Risk Evaluation and Management

In industrialized societies, the question “How safe is safe enough?” has emerged as a major policy issue of the 1980s. The frequent discovery of new hazards and the widespread publicity they receive is causing more and more individuals to see themselves as the victims, rather than as the beneficiaries, of technology. These fears and the opposition to technology that they produce have perplexed industrialists and regulators and led many observers to argue that the public’s apparent pursuit of a “zero-risk society” threatens the nation’s political and economic stability (Harris, 1980; Wildavsky, 1979).

Technology that is developed for benefits and profits will always produce risks and costs. Although technology assessments evaluate and compare these consequences, balancing the trade-offs and distributing the contrary effects require political decisions, which traditionally have been carried out by “stakeholders” representing special interests. With the emergence of new “public” interest organizations, the risk management system has been confronted with additional voices seeking ways to have their points of view included in risk management decision making. Hence, intervention in the form of organized opposition to, or support of, specific technologies, installations, or products—sometimes expressed in the name of the public—has become increasingly pivotal in the outcome of policy debates (Nelkin, 1979).

On their way to social acceptance, many technologies experience periods of public concern, controversy, and social opposition. Sometimes these responses to technologies are short episodes, motivated by naive fears or esoteric values of a small segment of the public. As in the case of religious opposition to the implementation of lightning rods (see Mazur, 1981), these episodes may be little more than amusing footnotes in the history of an ultimately successful technology. In other cases, the public responses may swell to a strong social movement against the technology, leading to its disruption, halting, or eventual abandonment. Water fluoridation, nuclear power, and the SST are recent examples of technologies suffering this fate (see Lawless, 1977; Mazur, 1981; Nelkin, 1978).

Perhaps one of the most central public policy issues in the past two decades has been focused on explorations of the ways in which appropriate relations may be fostered between modern industrial societies and their environments. Two general strategies may be followed in this exploration, one tightly focused, the other wider-ranging. The narrow-gauge strategy typically examines in depth the details of a specific environmental variable. In contrast, the broad-gauge strategy seeks to identify the major systems and to examine the interplay and linkages among them. Our investigation of the perception and management of technological hazards has been guided by the broader strategy, involving a conceptual framework that examines the interplay among personal, societal, and environmental systems (Figure l)(Craik, 1972).

Animals don’t take risks. Only human beings do. Risk is an ineluctably social construct. What counts as a risk, what weighs in the balance as a comparatively greater risk than some other, or as a risk worth taking in view of the expected benefit, or indeed, as a risk one is obliged to take despite the expected harm—all of this depends finally on some system of values within which such assessments can come to be made. And value systems are socially constituted, socially learned, and socially enforced. Though animals certainly respond to dangers and may behave in ways which put them in danger, they are incapable of taking risks. For whatever may be the stereotypical instinctive mechanisms or biological norms which have been selected out by species evolution as enhancing species survival, or whatever animals are capable of learning in order to cope with novel or species atypical threats, none of this yet counts as a system of values in terms of which something can come to be assessed as a risk, and judged to be acceptable or not. Being in danger or threatened by harm is not the same as being at risk—a distinction I hope to make clearer as I proceed. To put it differently, being at risk entails being in danger, but danger does not entail risk; the terms are not coextensive. Instinctive or learned responses to danger or threat among animals are, in this sense, neither risk taking nor risk avoiding. The category doesn’t apply to animal life. Human beings, by contrast, are a risk-taking and risk-avoiding species. For humans, the category of risk is indissolubly bound up with that of culture and with history, with the distinctive modes of cognitive praxis which are characteristic of human life. That is to say, risk is defined in relation to perceived values; and these in turn are possible only for a self-conscious creature who is capable of estimating present actions either in the light of norms or general principles, or of future outcomes.

This chapter is about comparative risk assessment, a relatively new term in an evolving area of research. Although comparative risk assessment (CRA) does not have a widely accepted precise definition or a textbook methodology, it is used in various forms in both personal and public decision making. Such decision making involves the identifying, weighing, and balancing of often uncertain advantages and disadvantages of alternate choices of action.

