Environmental Risk and Democratic Process: A Critical Review Daniel J. Fiorino* INTRODUCTION Can a technological society remain a democratic one? The rapid pace of technological change in contemporary society has become almost a cliche. The nuclear power debate has been with us for years. Entirely new fields like genetic engineering pose substantial scientific uncertainties where a political consensus has not even begun to emerge. Affected publics demand a role in technically based decisionmaking, while administrators grope for ways to involve them constructively. Science and expert knowl- edge have not taken the politics out of technically-based policy issues, as many observers expected only decades ago. Instead, the increasing involvement of technical experts in policy disputes has politicized expertise.' A major question is whether democratic institutions and processes can keep pace with these changes. This article argues that our thinking, research, and experience are lagging badly be- hind the demands that environmental risk policymaking presents in a democratic society. We face a participatory dilemma, in which people's expectations about their capacity to influence de- cisions are not matched by political and institutional realities. To the extent that public participation in risk policymaking occurs at the national level, it typically is mediated through interest groups or participation professionals. I argue that the solution lies in reassessing our institutions for managing risk and in designing * Senior Policy Advisor to the Assistant Administrator for Policy, Planning, and Evalua- tion. U.S. Environmental Protection Agency, 401 M Street, S.W , Washington, D.C. 20460. Research and writing for this paper were completed while the author was a Senior Fellow at the Conservation Foundation. The opinions expressed in this article are those of the author and not necessarily those of the U.S. Environmental Protection Agency or the Con- servation Foundation. This is a revised version of a paper originally presented at the Sym- posium on Decision-Making and Environmental Risks: Economic and Political Issues, University of Wisconsin at Madison (April 1988). The author is indebted to Michael Kraft. Terry Davies, and Fran Irwin for comments on a previous draft. 1. For a discussion, see Nelkin, The Political Impact of Technical Expertise, 5 Soc. STUD. Scs. 35 (1975). 501 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 more effective procedures and mechanisms for citizen participa- tion. Although my focus is on participation at the national level, the discussion applies at the local level as well, where par- ticipatory mechanisms may have an even greater probability of success. Participation has been a recurring theme in American political history. Even today, American politics can be understood as a se- ries of historical adjustments to demands for greater participa- tion. Lowi explains these demands over the last century as a response to cycles of expansion in American government. Each expansion of government has created a "crisis of public author- ity," followed by "demands for expansion of representation."2 The growth in federal regulatory powers that began with the In- terstate Commerce Act of 1887 fueled the Progressive and Popu- list movements, which in turn led to Congressional rule reforms, direct election of Senators, and changes in nominating and voting procedures. Early in this century, the progressivist expansion of government was accompanied by women's suffrage; widespread adoption of the initiative, referendum, and recall; the commission form of city government; and the first steps toward formal inter- est representation at the national level (e.g., separate Depart- ments of Labor and Commerce and the farm bureau movement). The New Deal and the aftermath of the Second World War accel- erated the expansion of government, and with it demands for broader participation. Since then, the growth of federal power has been in the administrative sphere, where the demands for participation also have been strongest. It is not hard to argue that institutions managing environmen- tal risk in American society face a crisis of public confidence. Confidence in public and private institutions is low.3 Approaches to solving environmental risk problems work through cycles of ineffectiveness and irrelevance; one set of solutions often only creates another set of problems.4 Attempts to site hazardous 2. T. Lowi, THE END OF LIBERALISM: THE SECOND REPUBLIC OF THE UNITED STATES 61 (1979). 3. This lack of confidence is described in S. LIPSEr & W. SCHNEIDER, THE CONFIDENCE GAP: BUSINESS, LABOR, AND GOVERNMENT IN THE PUBLIC MIND 13-40 (1983). The same authors discuss more recent data in Lipset & Schneider, The Confidence Gap During the Rea- gan Years, 1981-1987, 102 POL. SCI. Q. 1 (1987). 4. Cycles in hazardous waste policy are discussed in Mazmanian & Morell, The Elusive Pursuit of Toxics Management, 90 PuB. INTEREST 81 (1988). 502 Risk and Democratic Process waste facilities stall in the face of intense local opposition.5 Ef- forts to set regulatory standards for toxic chemicals drag on in a spiral of legal challenge, political opposition, and analytical de- bate.6 Cost-benefit analysis, the dominant analytical model that the federal government uses to make decisions about environ- mental risk, is criticized on ethical, philosophical, methodological, and political grounds.7 Despite official efforts to establish the credibility and objectivity of the scientific analyses underlying risk decisions, methods for assessing risk remain a source of contro- versy.8 Research demonstrates fundamental differences between lay perceptions of risk and the perceptions of the experts who shape public policy.9 It seems that each attempt to take the poli- tics out of regulatory science only expands the arena of political debate. The risk community has focused its attention on the technical and economic aspects of policymaking. Yet the challenges to ef- fective risk management may not be so much technical or eco- nomic as political. By political, I mean the ways people view their relationship to institutions making collective decisions about en- vironmental risk and their capacities for influencing those deci- sions. Yet the literature on environmental risk-whether it assumes the label of risk assessment, management, or communi- cation-often ignores this aspect of risk problems and their solu- tions. Much of the research on risk displays "an uneasy 5. Matheny & Williams, Knowledge vs. NIMBY: Assessing Florida s Strategy for Siting Hazard- ous Waste Disposal Facilities, 14 POt'Y STUD. J. 70 (1985); Ristoratore, Siting Toxic Waste Dispo- sal Facilities in Canada and the United States: Problems and Prospects, 14 Pot'Y STUD. J. 140 (1985); and Glasberson, Coping in the Age of NIMBY, N.Y. Times, June 19, 1988, § 3, at 1, col. 3. 6. For a discussion of litigation and its effects, see Harter, Negotiating Regulations: A Cure for Malaise, 71 CEO. L.J. 1 (1982). 7. The critical literature on cost-benefit analysis is substantial. For examples, see K.S. SHRADER-FRECHETrE, RISK ANALYSIS AND SCIENTIFIC METHOD: METHODOLOGICAL AND ETH- ICAL PROBLEMS WITH EVALUATING SOCIETAL HAZARDS (1985) and Zinke, Cost-Benefit Analysis and Administrative Legitimation, 16 Pot'Y STUD. J. 63 (1987). 8. For a discussion of the effects of assumptions and inferences on the results of scien- tific risk assessments, see Whittemore, Facts and Values in Risk Analysis for Environmental Toxi- cants. 3 RIsK ANALYSIS 23 (1983) and Nichols & Zeckhauser. The Perils of Prudence: How Conservative Risk Assessments Distort Regulation, 10 REG. 13 (1986). Similar issues are dis- cussed from a cross-national perspective in R. BRICKMAN, S. JASANOFF & T. ILGEN, CON- TROLLING CHEMICALS: THE POLITICS OF REGULATION IN EUROPE AND THE UNITED STATES 187-217 (1985). 9. Slovic, Perception of Risk, 236 SCIENCE 280 (1987) and Wilson & Crouch, Risk Assess- ment and Comparisons: An Introduction, 236 SCIENCE 267 (1987). 1989] 503 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 relationship with people, tending to treat them as a variable to be considered in analysis, but not as legitimate contributors to deci- sions."'0 We refine, polish, and perfect our formal models for determining acceptable levels of risk, despite evidence that the assumptions and methods bear little relationship to the lay pub- lic's conceptions of problems. We test techniques for communi- cating risk information to the public, but conduct almost no research on mechanisms for the lay public to communicate with government officials and technical experts. We should not be surprised, as one book on risk analysis has observed, that "Citi- zens in a democratic society will eventually interfere with deci- sions in which they do not feel represented."" Part I of this article reviews selections from the literature that consider the fit between risk policymaking and the .democratic process. A principal influence on our current conceptions of participa- tion in the federal administrative process was the participation movement of the 1960s and 1970s. Yet that experience is of lim- ited value in addressing the crisis of confidence that afflicts risk policymaking today. Although the public participation movement promoted substantive democratic values, it did very little to pro- mote the procedural ends of democracy, and may even have cor- roded them, by reinforcing American tendencies toward adversarialness and confrontation. It substituted litigation for discussion, joint problem-solving, and a search for common ground. It reinforced the idea that effective participation oc- curred through the actions of organized interest groups that mo- bilize resources and concentrate influence. In sum, participation took the only form it could be expected to take in a society grounded firmly in the principles and political realities of interest- group liberalism, with consequences that shape our perceptions of the meaning and value of participation today. I consider our contemporary experience with participation and its legacy in Part II. Is there an alternative conception of participation that can help to reconcile risk policymaking with democratic process? Part III draws upon the literature on democratic participation to suggest an alternative approach. This literature offers no easy way out of our participatory dilemma. It suggests that agencies take an ac- 10. Otway, Experts, Risk Communication, and Democracy, 7 RisK ANALYSIS 125, 126 (1987). 11. B. FiSCHHOFF, ACCEPTABLE RISK 148 (1981). 504 Risk and Democratic Process tive rather than a passive approach to participation. It describes an ethic that will require agencies and risk researchers to treat individuals as citizens rather than as subjects. It is a risky ap- proach in itself, because it would grant to citizens as amateurs a greater role in making policy. It suggests a role for analysis, but to inform the political process and not to replace it. It proposes that institutions seek ways to transform conflict rather than to ex- pect and thus to reinforce it. The last section in Part III proposes a participatory ideal as a basis for adapting and designing institu- tional mechanisms and procedures. In Part IV, I propose that we design participation programs as carefully as we now design our analytical programs. In theory, research, and practice, there is a need to reassess assumptions, conduct and evaluate experiments, and adapt existing institu- tions. The sooner we think as carefully and critically about polit- ical values as we do about scientific and economic ones, the better our capacities will be for coping with the participatory dilemma in risk policymaking. I. RISK POLICY AND DEMOCRATIC PROCESS: A SURVEY OF THE LITERATURE How has the literature treated the relationship between risk policymaking and the democratic process? What can it offer in an effort to improve mechanisms for citizen participation? The liter- ature on risk assessment, management, and communication is substantial. This review examines work that considers the rela- tionship between democratic institutions and processes and pub- lic policies for analyzing and managing environmental risks. This section is partly a definition of the problem, because the problem has been defined in many ways: as a matter of resolving the ten- sions between experts and nonexperts; as a problem of informing the lay public on technically complex issues; as the need to re- solve technical or scientific controversy in a way that is acceptable to affected parties, yet true to the standards of science; as the challenge of drawing community values into deliberations over scientific questions that cannot be answered scientifically; and as the dilemma of accommodating the rationality of formal risk as- sessment and evaluation with the political and ethical require- ments of democratic process, among others. This review of the literature is organized according to seven perspectives. These perspectives are proposed as a convenient 1989]1 505 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 way to view the literature, and not as a general classification scheme. Each perspective is defined by a key question that I asso- ciate with each. The seven perspectives are: A. Public Deliberation: How do risk managers engage the public in deliberations over risk issues so that social and community val- ues will inform administrative decisions? B. Resolving Disputes Among Experts: How should we resolve technically-based public policy disputes when the experts disagree? C. The Analytical and Ethical Critique: How do we reconcile the methodological requirements and biases of risk analysis with the ethical and political values of a democratic society? D. Risk Perception: How do people perceive risk, and what are the psychological, social, and cultural determinants of those perceptions? E. Risk Communication: How can the uncertainties and complex- ities of risk analysis be conveyed clearly, objectively, and effec- tively to the lay public? F. System Responses to the Need for Participation: How have political systems responded to growing demands for public participation in or control over technological decisions? G. Administrative Process and Culture: How do other administra- tive systems reconcile scientific with political authority, and what are the lessons for American policy and practice? A. Public Deliberation 1. Description This perspective focuses on how risk managers engage the pub- lic in deliberations so that social and community values will in- form administrative decisions. It is reflected in the institutional position of the Environmental Protection Agency since William Ruckelshaus returned as Administrator in 1983.12 EPA was influ- enced by the National Academy of Sciences (NAS) report on risk assessment, issued that same year, especially its distinctions be- tween risk assessment and management.1 3 Like the NAS report, 12. See Ruckelshaus, Science, Risk, and Public Policy, 221 SCIENCE 1026 (1983); Risk, Sci- ence, and Democracy, IssUES IN SCI. AND TECH., Spring 1985, at 19; and Russell, Environmental Protection: Laying the Foundation for the Year 2000, 4 ENVTL. F., Feb. 1986, at 7. Is. NATIONAL RESEARCH COUNCIL, RISK ASSESSMENT IN THE FEDERAL GOVERNMENT: MANAGING THE PROCESS (1983). For an EPA statement of this approach, see EPA, RISK ASSESSMENT AND RISK MANAGEMENT: A FRAMEWORK FOR DECISION MAKING (1984). 