Is Consensus an Appropriate Basis for Regulatory Policy?

Is Consensus an Appropriate Basis for Regulatory Policy? Cary Coglianese April 2001 RPP-2001-02 Regulatory Policy Program Center for Business and Government John F. Kennedy School of Government 79 John F. Kennedy Street, Weil Hall Cambridge, MA 02138 Citation This paper may be cited as: Coglianese, Cary. 2001. “Is Consensus an Appropriate Basis for Regulatory Policy?” Regulatory Policy Program Working Paper RPP-2001-02. Cambridge, MA: Center for Business and Government, John F. Kennedy School of Government, Harvard University. Comments may be directed to the author at Weil Hall, Center for Business and Government, John F. Kennedy School of Government, Harvard University, 79 JFK Street, Cambridge, MA 02138; Email [email protected]. Regulatory Policy Program The Regulatory Policy Program at the Center for Business and Government provides an environment in which to develop and test leading ideas on regulation and regulatory institutions. RPP’s research aims to improve the global society and economy by understanding the impacts of regulation and creating better decisions about the design and implementation of regulatory strategies around the world. RPP’s efforts are organized around the following three core areas: regulation, markets, and deregulation; regulatory instruments; and regulatory institutions and policymaking. The views expressed in this paper are those of the author and do not imply endorsement by the Regulatory Policy Program, the Center for Business and Government, the John F. Kennedy School of Government, or Harvard University. For Further Information Further information on the Regulatory Policy Program can be obtained from the program’s director, Jennifer Nash, Center for Business and Government, John F. Kennedy School of Government, 79 JFK Street, Cambridge, MA 02138, telephone (617) 384-7325, telefax (617) 496-0063, Email [email protected]. Is Consensus an Appropriate Basis for Regulatory Policy? Cary Coglianese* April 2001 RPP-2001-02 Abstract Observers of environmental policy increasingly urge regulators to build consensus before making policy decisions. By seeking consensus, regulators are supposed to be able to reduce conflict, increase compliance, improve public policy, and promote public participation. Yet consensus-building markedly shifts the prevailing norms of governance in the United States by de-centering the role of agency officials, making them facilitators or negotiation partners rather than central, accountable decision makers charged with seeking solutions that advance the overall public interest. A shift to consensus as the basis for regulatory policy also creates or exacerbates at least six pathologies in the policy process: tractability over importance, imprecision, lowest common denominator, increased time and resources, unrealistic expectations, and additional sources of conflict. The widespread establishment of consensus as the goal of regulatory policy making would constitute a shift in the prevailing mode of governance that is neither necessary nor wise. *Associate Professor of Public Policy and Chair of the Regulatory Policy Program This paper is forthcoming in Eric W. Orts and Kurt Deketelaere, eds., Environmental Contracts: Comparative Approaches to Regulatory Innovation in the United States and Europe, Kluwer Law International, 2001. Is Consensus an Appropriate Basis for Regulatory Policy? Cary Coglianese Harvard University Regulators in the United States are increasingly urged to build consensus before making policy decisions. The process of consensus-building aims to create an explicit agreement over the substance of regulatory policy among the individuals and groups who will be affected by the policy. This craving for consensus was institutionalized in 1990 with the passage of the Negotiated Rulemaking Act1 which authorizes agencies to establish formal negotiation processes over the terms of proposed regulations. Since that time, more than two dozen other federal statutes either compel or strongly encourage agencies to use consensus-based procedures.2 5 U.S.C. §§ 561-570 (1996). The Negotiated Rulemaking Act was permanently reauthorized by the Administrative Dispute Resolution Act of 1996, Pub. Law No. 104320 (1996). 1 The number of statutory provisions requiring the use of consensus-based decisionmaking continues to grow. For a list of a selected statutes requiring the use of negotiated rulemaking proceedings, see Coglianese (1997:1268 n. 75). For additional legislation requiring or encouraging consensus-based decisions, see the Transportation Equity Act for the 21st Century, Pub. Law No. 105-178 (1998); Higher Education Amendments of 1998, Pub. Law No. 105-244 (1998); VA, HUD, and Independent Agencies Appropriations Act, Pub. Law No. 105-276 (1998); Omnibus Supplemental Appropriations, Pub. Law No. 105-277 (1998); Carl D. Perkins Vocational and Applied Technology Education Amendments of 1998, Pub. Law No. 105-332 (1998); Traumatic Brain Injury Programs Authorization, Pub. Law No. 104-166 (1996); VA, HUD, and Independent Agencies Appropriations Act, Pub. Law No. 104-204 (1996); Sustainable Fisheries Act, Pub. Law No. 104-297 (1996); Omnibus Parks and Public Lands Management Act of 1996, Pub. Law No. 104-333 (1996); Fisheries Act of 1995, Pub. Law No. 104-43 (1995); Goals 2000 Educate America Act, Pub. Law No. 103-227 (1994); Marine Mammals Protection Act Amendments of 1994, Pub. Law No. 103-238 (1994); Federal Aviation Administration Authorization Act of 1994, Pub. Law No. 103305 (1994); Federal Acquisition Streamlining Act of 1994, Pub. Law No. 103-355 (1994); Yavapai-Prescott Indian Tribe Water Rights Settlement Act of 1994, Pub. Law No. 103-434 (1994). In addition, the National Technology Transfer and Advancement 2 The desire for consensus has also taken an especially strong hold in efforts to adopt new approaches to environmental regulation. Nearly every major commission report and panel study issued on environmental policy in recent years has called for greater reliance on consensus-building.3 In recent years, the United States Environmental Protection Agency has launched several consensus-based projects, such as the Common Sense Initiative and Project XL. Furthermore, a range of natural resource initiatives in the areas of ecosystem management and habitat conservation have also relied on collaborative approaches to policy-making. We are living, some might have it, at the dawn of an age of consensus. In the face of a prevailing enchantment with what Louis Jaffe (1937) once called the 'beauties of co-operation,' it is worth pausing to reflect on whether consensus really is an appropriate mode of making public policy. Much of what has been written so far on consensus in regulatory policy-making focuses on its advantages, with comparatively little attention having been paid to any potential hazards of consensus as a decision rule. In this chapter, I seek to broaden attention to the implications of efforts to increase the search for and reliance on consensus in the making of regulatory policy. Drawing on the Act of 1995, Pub. Law No. 104-113 (1996), declares that 'all Federal agencies and departments shall use technical standards that are developed or adopted by voluntary consensus standards bodies...as a means to carry out policy objectives or activities' unless doing so would be 'inconsistent with applicable law or otherwise impractical.' Reports issued by the National Performance Review (1993), Carnegie Commission on Science, Technology and Government (1993), National Academy of Public Administration (1995), President's Council on Sustainable Development (1996), Joint Presidential/Congressional Risk Commission (1997), and the Enterprise for the Environment (1998) -- to name just a few -- all recommend expanding the use of various forms of consensus-building. 