Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Introduction When government agencies must decide whether to evacuate people from areas where toxic substances are leaching from waste dumps, set standards for exposure to suspected carcinogens, or de- cide whether to license nuclear power plants despite some low prob- ability of rupture in a future earthquake, democratic societies are faced with difficult choices. The usual criteria of consensus or social acceptability are insufficient to resolve such issues of modern tech- nology. The decisions also need to be scientifically informed, because some choices set in motion physical or biological processes whose results, if they could be foreseen, would be considered undesirable by most people. Only a few experts possess the best knowledge available to es- timate accurately the extent of the possible harm or the likelihood of its occurrence. But while great weight needs to be given to the specialized knowledge of these experts, democratic principles require that the decisions be controlled by officials, generally nonspecialists, who are answerable to the public. As Jefferson realized Tong ago, public decisions that require specialized knowledge raise questions about political power. I know of no safe depository of the ultimate powers of society but the people themselves; and if we think them not enlightened enough to exercise their control with a wholesome discretion, the remedy is 14
INTROD UCTION not to take it from them, but to inform their discretion. (Thomas Jefferson, letter to William Charles Jamris, September 28, 1820) 15 To remain democratic, a society must find ways to put specialized knowledge into the service of public choice and keen it from h~.omin the basis of Dower for on slits ~0 Because technological- decisions have implications for public health and for political power, they are often highly contentious and emotional. Participants, expert and nonexpert alike, have much at stake and are strongly motivated to work for the outcomes they favor. The ensuing political struggles are often frustrating for the participants. Nonexperts are frustrated by the inaccessibility of the knowledge they need to inform their opinions and by presentations of needed knowledge that are oversimplified, overly technical, or condescending in tone. Technical experts are frustrated when their explanations of available knowledge are met with apathy, disbelief, or anger. Government and corporate officials- are frustrated when their discussions of technological alternatives are met by expressions of public mistrust and accusations of malevolence. Environmental activists are frustrated by requirements that they argue positions that are based on human and environmental values in the language of science and technology and by lack of sufficient resources to make technical arguments strongly. Participants come to see the debates in different ways, depending on their positions in them and the frustrations they have experienced (Dietz et al., 1989; Edwards and van Winterfeldt, 1986; Lynn, 1986 Otway and von Winterfeldt, 19823. Many, especially in the scientific and technical community and in government, have defined the un- derlying problem in terms of "public understanding of risk," "risk perception," and "risk communication." They believe that what is needed is for people to better understand or more accurately perceive the potential costs and benefits of certain technological op- tions. To accomplish this, they argue that scientists, governments, and the mass media need to do a better job of risk communication, by which they mean explaining the choices and their likely conse- quences to nonexperts. They argue that increased efforts of this kind would make conflicts about technological choices easier to resolve and would enable the society to make better choices about protect- ing public health, safety, and environmental quality. For reasons elaborated throughout this report, we believe this concept of risk communication and decision making is incomplete and, in important
16 IMPROVING RISK COMMUNICATION respects, misleading; it supports misconceptions about the risk com- munication process and raises unrealistic expectations about what risk communication can accomplish. THE NEW INTEREST IN "RISE COMMUNICATION Interest in "risk communication" is quite recent.1 That interest is evident in a recent explosion of conferences, seminars, articles, and books with the term "risk communication" in their titles (Bean, 1987; Covello et al., 1987b, 1988; Davies et al., 1987; Fischhoff, 1987; Lind, 1988; Otway, 1987; Plough and Krimsky, 1987; Zimmerman, 19873. It reflects increased attention, especially in some agencies of the federal government but in other organizations as well, to the task of informing the general public about the nature of the health, safety, and environmental risks associated with personal and societal choices. The new concern with informing the public has several motivating sources, not entirely consistent with each other, including (1) a requirement for or desire by government to inform, (2) a desire by government or industry officials to overcome opposition to decisions, (3) a desire to share power between government and public groups, and (4) a desire to develop effective alternatives to direct regulatory control. Moreover, the term risk communication has different meanings to different users. Requirement for or Desire by Government to Inform Sometimes government is required to inform the public. A series of federal laws, beginning with the Administrative Procedures Act of 1946 and continuing with the Freedom of Information Act, the National Environmental Policy Act, and the "Community Right to Know" provisions of Title IT! of the Superfund Amendments and Reauthorization Act of 1986, recognizes the public's right to be informed about certain hazards and risks, even if they have no part in the decision-making process. Many of these laws have been reinforced by federal court decisions and presidential executive orders. These actions emphasize the government's responsibility to be accountable to the people; they state as national policy that regulators must explain why one course was chosen rather than another and that the public has a right to see and challenge the basis for the decisions. Thus agencies are required to send messages to the public about the reasons for their decisions and to solicit messages of comment from citizens. The term risk communication is sometimes used to describe these messages.