506 Risk and Democratic Process however, the EPA position recognizes that scientific and method- ological assumptions and inferences can greatly affect the results of a risk assessment. EPA's strategy has been to handle these var- iations in two ways-by establishing guidelines that explain and standardize the risk assessment process as much as possible, and by subjecting the assumptions and range of uncertainties underly- ing a risk assessment to public scrutiny.14 EPA's approach to managing risk has stressed public education, participation, and deliberation. Public risk managers have an ob- ligation to conduct the scientific assessment, present the evidence with all of the uncertainties and limitations, lay out the policy choices and the implications of each, and bring the affected com- munity into the process of choice. Inherent in this approach is a Jeffersonian faith in the capacity of the public to take part or be taught to take part in decisions about risks.' 5 The role of public officials and their expert advisors is to inform public judgment, but not to replace it. The major characteristics of this approach, then, are maintaining the distinctions between risk assessment and risk management, while recognizing and conveying the un- certainties; communicating risk information to the public, both to build citizen competence and to draw the public into specific risk management decisions; and involving the affected public as early and openly as possible in the decision process. An excellent statement of a public deliberation perspective is Robert Reich's essay on "Public Administration and Public Delib- eration."'6 Reich examines two decision models that have domi- nated administrative theory and practice in recent decades. The first model, "interest-group mediation," descends from pluralist political science and was more influential in the 1960s and early 1970s. In this conception, the administrator is a referee who brings affected interests together to reconcile their demands and preferences. The second, "net-benefit maximization," descends from decision theory and has been more influential in the late 1970s and 1980s. The administrator is an analyst who defines policy options, estimates their consequences, compares them 14. Russell & Gruber, Risk Assessment in Environmental Policy-Making, 236 SCIENCE 286 (1987). 15. Russell, Environmental Protection for the 1990s and Beyond, 29 ENVIRONMENT, 12 (1987). 16. See Reich, Public Administration and Public Deliberation: An Interpretive Essay, 94 YALE L.J. 1617 (1985). - 19891 507 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 given a set of objectives, then selects the one offering the greatest net benefit or social utility. Although he accepts their place in the administrative process, Reich criticizes both models. Neither, he argues, can legitimize administrative decisions by sustaining public confidence in the ac- countability and responsiveness of administrators. By accepting public preferences as given, merely to be revealed through group interaction or the operations of analytical models, neither ap- proach allows the public to deliberate about shared values or the future of the community. Administrators can promote and lead this process of public deliberation by acting as teachers and guides-by articulating visions of the future and exposing people to an array of issues and problems. This is an ambitious task, as Reich acknowledges. Genuine deliberation is not easy; it de- mands time, attention, and facing up to painful choices. By pro- moting social learning and debate, administrators may generate rather than reduce controversy. So deliberation should be a con- scious strategy, one reserved for "decisions that are especially bound up with social values, or that are likely to have important effects on future generations."'7 2. Analysis The public deliberation perspective presents the clearest vision of a participatory process in the risk literature. It resembles, in many aspects, the arguments of the participation theorists dis- cussed in Part III. It presents. the government risk manager as more than a passive mediator of interest group interactions or an analyst adopting the results of analytical calculations. The task of the administrator is to lead: to educate, inform, listen, and draw out community values. What this work does not offer is practical solutions. We get a vision, but not an institutional infrastructure for implementing it. B. Resolving Conflicts Among Experts 1. Description This refers to a broad category of work. Some examples illus- trate its relevance to the discussion of democratic process. Nelkin examines two controversies in which experts were drawn into technical policy disputes-the siting of a nuclear power plant in 17. Id. at 1640. 508 Risk and Democratic Process upstate New York and the proposed construction of a new runway at Boston's Logan International Airport.' 8 In both cases, she concludes that while expert advice can clarify technical con- straints and choices, it also can increase conflict. The experts' disagreement stimulated political activity among interested par- ties. Perceptions of technical credibility were based less on tech- nical competence or validity of the testimony than on which position an expert supported. Put simply, she found that people trusted the experts who supported their position. In another example of this perspective, Brooks considers two ways of viewing the relationship between technical issues and pol- icy decisions. One assumes that the technical/scientific and polit- ical/value components of disputes can be separated. The science court proposal reflects this view.' 9 A second view, which he adopts, recognizes that such disputes are "inherently value- laden" and can be resolved only by "mixing together experts and generalists and forcing a continuing dialogue among them."20 Experts must take part in these decisions, because they have the knowledge and methods to estimate the likely range of conse- quences. However, participation by the lay public is necessary "to represent societal values to the experts and to clarify the nec- essary choices that the political process must make."21 Lay partic- ipation is desirable both to sustain legitimacy and public acceptance for policies and as "an intrinsic political value."22 The challenge to society, Brooks argues, is to devise institu- tional mechanisms for reconciling technical and value considera- tions. These mechanisms should enable representatives of the lay public to become immersed in the issues and be insulated from lobbying and media pressures. They should permit access to technical experts and allow reasoned discussion with them. And these mechanisms should be applied only to selected issues. Pop- ular referenda clearly fail these criteria. Two mechanisms that could prove valuable, Brooks suggests, are lay juries or citizens' commissions. The jury model deserves special consideration, be- 18. See Nelkin, supra note 1. 19. Brooks, The Regulation of Technically Intensive Public Policy Disputes, 46 Sci.. TECH. & Hum. VALUES 39, 40 (1984). For an earlier example of this perspective, see Mazur, Disputes Between Experts, II MINERVA 243 (1973). 20. Brooks, supra note 19, at 40. 21. Id. at 46. 22. Id. 5091989] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 cause it offers some guarantee of representativeness, the opportu- nity for reasoned deliberation and interaction with experts, and insulation from outside pressures. In the end he suggests that a hybrid of the jury and citizens' commission models could be ap- propriate. Yet he remains cautious about appeals to the lay pub- lic, who should be consulted only when "appeal from the experts becomes necessary."23 2. Analysis Where the public deliberation perspective articulates a vision, this perspective poses a problem: What do we do when the ex- perts disagree? Posing the question this way makes two impor- tant assumptions. One is that expert consensus exists independently of political circumstances. Yet the definition of technical consensus is itself a political question. Expert disagree- ments can simply reflect the degree and nature of political contro- versy on an issue. Nelkin's two case studies describe the interrelationships between science and values on intensely con- troversial issues. When scientific credibility depends more on the position the expert takes than on the perceived validity of the sci- entific argument itself, then it is difficult to argue that politics be- gins where technical consensus ends. A second problem with this perspective is the assumption that public values are relevant only on the residuals that remain after the experts have done their work. Often the important questions relate to the design of tech- nical analyses and the assumptions and inferences that the ex- perts use in reaching their conclusions.24 C. The Ethical and Methodological Critique 1. Description Another set of writers has explored how to reconcile the as- sumptions in risk analysis with the ethical and political values of a democratic society. An example is Shrader-Frechette's critique of the scientific, logical, epistemological, and ethical presupposi- 23. Id. at 49. 24. For discussions of these issues from different perspectives, see Lynn, The Interplay of Science and Values in Assessing and Regulating Environmental Risks, 55 Sci. TECH. & Hum. VAL- UES 40 (1986); Majone, Science and Trans-Science in Standard Setting, 46 Sc. TECH. & HUM. VALUES 15 (1984); and Slovic, Informing and Educating the Public About Risk, 6 RISK ANALYSIS 403 (1986). 510 Risk and Democratic Process tionS2 5 of formal models for defining acceptable risk. Although she accepts and defends the use of these analytical models, she argues that the dominance of risk analysis by experts who view it as a wholly scientific enterprise has allowed a number of method- ological and ethical biases to enter into risk policymaking. She conceives of risk analysis not as a principally scientific or objective process, but a "normative, policy-oriented, enterprise with signifi- cant scientific elements."26 She aims not to reject risk analysis but "to accommodate democratic values within analytic assessment."27 Shrader-Frechette proposes two possible approaches to recon- ciling the scientific and value elements in risk issues. One is an analytical approach that weighs risks, costs, and benefits to ac- count for public values not normally incorporated in formal mod- els-such as the importance attached to high consequence/low probability events that are catastrophic or socially disruptive. Her second proposal is to bring experts and the lay public together into a "technology tribunal" to consider both science and value questions in a risk controversy.28 Procedurally, the tribunal would function like a science court. The difference is that the tri- bunal would include lay as well as expert participants and would not separate science and value questions. Composed of anywhere from a few dozen to a few thousand people, the tribunal would issue a decision that was binding in itself or could be used by gov- ernment authorities as the basis for a decision.29 Flores and Kraft also use ethical, political, and methodological criteria to assess what they term "synthetic" approaches to deter- mining the acceptability of risk.30 Synthetic approaches use ana- lytical models and empirical data within an interpretative framework, and include risk/cost/benefit analysis, the revealed and expressed preferences approaches, and the natural standards approach to making risk decisions. Despite the advantages of these approaches as decision tools, they fail to account for several ethical and political issues. Similarly, Zinke argues that cost-ben- 25. K.S. SHRADER-FRECHETTE. supra note 7, at 197. 26. Id. at 202-203. 27. Id. at 204. 28. Id. at 209-213. 29. Id 30. A. Flores & M. Kraft, Determining the Acceptability of Risk in Regulatory Policy: Ethics. Politics. and Risk Analysis (1987) (draft paper). 1989] 511 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 efit approaches tend to "undermine the legitimacy of administra- tive and regulatory processes."3' By eliminating several ethical, moral, and political factors, he argues, a cost-benefit approach reduces the credibility of agencies and their decisions. It also reduces opportunities for public debate and increases the likeli- hood that experts will manipulate public choices. 2. Analysis This category of work is characterized by its reservations about the use of formal analytic models. More than other risk literature, it deals explicitly with the need to reconcile formal analysis (such as cost-benefit methodology) with the requirements of democratic process. It emphasizes that it is not only scientific experts who are framing public choices but experts in decision models as well. These writers argue that the political and ethical rationality of de- cisions are as important as their technical and economic rational- ity. Although they accept that formal analytic models should play a role, their criticisms suggest conflicts between formal analysis and democratic values. At a minimum, formal models make deci- sions less accessible by imposing another layer of specialized method on top of the scientific assessments of risk. But there may be an even more basic problem: Formal analytic models purport to resolve the value rather than the factual issues in a risk contro- versy. In addition, formal models substitute analytic calculations for the discussion and debate that I later argue are important fea- tures of a democratic process. D. Risk Perception 1. Description This approach includes studies of perceptions of risk and of the psychological, social, and cultural determinants of those percep- tions. Although it originally stressed psychological aspects of risk perceptions, it more recently has begun to consider social and cultural factors. Principal conclusions that relate to this discus- sion include: a. Risk means different things to different people. Experts rely on formal assessments of risk as a statement of probability times magnitude, with the result stated in measurable terms, such as expected fatalities. The lay public thinks more intui- 31. Zinke, supra note 7, at 63. 