3 experiences of several recent consensus-building processes, I examine what is to be gained from the institutionalization of consensus-building -- as well as what is to be lost. Even though I seek to raise questions about consensus in this chapter, I recognize that it does, in principle, hold a certain allure. Reaching consensus implies that people have worked out their differences and come to a collective decision. Consensus conjures up notions of teamwork, community, and harmony, all attractive ideas in themselves. Yet as alluring as consensus may be in principle, any widespread institutionalization of consensus-building as a basis for policy-making would mark a significant shift in prevailing modes of governmental decision-making in the United States. Such a shift, I argue here, appears neither necessary nor wise. It is not necessary because the benefits attributed to consensus-based processes can be obtained from other forms of public participation which do not revolve around a quest for consensus. It is not wise because reliance on consensus as a decision rule exposes policy-making to new sources of failure and fosters unrealistic expectations for governance in a complex political system. I. Consensus and Regulatory Policy At the outset, it is important to be clear about what consensus means. Contemporary policy lingo can create confusion about the different ways to involve nongovernmental actors in policy-making. A host of related terms are casually tossed about: 'stakeholder involvement,' 'outreach,' 'partnerships,' 'consultation,' 'public participation,' 'constructive engagement,' 'collaboration,' 'regulatory negotiation,' 'policy deliberation,' 'consensus.' Far too seldom are these terms defined with any precision. In order to evaluate the relative contribution of consensus, it is necessary to clarify what it means and how it might differ from other, related terms. Consensus commonly means unanimity or, at a minimum, something that everyone can 'live with,' even if it is not the ideal policy that everyone would want. The Negotiated Rulemaking Act defines 'consensus' as a 'unanimous concurrence' of the interests represented on a negotiated rulemaking committee, or any lesser agreement that has been unanimously agreed to by the committee.4 The achievement of such a unanimous agreement is the defining feature of negotiated rulemaking. Before convening a negotiated rulemaking committee, agencies are required to consider whether a committee could be formed consisting of 'persons who. . .are willing to negotiate in good faith to reach a consensus.'5 Once formed, the committee is legally obligated to 'attempt to reach a consensus.'6 Outside of federal negotiated rulemaking, 'consensus' has also been stipulated by statute and practice to mean unanimity.7 The EPA established the Common Sense Initiative (CSI) in 1994 as a vehicle for 'reinventing' environmental regulation in six industrial sectors. The Initiative, which ran until 1998, was overseen by an advisory 4 5 U.S.C. § 582. 5 5. U.S.C. § 583 (a)(3). 6 5 U.S.C. § 586 (a). See Omnibus Consolidated Appropriations Act of 1997, Pub. Law. No. 104-208 § 201(a)(6) (1996) (defining 'consensus' in the procedures for the Upper Klamath Basin Working Group to mean 'unanimous agreement by the Working Group members present'). Where 'consensus' has been defined in state negotiated rulemaking legislation, unanimity has also been the characteristic feature. See Mont. Code Anno. § 2-5-103 (1995); Tex. Govt. Code § 2008.056 (1997); Neb. Rev. Stat. § 84-923 (1994). 7 committee comprised of representatives from the various industrial sectors, environmental groups, and federal, state, and local government. The operating principles of the advisory committee, called the Common Sense Initiative Council, declared that the Council would 'operate by consensus decision-making,' which meant that decisions were 'reached when all Council members at the table can accept or support a particular position' (US EPA 1996). In a similar way, some of the EPA's Project XL initiatives have conceived of consensus in terms of the unanimous agreement of interests involved in the consultation processes that accompany these projects. Under Project XL, the EPA may waive certain regulatory requirements for individual firms that can demonstrate that alternative technologies or processes would allow them to achieve superior environmental performance. The EPA will approve Project XL waivers only after negotiating an agreement with the regulated firm and gaining the support of local community and environmental groups.8 Understood to require unanimity among designated interests, consensus specifies a decision rule. It therefore can be distinguished from other terms -- such as 'stakeholder involvement,' 'constructive engagement,' and 'public participation' -- which describe deliberative efforts that are not dependent on a particular decision rule. Stakeholders can be 'involved' by giving them a chance to be heard, giving them a vote, or giving them a veto. Only the last of these involves consensus. Consensus-based processes are those deliberative efforts that seek an agreement among all the participants. Conceived this The EPA has not required formal unanimity among affected parties for all of its XL projects; however, serious opposition to an XL proposal will usually mean EPA will not approve it. For a description of the consensus-based endorsement processes used in the Intel and Merck XL projects, see US EPA (1998a). 8 way, the term 'collaboration' can be used synonymously with consensus. For example, the Environmental Protection Agency, in a recent draft manual, defined collaboration as 'a joint endeavor, a sharing in the process, and its goal is working together towards an agreement -- consensus' (US EPA 1998b:8).9 Unfortunately, consensus is sometimes characterized in such a way that it might appear to be the only alternative to governmental fiat. Such characterizations are obviously misleading. Regulatory agencies can (and do) incorporate extensive public consultation into their decision-making processes without needing to strive for consensus.10 Regulators can infuse the regulatory process with public deliberation in three conceptually distinct ways: • Feedback. When public participation serves as a check on decisionmaking that the government has already initiated, we can consider such public involvement to be a form of feedback. The bare bones, notice-and-comment rulemaking procedures of the Administrative Somewhat confusingly, 'collaboration' is also sometimes used to refer to meaningful deliberation. For example, the Presidential/Congressional Commission on Risk Assessment and Risk Management (1997:17) has stated that in its view '[c]ollaboration does not require consensus, but it does require that all parties listen to, consider, and respect each other's opinions, ideas, and contributions.' 9 Policies based on consensus are certainly responsive to public input, but policies can be equally responsive -- if not more so -- without being based on consensus. Since it is normally impossible for policy-makers to secure the agreement of all the firms and individuals affected by a policy, especially in the area of environmental policy, consensus-based processes are inherently limited in terms of who participates. To the extent that agencies base their decisions solely on agreements reached between a limited segment of the affected public, consensus-based policy may very well turn out to be less responsive than decisions that follow a wide-ranging, open deliberation which is unconstrained by the need to reach agreement. 10 Procedure Act11 reflect this notion of public involvement. The Act merely requires agencies to provide notice of a regulatory proposal and to give members of the public an opportunity to comment on it before a regulation becomes final. • Input. Unlike feedback, input occurs before or contemporaneously with governmental decision-making. Members of the public provide their views as the problem is being framed and possible policy solutions are being developed. Such input can be sought by the government itself or it can be initiated by the parties. It can also be collected individually from each interested party in one-on-one conversations with government officials. Or it can be gathered collectively, in roundtables or dialogue sessions which allow multiple parties to engage each other in a conversation. However conducted, input processes aim to help government officials make more informed decisions. • Consensus. A consensus-based process will typically involve collective input, but the ultimate aim of the conversation is different. With consensus, the goal is to establish agreement among all of the participants, with the expectation that the government will use that agreement as the basis for its policy decision. 11 5 U.S.C. § 553. In questioning consensus, I am not challenging the idea that public feedback or input is desirable. Rather I am asking whether we should design policy-making in such a way as to seek agreement before setting public policy. Should we, in other words, structure domestic regulatory policy-making along the lines of NATO or UN Security Council decision-making? As I have already noted, it is increasingly suggested that we should. A recent report by the Keystone National Policy Dialogue on Ecosystem Management (1996), for example, asserted that '[c]learly, consensus is the most desirable outcome of a collaborative process.' A US EPA (1997) paper describing the Common Sense Initiative stated that the initiative's 'first priority is to craft agreements that parties support, accept, or are neutral on.' Yet giving high priority to consensus as a prerequisite to policymaking would markedly shift the prevailing norms of governance in two ways. First, an emphasis on consensus would 'de-center' the state. The government would no longer be, in practice or in theory, the central, accountable decision-maker but instead would become just a facilitator of bargaining between interest groups, or at most just another player in that bargaining game (Werhan 1996).12 Second, a focus on consensus would shift the aim of policy-making away from that which will serve the public interest to that which will be agreeable to those interests that are well represented in the political process (Funk 1997). Negotiators and facilitators, after all, are not analysts seeking to craft the To be sure, policy-making in the United States has long involved bargaining (Dahl 1956) and much regulatory litigation is resolved through negotiated settlement agreements (Coglianese 1996). However, the reliance on formal agreements to develop and implement regulatory policy has at least until now remained rather limited. 