INTROD UCTION 17 Some government officials provide information not required by law. Regulatory officials sometimes do this because they believe that people would benefit from specific knowledge. For instance, the U.S. Environmental Protection Agency (EPA) over the past few years has made an effort to inform householders about the hazard of radiation exposure from indoor radon. And public health officials, responding to their general mandate, have long offered information to citizens about the health risks of dietary and sexual habits, drug and alcohol use, and other personal activities. Provision of such information is what some public health officials mean by risk communication. - Des~re to Overcome Opposition to Decisions Over the past 30 years public participation in debates on tech- nological issues has intensified. More groups have become involved, including workers potentially at risk from hazardous activities, reg- ulatory organizations, citizens' and environmental groups, the press, and the courts. The proponents of controversial technological options or decisions, most frequently in government or industry, often meet intense political opposition. Frequently, groups of citizens who op- pose particular technological projects have delayed or stymied those projects with lawsuits, mobilization of congressional opposition, or public demonstrations. When a government or industry official has the benefit of extensive scientific study and the opposition seems simply to disregard the technical evidence, the official can come to see "the public" as irrational. Government and industry officials who see the issues this way are likely to define the conflicts that surround them as debates between the informed and the ignorant or, worse, between the rational and the irrational. Such officials are tempted to look for ways to influence the members of the opposition, either by more actively presenting a straightforward account of the knowledge they have available or by carefully packaging or even distorting that knowledge to achieve a persuasive effect. The use of information to overcome political opposition makes some notion of risk communica- tion attractive to many proponents of controversial technology; it is, in fact, what they mean by the term. Desire to Share Power Between Government and Public Groups Government officials have sometimes seen in risk communication a way to reduce conflict with segments of the public by sharing power. In this situation an agency takes the role of technical analyst and
18 IMPROVING RISK COMMUNICATION adviser, gathering and summarizing the information relevant to a decision at hand and explaining that information to various political actors. The agency's role might be to inform a public debate, for instance, in a legislative decision on siting of a hazardous facility. Or, if the agency is legally required to make the decision itself, it can provide information to the public and use the ensuing debate to help arrive at a decision that it judges to be both defensible within its legal mandate and maximally acceptable to the interested parties involved. Such an approach was used, albeit unsuccessfully, by the EPA in a controversial case in 1983. Prior to making a regulatory decision about an Asarco Corporation smelter that was releasing arsenic into the air, the agency presented the people of Tacoma, Washington, with the best information it had available about the risks and ben- efits of three possible outcomes: continued operation of the smelter, operation with pollution controls added, and closing of the smelter (Krimsky and Plough, 19-88~. EPA intended that the ensuing dia- Togue would help the community arrive at its own preference and inform EPA so it could make a defensible decision that would also satisfy local opinion. Administrator William Ruckelshaus justified his action, which depended critically on the success of the agency's efforts to provide information, with an appeal to Jefferson's advice to inform the public's discretion (Ruckelshaus, 1983~. The incident led to a heated controversy in which EPA was accused by some of an eva- sion of its responsibility and by others of attempting to manipulate the public by presenting an incomplete set of options. Although the smelter was shut down by Asarco before the public process ran its course, Ruckelshaus's goal of achieving consensus appears unlikely to have been attained. Desire to Develop Effective Alternatives to Direct Regulatory Control Government officials sometimes wish to persuade individuals to protect their health by personal action rather than to adopt regula- tory policies that require health-protective actions. The new interest in risk communication in government partly reflects a search for al- ternatives to direct regulatory control of health hazards, which was accelerated in the 1980s by the Reagan administration's philosophi- cal opposition to regulation. Government agencies have sought ways to control hazardous substances or activities short of banning them