In recent years, the public has become increasingly aware of and concerned about the risks associated with technology and has placed much of the responsibility for monitoring, controlling, and reducing these risks on government agencies. As a result, legislators and regulators are faced with complex, controversial decisions requiring the assessment and balancing of imperfectly known costs, risks, and benefits.

Approximately 50,000 people die every year in automobile accidents, but no life has yet been lost in the United States due to a nuclear accident. Despite these facts, many people show little concern for automobile safety issues, but oppose nuclear power because they believe it is unsafe. It is clear that an individual’s attitudes toward the risks associated with different technologies will vary significantly from case to case. An understanding of the factors that influence these attitudes could be helpful to public policy makers, especially those who are managing the introduction of new technologies.

Many important decision problems involve risks to the health and safety of members of the public. These risks include possible fatalities, sicknesses, and injuries, as well as psychological effects. For appraising alternatives in such contexts, decision makers need to address public risks in addition to other economic, social, and environmental consequences.

Each year an estimated 17–31% of the U.S. mortality rate is associated with undesired side effects of technology (Harriss, Hohenemser, and Kates 1978). The productivity loss from technology-related illness, death, and pollution is equivalent to 3–6% of the gross national product (GNP). When combined with the cost of private and public sector efforts to prevent and mitigate such losses, the undesired side effects of technology amount to 7–12% of the GNP (Tuller 1985). Even so, these estimates of the societal burden of technological hazards are incomplete, and do not include, for example, a number of newly recognized hazards.

Ten years ago, it was commonplace for economists to calculate the value of human life as the lost economic productivity associated with a shortened life-span.¹ As is well known, such a view has been shown to be grossly inadequate. Most obviously, it leads to counter-intuitive results, such as that the value of the life of a 65-year-old laborer is equal to the sum of his remaining earnings until retirement or that the value of the life of a small child is near zero, since her future earnings are discounted at a market rate of interest.

In the past few years, a number of federal regulatory programs have come to make use of quantitative risk assessments to provide information for policy decisions. But, current methods of risk assessment are flawed in a number of respects. Risk assessments are costly in time and effort; they often result in vague and unusable estimates; and it is difficult to separate the good ones from the bad ones because the probabilistic predictions are not observed for many years. Because of these facts, it is difficult for the assessors themselves to identify and learn from their mistakes.

The risk evaluation phase of risk analysis is concerned with the appraisal of the sociopolitical significance of an estimated risk. This significance may reflect its acceptability (or, in some instances, “tolerability,” when a risk exists that is not actually acceded to, but for whose mitigation no practicable approach is evident) or nonacceptability in relation to various possible criteria and/or as perceived by various groups. It may reflect the price paid in lost benefits if the risk is not accepted and the responsible hazardous activity is eliminated. Finally, it may reflect the costs required to mitigate the risk. This chapter reviews the literature on criteria or standards against which a hazardous activity’s risk can be decided to be acceptable or to require mitigation, and/or against which the residual risk of the mitigated hazard can be compared for acceptability. It is recognized, however, that the establishment of such criteria or standards suffers from severe difficulties, and that these difficulties underlie, at present, many decisions on highly beneficial but also highly hazardous activities in our society. Nuclear power is the paramount example; many others that have exercised Congress, regulators, activist groups, and the general public could also be noted.

The last decade has seen a very rapid expansion of activity by the federal government with regard to societal hazards and risks. Although federal regulatory decisions have tended to avoid quantifying the question “How safe is safe enough?” there is growing attention, on the federal level, to efforts to quantify risks and to examine various policies concerning “acceptable” or “assumable” risk. However, matters pertaining to many hazards and risks fall under the direct control of state, county, or municipal government. While there exists much current effort on risk quantification and risk management at the federal level, little information is available on the extent and magnitude of risks subject to local control. In addition, little has been done to examine the ramifications of various possible risk management policies at the state or municipal level.