12 best solutions to public problems. Indeed, sometimes they are skeptical of whether policy-makers ought to strive to make correct decisions at all.13 As a result, when reformers describe consensus as 'the most desirable outcome' and speak of securing agreement as 'the first priority,' they signal a significant shift away from norms which heretofore have made the development of sound public policy and the advancement of social values the most important priorities for government decision-makers. II. Is Consensus Necessary? Since the frequent pursuit of consensus would mark a shift in the prevailing practice of governance in the United States, it is helpful to consider first whether making such a shift is even necessary. The argument for making such a shift hinges on various claimed benefits that consensus can bring to policy-making, namely that consensusbuilding holds instrumental advantages over processes based on feedback or input. Philip Harter (1997), for example, has argued for consensus over what he calls 'consultative processes' -- that is, deliberative processes which do not seek consensus. The crux of his argument is that processes of consultation lack many of the benefits attributed to consensus.14 Consensual processes, it is often claimed, will reduce conflict, increase For example, Philip Harter (1983:475) has suggested that agencies should not think of themselves as seekers of correct decisions: 'The agency retains a wide range of discretion and is called on to make choices that are inherently political. Political choices, however, have no "right" or "wrong" or even "rational" answer.' 13 Harter (1997:1411,1420) has argued that 'the dynamics of the process change markedly if either the definition of consensus is modified to require less than unanimity 14 compliance, improve public policy, and promote public participation. In this section, I consider whether the goal of reaching a consensus is necessary in order to achieve these principal benefits attributed to consensus-building. For each advantage attributed to consensus-building, I conclude that the same benefits come from (or can come from) something other than a quest for consensus. A. Reduced Conflict Perhaps the most intuitive benefit attributed to consensus is the reduction of conflict. If all the interested parties can come to an agreement, or at least commit not to disagree, then it would appear that conflict should be eliminated. The intuition is that affected organizations will not subsequently challenge a policy with which they have concurred. Coming to a consensus, however, is certainly not the only way to avoid conflict. A potential conflict arises, by definition, when a policy does not satisfy the interests of affected individuals and groups. Government officials can avoid conflict by learning the interests of the various parties and seeking to craft a policy that addresses these interests. Most sophisticated government officials already do this all the time. That helps explain or, if no attempt is made to reach full closure, to require a commitment to adhere to the agreement' and that 'many of the benefits of the process are lost' by a reliance on consultation instead of consensus. Cornelius Kerwin (1999:183) has similarly noted that 'advocates of negotiated rulemaking are skeptical of partial substitutes and decry the loss of commitment that goes with them.' why, contrary to popular conceptions, most policy decisions are not challenged in court.15 To avoid litigated conflict, government officials need not fully satisfy all the interests of all affected individuals and organizations. Such a task is almost always impossible anyway. Rather, they need only design a policy that those affected are willing to 'live with.' Proponents of negotiated rulemaking have pointed to the averaging provision in EPA's reformulated gasoline regulation as an example of a key conflict-avoiding innovation that came about because of a pursuit for consensus (Weber 1998).16 The provision gave refiners more flexibility by allowing them to meet fuel standards on average over entire stocks of fuel rather than gallon by gallon. In return for this flexibility, refiners were required to meet average standards that were 10 percent more stringent, thus allaying some environmentalists' concerns. Philip Harter, one of the leading defenders and practitioners of negotiated rulemaking, has argued that the averaging approach, combined with a somewhat more stringent standard, was a significant innovation that EPA would not have developed had it not needed to find consensus.17 Environmental regulators, though, did not need a consensus-driven process to lead them to conceive or adopt this averaging plan. The emissions trading policies that Despite assertions made by many knowledgeable observers that almost all EPA regulations are challenged in court, only a minority ever are (Coglianese 1997). 15 16 For the text of the reformulated gasoline rule, see 54 Federal Register 7716 (1994). Harter advanced this argument in remarks delivered to the administrative law section of the Association of American Law Schools at its meeting on January 9, 1998, in an effort to continue to advocate the use of negotiated rulemaking notwithstanding research showing that negotiated rulemaking has failed to achieve its purposes of saving time and reducing litigation. 17 EPA adopted more than a decade before the reformulated gasoline rule all relied essentially on averaging.18 The EPA used averaging for fuel standards at least since the 1970s, and such an approach was integral to the EPA's phasedown of leaded gasoline in the 1980s (Hahn & Hester 1989). Moreover, agency officials hardly needed to conduct formal negotiations to realize that environmentalists would more readily support averaging if it was accompanied by lower standards. The EPA had adopted essentially the same kind of averaging approach, with a corresponding 20% reduction for the environment, four years earlier in its regulations governing the trading and banking of emissions from heavy duty diesel engines.19 In this earlier case, as with other marketbased policies, EPA officials developed the averaging policy through routine, informal consultations with industry and environmental groups, not through any formal negotiation that sought consensus. Attentive and savvy government officials pursue such innovative arrangements and other strategies to reduce conflict even without formally negotiating agreements. Having groups sign on to an agreement may well be one way to determine what the parties can accept and to design policies that resolve conflicts, but it is by no means the only way. Moreover, in the normal course of policy-making interest groups do not pursue every potential conflict they may have with an agency. A considerable amount of 'lumping it' always occurs and parties will often forgo their opportunity to contest each minute aspect of a policy with which they might disagree (cf. Felstiner, Abel, & Sarat For a review of emissions trading and other market-based policies, see Hahn & Stavins (1991). Administrative lawyers should already be quite familiar with emissions averaging, as such an approach undergirded the EPA's 'bubble' rule which was challenged in Chevron v. Natural Resources Defense Council, 467 U.S. 837 (1984). 