INTROD UCTION 19 (as they have done with high-dose vitamin preparations), restricting or taxing them (as with alcohol), or requiring control measures (as with seat belts). Some of the alternatives involve replacing regulatory prohibitions and financial penalties imposed on those who produce hazardous technologies with reliance on informed discretion of the users of those technologies. For instance, the early 1980s brought a shift in the government's treatment of most motorists' unwillingness to use seat belts. A regulatory requirement that manufacturers equip cars with air bags or other "passive restraints" to protect passengers who fad! to fasten seat belts was replaced by a campaign to persuade, relying on paid and public service advertising. The government even supported research on better ways to convince people to use seat belts (Geller, 1983~. Now, after many years, there appears to be an increase in the use of seat belts, although in some cases this may be due to state laws mandating their use. Such persuasion programs are adequate as alternatives to regulatory constraint only if two co-n- ditions are met: if persuasion is accepted as a technique of public policy and if persuasion is about as effective as direct regulatory control. To some an important aspect of risk communication is the use of messages to induce people to protect themselves. A NEW DEFINITION OF RISE COMMUNICATION Although the motives for and meanings of risk communication described above are very different in some ways, each emphasizes a particular kind of message: a message that is developed by techni- cal experts; that describes or characterizes hazards, risks, or risk- reducing actions; and that is addressed to nonexperts. To many who use the term, risk communication means simply the development and delivery of this kind of one-way message. This widespread usage is illustrated in the foreword to the published proceedings of the first National Conference on Risk Communication, attended by 500 peo- ple in Washington, D.C., in 1986. William Reilly, the president of the Conservation Foundation, observed: [In] the conflict or confusion over risk questions . . . often the commu- nication process is at fault or, at the least, exacerbates the problem. Risk communicators simply do not do a good job of getting their message across. (Reilly, 1987:'rii) This very typical formulation equates risk communication with the delivery of certain kinds of messages one-way messages from government or other risk communicators to the general public about
20 IMPROVING RISK COMMUNICATION the nature of risks. It defines the success of risk communication from the point of view of the senders of those messages, in terms of "getting the message across." The image is of experts enlightening or persuading an uninformed and passive public. We consider this formulation of the problem to be incomplete in critical ways. Increased efforts "to get the message across" by describing the magnitude and balance of the attendant costs and benefits or by telling people which option provides the greatest net benefit to society will have little effect for several reasons. First, costs and benefits are not equally distributed across a society. Those who bear more than a proportionate share of the costs of one of the options want to convince others that the selection of that alternative would be unfair to them. Other political participants want to make similar arguments on their own behalf or to consider the arguments of all the interested parties. Thus an important aspect of conflicts about technological issues is that these are often conflicts between different interest groups. These conflicts cannot be resolved simply by knowledge about the likely effects of each alternative on the society as a whole or on various groups. Second, people do not agree about which harms are most worth avoiding or which benefits are most worth seeking. They want to argue for the protection of what they value and to consider which values are most worth preserving or advancing in each decision con- text. Because conflicts about technological issues pit values against each other, it is impossible to calculate net benefit to society or even to subgroups of the society-on any scale that will satisfy all the participants. Values need to be debated and weighed in a political process. Third, citizens of a democracy expect to participate in debate about controversial political issues and about the institutional mech- anisms to which they sometimes delegate decision-making power. A problem formulation that appears to substitute technical analysis for political debate, or to disenfranchise people who lack technical training, or to treat technical analysis as more important to decision making than the clash of values and interests is bound to elicit re- sentment from a democratic citizenry. Because of such reactions to them, problem formulations that attribute technological conflict to widespread public ignorance only exacerbate the conflict. We do not deny or minimize the importance of scientific and technological knowledge to informed public decisions about technol- ogy. In fact, we strongly endorse the proposition that understanding