512 Risk and Democratic Process tively about risk, based on a more complex set of social and cultural influences.32 b. People invoke a subjective immunity to understate and screen out familiar, low-frequency risks. This is a rational coping mechanism in modem society, if only because we know so much more about risk than we did even a few de- cades ago. This sense of subjective immunity is expressed in such attitudes as personal invulnerability (It can't happen to me), fatalism (Whatever happens was meant to happen), or skepticism (X smoked for forty years and never developed cancer). c. People tend to ignore or discount discrete, familiar, volun- tary, and low-probability risks. Unknown, dread, cata- strophic, or socially-disruptive events, or those posing consequences for future generations, inspire much higher levels of concern, especially in contrast to the experts' em- phasis on expected fatalities or other measurable conse- quences. An event such as Three-Mile Island caused no immediate fatalities and is not expected to result in any can- cers, but through its "signal potential" can have significant social consequences.3 3 d. Feelings of control influence perceptions about the accepta- bility of risk. Attribution theory and research on stress sug- gest that "a generalized expectation of being in control reduces the experience of stress."3 4 People are less likely to invoke their sense of subjective immunity if hazards are seen to be imposed by outside forces over whom they have no control (e.g., a hazardous waste siting decision vs. a concern about radon). Negative reactions are even higher when the perceived source of the risk is seen to benefit from the action. e. Initially, people will react strongly to information that is in- consistent with their existing views, just as they are more likely to accept information that is compatible with those views. Research suggests, however, that under certain cir- cumstances people can move beyond their initial reactions and deal more openly with additional information. f. People react to risk as members of a community, where norms about the acceptability of risk "are debated and so- 32. This discussion draws on M. DoucLAs, RISK ACCEPTABILITY ACCORDING TO THE So- CIAL SCIENCES (1985); Fischhoff, Slovic & Lichtenstein, Lay Fables and Expert Foibles injudg- ments About Risk, 36 AM. STATISTICIAN 240 (1982); M. DOUGLAs & A. WILDAVSKY, RISK AND CULTURE: AN ESSAY ON THE SELECTION OF TECHNOLOGICAL AND ENVIRONMENTAL DANGERS (1982); Kasperson, Six Propositions on Public Participation and Their Relevance for Risk Communi- cation, 6 RISK ANALYSIS 275 (1986); and Slovic, supra note 9. 33. Slovic, supra note 9, at 284. 34. M. DOUGLAS, supra note 32, at 34, (citations omitted). 1989] 513 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:501 cially established."3 5 In addition to reflecting cultural influ- ences, reactions to risk may be serving other social ends, such as "to control uncertainty about human behavior, to re- inforce norms, and to facilitate coordination."3 6 Rayner and Cantor use research on cultural and social factors to challenge two assumptions that dominate the literature-that risk exists as a measurable and definable phenomenon, and that the job of the societal risk manager is to minimize the probability and/or magnitude of unwanted consequences, at socially accepta- ble costs. The key question, they assert, is not " 'How safe is safe enough?' " but " 'How fair is safe enough?' " People are con- cerned about matters of process and fairness, such as the proce- dures for obtaining the collective consent from people bearing the consequences of an action; or the principles used to appor- tion the liabilities; or whether or not the "institutions that make the decisions that manage and regulate the technology [are] wor- thy of fiduciary trust."38 2. Analysis This research documents striking differences between expert and lay attitudes toward risk. While the experts were thinking in technocratic terms, the public was thinking in terms of social and political values. While the experts were treating the public as subjects, the lay public was asserting its identity as citizens. The risk perception research underscores the narrowness of the for- mal assessments of risk and the analytic models that are based upon it. Indeed, I would argue that these apparently "nonratio- nal" attitudes about risk-about the concern for such values as fairness, social stability, control over people's lives, due process, social consent, the ethics of a decision to expose a group of peo- ple to risk-are as important an assertion of democratic values as more visible signs of organization and protest. The persistence of these procedural and substantive political values in lay judgments about risk are an important theme in later sections of this article. 35. Id. at 69. 36. Id. at 92. 37. Rayner & Cantor, How Fair is Safe Enough? The Cultural Approach to Social Technology Choice, 7 RisK ANALYSIs 3 (1987). 38. Rayner & Cantor, supra note 37 at 4. 514 Risk and Democratic Process E. Risk Communication 1. Description From this perspective, the important question is how to com- municate risk information clearly, objectively, and effectively to the lay public. Some of the research on risk communication re- lates public beliefs and attitudes to categories of risk problems, channels of communication, and strategies for presenting infor- mation. Other work presents case studies of particular hazards- of how they were communicated to the lay public, and with what effect.39 In many forms, risk communication has been with us for years. Often, it has been the principal strategy for reducing exposures to risk: to warn of hurricanes, floods, or other natural disasters; to provide directions on the safe use and handling of pesticides or solvents; or to persuade people to behave in ways that will reduce their vulnerability to hazards (cigarette label warnings or seat belt campaigns).40 More recently, EPA and other agencies have rec- ognized the role that information and education can play in influ- encing people to reduce exposures to such risks as radon, lead in home plumbing, ultraviolet sunlight, or home use of pesticides. For these problems, risk communication can offer a more effec- tive risk reduction strategy than can more traditional regulatory approaches. The other objectives of risk communication research relate more directly to the subject of citizen participation. Clearly, if people are going to take part in decisions about risk, they need information on issues, positions, and choices. To the extent that risk communication research is directed toward participation in policy deliberations, it can contribute to citizen participation. One deficiency in the literature, however, is its nearly exclusive focus on communication from experts and government officials to the affected public. Communication can occur two ways, and re- search and practice should also consider how the lay public can 39. For examples, see the case studies in S. KRIMSKY AND A. PLOUGH, ENVIRONMENTAL HAZARDS: COMMUNICATING RISKS AS A SOCIAL PROCESS (1988). 40. S. HADDEN, READ THE LABEL: REDUCING RISK BY PROVIDING INFORMATION (1986); Covello, von Winterfeldt & Slovic, Communicating Scientific Information about Health and Envi- ranmental Risks: Problenu and Opportunities from a Social and Behavioral Perspectives, in RISK COMMUNICATION 109 (1986). 1989] 515 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 convey to the decisionmaker and experts a sense of public senti- ments, concerns, preferences, and values.41 Much of the work on risk communication reflects the tendency in the risk literature to view the lay public as subjects rather than as citizens. In one recent article, for example, the authors assert that "Two problems with the public [are] their desire for 'zero risk' and their 'thirst for certitude.' "42 More recent work, how- ever, has viewed the public not as a problem to overcome, but as a legitimate source of judgment and information for experts and policymakers. Plough and Krimsky in particular argue that we should broaden conventional definitions of risk communication to account for its symbolic, cultural, and experiential dimen- sions.43 Adopting a popular epidemiology perspective, Brown describes a case in which the lay public communicated risk infor- mation to scientific experts and government officials, thus chal- lenging an elite-centered view of risk communication.44 Kasperson examines the relationship between risk communica- tion and public participation, stresses the role of community and group influences, and observes that participation and communi- cation programs are rarely the subject of careful evaluation.45 2. Analysis Effective risk communication can build capacities for citizen participation. By defining categories of communication problems based on the message, the source,.the channel, and the recipients of the information, the literature provides a framework for under- standing the process of risk communication and diagnosing the causes of problems. It describes the uncertainties, ambiguities, and complexities of risk problems and the need to account for them when presenting information. The case studies offer a growing body of experience on which strategies and techniques succeed or fail in different contexts. But this body of work is valu- 41. EPA's Science Advisory Board recently made a strong case for research on "effec- tive, multi-way communication" and on its "social and cultural contexts." See "Appendix E: Strategies for Risk Reduction Research," the report of the subcommittee on Risk Re- duction of the Research Strategies Subcommittee 32 (September 1988). 42. Keeney & von Winterfeldt, Improving Risk Communication, 6 RiSK ANALYSIs 417, 420 (1986) (footnote omitted). 43. The Emergence of Risk Communication Studies 12 Sci. TECH. & HUM. VALUES, 6-8 (1987). 44. Brown, Popular Epidemiology: Community Response to Waste-Induced Disease in Woburn. Massachusetts, 12 Sci. TECH. & HUM. VALUES 78 (1987). 45. Kasperson, supra note 32. 516 Risk and Democratic Process able only if it recognizes the legitimacy of lay perceptions and the values they reflect, builds firmly on the risk perception research, and incorporates the need for communication to, as well as from, experts and risk managers. The normative principles that guide this research, especially the conception of "citizen" on which it is based, will determine its value to the study of democratic process in risk policymaking. F. System Responses to Demands for Participation 1. Description Some of the risk research has examined how political systems respond to growing demands for democratic participation in technical decisions. This work adopts a system-wide, comparative perspective, focusing on the Western democracies. It places the issue of environmental risk in the larger context of technology and society, recognizing the issue not only as one of the accepta- bility of specific technologies, but of fundamental values regard- ing technological change and social processes for coping with that change. In one of the better statements of this perspective, Nelkin and Pollak observe: Contemporary resistance to technological change, however, re- flects broader questions: it is marked by hostility to public bu- reaucracies and resentment of impersonal, expertise- dominated policies. The issue is no longer simply the impact of technology, but the locus of control over major public deci- sions. . .. Given this shift from technical concerns to questions of values 'too important to be left to experts,' competence is suspect and is no longer accepted as a sufficient basis for deci- sion-making authority.46 Because this research is comparative, it documents how polit- ical traditions and institutions affect national responses to de- mands for broader participation in technical decisions. The Swedish experience with study circles illustrates this point. Oper- ating within a tradition of consensus and compromise on social and political issues, the government assumed that increased pub- lic understanding of energy and nuclear issues would reveal an underlying consensus in favor of a proposed expansion in nuclear generating capacity. When the government's nuclear program 46. Nelkin & Pollack, The Politics of Participation and the Nuclear Debate in Sweden, the Nether- lands, and Austria, 25 PUB. POL. 333, 353 (1977). See also D. NELKIN, TECHNOLOGICAL DECI- SIONS AND DEMOCRACY: EUROPEAN EXPERIMENTS IN PUBLIC PARTICIPATION (1977). 1989] 517 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 became a major social issue in the early 1970s, one response was to invite major social and political institutions to organize energy study circles. The study circles began in late 1973 and lasted for about a year. Altogether, some 10,000 circles were organized and included an estimated 80,000 participants. Although the govern- ment viewed this as a means of broadening the awareness of blue- collar and other groups who had not been active on the nuclear issue, the participants generally were well-educated, well-in- formed, and politically active. An evaluation of the effects of the study circles by the National Board of Civic Information in the Fall of 1974 revealed very slight differences between participants and non-participants in their at- titudes toward nuclear power. In addition, the circles do not ap- pear to have led to noticeable improvements in public understanding. The Board's survey suggested that "many partici- pants were more confused after they took part in the study circles than they were before."47 Another set of surveys, following a se- ries of four public hearings held late in 1974 and 1975, showed a modest shift toward public support for the government's policies. The study circles were a particularly Swedish approach to the problem. The leadership expected that the social conflict over nuclear power would decline once information and education al- lowed an underlying consensus to emerge. That the consensus did not emerge suggests something important about nuclear power and the scientific and political controversies it reflects. The two other systems Nelkin and Pollak studied-the Nether- lands and Austria-reveal different kinds of system responses. In the Netherlands, there is greater experience with the manage- ment of social and political diversity. Political institutions are designed to deal explicitly with multiple demands of different in- terests by recognizing and incorporating them into policy formu- lation. The Austrian response shows more of a balance in expectations and institutional responses. Nelkin and Pollak document many of the same experiences as the OECD report, "Technology on Trial."48 It surveys the mech- 47. Nelkin & Pollack, supra note 46, at 344. The Swedish case is also discussed in K.G. NICHOLS, TECHNOLOGY ON TRIAL: PUBLIC PARTICIPATION IN DECISION-MAKING RELATED TO SCIENCE AND TECHNOLOGY 29 (1979). A useful overview and classification of participatory mechanisms is Nelkin & Pollack, Problems and Procedures in the Regulation of Technological Risk, in MAKING BUREAUCRACIES WORK 259 (1980). 