. v. Federal Maritime Comm'n, 584 F.2d 519 (D.C. Cir. 1978) (overturning approval of industry agreement because of reliance on matter outside record).

[FN420]. See generally Gellhorn & Robinson, Rulemaking 'Due Process': An Inconclusive Dialogue, 48 U. CHI. L. REV. 201 (1981); Nathanson, Report to the Select Committee on Ex Parte Communications in Informed Rulemaking Proceedings, 30 ADMIN. L. REV. 377 (1978); Preston, A Right of Rebuttal in Informal Rulemaking: May Courts Impose Procedures to Ensure Rebuttal of Ex Parte Communications and Information Derived from Agency Files After Vermont Yankee?, 32 ADMIN. L. REV. 621 (1980); Robinson, The Federal Communications Commission: An Essay on Regulatory Watchdogs, 64 VA. L. REV. 169, 228-30 (1978) (cutting off the flow of informal information would frustrate effective decisionmaking); Note, Ex Parte Communications During Informal Rulemaking, 14 COLUM. J.L. & SOC. PROBS. 269 (1979); Note, Due Process and Ex Parte Contacts in Informal Rulemaking, 89 YALE L.J. 94 (1979).

[FN421]. Recommendation 77-3 of the Administration Conference of the United States, 1 C.F.R. §  305.77-3 (1983) (general prohibition on ex parte contact in informal rulemaking would be unwise).

[FN422]. 657 F.2d 398 (D.C. Cir. 1981).

[FN423]. Id. at 400-02.

[FN424]. See Home Box Office, 567 F.2d at 57 (implying that ex parte contact is limited only after NPRM is published); Iowa State Commerce Comm'n v. Office of Fed. Inspector, 730 F.2d 1566, 1576 (D.C. Cir. 1984) (same); J. O'REILLY, ADMINISTRATIVE RULEMAKING §  4.02, at 69 (1983) (barriers to ex parte contact do not apply to prerulemaking contacts); id. §  9.04, at 183 (case law on ex parte communications does not restrict contact before rule is proposed by agency).

[FN425]. 647 F.2d 1189 (D.C. Cir. 1980), cert. denied, 453 U.S. 913  (1981).

[FN426]. Id. at 1218.

[FN427]. Lead Indus. Ass'n v. OSHA, 610 F.2d 70 (2d Cir. 1979) (finding consultant reports to be within exemption 5 of the FOIA).

[FN428]. 647 F.2d at 1220 (footnotes omitted).

[FN429]. Id.

[FN430]. For a slightly different proposal, see Note, supra note 124, at 1888-89 (1981) (suggesting that ex parte problem could be avoided in negotiated rulemaking by judicial review limited to question of whether party claiming harm was represented adequately in negotiations).

[FN431]. Id.

[FN432]. See Carlin Communications v. FCC, 749 F.2d 113, 118 n.9 (2d Cir. 1984) (no violation of agency rules or general principles of administrative law where ex parte contacts summarized in record of notice and comment rulemaking).

[FN433]. 5 U.S.C. §  706 (1982).

[FN434]. Id. §  706(2)(A).

[FN435]. See, e.g., Occupational Safety and Health Act, 29 U.S.C. §  655(f) (1982) (impliedly subjecting rules to APA review by exempting emergency standards from APA, while imposing additional requirement that agency determinations in rulemaking must be supported by 'substantial evidence in the record taken as a whole'); National Traffic and Motor Vehicle Safety Act of 1966, 15 U.S.C. §  1392(b) (1982).

[FN436]. Motor Vehicle Mfrs. Ass'n v. State Farm Mut. Ins. Co., 463 U.S. 29, 43 (1983). The Motor Vehicle case plowed new ground with respect to applying the arbitrary and capricious standard to a decision to rescind a rule, but it did nothing new or controversial with respect to the meaning of the standard itself. See also Levin, supra note 394 (offering analytical framework for arbitrary and capricious analysis).

[FN437]. Id. at 48.