INTROD UCTION 21 science in general and the likely consequences of particular techno- logical choices should be more widespread. But we emphasize just as strongly the fact that technological choices are value laden. Non- experts need to gain technical knowledge, but technical experts and public officials also need to learn more about nonexperts' interests, ~ . values, and concerns. In a democracy communication is an essential part of all societal decisions. The participants individuals, groups, and institutions- express their concerns and viewpoints, present facts and arguments to support them, and listen to what other participants have to say. At various points in this ongoing process, elected officials and pub- lic servants act in the name of the society, sometimes adding their own messages to those already current. The communication con- tinues, with concerns and viewpoints about government actions and messages as well as about the original issues being addressed. We see risk communication as a particular instance of this sort of democratic dialogue. Thus we have come to use the term risk communication differently from its common current usage.2 Risk communication is an interactive process of exchange of information and opinion among individuals, groups, and institutions. It involves multiple messages about the nature of risk and other messages, not strictly about risk, that express concerns, opinions, or reactions to risk messages or to legal and institutional arrangements for risk man- agement. As we will establish in Chapter 4, risk communication is successful only to the extent that it raises the level of understanding of relevant issues or actions and satisfies those involved that they are adequately informed within the limits of available knowledge. Risk communication is a component of risk management, which is the selection of risk control options. It is the process that provides the information on which government, industry, or individual deci- sion makers base their choices. Successful risk communication does not guarantee that risk management decisions will maximize genera] welfare; it only ensures that decision makers will understand what is known about the implications for welfare of the available options. The above definition of risk communication differs critically from many common uses in distinguishing between communication, which is an interactive process, and messages, which flow one way. Among people responsible for designing messages about risk, there is a temp- tation to confuse the task of message design and dissemination with the entire communication process and to equate the success of their messages in producing the effect desired with the success of risk .
22 IMPROVING RISK COMMUNICATION communication. We have chosen a definition that takes a broader perspective than that of any single participant in the process in or- der to emphasize the difference between the disparate activities en c] goals of the many participants and the social purposes of the risk · - communication process. Risk communication includes all messages and interactions that bear on risk decisions. Thus risk communication includes announce- ments, warnings, and instructions moving from expert sources to nonexpert audiences" the kinds of messages Reilly refers to. But it also includes other kinds of messages about risk information and information sources, about personal beliefs and feelings concerning risks and hazards, and about reactions to risk management actions and institutions. Not all these messages are strictly about-risk, but all are material to risk management. Our use of the term risk communication also pays explicit at- tention to the social interaction and debate that are essential to democratic political choice and that often contribute to personal decisions about hazardous activities. Risk communication includes messages moving in various directions-not only from experts to nonexperts but also from nonexperts to each other, from nonexperts to experts, and especially the messages of political participation, from citizens to public decision makers. Decisions in government depend on dialogue between the decision maker and staff within the responsible agency and between the decision maker and various po- litical participants, who influence the decision maker's view of the risks and the risk management options. Messages about nonexperts' perceptions of fairness, legal constraints, feelings of outrage, and the mobilization of interest group pressure are among the important ele- ments of the risk communication process, along with messages about the risks themselves. Even with personal risk decisions, choice often depends on a dialogue in which technical knowledge may not be the dominant influence. Decisions to stop smoking, for instance, have often been influenced more strongly by the expressed value prefer- ences of the smoker's children than by experts' messages about health consequences. As with other communication in a democracy, the intent of the participants in risk communication is sometimes political. That is, messages about risk are sometimes intended to influence the beliefs or actions of those to whom they are addressed. Risk communication, then, must be understood in the context of decision making involving hazards and risks, that is, risk management. Communication about