48. K.G. NICHOLS, supra note 47. 518 Risk and Democratic Process anisms and procedures that OECD countries have used to promote participation in technical decisionmaking, based on whether the objective was to inform the public, inform policy- makers, reconcile competing interests, or achieve more collabora- tive decisiopmaking. Although the OECD report was published over a decade ago, its conclusions remain valid: "Government agencies have generally reacted to participatory demands," it concludes, "not anticipated them."49 The report adds that "the need for new approaches and more flexible institutions for public participation in the development and formulation of policies re- lated to science and technology appears to be especially crucial."50 Although this work has focused on risks from nuclear power and other large-scale energy technologies, it illuminates other en- vironmental risk issues as well. The nuclear controversy high- lights the same kinds of value conflicts we can observe in other risk controversies-preservation versus development, intangibles versus tangible goods, democracy versus technology. It has pit- ted experts against experts, experts against the lay public, gov- ernments against citizens, industry against communities. The controversy over nuclear power has introduced ideological fault lines that will structure the debate over risk issues for years to come. The issues are not just risk, or public health, or environ- mental quality, but a combination of these and other more funda- mental concerns, as Douglas and Wildavsky have argued.5 ' 2. Analysis This literature illustrates how institutional and cultural factors can shape national responses to highly visible, technically-based controversy. The conflict over nuclear power can be seen as a challenge to control by political, administrative, and technical elites. Even in Sweden, with its tradition of consensus and com- promise, the nuclear issue introduced social divisions that could not be accommodated within the normal political structure. The same was the case in other Western democracies. This research 49. Id. at I 13. 50. Id. 51. See M. DOUGLAs & A. WILDAVSKY, supra note 32. For a detailed analysis of elite per- ceptions of nuclear power and the effects of political ideology on them, see Rothman & Lichter. Elite Ideology and Risk Perception in Nuclear Energy Policy, 81 AM. POL. ScI. REV. 383 (1987). 1989] 519 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 also emphasizes that the boundary of scientific consensus on a given issue is itself a political question. Krimsky's discussion of guidelines for research on recombinant DNA makes the same point: What began as a scientific matter whose resolution would be left to technical experts quickly evolved into an issue requiring political solutions. The creation of the Cambridge Experimenta- tion Review Board (CERB) was only one manifestation of this metamorphosis from a technical to a political issue.52 G. Administrative Process and Culture 1. Description This area of research -also adopts a comparative approach, whose objective is to describe how scientific and political author- ity are reconciled in different national administrative systems, and to explain what effects these differences have had on decisions to regulate toxic chemicals. Unlike the nuclear power issue, these decisions have been made largely within existing political processes. They have not presented the kinds of demands for sys- tem responses that we can observe on the nuclear power issue. This research documents the generally acknowledged charac- teristics of national administration in the United States.53 These characteristics include a popular distrust of bureaucratic power, fragmented administrative authority, and skepticism toward ex- perts. Agencies must contend with vigorous Congressional over- sight, active intervention by the courts, and a public and adversarial decision process. Because they lack a clear basis for their authority, administrators turn to science and expert knowl- edge as a source of legitimacy. Yet scientists cannot make value judgments, which are reserved for the political process. So the scientific process of risk assessment is separated from the political process of risk management. Because regulatory science involves assumptions, choices, and uncertainties, administrators seek to make the process of scientific assessment as neutral and value- free as possible. 52. A discussion of the CERB can be found in Krimsky, Regulating Recombinant DIA Re- search, in CONTROVERSY: POLITICS OF TECHNICAL DECISIONS 251 (D. Nelkin ed. 1977). 53. This discussion is based upon Brickman, Science and the Politics of Toxic Chemical Regu- lation: US. and European Contexts, 9 ScI. TECH. & Hum. VALUES 107 (1984); S. JAsANOFF, RISK MANAGEMENT AND POLITICAL CULTURE: A COMPARATIVE STUDY OF SCIENCE IN THE POLICY CONTEXT (1986); D. VOGEL, NATIONAL STYLES OF REGULATION: ENVIRONMENTAL POLICY IN GREAT BRITAIN AND THE UNITED STATES (1986); and Majone, supra note 24. 520 Risk and Democratic Process In the United States, administrators use science to bolster their authority and to define the choices they and the affected public must make. European administration takes place within a differ- ent context, in which there is greater centralization of administra- tive power and more public deference to bureaucratic authority and expert knowledge. Access to the courts is more restricted. Parliaments have less reason to interfere in administrative delib- erations than does Congress. Administrators defer to scientific expertise as expressed in elite, standing bodies, such as the French Ecotoxicity Advisory Committee. Processes for making decisions about toxic chemicals are relatively closed to participa- tion by non-elites, both scientific and administrative, and access to information is far more restricted than in the United States. The consensus of scientists is more likely to be accepted on its own terms, and thus less likely to be drawn into the political debate. An illustration is Jasanoff's analysis of how three kinds of ad- ministrative cultures identify chemical carcinogens.4 She associ- ates one approach with West Germany and the International Agency for Research on Cancer, where scientific issues are dele- gated to expert panels who apply the technical evidence to their own criteria and explain their conclusions based on that evidence. In this approach, both science and value questions are resolved by expert panels. Canada and Great Britain exemplify a second approach. Conclusions about cancer risk result from a mixed sci- entific and administrative process, "in which uncertainty is not al- ways publicly analyzed."55 The government acts when political pressures and the evidence seem to require it. The United States illustrates a third approach, in which the distinctions between sci- entific evidence and political judgment are drawn most clearly. Scientific assessments are the product of a public process and fol- low explicit guidelines. Political officials resolve issues that re- main after the scientific analysis is complete. The third approach, she observes, is least likely to lead to closure on decisions about identification of carcinogens. Yet she concludes that the United States approach is more open and accountable, allowing more ex- tensive public scrutiny and participation. 54. S. JASANOFF, supra note 53. 55. Id. at 80. 1989] 521 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:501 The differences in approach, she observes, suggest a -trade-off between "administrative efficiency and scientific credibility" in the European systems and "analytical rigor and informed public participation" in the American.56 She concludes: "The indeci- siveness of the U.S. approach, for example, may be bearable to many because what matters most in risk management is the pro- cess, not the outcome."57 2. Analysis The American process clearly is more open and participatory than its European and Canadian counterparts. It is plausible to argue that the more open and participatory process in the U.S. impedes consensus and delays regulatory action on toxic chemi- cals. Closed decisions generally are made more efficiently than open ones. But is delay inevitably the consequence of a more participatory process? Can a more democratic process promote consensus rather than impede it? In the next part of this article, I argue that the problem we face may not be too much participa- tion, but the wrong kind of participation-participation that is too adversarial, too subject to legal maneuver and challenge, too de- pendent on the actions of organized interest groups, with too lit- tle opportunity for constructive discussion and cooperative problem-solving. What this group of writers may help to demon- strate is that participation in the contemporary American setting has had its costs. The question is whether delay and open conflict are the inevitable consequences of broad participation in risk decisionmaking. This question is taken up in the remaining sec- tions of this article. H. Conclusions One purpose behind the preceding discussion was to present a review and critique of the literature as its bears on the challenge of reconciling risk analysis with democratic values and process. Several conclusions can be drawn at this point. One is that this challenge is a recurring theme in the literature. Whether it is one writer invoking a Jeffersonian faith in public understanding, or another exploring the interaction between science and politics, or a third proposing a technology tribunal made up of lay as well as 56. Id. at 81. 57. Id. 522 Risk and Democratic Process expert representatives, an awareness of the technical and political dichotomies in risk analysis is always present. A second conclu- sion is that there are no easy answers. The literature is long on definitions of the problem but short on practical institutional so- lutions. Although we see some speculation on possible institu- tional mechanisms-such as in the parallel with the lay jury or in the technology tribunal-it is at a general level rather than at the level of tested, implementable mechanisms. Third, the clearest vision of what a democratic model of risk analysis should look like comes through in the public deliberation perspective. Those writers' visions comes closest to the democratic ideal of the par- ticipation theorists presented in Part III. The second purpose behind the literature review was to define the basis for the discussion in the following sections on citizen participation and democratic theory. The argument is that there are two possible sources of direction in attempting to reconcile risk analysis with democratic values and process. The first is our experience with citizen participation in administrative decision- making, especially in the environmental area. The second is con- temporary democratic theory. Parts II and III address these two sources of direction in turn. These.two sections and the conclud- ing one suggest that a technological society can remain a demo- cratic one only by remaining conscious of democratic values and by searching for institutional measures that will promote those values in social decisionmaking. II. CITIZEN PARTICIPATION IN THE ADMINISTRATIVE PROCESS The term "citizen participation" conjures up diverse images. To some people, it is synonymous with computer mailing lists, outreach meetings, well-publicized hearings, ind slickly-packaged information brochures. To others, the term evokes images of raucous public meetings, rising costs, lawsuits, and delay. To still others, the term is a symbol for rallying opposition to govern- ment and corporate insensitivities, or a strategy for mobilizing otherwise disinterested publics. To the government administra- tor, participation can mean a nuisance or a strategy, to the public affairs staff an opportunity, to the public interest group a tactic, and to newly-organized groups a symbol. Few terms in our con- temporary political lexicon have been used with so little semantic precision. 1989] 523 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:501 Citizen participation in the administrative process will be only as good or useful as the meanings we give to it. I argue that the meanings we have given the term in recent decades have served neither our democratic ideals nor the goals of effective policy- making. At a theoretical level, our conceptions of participation depend on interest-group, pluralist thinking. At a practical level, our conception of participation is narrow, superficial, and biased. Until we can devise new approaches, based on sounder theoreti- cal and practical foundations, it will be difficult to achieve more effective lay participation in risk policymaking. As one source for such a foundation, I turn to the literature on democratic partici- pation. This literature offers both a diagnosis of the roots of our participatory dilemma and a set of principles for reinvigorating our approach to citizen participation. A. Citizen Participation: The Concept That Lost Its Way As the scope of national administration expanded over the last several decades, so has the need for participation. In Lowi's terms, the administrative state required new institutions and processes to accommodate demands for greater participation. A useful way to review the recent history is to follow Walter Rosen- baum's distinction between old, new, and newer forms of participation58 The "old" or traditional standard is exemplified in the Admin- istrative Procedure Act of 1946 (APA). 59 The APA expressed a lawyer's vision of due process, with its emphasis on public notice, right to comment, opportunity for hearings, and other procedural requirements. It defined a structure and set of procedures, while leaving the initiative for participation with private entities. Partic- ipation was a privilege available to parties with the organization and resources needed to exercise it. More recent legislation, such as the Freedom of Information Act of 1966, reflected a similar 58. Rosenbaum, The Politics of Public Participation in Hazardous Waste Management, in THE PoLITmCs OF HAZARDOUS WASTE MANAGEMENT 176 (1983). This discussion also draws upon CITIZEN PARTICIPATION IN AMERICA (S. Langton ed. 1978); Kloman, A Mini-Sympo- sium: Public Participation in Technology Assessment, 35 PUB. ADMIN. REV. 67 (1975); and Sewell & O'Riordan, The Culture of Participation in Environmental Decisionmaking, 16 NAT. RESOURCES J. I (1976). A valuable discussion of participation and technological risk can be found in M. Kraft, Managing Technological Risks in a Democratic Polity: Citizen Participation and Nuclear Waste Disposal (March 1987) (prepared for presentation at the Annual Meeting of the American Society for Public Administration, Boston). 