18 19 55 Federal Register 30,584 (July 26, 1990) 1980-81). An organization's disagreement with an agency must be sufficiently great to justify the costs of a subsequent challenge as well as the risks of creating an adverse decision or prompting those with opposing interests to raise challenges of their own (Cooter & Rubinfeld 1989). For this reason, consensus is not only unnecessary for avoiding conflict, but it can be counterproductive as well. Processes that seek out consensus can actually make conflict more protracted. In order for representatives of industry and citizen groups to 'sign on' to a policy they need to achieve an outcome which is tangibly better than what they would otherwise have received. All things being equal, it is always harder to convince representatives of constituent-based organizations that they should affirmatively endorse a policy rather than simply forego raising objections or legal challenges to that same policy. Thus, it is not surprising that conflicting views have arisen over the meaning of 'superior environmental performance' in EPA's Project XL, with environmentalists insisting that to gain their support firms must achieve performance superior to what their facilities have already achieved (which is often cleaner than the regulations allow) and not merely performance superior to what is legally allowable. Industry also finds it harder to make affirmative endorsements than to offer tacit acceptance. Representatives of a utility trade association pursued difficult negotiations to secure a preliminary agreement with the Nuclear Regulatory Commission on proposed regulations, only later to balk at formally 'signing on' to the agreement in the face of objections from some of the trade association's corporate members (Harrington 1994). Conflict can best be avoided if state officials listen carefully to the concerns of affected parties and craft policies that address these concerns. Once organizations are put on the spot and expected to affirm a policy, they will likely demand an even better outcome for their interests. If many of the participants in a negotiation act in this way, it will prove still more difficult under a consensus process to resolve the conflicts between groups. This helps explain why consensus-based processes tend to consume more time and resources for everyone involved. B. Increased Compliance A related benefit sometimes attributed to consensus-building is that it will increase compliance by the regulated industry. People tend to be more likely to follow through on those things for which they claim ownership. So if a regulated industry signs onto a consensus policy, it would be predicted to be more likely to comply with that policy, as well as to comply more quickly and more fully than it otherwise would. The existence of subsequent litigation over negotiated rulemakings seems to draw this prediction into question, for it shows that a 'buy in' to a consensus agreement does not necessarily mean complete 'buy in' to the final policy. Environmental regulations developed using consensus-based procedures have resulted in more litigation than have comparable regulations promulgated using other forms of public participation (Coglianese 1997). Of course, litigation is not direct evidence of the specific compliance effects of a consensus decision rule, but unfortunately researchers have yet to study systematically whether consensus-based policies elicit greater compliance. That said, there are plausible grounds to question whether consensus is all that significant in promoting industry compliance with a policy decision. Compliance rates are affected by any number of factors, such as industry knowledge of a standard, the cost of complying with the standard, the probability that noncompliance will be detected, and the penalties for noncompliance. When a regulation is backed by effective monitoring and the possibility of penalties of ten to twenty-five thousand dollars per day (as are many environmental regulations), it is not unreasonable to ask whether the mere fact that the policy emerged from a consensus process should matter at all. Lead has been phased out of gasoline, passive restraints have been installed in cars, and cigarette smoking in some states has been virtually eliminated from public buildings, all without attempts at consensus-building preceding the policy decision. Effective compliance can certainly be achieved without efforts at building consensus. At most, the 'buy in' that accompanies consensual decision-making process may increase compliance at the margins. Yet we still have to wonder whether even this marginal effect exists. The 'buy in' that some consensus processes require, after all, is simply a willingness to let the policy move forward. A decision that an industrial sector can 'live with' is not the same as a decision it affirmatively likes. It is not altogether clear why organizations which are simply willing to 'live with' a policy developed through negotiation would have all that much more incentive to comply with it than with a policy that was not negotiated. It is also far from clear that any effects of a 'buy in' by those sitting around the table in Washington, D.C., will necessarily carry through to the individuals across the country who have, at the ground level, the day-to-day responsibility for complying with government regulation. When consensus is used as the basis for industry-wide or sectorwide regulation, can we expect that the plant-level managers who carry out the implementation and monitoring of the regulation will take 'ownership' of that regulation? Will they even know that the regulation was negotiated with representatives from their industry's trade association in Washington? More empirical research obviously needs to be done to answer these questions, but it clearly is not self-evident that the 'buy in' to a consensus regulation will have any substantial effect on compliance. C. Improved Policy Even if the 'buy in' prediction does not hold, consensus might still affect compliance indirectly by the kind of policy that is likely to emerge from a consensus process. Perhaps consensus-based policies are simply better and more sensible because they are based on better information and a more realistic understanding of the specific demands of an industry and concerns of affected citizens. Proponents of consensusbuilding have increasingly argued that such processes will lead to better policy decisions.20 The intensive discussions with regulated industry, as well as the give-andtake dialogue with the other parties, brings detailed information to the table that should lead to a better policy decision. A consensus-based process, it is argued, takes advantage of the collective wisdom of those who are sitting around the negotiating table, as opposed to relying mainly on agency staff's best guess of plant conditions or other technical Harter (1997:1418) has asserted that '[a] negotiated rulemaking forces the parties to bring an enormous amount of practical information to the table and hence expands the data base on which to build a regulation. The practical insight contributed by those with first-hand experience also allows agency staff to focus resources on areas with the greatest potential payback.' 20 aspects of industry operations. Consensus-building efforts are therefore thought to promote learning and yield more informed decisions. Empirical evidence to sustain the claim that consensus-based processes yield systematically better policies has yet to emerge.21 Moreover, as I discuss in Part III of this chapter, several pathologies can afflict consensus-based processes which will lead to inferior policy results. The existence of these pathologies alone provides reason to doubt whether consensus-based processes will tend to lead to better policies. Yet there are still other reasons to question whether structuring a dialogue around a quest for consensus will indeed yield full disclosure and debate of policy issues. The fact that the group is charged with the task of achieving a consensus may actually inhibit some participants from raising important issues which seem at the time likely to hinder consensus-building. In his study of groupthink in government, Paul 't Hart (1994:293) writes that when policy decisions are based on consensus some participants 'may refrain from voicing their concerns, either by self-discipline and a desire not to shatter group harmony (suppression of doubts) or following direct hints by the leader (compliance) or by fellow group members (mindguards; peer pressure). When consensus is no longer required, group discussion can be more open.' In regulatory negotiations, such inhibition does occur. In one case, for example, an EPA official told me that he knew industry was overlooking an entire category of equipment in setting consensus-based standards for equipment leaks, Langbein and Kerwin (1998) report findings that show participants in negotiated rulemakings tend to rate the quality of the final outcome higher than do participants in other rulemakings. However, it is far from clear that ratings by participants are unbiased measures of the actual quality of the policy outcomes achieved in these cases. Cognitive dissonance seems likely to explain the favorable outcome ratings participants give to negotiated rulemakings, as these proceedings involve considerably more time and effort on their part. 21 but that he never said a word about it during the negotiations. In another illustrative case, a citizen member of the Intel Project XL negotiation group reportedly signed the final agreement, but only reluctantly after 'feeling pressure from all sides' (US EPA 1999:Appendix 1). Often what decision-makers need is conflict to illuminate policy issues most fully. The full articulation of opposing views may provide more useful information on which to construct public policy than the truncated discussion that can develop when individuals feel pressured to achieve consensus. Nevertheless, even if consensus-building processes do yield better, more informed decisions, the question remains whether this benefit derives from, or depends on, consensus itself. On its face, it is the deliberation -- not the consensus -- which advocates claim yields the additional information needed to craft better policy decisions (Freeman 1997:40). Consensus-based procedures certainly can