INTROD UCTION 23 risk deserves special attention because the highly technical nature of the subject matter makes it more difficult than communication about other controversial issues. Risk decision makers, including individuals managing personal hazards and participating in public decisions, need to seek and interpret complex technical information from scientific disciplines in which they have not been trained. They must communicate with, and to some extent rely upon, the experts who generate that information. Because the attendant choices are controversial, affecting important economic interests and strongly hell! values, participants in the decision process, including experts and their employers, have incentives to appeal to emotions, distort facts, and otherwise use communication to influence the ultimate choice in the directions they desire. Thus there are no participants in debates on technological issues on whom nonexperts and public officials can rely unquestioningly for unbiased information. RISE MESSAGES AS PART OF THE RISE COMMUNICATION PROCESS Risk messages, because they flow in only one direction, are only part of the interactive risk communication process. Risk messages in- clude verbal statements, pictures, advertisements, publications, legal briefs, warning signs, or other declaratory activities that describe, characterize, or advocate positions or actions regarding risks, haz- ardous technologies or activities, or risk control options. Each risk message has an identifiable source and is addressed to one or more audiences. Risk messages come from a variety of sources: physicians, jour- naTists, regulatory agencies, manufacturers, environmental groups, health officials, and various self-appointed advisers. The messages are sometimes merely descriptive of risks and scientific studies of them; at other times the messages also describe the broad context within which a specific hazard or risk is found, the developments that preceded its occurrence, comparison of it to other hazards or risks, or the presentation of information about a risk along with information about the attendant benefits and the risks and benefits of alterna- tives. As mentioned above and as discussed-in more detail in Chapter 4, risk messages may be constructed to inform their recipients or to influence them. A large theoretical and empirical literature on communication, social influence, and persuasion provides considerable knowledge for
24 IMPROVING RISK COMMUNICATION anyone who wants to design effective risk messages. However, this knowledge is sufficient only to identify important principles, barri- ers, pitfalls, opportunities, and so forth. It is not adequate to inform many of the specific choices message designers make about charac- terizing particular risks for particular audiences. Lessons from the communication literature have been applied with some success in a range of areas, some of which involve efforts to induce individuals to reduce risks to themselves from cigarette smoking (McAlister, 1981) and heart disease (Maccoby and Solomon, 1981~.3 The following, necessarily brief, summary gives some idea of the concepts and gen- eral conclusions developed in this research tradition. Researchers typically discuss the message content, the source of the message, the channel by which the message is transmitted, and the audience or recipients of the message (HovIand et al., 1953; McGuire, 1985~. Of key importance to the effect of a message are the charac- teristics of the intended audience. The important attributes of the audience include its makeup in terms of cultural background, shared interests, concerns and fears, social attitudes, and its facility with language. A message that has a desired effect on one audience may have little effect on another. Messages in scientific language are likely to mean little to nonscientists, whereas messages about risk in everyday language may be unimpressive to scientists. Risk messages can be carried by a variety of media: face-to-face interaction, direct mailings, advertising, hot lines, presentations to groups, press conferences, television or radio interviews, newspaper or journal articles, and so on. Each medium has its advantages and limitations-for example, television reaches many people but needs visual material and is typically presented in short segments, and newspapers rely on the written word and can present longer, more complex messages but are less vivid and immediate in emotional impact. In general, the characteristics of each channel affect the type of message that can be effectively transmitted. The characteristics of the source of a message often affect the way audiences respond to it. Among the key factors influencing the way recipients judge a message are the degree of expertise the re- cipients believe the source to possess and the degree of trust the recipients have in the source. The term "credibility" is used by researchers in this field to refer to an attribute of a source that derives from a combination of expertise and trust, as seen by the audience. It is possible for a source to be credible to some recipients
INTROD UCTION 25 and not others or on some issues but not others (McGuire, 1985~. Thus a locally respected old farmer may be credible to neighboring farmers as a source of information on pesticide risks but may not be credible to the officials who convene a regulatory hearing. Similarly, the scientific representative of a hazardous waste disposal company may be credible to a federal regulator but not to the neighbors of a proposed waste site. The officials do not credit the farmer because of lack of technical expertise; the neighbors do not credit the company's scientist because of lack of trust. Where there is widespread mistrust of public sources of informa- tion, people often rely on word-of-mouth or other local sources, even if their informants are less expert than those available through public sources. Because of this practice, public agencies can sometimes be more effective in delivering technical information to individual citi- zens by using trusted sources as intermediaries than by designing and disseminating messages themselves (Stern and Aronson, 19843. Pub- lic officials can also listen to