[FN438]. Id. at 50; accord Industrial Union Dep't v. American Petroleum Inst., 448 U.S. 607, 631 n.31 (1980).

[FN439]. 448 U.S. at 631.

[FN440]. For the history of benzene regulation by the OSHA, see part III.B of the article.

[FN441]. 463 U.S. at 52.

[FN442]. Id.

[FN443]. Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council, 435 U.S. 519 (1978).

[FN444]. See Stewart, supra note 21, at 1799-1800 (no logical reason to preserve current concepts of judicial review if rulemaking is a political process with adequate representation of affected interests, but courts unlikely to change approaches); Note, supra note 124, at 1885-86 (suggesting use of negotiation ought to relax judicial review standards); Wald, supra note 381, (expressing wariness of requiring courts to consider adequacy of representation and opining that judicial review of negotiated rules will not change much).

[FN445]. The FAA justification for the negotiated flight and duty time rule and the justification of the EPA for the NCP and pesticide exemption rulesare good examples. See parts IV.H., V.C., and V.D of the article.

[FN446]. Requirements for support in the record, contained in substantive statutes, reinforces the need to give attention to the record support for a negotiated rule.

[FN447]. The agency's contemporaneous rationale must support the rule. Levin, supra note 394, at 265. The agency's reasoning also must be logical. Id. at 255, 259.

[FN448]. Consideration of all relevant factors is one requirement of the arbitrary and capricious standard. Levin, supra note 394, at 250-51.

[FN449]. Harter, supra note 7, at 65 (referring to Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402 (1971)).

[FN450]. Id. at 66.

[FN451]. Id.

[FN452]. Use of negotiated rulemaking obviously cannot expand an agency's authority beyond what was delegated to it by Congress. See Levin, supra note 394, at 268 (independent judgment by court appropriate to determine whether agency adhered to limits and conditions of delegation; more lenient standards of judicial review appropriate for other questions).

[FN453]. Id. at 250-51. Determining that the agency has complied with congressionally mandated criteria is closely related to determining that the agency has stayed within its authority. Id. at 251. The congressional mandate is not alterable by the particular process used by the agency.

[FN454]. See id. at 253-56 (cautioning that judicial review should allow agency broad discretion after court has ascertained compliance with statutory mandate).

[FN455]. The 'hard look' approach is one application of the arbitrary and capricious standard forcing careful agency examination of alternatives and complete rationale in support of alternative selected. See id. at 259 (characterizing 'hard look' line of cases).

[FN456]. 5 U.S.C. app. § §  1-15 (1982).

[FN457]. Pub. L. No. 92-463, 86 Stat. 770 (codified as amended at 5 U.S.C. app. § §  1-15 (1982 & Supp. III 1985)).

[FN458]. See Parritt & Wilkinson, Open Advisory Committees and the Political Process: The Federal Advisory Committee Act After Two Years, 63 GEO. L.J. 725, 731-35 (1975).

[FN459]. 5 U.S.C. §  552 app. I (1982). See Perritt & Wilkinson, supra note 458, at 735 (describing original exceptions to open meetings requirement).

[FN460]. 5 U.S.C. §  552b (1982).

[FN461]. Section 10(d) of the amended FACA reads:

Subsections (a)(1) and (a)(3) [requiring meetings to be open and affording interested persons the right to attend meetings and to file statements] of this section shall not apply to any portion of an advisory committee meeting where the . . . agency to which the advisory committee reports, determines that such portion of such meeting may be closed to the public in accordance with subsection (c) of section 552b of title 5, United States Code. Any such determination shall be in writing and shall contain the reasons for such determination.

[FN462]. See generally Note, Government in the Sunshine Act: Opening Federal Agency Meetings, 26 AM. U.L. REV. 154, 172-95 (1976) (discussing 10 Sunshine Act exemptions).

[FN463]. See Braniff Master Executive Council v. Civil Aeronautics Bd., 693 F.2d 220 (D.C. Cir. 1982) (release of transcripts, rather than invalidation of agency's substantive action, is remedy for disregard of Government in Sunshine Act).