demand a lot more time and resources on the part of participants than other procedures. If this same amount of time and effort were devoted to policy deliberations that did not aim at consensus, it seems quite plausible that the informational benefits would be the same, if not better. As we know, alternatives which do not aim at consensus do exist. To the extent that public officials employ deliberative processes that lead interested parties to identify their areas of agreement and disagreement, this kind of intensive deliberation can provide comparable, if not even superior, results in terms of contributing to better public policy (cf. Reich 1985). D. Expanded Participation It has sometimes been suggested that a consensus rule, along with a commitment by the agency to implement the consensus decision, is needed in order to attract people to participate in deliberative processes.22 Consensus, according to this argument, amounts to something like a 'field of dreams.' If you seek to build consensus, the players will come to the table. If consensus is not the main goal, few will engage in the sport. While it is certainly the case that in order to have a policy deliberation individuals must be motivated to deliberate, a quest for consensus is by no means necessary in order to achieve that motivation. A few years ago I helped facilitate a pilot workshop convened by the National Performance Review and EPA's Region 1 office to which members of industry, the environmental community, and local government were invited. This was to be the first in a series of workshops across the country through which EPA would 'get in touch' with the grassroots.23 Although the goals of this particular meeting were not specified at all, and only short notice was given of the meeting, well more than fifty participants crowded into the workshop room for the day long event. They needed not the slightest assurance that they would be there to reach a consensus which the EPA would implement, but rather took advantage of this opportunity to communicate their concerns with regulators. For example, Jody Freeman (1997:92) argues that 'agencies must presumptively commit to agreements developed in these processes, and courts must presumptively defer to them. If not, participants will never engage in the prolonged negotiation and planning required to produce either a consensus rule or an FPA.' In a similar vein, Harter (1997:1411) claims that '[t]he dynamics of the process change markedly if either the definition of consensus is modified to require less than unanimity or, if no attempt is made to reach full closure, to require a commitment to adhere to the agreement.' 22 For a discussion of the lack of focus in the Clinton Administration's early partnership effort, see Sparrow (2000). 23 The possibility of influencing policy decisions is what drives participation in policy-making. It is well documented that the past thirty years have seen considerable growth in interest group representation in Washington, D.C. (King & Walker 1991). These groups and their representatives have come to play the game even though consensus has only recently become a trendy -- even though still relatively infrequent -way to play. Interest group representatives already have the incentives to engage in policy deliberations without a consensus decision rule. It turns out, in fact, that most of what the EPA considers its 'stakeholder involvement' projects do not formally adopt a consensus rule. What group representatives need in order to be motivated to play is an assurance that some decision is going to be made and that their input can help influence that decision.24 Agencies can provide that assurance without requiring the parties to search for consensus and agreeing to implement a consensus that develops. Agencies can -- and do -- show that they have an open mind, that they have a decision yet to be made, and that they are genuinely interested in learning about the various perspectives (Applegate 1998). Participation can be sought in a collective forum which allows for a give-and-take dialogue as long as the agency demonstrates a willingness to listen and, as best it can while still pursuing its public mandate, to incorporate what it learns into its decisions. As I discuss in Part III, consensus rules can sometimes create unrealistic expectations about what an agency is able or willing to deliver, and when these expectations are dashed it can actually lead to less willingness in the future to engage in public deliberations. Ironically, when a consensus decision rule is adopted, it may actually turn groups away from participating. Some of the sector projects in the Common Sense Initiative floundered for lack of interest once players saw that little would be accomplished given the requirement for unanimity imposed on the initiative. 24 Policy-making by consensus can also lead, ironically, to the exclusion of some affected parties from the decision-making process in order to facilitate agreement (Beardsley, Davies & Hersh 1997; Rossi 1997; Harrison 1999). The best way for an agency to motivate outside groups to participate in policy deliberations seems to be to set realistic expectations, meet them, and demonstrate on a repeated basis that the deliberations have an impact on -- even if they do not control -- the agency's final decisions. Promises about consensus cannot stack up against an agency culture that takes deliberation and consultation with affected parties seriously, even though the agency maintains its position as the ultimate decision-maker. When it comes to encouraging public participation -- as with reducing conflict, increasing compliance, and developing better public policy -- a consensus decision rule simply is not needed. Alternative forms of engagement with the public can yield the same benefits that have been attributed to consensual policy processes. Advocates of consensus-building have overstated the need for consensus by de-emphasizing alternative forms of public participation that do not depend on consensus. III. The Pathologies of Consensus Just as the recent literature on regulatory consensus-building has overstated the need for consensus, it has also tended to understate the hazards of consensus as a decision rule. Perhaps this has been because those writing about consensus have largely sought to convince legislators and regulatory agencies of the value of formal negotiated proceedings -- and possibly also of the need for professional facilitation services which some of those writing in this area provide (Rabe 1988). For example, Philip Harter, a leading advocate for and practitioner of regulatory negotiation, argues that consensusbuilding leads to 'regulatory actions [that] are often simply better by virtually any measure. These are, indeed, powerful tools' (Harter 1997:1423). The generally one-sided view that emerges from contemporary advocacy of consensus-building in the regulatory process, however, is not reflected in the broader literature on group decision-making and policy deliberation. There is, for instance, considerable work in social psychology examining both the strengths and weaknesses of group decision-making, one of the latter being the potential for 'groupthink' (Janis 1972; Hart 1994). In addition, the work of Jane Mansbridge (1980:163-82,252-69) on democratic decision-making illuminates certain strengths of consensus decision-making for small groups with relatively homogeneous interests, but also stresses substantial disadvantages of consensus-building used more widely. Mansbridge studied participatory decision-making in both a small workplace and the government of a small New England town and found that decision-making by consensus presented numerous disadvantages. It demanded considerably more time, opened decisions up to more frequent revision, generated ambiguity and imprecision, and sometimes resulted in deadlock or social coercion. She also confirmed that consensus tended to bias decisions in favor of the status quo and strengthened those who were already powerful. Mansbridge (1980:293) concluded that even though consensus may be suitable for the governance of small groups of individuals who have ongoing relationships and common interests, it is not suitable for governance of large nation-states or in highly conflictual settings. The disadvantages of consensus in the context of regulatory policy have not received comparable consideration in the literature. However, experience is showing that the same kinds of pathologies elsewhere associated with consensus decision-making also find their way into the regulatory process. Even those who otherwise support consensusbuilding acknowledge that decisions agreed to by select groups of policy actors need not necessarily comport with the public interest (e.g., Langbein & Kerwin 1998; Weber 1998). A consensus decision rule can create or exacerbate at least six pathologies in the policy process: (a) tractability having priority over public importance; (b) regulatory imprecision; (c) the lowest common denominator problem; (d) increased time