59. Administrative Procedure Act of 1946, 5 U.S.C. § 551-558 (1982). 524 Risk and Democratic Process philosophy. With its detailed procedures granting access to deci- sions and information, American national administration is one of the most open among the Western democracies. The "new" participation emerged with the social programs of the 1960s. Under the Economic Opportunity Act of 1964, the Of- fice of Economic Opportunity was instructed to achieve "maxi- mum feasible participation" in communities affected by its programs. The Demonstration Cities and Metropolitan Develop- ment Act of 1966 directed the Department of Housing and Urban Development to effect the "widespread" participation of the com- munity.co In this conception, participation was a right, rather than a privilege, and it was directed toward target groups who tended to be the programs' beneficiaries. These programs did not always have their intended effects. In both the urban renewal and anti-poverty programs, participation served the needs of agencies and clientele groups more than the target groups in the community.61 In their review of the participa- tion literature in the late 1970s, Checkoway and Van Til 6 2 note that agencies often used participation as a strategy for achieving organizational ends, without effecting any transfer of power. Their review of the relevant research also demonstrated two other tendencies in participation programs-the need for groups to organize to be able to assert influence, and the emergence within these organizations of leaders and contrasting interests, with the leadership often asserting its own interests over those of the followers. Participation assumed strategic or tactical value as a means of promoting bureaucratic and professional interests. The "newer" form of participation emerged in the 1970s, in response to several forces. One major force was the National En- vironmental Policy Act of 1969 (NEPA). NEPA was not a partici- pation statute per se. But two of its provisions, used together, transformed the role that noneconomic, environmental factors played in agency decisionmaking. One was the requirement that agencies prepare detailed environmental impact statements (EIS) 60. Economic Opportunity Act of 1964, Pub. L. No. 88-452, 78 Stat. 508 (1964); Dem- onstration Cities and Metropolitan Development Act of 1966, Pub. L. No. 89-754, 80 Stat. 1255 (1966). 61. Krause, Functions of a Bureaucratic Ideology: "Citizen Participation," in PARTICIPATORY DEM4oCRAcy 420 (1971). 62. Checkoway & Van Til, What Do We Know About Citizen Participation? A Selective Review of Research, in ClTIZEN PARTICIPATION IN AMERICA 25 (S. Langton ed. 1978). 19891 525 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 for major actions having an effect on the environment. The sec- ond was a provision authorizing judicial review of agency compli- ance -with the EIS requirement. The first established a decision rule and the second gave it bite. The strategy was not new, be- cause analytical requirements had long been a means of control- ling agencies. The difference with NEPA was that it "served as an instrument of popular rather than executive control over agency decisions."63 Environmental advocates used the law and a sympathetic fed- eral judiciary to oppose public works projects as well as private projects that depended on federal action." NEPA provided ac- cess for a new category of interests, and eventually it redefined the premises on which agencies -made decisions affecting the envi- ronment. Although NEPA may have been important in promot- ing substantive democratic values, I argue that it did not promote the procedural values of democracy. It reinforced a conception of participation as confrontation and opposition. It strengthened the reliance on litigation as a political tool. It established the pre- cedent for elaborate written documentation that became "an in- strument of legal and political warfare."65 A second force was a "reformation" of administrative law in the late 1960s and early 1970s. -Led by the Court of Appeals for the District of Columbia Circuit, the federal courts articulated the conception of administrative law as, in Stewart's words, "a surro- gate political process to ensure the fair representation of a wide range of affected interests in the process of administrative deci- sion."" Stewart explains- this-new conception as an effort to pro- tect new classes of interests that required protection under an expanded government. By redefining the rules of standing to in- clude noneconomic interests, broadening access to agency deci- sion procedures, and demonstrating a -greater willingness to intervene in agency actions, the courts laid the legal groundwork 63. Andrews, Economics and Environmental Decisions, Past and Present, in ENVIRONMENTAL POLICY UNDER REAGAN'S EXECUTIVE ORDER 43, 51 (V.K. Smith ed. 1984). See also the Na- tional Environmental Policy Act, 42 U.S.C. § 4321 (1982). 64. R. -LIROFF, A NATIONAL POLICY FOR THE ENVIRONMENT: NEPA AND ITS AFTERMATH 142-152 (1976). 65. Bardach & Pugliaresi, The Environmental Impact Statement vs. The Real World, 49 Pun. INTEREST 22, 24 (1977). 66. Stewart, The Reformation of American Administrative Law, 88 HARV. L. REV. 1667, 1670 (1975). These changes in administrative law are also discussed in DeLong, Informal Rulemaking and the Integration of Law and Policy, 65 VA. L. REV. 257 (1979). 526 Risk and Democratic Process for the participation movement. Although this reformation ex- tended across all categories of administrative action, it was proba- bly most pronounced in the environmental field.6 7 Statutory innovations helped to make the 1970s the decade of participation. Some sixty percent of the participation provisions contained in existing federal legislation were enacted in this dec- ade, including many in environmental statutes.68 A prime exam- ple was Section 101(e) of the Federal Water Pollution Control Act (FWPCA) of 1972 , which established procedures for EPA and the states to prepare and implement plans for controlling sources of water pollution. This provision directed that "public participa- tion . . . shall be provided for, encouraged, and assisted by the Administrator and the States." This section of the FWPCA, popu- larly known as the "208 program," was described at the time as ''one of the strongest requirements for participatory democracy in the entire federal statute book."69 The 208 programs often es- tablished an elaborate apparatus for public involvement. North Carolina's program encompassed three phases, five major prod- ucts, and seventeen participatory methods, from newsletters and slide shows to reviews of draft plans and "goal-setting workshops."70 But the increased participation was limited to a select group: over one-half of those who took part in the North Carolina pro- gram represented an organized interest group, and the partici- pants as a whole were above average in socioeconomic status. Other research has noted the atypical nature of participants in such programs and the high levels of interest group involvement. Studies have also found differential rates of participation; with political activity related directly to income, education, and status.7 ' 67. See H. Leventhal, Environmental Decisionmaking and the Role of the Courts, in ENVIRON- MENTAL LAW REVIEw-1975 545 (H.F. Sherrod ed. 1975). 68. See Rosenbaum, supra note 58, at 182. 69. CONSERVATION FOUNDATION, TOWARD CLEAN WATER: A GUIDE TO CITIZEN AcTION 27 (1976). 70. Godschalk & Stiftel, Making W1aves: Public Participation in State Water Planning, 17 J. APP. BEHAV. Sci. 597, 603 (1981). See also S. KAMIENIECKI, PUBLIc REPRESENTATION IN EN- VIRONMENTAL POLICYMAKING: THE CASE OF WATER QUALITY MANAGEMENT (1980). 71. The evidence is discussed in Sewell & O'Riordan, supra note 58; Godschaik & Stiftel, supra note 70; and Checkoway & Van Til, supra note 62. 5271989] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 B. The Decline of the Participation Movement The fortunes of the participation movement declined dramati- cally after 1981. The Reagan Administration systematically re- duced resources and dismantled participation programs. With the "New Federalism" as a rhetorical justification, responsibility for involving the public fell to state and local governments. Pro- grams that appeared to have been embedded in law and public expectations came "to epitomize administrative marginality."7 2 There are several explanations for the sudden decline. Con- gressional support for the participation programs was often rhe- torical, and it could evaporate quickly when it conflicted with other Congressional interests or objectives. In addition, the stat- utory directives typically were so vague that Congress had diffi- culty holding the agencies accountable. The Office of Management and Budget had long opposed the participation pro- grams, and with the new administration's backing they were able to cut funding. The energy crisis of the 1970s also undermined support for the programs, which often were seen as a source of delay and paralysis, especially for developing such hard technolo- gies as nuclear power.73 Another explanation for the decline is that "citizen participa- tion" had come to be equated with involvement by environmental and other public interest groups. The participation provisions in statutes often were the result of agreements between environ- mental groups and their Congressional allies, designed to guaran- tee the former's access to agency decisions. The list of plaintiffs that prevailed in the major administrative law rulings of the era includes several public interest organizations: Citizens to Pre- serve Overton Park, Calvert Cliffs' Coordinating Committee, and Scenic Hudson Preservation Conference.74 The administration equated support for the participation programs with support for the political opposition. "Participation" was viewed less as an ideologically-neutral reference to a democratic process than a 72. Rosenbaum, supra note 58, at 177. 73. For a discussion of participation in energy policymaking and its relationship to is- sues of size, control, and decentralization, see Orr, U.S. Energy Policy and the Political Econ- omy of Participation, 41 J. POL. 1027 (1979). 74. The cases are Citizens to Preserve Overton Park v. Volpe, 401 U.S. 402 (1971); Calvert Cliffs' Coordinating Comm., Inc. v. United States Atomic Energy Comm'n, 449 F.2d 1109 (D.C. Cir. 1971); and Scenic Hudson Preservation Conference v. Federal Power Comm'n, 354 F.2d 608 (2d Cir. 1965), cert. denied. 384 U.S.,941 (1966). 528 Risk and Democratic Process strategy for asserting noneconomic, nontraditional interests in administrative processes. Other factors reinforced the political vulnerability of participa- tion. The Reagan Administration arrived in 1981 committed to reducing the burdens and costs of federal regulation. Its view was that environmental standards were too strict, imposed excessive costs on the economy, and hampered economic development and technological innovation. These views coincided with rising aca- demic and political interest in cost-benefit analysis. With Execu- tive Order 12291, issued in February of 1981, the administration adopted the net-benefit model as the preferred basis for setting health and environmental standards.75 Participation was at best irrelevant to such an approach, and at worst inimical to it. Regu- latory reform advocates came to view active interest group partici- pation as politically and philosophically inconsistent with a net benefits approach. If the 1970s were the decade of participation, the 1980s have been the decade of analysis. Whatever the cause of its demise, the participation movement of the 1960s and 1970s left its legacy. Our conceptions of partici- pation at the national level today reflect the practices and exper- iences of the last few decades. The most important element in that legacy is that it reflects and reinforces the tenets of interest- group pluralism. Effective participation requires organization, re- sources, and professional representation. Participation at the na- tional level typically is indirect, because it is mediated through voluntary associations organized to assert related interests. Often participation is equated with opposition. The public is placed in a reactive posture: comments are noted, views are heard, opportu- nities are presented. But we rarely see a sharing of power or the codetermination of policy. The process concedes a marginal role to the individual citizen. Genuine influence is granted reluc- tantly, minimally, and to interests with the capacity to obstruct decisions later. Participation is biased, because it draws on groups with the needed information, competence, and resources. It is skewed in the way that it solicits the participation of those with the most to lose or the greatest intensity of feeling on an issue. As Barber observes of a liberal polity generally, our model 75. For discussion of the Executive Order and the regulatory relief program of the early 1980s, see SMrrn, ENVIRONMENTAL POLICY UNDER REAGAN's EXECUTIVE ORDER supra note 63; Andrews, Deregulation: The Failure at NEPA, in ENVIRONMENTAL POLICY IN THE 1980s: REAGAN'S NEW AGENDA (1984). 1989) 529 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 of participation assumes conflict.76* And so people expect that participation will increase rather than transform conflict. III. PARTICIPATION THEORY AS DIAGNOSIS AND PRESCRIPTION This and the following section outline an alternative approach to participation, one based upon the writing of what I will call the "participation theorists." Before discussing these writers' vision of participation, however, I want to discuss the roots of the par- ticipatory dilemma to which each of them reacts. A. The Roots of a Participatory Dilemma Political participation has not always been highly valued in modem democratic theory. Writing in 1970, Carole Pateman painted a picture of a mainstream theory that justified a limited conception of participation.7 7 The selection of leaders through elections was seen as sufficient for ensuring democratic control over decisions. Pateman attributes this view to two influences. The first was the rise of totalitarianism in Europe, and in particu- lar the collapse of the Weimar Republic in Germany. The second influence on theorists was behavioral research that revealed low levels of political awareness, information, and support for demo- cratic principles among certain groups. These tended to be groups that were less active politically. Mainstream theory justi- fied their lack of participation, because it viewed their involve- ment as a threat to democratic stability. The expansion of the administrative state challenged this con- ception of participation as limited to electoral control. I argued in the previous section that participation in American administra- tion has both reflected and reinforced the pluralist model. As government expanded into new areas of social policy, new forms of political representation emerged to fill what Lowi describes as the inevitable crisis in public authority. This representation took the only form it could be expected to take in an administrative system firmly grounded in the interest group model. The rise of environmental, consumer, and other public interest groups can be seen simply as a balancing of the historical scales. For de- 76. B. BARBER, STRONG DEMOCRACY: PARTICIPATORY POLITICS FOR A NEw AGE 3 (1984). 