[FN464]. 1 C.F.R. § §  101.6.1001-101.6.1035 (1984).

[FN465]. Id. §  101.6.1004 (1984); see 48 Fed. Reg. 19,324  (1983) (preamble to GSA regulations).

[FN466]. Id. at p h(1).

[FN467]. Id. at p h(2).

[FN468]. Id. at p j.

[FN469]. 48 Fed. Reg. 19,324 (1983).

[FN470]. Id. at 19,325.

[FN471]. These concerns led the ACUS to recommend congressional action to relax requirements of the FACA for negotiated rulemaking. See p 2 of ACUS Recommendation 82-4, 47 Fed. Reg. 30,708, 30,709 (1982).

[FN472]. Part II.B.1 of the article discusses the dynamics of regulatory negotiation.

[FN473]. The quoted phrases are the predicates for satisfying the definition of advisory committee, contained in §  3 of the Act. 5 U.S.C. App. 2 (1982).

[FN474]. Such a delegation would present other legal problems, discussed in part VIII.B of the article.

[FN475]. Some support for this position and other theories supporting closed meetings of caucuses and subgroups can be found in cases decided under the Sunshine Act. See, e.g., FCC v. ITT World Communications, 466 U.S. 463 (1984) (informal background sessions or negotiations are not 'meetings').

[FN476]. See National Anti-Hunger Coalition v. Executive Comm. of the President's Private Sector Survey on Cost Control, 557 F. Supp. 524 (D.D.C.) (task forces chaired by members of presidential commission not advisory committees), aff'd, 711 F.2d 1071 (D.C. Cir. 1983); Consumers Union of the United States v. Department of Health, Educ. & Welfare, 409 F. Supp. 473 (D.D.C. 1976) (presentation by industry group on voluntary, industry sponsored proposal not advisory committee where agency lacked authority to regulate subject), aff'd, 551 F.2d 466 (D.C. Cir. 1977); Nader v. Baroody, 396 F. Supp. 1231 (D.D.C. 1975) (White House meetings with selected private groups not advisory committee).

[FN477]. Compare Ader v. Dunlop, 370 F. Supp. 177 (D.D.C. 1973) (FACA does not permit closing all meetings to protect deliberative consulations among members) with Aviation Consumer Action Project v. Washburn, 535 F.2d 101 (D.C. Cir. 1976) (approving closure of portions of 3 of 20 meetings because intra-agency memoranda within exemption 5 of FOIA had been discussed). See generally Perritt & Wilkinson, supra note 458, at 743-47 (advocating applicability of FOIA exemption 5 to protect deliberative process of advisory committee).

[FN478]. 48 Fed. Reg. 19,325 (1983) (preamble to GSA regulations).

[FN479]. 557 F. Supp. 524 (D.D.C. 1983).

[FN480]. Id. at 529.

[FN481]. Id.

[FN482]. Id.

[FN483]. 466 U.S. 463, 465 (1984).

[FN484]. As explained above, exceptions to the FACA requirement for open meetings are linked to exceptions to the Sunshine Act open meeting requirement.

[FN485]. FCC v. ITT World Communications, 466 U.S. at 469 (quoting Senate Report on Sunshine Act).

[FN486]. Id. at 470.

[FN487]. See 29 U.S.C. §  159 (1984) (providing for elections to select exclusive representative for appropriate unit of employees).

[FN488]. See 45 U.S.C. §  152, Fourth (1984) (providing for elections to select exclusive representative for class or craft of employees).

[FN489]. FED. R. CIV. P. 23.

[FN490]. See T. SULLIVAN, supra note 59, at 40-47 (distinguishing different types of negotiation).

[FN491]. See p 7 of ACUS Recommendation 85-5.

[FN492]. A similar conclusion has been reached with respect to negotiation of state regulatory issues. See J. Brock, Developing Systems for the Settlements of Recurring Disputes (Sept. 1984) (unpublished manuscript available from The Mediation Institute, 605 First Ave., Suite 525, Seattle, WA 98104, (206) 624-0805).

[FN493]. 50 Fed. Reg. 52,895 (1985).