and expense; (e) unrealistic expectations; and (f) new sources of conflict. To be sure, not every consensus-based process will suffer from these pathologies, and just as surely some non-consensual processes will. Nevertheless, each of these problems, elaborated in the sections that follow, derive from a quest for consensus and their risk is increased when decision-making procedures effectively hand each participant a veto over the policy decision. A. Tractability over Importance The first pathology created by a focus on consensus relates to the nature of the issues selected for consideration. Consensus-based processes increase the likelihood that the wrong issues will receive attention. Instead of devoting time and resources to the issues of most importance to the public, a focus on consensus tends to lead to the selection of the most tractable issues, the ones most amenable to agreement. That such a selection process occurs is evident in the paucity of cases in which consensus has been used to develop federal regulations. Proponents of negotiated rulemaking have never claimed that consensus-building would be appropriate for more than about five percent of all agency rulemakings, and in practice the use of the procedure has been exceedingly rare (less than one-tenth of one percent of all regulations). The small fraction of rules that agencies have selected for negotiated rulemaking has not been comprised of the rules with the largest impact on the public. For example, only five negotiated rulemakings have been classified as 'major' or 'significant' rules according to the standards set forth by executive order (Coglianese 1997). The Negotiated Rulemaking Act sets forth standards for selecting rules for negotiation, most of which guide agency officials to select rules that are most likely capable of resulting in a consensus. Among other things, agency officials contemplating negotiated rulemaking are required to determine that there is '[a] reasonable likelihood that a committee will reach a consensus on the proposed rule within a fixed period of time.'25 Standards such as these place a primacy on tractability over social importance. An emphasis on consensus not only leads to the selection of more tractable policy matters for negotiation to begin with, it also leads to a selection of the more tractable issues within the negotiating proceedings themselves. The Quincy Library Group, a high profile group organized to develop a consensus over forest policy in California, focused on those issues where agreement was possible. As two participants acknowledged, 'true consensus' as a decision rule 'greatly limits the range of issues the group can take on' 25 5 U.S.C. § 563(a)(4) (1994). (Terhune & Terhune 1998). The subset of issues addressed in consensus-building are typically tractable ones, not necessarily the ones that are most important. The problem with tractability is also evidenced in the recent report of the Enterprise for the Environment (1998) initiative. In 1996, former EPA Administrator William Ruckelshaus convened this initiative, dubbed 'E4E' for short, in order to bring together leaders from industry, government, and the environmental community to forge a consensus about how to improve environmental policy in the United States. The project initially sought agreement on a diagnosis of the problems in the current system of environmental protection and on a set of concrete legislative solutions. Not long into the discussions, however, it became apparent to the participants that consensus would never be achieved on either the specification of current problems or the precise form of legislative proposals. Even though an illumination of existing problems and specific legislative fixes was surely what was needed, the group shifted its goals to what was more attainable (but ultimately less valuable): agreement on a broad 'vision' of an ideal environmental protection system.26 It seems a truism that consensus will be more difficult to achieve on the most vexing problems. In evaluating policy 'experiments' which rely on consensus-building, one therefore needs to be mindful of potential selection bias. This is especially true with initiatives that rely on volunteers. Not only are the firms that volunteer for initiatives such as Project XL already likely to be ones that have better-than-average relationships with their communities, the types of problems that they initially choose to address are not 26 For a discussion of the E4E process and report, see Coglianese (1999). likely to tend towards the most complex or challenging.27 This raises a question. If sitespecific regulation of the kind envisioned by proponents of Project XL really can improve efficiency and innovation, should resources go to projects based on whether they do (or do not) draw initial support (or lack opposition) from various citizen groups? At least in the early years of the XL initiative, a community group could write a strongly worded three-paragraph letter asking the EPA 'to scratch this company from the list immediately'28 -- and the project would never go forward. The EPA now assures XL applicants that a single community group cannot kill an application, but the agency nevertheless maintains that stakeholder support is a critical criterion for selecting projects.29 Certainly there will still be opportunities to make cost-saving environmental improvements in the projects that garner community support and are selected for Project XL or other reform efforts. However, to the extent that consensus (or even just the absence of controversy) drives decision-making, it is likely that the opportunities for gains will be smaller than if other selection criteria were used.30 Blackman and Mazurek (1999) have noted that the EPA's Project XL has been biased against the most complex and innovative projects. 27 The quoted language comes from a letter from Denny A. Larson, Communities for a Better Environment, to President William Clinton, dated Oct. 26, 1995, in which Larson expressed his organization's opposition to an XL proposal submitted by Citgo Corporation. The EPA did not pursue Citgo's proposal. 28 29 Notice of Modifications to Project XL, 62 Fed. Reg. 19872 (April 23, 1997). Of course, in saying this, I am mindful that community support (or lack of significant opposition) could be thought necessary given the legal vulnerability of XL agreements and site-specific rulemakings (Caballero 1998). However, I am also mindful of a small survey of participants in four XL projects which resulted in at least one interesting finding: nearly half of the respondents thought that the deliberation process neglected issues that should have been addressed (US EPA 1998a:32). Had the EPA not 30 B. Imprecision Just as a focus on tractability makes consensus easier to achieve, so too does imprecision or ambiguity. In her study of democratic decision-making, Mansbridge (1980:167) found that '[c]onsensual decision making. . .generates imprecision. In order to reach unanimous agreement, groups formulate their collective decision so as to blur potential disagreements.' The E4E process, mentioned earlier, is a stark example of the pathology of imprecision that comes from a quest for consensus. The final Enterprise for the Environment (1998:3) report described the project's resulting consensus in terms with which no one could seriously disagree: '[T]he environmental protection system of the next century must become as efficient and low cost as possible without compromising environmental progress.' Elsewhere the report offered other platitudes as recommendations: policy-makers should 'adapt and adjust policies, strategies, and systems based on experience and new information;' they should 'generate, disseminate, and rely on the best-available scientific and economic information;' and society should 'place authority, responsibility, and accountability at the appropriate level of government.' Of course, no one would seriously urge otherwise, although different people do disagree about specifically how best to achieve better environmental protection at lower cost. Rather than seek consensus for its own sake, what was needed was to illuminate areas of disagreement and to conduct further analysis that might better inform decision-making. placed such a priority on securing stakeholder support, there would presumably have been less pressure to truncate the deliberative processes in these cases. Admittedly, the language found in a consensus-based policy report reflects an extreme case, and the regulations and environmental agreements that have been crafted using consensus-based processes have been clearer and more specific. However, the pressure always exists that, in order to secure an agreement, negotiators will adopt abstract or unclear language. It will usually be easier to achieve consensus at higher levels of abstraction, and it is always less time-consuming and less controversial to adopt imprecise language (Diver 1983). Adopting abstract principles and vague standards may serve to secure agreement in the face of conflict, but