77. C. PATEMAN, PARTICIPATION AND DEMOCRATIC THEORY 1 (1970). See also R. COBS & C. ELDER, PARTICIPATION IN AMERICAN POLITICS 1 (1972); P. BACHRACH, THE THEORY OF DEMOCRATIC ELITISM: A CRITIQUE (1967) (see particularly ch. 2). 530 Risk and Democratic Process cades, groups with a tangible economic stake in administrative de- cisions had enjoyed access. Now, with substantive levers such as NEPA, and with the reformation of administrative law that rede- fined rules of standing and other doctrine, representatives of dif- fuse interests are asserting noneconomic values. Citizen participation came to be equated with interest group participa- tion. Litigation became an effective tool in the competition among interests. Citizens did not participate-they joined or otherwise supported interest groups that participated on their be- half. When individuals did have the opportunity to take part in decisions, it was typically in the context of a public hearing or meeting, in which effective influence and genuine participation were unlikely.78 The participatory theorists react to these conceptions of demo- cratic influence and control. Pateman reacts to mainstream dem- ocratic theory, its fears about democratic stability, and its reliance on electoral control. Thompson proposes "citizen theory" as an antidote to "elite theory," which focuses on leaders rather than citizens and "reinforces the potent historical pressures toward centralized bureaucratic power that makes citizens feel remote from politics and that discourage citizenship."7 9 Barber chal- lenges traditional liberal institutions and philosophy on several grounds-from its assumption of conflict as the basis of all polit- ical relations to its overreliance on representation.80 Mansbridge outlines a model of "unitary" democracy as a complement to "ad- versary" democracy and its expression through secret voting, majority rule, and equality defined as equal protection of the laws.8 ' Olsen's "participatory pluralism" is presented as an al- ternative to "sociopolitical pluralism." 82 Participatory theorists argue that new forms of participation are needed in a world in which people increasingly lack control over 78. For a discussion of the limitations of the public hearing as a participatory mecha- nism, see Checkoway, 7he Politics of Public Hearings, 17 J. APPL. BEHAV. Sc. 566 (1987); Heberlein. Some Observations on Alternative Mechanisms for Public Involvement: The Hearing, Pub- lic Opinion Poll, the Workshop and the Quasi-Experiment, 16 NAT. RESOURCES J. 197 (1976); and Milbrath, Citizen Surveys as Citizen Participation Mechanisms, 17 J. APPL. BEHAV. Sc. 478 (1981). 79. 1). THOMPSON, THE DEMOCRATIC CITIZEN: SOCIAL SCIENCE AND DEMOCRATIC THE- ORY IN THE TWENTIETH CENTURY 25 (1970). 80. B. BARBER. supra note 76, at Xi-XV. 81. J. MANSBRIDGE, BEYOND ADVERSARY DEMOCRACY (1980). 82. N. OLSEN, PARTICIPATORY PLURALISM ch. 2 (1982). 1989] 531 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 social decisions that affect them. Even in their work and social lives, people typically are part of large organizations that offer lit- tie opportunity for access to decisionmakers or influence over policy. Historical trends toward broader participation intensify the problem; participation by greater numbers of people reduces the relative influence any one of them may have. It also increases reliance on elected representatives, because the size of the popu- lation being represented has expanded. Other reform measures designed to improve the competence of government as its func- tions expand further remove the individual from control over de- cisions. The institution of a career civil service, for example, established merit and "neutral" competence as the basis for ap- pointing administrative officials. By creating a new appointed elite, however, some participation theorists would argue that the civil service further concentrated power and reduced the oppor- tunities for citizen influence.8 3 We can formulate a version of the participatory dilemma as it applies specifically to risk policymaking: The increasing complex- ity of government policy issues, in particular the reliance on sci- entific method and data, reinforces broader trends that reduce individual influence. Now a scientific and technical elite plays an influential role in making social decisions. Quantitative models are commonly-used elements in the deci- sion process. Policy issues, including environmental risk, are cast in technical terms. Resolution of technical controversy is en- trusted to experts, because the lay public cannot grasp the basis for decisions and their consequences. "Expert" perceptions of problems (e.g., the nature of environmental risk) are judged to be more rational and more consistent with society's interests than the "subjective" judgments of the less technically sophisticated public. It is not long before the citizen is depicted as more of an obstacle to overcome than an interest or collective judgment to serve. In a political culture that placed a low value on citizen influ- ence-where people viewed themselves more as subjects than cit- izens-this concentration of power might not present a problem. But American society is one in which the need for participation 83. T. COOK & P. MORGAN, PARTICIPATORY DEMOCRAcy 3 (1971). One can also argue that the civil service expands participation for those who become career officials. For this point of view, see Rourke, Bureaucracy in the American Constituional Order, 102 POL. Sc. Q. 217. 230 (1987). 532 Risk and Democratic Process and citizen influence is high. In The Civic Culture, Almond and Verba studied the political cultures of five democratic nations- the United States, Britain, Germany, Italy, and Mexico. Political culture is defined in this study as "the particular distribution of patterns of orientation toward political objects among the mem- bers of the nation."84 A central element in the political culture is how people view their relationship to political authority and their capacity for influencing it. When people feel that they can influ- ence government decisions affecting them, they are described as high in "subjective competence." The data from the survey sug- gest a strong relationship between perceptions of subjective com- petence, high levels of actual participation, and effective democracy. Almond and Verba note that "the general pattern of authority in American social systems, including the family, tended to stress political competence and participation rather than obedience to legitimate authority." Similarly, "the opportunity to participate in political decisions is associated with greater satisfaction with that system and with greater loyalty to that system."85 The Civic Cul- ture data describe a pattern in which feelings of competence, op- portunities for participation, actual participation, and the legitimacy and stability of the political system are related and mu- tually reinforcing. The data also suggest problems when political realities do not match people's expectations. A sense of cynicism and powerlessness is likely to be stronger in the participatory political culture than the one in which people share less of a sense of subjective competence. Environmental policy research presents interesting compari- sons with these survey data, especially regarding the importance of feelings of competence and control. For example, research on public attitudes toward siting of hazardous waste facilities sug- gests a link between support for siting decisions, information about institutional controls, and the opportunity to influence sit- ing decisions and control measures.86 One study tested the effec- tiveness of compensation in altering people's perceptions of the 84. G. ALMOND & S. VERBA, THE Civic CULTURE: POLITICAL ATTITUDES AND DEMOCRACY IN FIVE NATIONS 14-5 (1963). 85. Id. at 38, 253. 86. See, e.g., Matheny & Williams, supra note 5; Kraft & Kraut, The Impact of Citizen Partici- pation on Hazardous Waste Implementation: The Case of Claremont County, Ohio, 14 PoL'Y STUD.J. 52 (1985). 19891 533 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:501 risks and benefits of accepting waste facilities in a community.87 Based on a survey of five Massachusetts towns, the study con- cluded that economic incentives (i.e., various forms of compensa- tion) had almost no effect in the community's willingness to accept a waste facility. The proposal having the greatest effect on attitudes was one allowing local public officials and citizens to conduct regular safety inspections. A survey of Wyoming resi- dents found that compensation proposals were less effective in persuading them to accept a facility than good information and the opportunity to participate in decisions.8 In another study, Elliott found that community control of a facility through a safety board with substantial authority helped in gaining public accept- ance of a waste site.89 The theorists' depiction of the participatory dilemma in mod- ern society can be summarized in these terms: The scope of gov- ernment expands, and with it the need to broaden opportunities for participation in political decisions. The expansion in govern- ment responsibilities creates the need for a professional, career service that invests power in an appointed elite. Broader political participation reduces individual influence while it expands the in- fluence of an elected elite. Finally, issues become more complex as they rely more on scientific method and must cope with the effects of new and rapidly changing technologies. This further reduces the capacity of the lay public and of elected representa- tives to affect decisions. Inevitably, society must rely on technical experts and administrative authority to make political decisions. And the spiral of alienation, apathy, distrust of politicians and political institutions continues. Rising education levels and mass communications make people more aware of their lack of power, so that they "are powerless relative to what they think they should be in an ideal democracy."90 87. Portney, The Potential of the Theory of Compensation for Mitigating Public Opposition to Hazardous Waste Treatment Facility Sitings: Some Evidence From Five Massachusetts Communities, 14 POL'Y STUD. J. 81 (1985). 88. Davis, Public Involvement in Hazardous Waste Siting Decisions. 19 POLITY 296, 302-303 (1986). 89. See Elliott, Improving Community Acceptance of Hazardous Waste Facilities Through Alterna- tive Systems for Mitigating and Managing Waste, I HAZARDOUS WASTE 397 (1984). 90. M. OLSEN, supra note 82, at 5. 534 Risk and Democratic Process B. The Participatory Theorists' Response Participation theory responds to this dilemma by offering a vi- sion of participation that is broader than the traditional mecha- nisms of elections and interest groups. Yet these theorists are also realists. However romantic Rousseau's vision of Swiss peas- ants gathering under an oak to transact public business may ap- pear, modern participation theorists recognize the implausibility of pure democracy, and they look to institutional forms that can be adapted to the demands of a modern state. "There is little wrong with liberal institutions," Barber has written, "that a strong dose of political participation and reactivated citizenship cannot cure."9 1 Barber's criteria for more participatory "institu- tional forms" are that they be realistic, workable, and compatible with existing institutions; offer appropriate safeguards (e.g., pro- tect minority rights); and contend with the obstacles posed by modernity-scale, technology, complexity, and the concentration of effective power in national institutions.92 Margolis' notion of "viable democracy" aims to establish institutions "capable of dealing with twentieth-century problems in democratic fashion."" Participation theory is optimistic about individual capacities to develop political skills, interests, and competence in democratic society. The two presuppositions underlying Thompson's citizen theory are "autonomy" and "improvability." Autonomy means that citizens are treated as the best judge of their own interests. Political, administrative, or technical elites may argue at times that people do not understand what is in their own interest. But their judgments cannot displace those of citizens; they can only assert that at some point in the future people will agree with a different point of view, and persuading people to come to that point of view is the function of leadership. The related notion of improvability is the "belief in the capacity of citizens in general to improve their judgment about what is in their interest . . . . Low political awareness, low levels of information, and lack of in- terest in issues are taken as a sign of deficiencies in social and 91. B. BARBER, supra nore 76, at xi. 92. Id. at 262. 93. M. MARGOLIS, VIABLE DEMOCRACY 157 (1979). 94. D. THOMPSON, supra note 79, at 14. 1-989] 535 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 political institutions, rather than as limitations inherent in individ- ual capacities. Discussion is highly valued in the participation literature. Its function in Thompson's citizen theory is "to help citizens to rec- ognize their own political interests; to create and reveal common interests; and to maintain peace and stability."95 With political talk, Barber asserts, "we can invent alternative futures, create mu- tual purposes, and construct competing visions of community."96 Mansbridge observes that face-to-face contact can "correct inac- curacies of perception, iron out differences, and create a spirit of community."97 Research on small group processes documents that discussion is far more effective than other, more passive forms of communication in changing attitudes, stimulating prob- lem-solving thought, and influencing action. The risk perception research suggests that time, interaction, and information can help people overcome their predispositions and adopt a more open stance on risk issues. Discussion emphasizes social interaction and strengthens associations with other groups in the political community. It reinforces the sense of autonomy and improvabil- ity of citizens, because it promotes individual reason, judgment, and choice. The theorists assert that participation engenders civic compe- tence by building democratic skills, overcoming feelings of powerlessness and alienation, and contributing to the legitimacy of the political system. Pateman bases much of her case on the argument that "we do learn to participate by participating and that feelings of political efficacy are more likely to be developed in a participatory environment."9 8 To build interest and capacities on a regional and national scale, she argues that people first need opportunities to build skills and confidence at more modest levels-in their neighborhoods, workplaces, and small institu- tions. Participation will breed participation, not only in politics but in other aspects of life as well. C. Toward a Participatory ideal The work of the participation theorists suggests the outline of a participatory ideal. It is an ideal because it defines characteristics 95. Id. at 86. 