[FN494]. Part II.B.2 of the article discusses the intensity of feeling and other determinants of interest group involvement in governmental decisionmaking.

[FN495]. The same approach has been used by the OSHA in the MDA negotiation and the FTC in the Informal Dispute Procedure negotiation. See parts VI.D & VI.F of the article.

[FN496]. Everyone with labor-management negotiation experience is familiar with the common phenomenon in the grievance dispute process where labor and management reach de facto agreement on resolution of a grievance but need an arbitrator's decision to 'take them off the hook.' In some cases, the disputing parties actually write the arbitrator's decision.

[FN497]. The APA requires this opportunity for comment. 5 U.S.C. §  553  (1982).

[FN498]. See p 2 of ACUS Recommendation 82-4.

[FN499]. See part VIII.G of the article.

[FN500]. S. 451, 99th Cong., 1st Sess., 131 CONG. REC. 51,348-53 (1985).

[FN501]. Id.

[FN502]. Id. at 51,351.

[FN503]. Section 7(h) of S. 451 contemplates termination of a negotiation group before it has an opportunity to review comments submitted in response to its recommended rule. Id. at 51,352.

[FN504]. Some agencies and commentators question the appropriateness of participation by the OMB, arguing that it is the agency's responsibility to obtain concurrence of the OMB in any proposed regulation. The whole point of negotiations, however, is to get all the interests likely to influence the substance of a regulation to communicate directly with each other. Under Executive Order 12,291, 46 Fed. Reg. 13,193 (1981), reprinted as note to 5 U.S.C.A. §  601 (West 1985 Supp.), the OMB has a major role to play. The cotton dust experience shows the risk of negotiating a consensus without taking the views of the OMB into account. Evidence from the benzene negotiations strongly suggests that industry's position was determined in significant part by what it thought the OMB would do with a unilateral proposal from the OSHA. Such perceptions should be tested and taken into account in the negotiations themselves, rather than outside. The FAA and the EPA experiences show that the OMB can support negotiations and can participate in various ways. At a minimum, agencies considering negotiated rulemaking in a particular proceeding should have a clear strategy for obtaining support from the OMB and for involving the OMB as negotiations proceed. This may be accomplished using facilitators or mediators as intermediaries.

[FN505]. See Harter, supra note 7, at 57-58; Note, supra note 124, at 1875- 76 (1981) (describing agency oversight model).

[FN506]. The cotton dust negotiation is a clear example. See part VI.C of the article.

[FN507]. Part II.B1 of the article explains the BATNA concept.

[FN508]. Post-NPRM review of comments occurred at the FAA and EPA as a continuation of pre-NPRM negotiations, but the point is that negotiations might be commenced after an NPRM is issued, as occurred in the cotton dust controversy.

[FN509]. In some respects, the benzene negotiations benefitted from such sharpening, because of the earlier litigation.

[FN510]. See part VII.E of the article.

[FN511]. See 5 U.S.C. §  552b(c)(10) (1982) (applied to advisory committees by FACA §  10, authorizing closure of meeting concerning determination of a matter on the record after opportunity for a hearing).

[FN512]. See, e.g., Food Chemical News v. Davis, 378 F. Supp. 1048  (D.D.C. 1974) (post-NPRM meetings included both industry and consumer groups that participated in earlier meetings); Consumers Union of the United States v. Department of Health, Educ. & Welfare, 409 F. Supp. 473 (D.D.C. 1976) (post- NPRM meetings requested by industry), aff'd mem., 551 F.2d 466 (D.C. Cir. 1977).

[FN513]. Part VI.D of the article discusses the MDA negotiations.

[FN514]. 1 C.F.R. §  305.82-4 (1985). Part VI.D discusses the MDA negotiations.

[FN515]. See parts VI.E and F.

[FN516]. This happened is the benzene negotiations, discussed in part III, and in the farmworker protection negotiations, discussed in part VI.E.

[FN517]. Part VI.G of the article summarizes the program of the ABA committee on administrative dispute resolution.

END OF DOCUMENT