doing so will constrain the usefulness of the public policy that emerges from consensual processes. C. Lowest Common Denominator By handing each participant a veto, consensus-based processes also make it more likely that the final outcome will amount to no more than the lowest common denominator acceptable to all the parties. Consensual decision rules have the effect of giving domestic policy-making the same structural form as international policy-making. It is common for multilateral international agreements to require no more than what is acceptable to the state with the most objections to regulation. For example, initial drafts of the ISO 14001 environmental management system standards would have required public accessibility of environmental data, third party certification, and sector-specific pollution standards -- requirements that some have argued are needed to make environmental management systems credible. However, these requirements were dropped in response to objections from the United States and Japan (Roht-Arriaza 1995). The problem with the lowest common denominator, of course, is that such a minimallyacceptable outcome will not be enough when a more dramatic decision is needed. In a recent study of negotiated rulemaking, Caldart and Ashford (1999:201) concluded that because industry representatives will not likely 'sign on' to any regulations that would force dramatic changes upon business, 'negotiated rulemaking's focus on consensus can effectively remove the potential to spur innovation.' D. Time and Resources The lowest common denominator problem, along with the pathologies of tractability and imprecision, arise because it takes time and resources to achieve consensus. Deliberation takes time for everyone to present their concerns and for others to respond, and consensus demands that the deliberation continue until everyone agrees (or at least agrees to 'live with' a decision). Time, in itself, is not inherently a pathology, at least not if the additional time yields valuable information and better results. All things being equal, though, the additional time it takes to develop a decision through consensus is certainly an important drawback, especially when it takes longer to reach closure on only the most tractable issues. Those who have participated in consensus-building processes complain about the amount of time and effort they take. In one study, participants in negotiated rulemakings were three times more likely to complain that the process took too much time, effort, and resources than were those respondents who participated in conventional rulemakings (Kerwin & Langbein 1997). One of the most common complaints about the EPA's now- defunct Common Sense Initiative was that it took a frustratingly long time to accomplish anything (Todd 1997). EPA's Project XL has generated similar reactions.31 Studies of negotiated rulemaking confirm that consensus-building fails to save time (Coglianese, 1997; Balla & Wright 1999). No one should expect that decision-making by consensus will help speed up the policy process. E. Unrealistic Expectations By making consensus the goal of participatory processes, public officials can give rise to unrealistic expectations about how much any agreement will affect the ultimate policy decision. Even though widespread reliance on consensus would have the effect of 'de-centering' the state by making it more of a player and facilitator than a decisionmaker, the agreements made through collaborative partnerships are usually not selfimplementing statements of policy. Government officials must still formally enact and implement these agreements. In so doing, the policy may change -- even slightly -- from the proposal on which the parties thought they had agreed. After a consensus is forged, maintaining that consensus throughout the remaining steps of the policy process can prove difficult. Other actors not party to the agreement, such as legislators, other interest According to the US EPA's (1998a:41) review of Project XL, '[m]ost stakeholders commented the process was too long or much longer than they expected or felt was warranted.' To be sure, not all of the time and expense associated with developing XL agreements have been due to the pursuit of consensus, and only some XL projects have sought to achieve consensus from a broad range of affected parties. However, the need to reach an agreement with EPA and other governmental agencies has still contributed to a more costly and time-consuming process than participants would like. 31 groups, and executive branch officials, may also try to take another bite at the apple (Kagan 1997). When this happens and the policy outcome diverges from the agreement, participants in the consensus proceeding will undoubtedly have certain expectations disappointed, expectations that would have been much less likely to have developed had the process simply sought public input to assist the agency in reaching its decision. In a study of several consensus-building initiatives at the National Marine Fisheries Service, sixty percent of the participants who were surveyed reported that they were dissatisfied with the results of the process (RESOLVE 1999). The study's authors found that much of the dissatisfaction arose because participants expected to control the outcomes much more than was realistically possible. It was also precisely such a case of dashed expectations that led to the underlying litigation in the first major appellate decision to interpret the Negotiated Rulemaking Act. In USA Group Loan Servicers, Inc. v. Riley,32 participants in a Department of Education negotiated rulemaking sued the agency claiming that it had reneged on commitments made during the negotiated rulemaking. The Seventh Circuit held that federal agencies could not be compelled to adopt a consensus agreement nor held to positions taken during negotiations because the agency retains the ultimate decision-making authority. To the extent that it will remain possible for government officials to enact policies that depart from the precise (or perhaps not-soprecise) understandings of those involved in policy negotiations, a process centered on the quest for consensus only sets up expectations that in the end probably cannot help but 32 82 F.3d 708 (7th Cir. 1996). be somewhat unfulfilled. In this way, an increased reliance on processes that aim for consensus could very well undermine trust and increase cynicism in the policy process. F. Additional Sources of Conflict The case of USA Group Loan Servicers shows that conflicts not only persist following consensus-building, but that they can even be engendered by the expectations such processes create. Consensus is not always attainable, and even when it is it may only temporarily hide underlying conflicts. Perhaps the most notable disappointment in terms of avoiding conflict has been the EPA's reformulated gasoline regulation. Heralded by some as a 'successful collaboration' (Weber 1998), this negotiated rulemaking hardly succeeded at all in eliminating conflict. The final rule elicited extensive criticism in the press and from the public, prompted four legal challenges and a petition for administrative review, and resulted in an adverse ruling by the World Trade Organization (Coglianese 1997). The reformulated gasoline regulation is not unique. As noted earlier, environmental regulations developed through consensus-based processes overall end up being challenged in court more frequently than do comparable regulations formulated through procedures that do not depend on consensus. Consensus-based processes create new sources of conflict that do not exist with other methods of policy-making. Conflicts first arise over who participates in the negotiations. A recent set of negotiated rulemakings at the Department of Housing and Urban Development (HUD) spawned what appears to be the first legal action filed to secure a spot on a negotiated rulemaking committee.33 HUD had originally named four public housing organizations to participate on negotiated rulemaking committees for regulations addressing subsidies and capital funds. After the housing organizations subsequently filed a petition against the agency over a separate matter, HUD officials declared that the organizations could no longer bargain in good faith and removed them from the negotiated rulemaking committees. The organizations filed for a court order reversing the agency's decision to remove them from the committee, claiming that HUD's action discriminated against them for exercising their fundamental right of petition. HUD subsequently capitulated and reinstated the organizations to the negotiated rulemaking committees, but the experience demonstrates one significant new source of conflict caused by a process designed around the search for consensus. In addition to conflicts over who gets to participate, processes structured around consensus can create conflicts over the meaning of any agreements that are reached and over whether final government decisions comport with those agreements. Disagreements also arise over the meaning of terms in consensus statements as well as over the implications of terms or issues that are absent from these statements.34 Just as with disputes over membership in the consensus process, neither of these additional sources of conflict arise outside the context of consensus-based processes. Council of Large Public Housing Authorities, Inc. v. US Department of Housing and Urban Development, No. 1:99CV00634 (Dist. D.C. March 25, 1999) (motion for a temporary restraining order). 