The explosive growth of federal responsibility for risk management in health and safety areas is generally well known. Table 1 illustrates the broad statutory responsibilities of three major federal agencies,¹ the Occupational Safety and Health Administration (OSHA), the Consumer Product Safety Commission (CPSC), and the Environment Protection Agency (EPA), none of which existed before 1970. An estimate² of the work load imposed by OSHA, CPSC, and EPA on the federal appellate judiciary³ is given in Table 2. In the case of the Courts of Appeal, no less than 1.3% of all cases with a published opinion involved OSHA, CPSC, or EPA. In view of the enormous controversy surrounding the passage of the laws administered by these agencies, it is perhaps not surprising that these three agencies would account for such a comparatively large share of the appellate work load so soon after their creation. Moreover, the percentages by themselves underestimate the amount of work involved because, at least in the cases involving challenges to standards, the courts have had to review massive⁴ records containing highly technical information from the fields of medicine, toxicology, engineering, and economics. Quite clearly, risk management problems have had a major impact on both the executive and judicial branches.

Public concern for a wide array of risks to health, safety, and environmental quality has increased significantly in the last 15 years as have governmental efforts to deal with those risks. More recently, scientific analysis of such technological risks, from nuclear power plant operation to toxic and hazardous materials in the environment, has become a growth industry, the result of which is a substantial literature on risk analysis. Most of this literature, however, focuses on technical and methodological issues of risk assessment; there has been a comparative neglect of processes of risk evaluation and more generally political variables affecting implementation of risk reduction policies. If methods of risk assessment are to be employed more effectively in governmental settings and if they are to help bring greater rationality to policy decisions, improved understanding of political as well as technical aspects of risk analysis is essential. In particular, greater knowledge of how political and institutional forces affect the conduct, use, and impact of risk analysis is necessary to speak confidently of effective strategies of change.

Many types of hazards confront individuals, families, communities, businesses, and nations. Individuals face immediate danger from accidents and acute disease. They face longer term danger in the form of chronic disease. Financial hazards loom in the form of unemployment, assets becoming worthless, large liabilities emerging, expenses for housing or children’s education, and the termination of earned income at retirement.

In a democracy, government is made legitimate by the consent of the governed. But figuring out what counts as consent where opinions are divided and why consent should be so valued poses deep problems, whether we are considering authority in general or are focusing our attention on a particular application of authority through government. Decisions about risk and safety that must be centralized, usually in regulatory agencies, provide some dramatic examples of this problem. The agencies must determine acceptable levels, distributions, and kinds of risk. Most of these agencies were established by legislation, but they were not established to do whatever they please. What should they do? How can they appeal to the consent of the governed to justify their decisions?

Decisions with consequences for human health, safety, and the environment are made by consumers in the marketplace, by workers in the factory, by business executives in production decisions, by administrators of regulatory agencies, and by ordinary citizens in their daily lives. In most cases these decisions are based primarily on common sense, ordinary knowledge, and nonscientific analysis. It is frequently impossible to identify any single decision maker as decisions evolve out of dynamic processes of social interaction. Interaction refers to, say, the simple communication between a consumer and seller as well as the complex bargaining between adversaries in an environmental dispute. All people participate in such interactive decision making about risks in their daily lives even though they do not think about it as “social interaction.”

In the Tigris-Euphrates valley about 3200 B.C. there lived a group called the Ashipu. One of their primary functions was to serve as consultants for risky, uncertain, or difficult decisions. If a decision needed to be made concerning a forthcoming risky venture, a proposed marriage arrangement, or a suitable building site, one could consult with a member of the Ashipu. The Ashipu would (1) identify the important dimensions of the problem, (2) identify alternative actions, and (3) collect data on the likely outcomes (e.g., profit or loss, success or failure) of each alternative. The best available data from their perspective were signs from the gods, which the priestlike Ashipu were especially qualified to divine. The Ashipu would then create a ledger. For each alternative, if the signs were favorable, they would enter a plus; if not, they would enter a minus. After the analysis was completed, the Ashipu would recommend the most favorable alternative. The last step was to issue a final report, etched upon a clay tablet (Oppenheim, 1977).