96. B. BARBER, supra note 76, at 177. 97. J. MANSBRIDGE, supra note 81, at 270. 98. C. PATEMAN, supra note 77, at 105. 536 Risk and Democratic Process that rarely would be met in practice, especially given current atti- tudes and legal arrangements. But even if we devised institu- tional mechanisms and procedures that fit some of these characteristics, if only in part, then we would have made progress toward resolving the participatory dilemma. In this participatory ideal, institutions, administrators, and their technical experts would recognize the lay public as citizens rather than as subjects. This change in attitude accepts that people are the best judge of their own interests and can acquire the political skills and knowledge to take part in collective decisions about en- vironmental risk. Science and technical expertise would inform citizens' judgments, but not replace them. Societal risk managers would attempt to educate public judgments and to lead citizens to a more "rational" or "enlightened" conception of their interests. The perspective in the risk literature that comes closest to reflect- ing this vision is that on public deliberation. Second, we would design institutions and procedures to allow for the direct participation of amateurs. At the national level, most participation now takes the form of individuals acting in their capacities as elected representatives, appointed administra- tors, interest group professionals, or technical experts. Even elected representation has become a career. Both the career civil service and the reliance on technical expertise are necessary ad- justments to the expanding scope and complexity of public ac- tion. But they diminish the opportunities for citizen influence. Participation theory seeks to involve people as amateurs rather than in their professional or career roles. "Direct" participation simply refers to the opportunity to influence decisions without having to rely on organized interest groups or professional in- termediaries, except in a staff capacity. An example would be the citizen on a commission recommending criteria for siting a waste facility. Brooks' discussion of the jury model and Shrader-Fre- chette's technology tribunal both incorporate forms of direct participation. Third, in a participatory ideal, citizens share in making collec- tive decisions. The ideal envisions participation that is more than therapeutic, oppositional, or plearling, but in which "citizens share in governing . . . ."9 Citizens act as authorities who deter- 99. D. THOMPSON, supra note 79, at 3. For a critique of participation as therapy, see Arnstein, A Ladder of Citizen Participation, 35J. AM. INST. PLANNERS 216 (1969). 1989]1 537 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 mine policy or as collaborators with government officials who codetermine policy. Participation in risk policymaking typically falls short of this ideal. It often occurs too late and is designed merely to identify potential opponents. At most, participation al- lows members of the public to protect their interests but not to share in the responsibilities and obligations of collective choice. Under this ideal, participation is not purely defensive or a last point for responding to administrative proposals, but a process for setting priorities and shaping decisions as they emerge. Fourth, a participatory mechanism or procedure should allow for face-to-face discussions over some period of time. Construc- tive talk, deliberation, discovery of shared values, the search for common visions-these are not the acknowledged virtues of the American administrative process. When people with different points of view talk, it is usually through their lawyers, who already are marking out positions for litigation. Or they communicate in writing, through a formal public comment period, where the most telling points are the ones exposing a legal vulnerability the op- position can exploit later.00 The ideal envisions settings that al- low people to talk, listen, debate, persuade, and be persuaded. This implies communications that are face-to-face and recurring. This view comes through strongly in the public deliberation work. It is one of the conditions Brooks defines for institutional mecha- nisms that can allow a synthesis of technical and value issues. The risk literature also recognizes that the opportunity to focus on is- sues, interact with others (including technical experts), and search for shared values can produce better decisions and perform im- portant psychological and social functions for participants. Finally, under the ideal, citizens participate on some basis of equality with administrative officials and technical experts. A par- ticipatory process or mechanism can accomplish this in several ways. One is in the terms of the relationship with authorities and experts. Do participants have the opportunity to define issues, question technical authorities, dispute evidence, and influence the agenda? Are they dealing directly with administrative officials who can exercise decision authority, or with staff who can only represent authoritative decisionmakers? A second basis for equality is in preparation and resources. Do participants have in- 100. See Fiorino, Regulatory Negotiation as a Policy Process. 48 PuB. ADMIN. REV. 764 (1988) (discussing conventional rulemaking). 538 Risk and Democratic Process dependent capabilities for gathering and assessing technical and analytical information and for acting on that information?'0 A third way that we can allow citizens to participate on a more equal basis involves reassessing our use of formal models for evaluating risks; cost-benefit and other models make risk decisions less ac- cessible to non-experts. They also attempt to substitute technical judgment for political judgment and analytical technique for de- bate and discussion. Analysis is necessary and is fully compatible with democratic process when it informs political judgment, but not when it replaces such judgment. The greater the reliance on such models, the less that citizens can act on a basis of equality with administrative officials and experts. D. Illustrative Applications of the Participatory Ideal Three brief examples illustrate how we can use these criteria to assess institutional mechanisms and procedures. The first, the ASARCO case, was Administrator Ruckelshaus' effort to apply the public deliberation perspective I discussed in Part 1.102 The issue was environmental standards for controlling emissions of inor- ganic arsenic from a smelter owned by the American Smelting and Refining Company near Tacoma, Washington. EPA had con- cluded that approximately four new cases of lung cancer could be expected annually at existing emission levels. Installation of "best available" control equipment at the smelter would reduce the estimate to about one case per year. The dilemma was that requiring the facility to install best available technology would close the plant, which was an economic lifeblood of the commu- nity. EPA took the issue to the community by announcing a series of three public workshops in the summer of 1983. The Agency began each workshop with an explanation of health risks and the procedures for estimating them. The audience then was divided into three discussion groups, with Agency officials facilitating the discussions. The issue became moot in 1985, when the smelter closed anyway for economic reasons. 101. Kasperson stresses that effective participation depends on people developing "in- digenous" technical and analytic resources as well as the institutional means for acting on that knowledge. See Kasperson, supra note 32, at 278. 102. For discussions and assessments of ASARCO, see Reich, supra note 16, at 1632; Baird, Tolerance for Environmental Health Risks: The Influence of Knowledge, Benefits, Voluntariness, and Environmental Attitudes, 6 RISK ANALYSIS 425 (1986);. Call, Arsenic, ASARCO, and EPA: Cost-Benefit Analysis, Public Participation, and Polluter Games in the Regulation of Hazardous Air Pollutants, 12 ECOLOGY L. Q. 567 (1985); S. KRIMSKY & A. PLOUGH, supra note 39, ch. 5. 1989]1 539 COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 The second example is regulatory negotiation. This is a pro- cess of policy formulation that brings representatives of affected interests together to reach consensus on the content and some- times the language of a proposed rulemaking. EPA has com- pleted seven such negotiations to date, and other federal agencies have completed about the same number. The negotiations take place within the notice-and-comment format of the Administra- tive Procedure Act. Well in advance of any negotiations, the Agency conducts a thorough convening effort to identify affected parties, define the issues and whether or not they are negotiable, and determine the feasibility of a negotiation. EPA participates as a party-at-interest in the negotiations; like any other party, it can block agreement or withdraw without prejudice at any point. Once constituted (under the Federal Advisory Committee Act), the committee has substantial control over its mode of operation, composition, use of resources, and the terms and timing of its dissolution. Decisions are made by consensus, which typically has meant "the concurrence of all interests represented" on the com- mittee. The Agency commits in advance to publishing the com- mittee's consensus as a proposed rule, so long as it is consistent with the Agency's statutory authority.0 3 The third example is exploratory efforts with "citizens panels."104 Modeled after the lay jury, citizens panels convene representatives of the lay public to hear testimony, question tech- nical and administrative experts, deliberate over the issues, and reach a conclusion. Participants are selected through stratified random sampling, although in some cases representatives of af- fected interest groups may also take part. A steering committee representing a cross-section of affected interests determines which experts will present evidence and views so the panel will receive a balanced treatment of the issues. The panel devotes two or more days to studying the problem, hearing testimony, ques- tioning experts, and evaluating the evidence. It then debates and discusses issues and reaches a decision or recommendation. The 103. A growing literature describes and assesses regulatory negotiation in detail. It in- cludes: Fiorino, Regulatory Negotiation as a Policy Process, supra note 100; Harter, Negotiating Regulations: A Cure for Malaise, supra note 6; Fiorino & Kirtz, Breaking Down Walls: Regulatory Negotiation at EPA, 4 TEMP. ENvrt.. L. & TECH. J. 29 (1985); Susskind & McMahon, The Theory and Practice of Negotiated Rulemaking, 3 YALEJ. ON REG. 133 (1985); and Perrint, Negoti- ated Rulemaking in Practice, 5 J. POL'Y ANALYSIS & MGMT. 482 (1986). 104. Crosby, Kelly & Schaefer, Citizens Panels: A New Approach to Citizen Participation, 48 Pus. ADMIN. REV. 170 (1986). 540 Risk and Democratic Process panel is a flexible tool that could include interaction with deci- sionmakers in preparing its decisions or recommendations. The major use of a citizens panel on an environmental issue occurred in 1984, when the Center for New Democratic Processes con- vened five regional citizens panels for policy recommendations on the impacts of agriculture on water quality in Minnesota. Three members from each regional panel then took part in a state-wide panel that offered policy recommendations to the state legislature. The three examples meet the participatory characteristics in va- rying degrees. In ASARCO, the Administrator tried to inform public judgment and build citizens' capabilities for understanding and making policy choices. EPA went directly to citizens in their private capacities, rather than only to interest group in- termediaries. Given the structure of the workshops and hearings, there was only limited opportunity for the public to engage in deliberations among themselves and with Agency officials. Reich cites evidence, however, that the residents of Vashon Island, who were exposed to emissions but not economically dependent on the smelter, developed more sympathy with the Tacoma resi- dents, so that "a feeling of citizenship began to infiltrate even the expressions of advocacy."105 The Tacoma experience was not effective in granting decision authority and in allowing citizens to participate on a basis of equality with administrative and technical professionals. The Ad- ministrator was criticized for bringing the issue to the community the way he did; a greater effort to share authority for the decision would have run into any number of legal and political obstacles. The workshops were insufficient to bring citizens to the point that they could fully grasp and debate the issues. Baird's survey found that the participants were not especially well-informed on the health risk assessment even after the workshops.'0 6 More time, a more structured format, and better use of group learning tech- niques probably would have been necessary to have substantially affected citizens' capacities. Regulatory negotiation presents a different case. It does not involve citizens as amateurs, because the members of the negoti- ating committees typically are professionals who represent organ- 105. Reich, supra note 16, at 1635. 106. Baird, supra note 102, at 434. 