34 For examples of the range of conflicts engendered by attempts to build consensus, see Coglianese (1997). 33 Conclusion A reliance on consensus, I have argued, introduces new sources of conflict and creates additional problems in the policy process. It leads to unrealistic expectations, increased time and resources, lowest common denominators, imprecision, and a focus on tractability over importance. We should therefore not engage in any wishful thinking about consensus-based processes. Even though public officials, scholars, and policy advocates seem to be converging on a new vision of policy-making based on consensus, we should not expect that simply by organizing policy around consensus, defined variously as requiring either unanimity or a decision everyone 'can live with,' we will indeed achieve a more timely, less conflictual, and higher quality system of regulation. The widespread establishment of a consensus-based approach to regulation would constitute a shift in the prevailing mode of governance in the United States, amounting to a 'decentering' of the state and a retreat from the public interest as the primary goal of government officials. In this chapter, I have argued that such a shift is neither necessary nor wise. All the purported benefits of consensus-building can be achieved through other participatory processes which do not aim for consensus. Moreover, basing decisionmaking on a search for consensus introduces new pathologies into the policy process. Enthusiastic calls for consensus, and particularly those efforts to compel agencies to employ consensus-building, are at best premature. Environmental regulators can -- and many increasingly do -- seek to engage in public deliberation in ways that do not impose consensus as a constraint on decision-making. In doing so, they can achieve the same kinds of benefits that have been attributed to consensual decision-making without introducing the pathologies of consensus as a decision rule. References J. Applegate, 'Beyond The Usual Suspects: The Use Of Citizens Advisory Boards In Environmental Decisionmaking' (1998) 73 Indiana Law Journal 903 S. Balla and J. Wright, ‘Consensual Rulemaking and the Time it Takes to Develop Rules,’ (1999) (Paper presented at the fifth National Public Management Research Conference, College Station, Texas). D. Beardsley, T. Davies, and R. Hersh, 'Improving Environmental Management: What Works, What Doesn't' (September 1997) 39 Environment 6 A. Blackman and J. Mazurek, The Cost of Developing Site-Specific Environmental Regulations: Evidence from EPA's Project XL (Resources for the Future Discussion Paper, Washington, D.C., 1999) T. Caballero, 'Project XL: Making It Legal, Making It Work' (1998) 17 Stanford Environmental Law Journal 399 C. Caldart and N. Ashford, 'Negotiation as a Means of Developing and Implementing Environmental and Occupational Health and Safety Policy' (1999) 23 Harvard Environmental Law Review 141 Carnegie Commission on Science, Technology and Government, Risk and the Environment: Improving Regulatory Decision Making (New York, 1993) C. Coglianese, 'Litigating within Relationships: Disputes and Disturbance in the Regulatory Process' (1996) 30 Law & Society Review 735 C. Coglianese, 'Assessing Consensus: The Promise and Performance of Negotiated Rulemaking' (1997) 46 Duke Law Journal 1255 C. Coglianese, 'Limits of Consensus' (April 1999) 41 Environment 28 R. Cooter and D. Rubinfeld, 'Economic Analysis of Legal Disputes and their Resolution' (1989) 27 Journal of Economic Literature 1067 R. Dahl, A Preface to Democratic Theory (The University of Chicago Press, 1956) C. Diver, 'The Optimal Precision of Administrative Rules' (1983) 93 Yale Law Journal 65 Enterprise for the Environment, The Environmental Protection System in Transition: Toward a More Desirable Future (The CSIS Press, Washington, D.C. 1998) W. Felstiner, R. Abel, and A. Sarat, 'The Emergence and Transformation of Disputes: Naming, Blaming, Claiming...' (1980-81) 15 Law & Society Review 631 J. Freeman, 'Collaborative Governance' (1997) 45 UCLA Law Review 1 W. Funk, 'Bargaining Toward the New Millennium: Regulatory Negotiation and the Subversion of the Public Interest' (1997) 46 Duke Law Journal 1351 R. Hahn and G. Hester, 'Marketable Permits: Lessons for Theory and Practice' (1989) 16 Ecology Law Quarterly 361 R. Hahn and R. Stavins, 'Incentive-Based Environmental Regulation: A New Era from an Old Idea?' (1991) 18 Ecology Law Quarterly 1 C. Harrington, 'Howard Bellman: Using "Bundles of Input" to Negotiate an Environmental Dispute,' in D. Kolb (ed.), When Talk Works (Jossey-Bass, San Francisco, 1994) K. Harrison, 'Talking with the Donkey: Cooperative Approaches to Environmental Protection' (1999) 2 Journal of Industrial Ecology 51 P. Hart, Groupthink in Government: A Study of Small Groups and Policy Failure (Johns Hopkins University Press, Baltimore, 1994) P. Harter, 'The Political Legitimacy and Judicial Review of Consensual Rules' (1983) 32 American University Law Review 471 P. Harter, 'Fear of Commitment: An Affliction of Adolescents' (1997) 46 Duke Law Journal 1389 L. Jaffe, 'Law Making by Private Groups' (1937) 51 Harvard Law Review 201 I. Janis, Victims of Groupthink: A Psychological Study of Foreign Policy Decisions and Fiascoes (Houghton Mifflin, Boston, 1972) R. Kagan, 'Political and Legal Obstacles to Collaborative Ecosystem Planning' (1997) 24 Ecology Law Quarterly 871 C. Kerwin and L. Langbein, An Evaluation of Negotiated Rulemaking at the Environmental Protection Agency (manuscript 1997) B. Kerwin, Rulemaking: How Government Agencies Write Law and Make Policy (2d edn., CQ Press, Washington, D.C., 1999) Keystone Center, The Keystone Center National Policy Dialogue on Ecosystem Management: Final Report (Keystone, Colo., 1996) D. King and J. Walker, Jr., 'An Ecology of Interest Groups in America,' in J. Walker, Jr., Mobilizing Interest Groups in America: Patrons, Professions, and Social Movements (University of Michigan Press, 1991) L. Langbein and C. Kerwin, Regulatory Negotiation Versus Conventional Rulemaking: Claims, Counter-Claims, and Empirical Evidence (manuscript 1998) J. Mansbridge, Beyond Adversary Democracy (University of Chicago Press, 1980) National Academy of Public Administration, Setting Priorities, Getting Results: A New Direction for EPA (Washington, D.C., 1995) National Performance Review, Improving Regulatory Systems (Washington, D.C., 1993) Presidential/Congressional Commission on Risk Assessment and Risk Management, Framework for Environmental Health Risk Management (Washington, D.C., 1997) President's Council on Sustainable Development, Sustainable America: A New Consensus for Prosperity, Opportunity, and a Healthy Environment for the Future (Washington, D.C., 1996) B. Rabe, 'The Politics of Environmental Dispute Resolution' (1988) 16 Policy Studies Journal 585 R. Reich, 'Public Administration and Public Deliberation: An Interpretive Essay' (1985) 94 Yale Law Journal 1617 RESOLVE, Inc., The National Marine Fisheries Service Take Reduction Team Negotiation Process Evaluation (National Marine Fisheries Service, Washington, D.C., 1999) N. Roht-Arriaza, 'Shifting the Point of Regulation: The International Organization for Standardization and Global Lawmaking on Trade and the Environment' (1995) 22 Ecology Law Quarterly 479 J. Rossi, 'Participation Run Amok: The Costs of Mass Participation for Deliberative Agency Decisionmaking' (1997) 92 Northwestern University Law Review 173 M. Sparrow, The Regulatory Craft: Controlling Risk, Solving Problems, and Managing Compliance (Brookings Institution Press, Washington, D.C., 2000) P. Terhune & G. Terhune, 'QLG Case Study' (workshop paper, 1998) J. Todd, Review of the Common Sense Initiative (US EPA, Washington, D.C. 1997) US EPA, Operating Principles of the U.S. Environmental Protection Agency Common Sense Initiative Council (Washington, D.C., 1996) US EPA, Consensus Decision-Making Principles and Applications in the EPA Common Sense Initiative (Washington, D.C., 1997) US EPA, Evaluation of Project XL Stakeholder Processes: Final Report (Washington, D.C., 1998a) US EPA, Better Decisions Through Consultation and Collaboration: A Manual on Consultative Processes and Stakeholder Involvement (Washington, D.C., 1998b) US EPA, Constructive Engagement Resource Guide: Practical Advice for Dialogue Among Facilities, Workers, Communities and Regulators (Washington, D.C., 1999) E. Weber, 'Successful Collaboration' (November 1998) 40 Environment 10 E. Weber K. Werhan, 'Delegalizing Administrative Law' 1996 University of Illinois Law Review 423