trusted intermediaries to learn if tasks might be clelegated or to save the time and expense of questionnaires or other analyses when less detail is sufficient. Risk messages are often designed to inform nonspecialists. Be- cause such messages involve complex and difficult concepts, present- ing clear and understandable information is a tremendous challenge for message designers. The source's choice of message content de- pends on what it believes the audience needs to know, on what recip- ients can be expected to understand, and on the action or response the source hopes to engender. Some risk messages are intended to influence the recipients' be- liefs or actions. Messages are more effective at producing behavior change when, in addition to producing understanding, they are spe- cific about any desired response and proximate in time and place to that response. Generally, single messages can be expected to have little effect on recipients' behavior, but organized programs of messages, in which different messages are aimed at different specific purposes, can be effective. As discussed above, considerable research has been devoted to the study of messages to change individual behavior, and the result- ing knowledge can help in designing more effective risk messages. But less is known about other aspects of risk communication. For instance, there has been little systematic study of ways to design more effective messages to express citizens' concerns to government
26 IMPROVING RISK COMMUNICATION or to influence the actions of organizations such as corporations or government agencies. SUCCESSFUL RISK COMMUNICATION A focus on risk communication implies a standpoint outside the process. It puts no particular actor or message source at the center. In this respect an emphasis on risk communication is different from one on risk messages. The source of a risk message is likely to define and assess the success of its messages according to its own criteria. It may choose to consider its messages successful when the recipients understand them, or when they believe them to be accurate, or when they do what the sender wants to influence them to do. Obviously, different sources may have conflicting goals for their risk messages. This is one difference between the success of-a single source's messages and the success of the risk communication process. It is possible to arrive at a meaningful idea of success for risk communication by considering a broad public purpose that successful risk communication serves. If a society values democratic decision making and well-informed, goal-directed individual choice, it makes sense from the societal standpoint to say that the purpose of risk communication is "to inform the discretion" of government officials, private organizations, and individuals. from that standpoint, risk communication is successful to the extent that it raises the level of understanding of relevant issues or actions and satisfies those in- volved that they are adequately informed within the limits of available knowledge. Informer] discretion for a government official is based on knowI- edge about the risks and benefits of the choices at hand; about the management situation, including legal and other constraints on choice; about the concerns and preferences of citizens and other polit- ical actors; and about the political environment. Corporate officials need much the same kinds of knowledge, although they need to pay special attention to the preferences of shareholders and can some- times afford to pay less attention to the preferences of the public. Government and corporate officials usually inform themselves about risks and benefits with the help of expert employees or consultants who interpret technical knowledge for them (Chapter 2 discusses what is involved in understanding risks). They inform themselves about citizen concerns and political matters by paying attention to
INTROD UCTION 27 elected officials, the mass media, and diverse other sources. Accord- ing to the above definition, the more accurate the official's under- standing of those matters, the better the risk communication system. Citizens are decision makers in their private lives and when they participate in political decisions. Thus a successful risk communica- tion process informs their discretion, too. Citizens inform themselves by interpreting risk messages from various sources, including experts, intermediaries such as journalists, public relations officials of public agencies and corporations, and even friends and neighbors. They evaluate or balance what they know in order to reach a judgment and to make decisions regarding risks, such as whether to protest, ignore, negotiate, or take protective action. The more accurate the citizens' understanding of the issues at hand, the better the risk communication may be said to be. Citizens are well informed with regard to personal choices if they have enough understanding to identify those courses of action in their personal lives that provide the greatest protection for what they value at the least cost in terms of those values. Citizens are well informed with regard to a public policy issue if they have enough understanding to evaluate which options provide the most protection at the least cost, both for themselves and for the things they believe the society should value. (In Chapter 4 we discuss the meaning of successful risk communication in more detail.) It is important to