1989] 541 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:501 ized interests. But when it is assessed under the other characteristics, negotiation appears to offer important benefits as a participatory process. Most of the parties give the process high marks as a forum to explore and discuss issues, discover shared concerns or objectives, trade off preferences, and understand the interests underlying the other parties' positions. They also ap- preciated the opportunity to work with senior EPA managers on the committee and in the smaller subgroups. They noted striking differences between the negotiation process and the typical inef- fectiveness of the written public comment period. Having equal access to information available to the other parties and operating on the principle of consensus allowed them to deal with each other and the Agency on some basis of equality. The parties ob- served reasoned discussion, lively debate, and a sense of collec- tive responsibility for their product-all of which are rare in the conventional rulemaking process.07 Negotiation meets several but not all of the characteristics of the participatory ideal. A properly designed citizens panel comes closest to meeting the ideal. It permits the direct participation of amateurs who can discuss and debate issues over a period of days. Panelists can have access to technical experts who advocate different points of view. Participants are likely to be treated as citizens, because they are assumed to have the capacity to understand issues and the implications of the available choices- for their own and the com- munity's interests. The opportunity to exercise or share decision authority will depend on the design of the panel. It would be neither legally possible nor desirable to delegate decision author- ity to a randomly-selected citizens panel. But citizens can gain the opportunity for access and influence through discussions and de- liberations with the people having the decision authority. One creative application of the citizens panel concept would be for risk policymakers, such as an EPA assistant or regional administrator or a state environmental commissioner, to take part in the delib- erations and discuss recommendations with the panel. The above examples apply to the formulation of policy, but there also are ways to promote procedural democratic values in the implementation of risk programs. For example, the safety panel that Elliott proposes in his study of public acceptance of waste facilities embodies a participatory approach to implement- 107. See Fiorino, supra note 100, at 768-770. 542 Risk and Democratic Process ing policy.108 The system of state and local emergency planning committees established under Title III of the Superfund Reauthorization Act also offers a structure for greater citizen con- trol of risk policies at the community level. Experience at the state level suggests that citizen access to information about indus- try practices and emissions can create a structure for achieving community-level agreements that reduce potential risks. 09 One of the challenges of implementing the Title III program in com- ing years will be to use it to achieve greater citizen participation in managing environmental risks at the local level. IV. TOWARD DEMOCRATIC PROCESS IN RISK POLICYMAKING The obstacles to greater participation appear almost to be in- surmountable. There is first the nature of risk policymaking it- self. Often the problems are invisible and exist only as scientific constructs. Effects of exposure to chemicals or other hazards may not appear for decades. When they do appear, cause and effect relationships are difficult to establish. The factual premises of de- cision depend on the uncertainties of science. The assumptions underlying the assessment of risk are a subject of controversy and can have a profound effect on the results. Agencies and their sci- entific advisors must hedge their conclusions and communicate complex information in terms that clarify, but do not oversim- plify. The methods for assessing risk are complex, specialized, and require their own terminology. Formal analytic models add another layer of specialization and complexity as well as a set of prepackaged value premises for determining acceptable risk. Even when decisions appear to have been made, legal challenges upset the conclusions and send agencies back to ponder a more acceptable approach."10 These problems are inherent in risk policymaking. In addition, there are the obstacles to effective participation in administrative policymaking generally. From the agency perspective, participa- tion typically occurs on an "as needed" basis, so that influence over a decision tends to occur in direct proportion to the power 108. See Elliott, supra note 89. 109. The national community right-to-know program is established in the Superfund Amendments and Reauthorization Act of 1986 (SARA), 42 U.S.C. H 11001-11005, 11021-11023, 11041-11050 (Supp. IV 1986). 110. For an analysis of the effects of judicial review in the air pollution program, see R.S. MELNICK, REGULATION AND THE COURTS: THE CASE OF THE CLEAN AIR AcT (1983). 1989] 543 COLUMBIA JOURNAL OF ENVIRONMENTAL LAW [Vol. 14:501 of an interest group to challenge it later. The most common mechanism for involving nonexperts and nonelites is the public meeting or hearing, which embodies most of the liabilities of our current approach to participation. Public involvement is left largely in the hands of public relations professionals who define the task in terms of public information, education, or persuasion and lack the leverage to adapt institutions and allow the lay public a more substantive role. Participation is seen in narrowly proce- dural or tactical terms-as a way of complying with the law, or of granting necessary concessions to organized interests, rather than as an expression of fundamental democratic values. There clearly are significant obstacles to achieving more citizen participation. Modern technological societies are poorly-suited to participatory democracy, as the participation theorists have been telling us. The problem, however, may be not that the chal- lenge is insurmountable, but that people have not thought hard enough about how to overcome it. Risk professionals seem will- ing to accept the improvability (to use Thompson's term) of toxic- ity testing, exposure modeling, or benefits estimation techniques, but not of mechanisms or procedures for more participatory deci- sionmaking. In our capacities for risk assessment and net-benefits analysis, we have made great strides over the last two decades. With some exceptions, such as the growing use of mediation and negotiation, one can say that the capacities for participation have fallen well behind the challenges posed by environmental risk problems.' One thesis worth considering is that of the participation theo- rists, who assert that current levels of low involvement are more the result of institutional deficiencies than an inherent lack of citi- zen interest or capacities. A recent analysis of voter turnout in nineteen industrial democracies provides some support for this view. Variations in turnout are related not to cultural factors (which in the United States would support a high level of partici- pation), but to the effects of institutions and electoral laws. "Where institutions provide citizens with incentives to vote," Jackman concludes, "more people actively participate; where in- 111. See G. BINGHAM, RESOLVING ENVIRONMENTAL DispuTrEs: A DECADE OF EXPERIENCE (1986). The case for mediation of technical.policy disputes is made in Ozawa & Susskind, Mediating Science-Intensive Policy Disputes, 5 J. POL'Y ANALYSIS & MGr. 23 (1985). 544 Risk and Democratic Process stitutions generate disincentives to vote, turnout suffers.""l2 The same may hold for administrative policymaking: If institutions and incentives were designed not to discourage but to encourage participation, then perhaps we would find more of it. Surely the current institutional infrastructure for participation, with its in- strumental focus, assumption of conflict, bias toward organized interests, and procedural orientation, does not present a suffi- cient test of citizens' willingness to take part in collective risk decisions. The existing literature on risk does not provide answers, but it defines issues and establishes an empirical foundation. The pub- lic deliberation perspective offers a vision of participatory deci- sionmaking and an appreciation of the need to sustain the credibility of risk institutions. From the perspective of resolving disputes among experts, we can see the importance of the synthe- sis of expert and generalist competencies and the interrelation- ships between the two. The critical writing on formal models demonstrates the need to reconcile those models with procedural and substantive democratic values. Research on risk perception can help in explaining public attitudes and beliefs about risk and in understanding the influence of social and political factors. Re- cent research appreciates the cultural, experiential, and subjective influences on peoples' judgments about risk."13 Because risk policymaking is so dependent on scientific method and language, effective risk communication is an essential tool for citizen partici- pation. Comparative studies of demands for participation and of how administrative systems balance scientific with political au- thority demonstrate different patterns of democratic adaptation to technical controversy. The participation movement of the 1970s may have been a nec- essary stage in our administrative development, but it is not the only model available for solving the participatory dilemma today. Participation theory offers an alternative vision of citizen partici- pation and a set of criteria for adapting existing institutions or designing new ones. Democratic theorists still will need to estab- lish a firmer grounding for a reinvigorated concept of citizen par- ticipation, in particular as it is applied to risk policymaking. I 112. Jackman, Political Institutions and Voter Turnout in the Industrial Democracies, 81 AM. POL. Sci. REV. 419 (1987). 113. See, e.g., Fessenden-Raden, Fichton & Heath, Providing Risk Information in Communi- ties: Factors Influencing What Is Heard and Accepted, 12 Sci. TECH. & Hum. VALUES 94 (1987). 5451989] COLUMBIA JOURNAL OF ENVIRONMENTAL LAw [Vol. 14:501 have suggested the outlines of that grounding here: The value placed on participation should depend on more than the latest political or administrative fashion. Its aims should be broader than the instrumental objectives of the agency or key clientele groups. Its ethical basis should reflect democratic values and the intellectual contributions of democratic theory, not just the need to satisfy opposition demands as they arise. Participation theory is not offered as a substitute for current approaches to participa- tion at the national level, but as a basis for moving beyond the interest group, adversarial, pluralist conception and for stimulat- ing institutional innovation and experimentation. At a practical level, we need more institutional policy analysis, which is "the study of government reform and its consequences." Although it takes many forms, in institutional policy analysis "the focus is always on government reform as an independent varia- ble."' i4 Agencies could design participatory experiments that es- tablish a sounder empirical basis for research and institutional innovation. Research should document not only the conse- quences of these experiments for policy, but also for people's sense of citizen competence and for the legitimacy of institu- tions.i5 Agencies will need to take the design of participatory institutions as seriously as they take the design of their analytical documents. This will require intellectual rigor, resources, and the capacity to learn through experience. At the level of individual programs, administrators should not assume that broad participation will translate directly into sup- port for policies. In its often exemplary, open planning programs of the 1970s, the Army Corps of Engineers probably had its ex- pectations raised by the "participation thesis" of the social psy- chologists and organization theorists, who asserted that effectively involving people in a decision would ensure their agreement with the substantive result. But people gave high rat- ings to the Corps' process while they continued to disagree with its projects.116 Even a longer-term participation program might have had a limited effect on evaluations of development projects, 114. See Gormley, Institutional Policy Analysis: A Critical Review, 6 J. Pot'Y ANALYSIS & McmT. 153 (1987). 115. For an evaluation of citizen participation programs, see Rosener, Citizen Participa- tion: Can We Measure Its Efectiveness?, 38 PuB. ADMIN. REV. 457 (1978). 116. D. MAZMANIAN &J. NIENABAR, CAN ORGANIZATIONS CHANGE? ENVIRONMENTAL PRo- TECTION, CITIZEN PARTICIPATION, AND THE CORPS OF ENGINEERS, 30-32, 166-67 (1979). 546 Risk and Democratic Process but it could have affected the public's views of the Corps as an institution and achieved a higher measure of confidence in the result. Either way, we can. only guess, because the Corps lost much of its interest in open planning when the results failed to live up to its expectations. More broadly, administrators will need to be more realistic in their expectations about what participation can accomplish. It will not offset the effects of bad decisions, unresponsive policies, or poor governmental performance. Survey data demonstrate that the declining confidence in American institutions over the last two and one half decades has largely been a response to events, especially a poor economy, Vietnam, Watergate, and the energy crisis." 7 People still care about results and substantive perform- ance. Participation is just one element in the complex relation- ship between citizens and their political institutions. But it is an important element that in the area of risk analysis has been ne- glected. In a political culture in which citizens value participation and are reluctant to delegate collective decisions to administrative and technical elites, it is difficult to separate process from sub- stance, or to assume that an analytically rational outcome will ob- viate the effects of deeply-held democratic values. The principle should be to avoid a "one-dimensional" conception of democracy in which, as Bachrach explains, interest is defined solely as a mat- ter of end results, or of "gains in material well-being, power, or status," and not also as a matter of process, or of "the personal satisfaction and growth attained from active engagement in the political process."' " It is easy to assume that the issues are too complex, the science too uncertain, the need for objective rationality too great to allow the lay public a substantive role in making risk decisions. But easy assumptions only beg hard questions. At some point, we will need to develop mechanisms for achieving more effective citizen participation, because only then can we find a democratic solution to the challenges of a technological society. To accomplish this, we first must recognize that the function of government is not only to reduce exposure to risks but to allow people as citizens to share in the tasks of governing. 17. S. LIPSET & W. SCHNEIDER, THE CONFIDENCE GAP: BUSINESS, LABOR, AND GOVERN- MENT IN THE PUBLIC MIND. supa note 3, at 375-412. 118. P. BACHRACH, supra note 77. at 38. 1989) 547