make several points about the definition of suc- cessfuT risk communication. First, success is defined in terms of the information available to the decision makers rather than in terms of the quality of the decisions that ensue. Successful risk communication does not always lead to better decisions because risk communication is only part of risk management. Risk managers, including public officials and private citizens, must also take into account their public responsibilities or personal values. It is possible to understand fully what is known about the likely consequences of each available option and yet to make a "bad" choice; if this occurs, it is not because of a failure of risk communication. Consider, for instance, the head of a federal agency who is constrained by law to ban a food additive even though the risk communication process has made it clear that there are no less harmful or costly alternative additives. Making the legally required decision does not mean the communication process failed; in the long run the process may even provide impetus for changing the law. Similarly, if someone understands but disregards information about the dangers of smoking, skydiving, or riding a
28 IMPROVING RISK COMMUNICATION motorcycle without a helmet, this may not mean risk communica- tion was at fault. People sometimes put themselves at risk with full knowledge, and observers attribute their acts to overriding values, willfulness, or addiction rather than to a failure of communication. Risk communication, even at best, can accomplish only so much. Second, s?`ccessfut risk communication need not result in con- sensus about controversial issues or in uniform personal behavior. Although such objectives often serve the producers of risk messages as criteria of success for those messages, they are not appropriate criteria for the risk communication process in a democracy. To say that success requires that the recipients do or believe what a partic- ular message source desires is to assume that that message source is a better judge of the recipients' interests than the recipients them- sel~res. Because people do not all share common interests or values, better understanding will not lead them all in the same direction. And it will not necessarily make choices easier for decision makers in government agencies or elsewhere. Third, according to the definition of success, the recipient must be able to achieve as complete an understanding of the information as he or she desires. In Chapter 4 we develop the reasons underlying this definition of successful risk communication. A risk communica- tion process that disseminates accurate information is not successful unless the potential recipients achieve a sufficient understanding. Thus the risk communication process must be judged by the level of knowledge on which decision makers act rather than by the level of knowledge reflected in particular messages or even in the full set of messages accessible to decision makers. It is common for accurate messages to be ignored, misunderstood, or rejected; it is also possible for several inaccurate messages from different sources to be compared with each other in such a way as to give the recipient a fairly accurate overall picture. Risk communication, then, is more than one-way transmission of expert knowledge to the uninformed. Certainly, messages describing expert knowledge to nonexperts play a critically important role in risk communication. They provide essential information that nonexperts cannot get from other sources. They are also essential because, by revealing expert dissent, they give nonexperts, including many gov- ernment officials, an important toot for checking against omissions or excesses in any one expert's analysis. Messages about expert knowI- edge are necessary to the risk communication process; they are not sufficient, however, for the process to be successful. Thus, although
INTROD UCTION 29 experts and the organizations that disseminate their knowledge are important participants in risk communication, nonexperts also play an important active part by expressing concerns to experts and by asking or pressuring them to provide analysis of aspects of risk that they may not yet have examined in detail. They play an essential role by participating in the debate about what values ought to be applied to knowledge about risks and how they ought to be applied. Citizens' dialogue with public and industrial risk managers, even when it does not directly address risk, can be critical to risk management deci- sions. The broad definition of risk communication is a reminder that public decisions about risk require debate about values and interests as much as about risks because risks cannot be weighed against each other without considering values. As we will see in the next chapter, they cannot even be understood without considering values. NOTES 1. For discussions of the recent interest in the subject of risk communica- tion, see Plough and Krimsky (1987) and Stallen and Coppock (1987~. 2. For a complete listing of the special terminology used in this report, see Appendix E. 3. Extensive reviews of the communication literature, covering well over 1000 sources, can be found in chapters of the 1985 Handbook of Social Psychology (McGuire, 1985; Moscovici, 1985; Roberts and Maccoby, 1985~. A review of much of this literature with a focus on risk communication has been completed by Covello et al. (1987b).