CONSIDERATIONS OF HEALTH BENEFIT-COST ANALYSIS FOR ACTIVITIES INVOLVING IONIZING RADIATION EXPOSURE AND ALTERNATIVES A Report of Advisory Committee on the Biological Effects of Ionizing Radiations Assembly of Life Sciences NATIONAL ACADEMY OF SCIENCES Washington, D.C. SUMMARY AND RECOMMENDATIONS SUMMARY It must be recognized that in discussing the possible usefulness of benefit/cost analysis for radiation regulation, we are dealing with data ranging from the known to the nebulous, of issues ranging from the practical to the purely ethical, and of problems ranging from local to global and from the recent past to the indefinite future. For some questions, benefit/cost analysis can provide unequivocal answers. For instance, if it were determined, statistically or theoretically, that mass mammography screening programs induce more cancers than they detect, then such programs should be halted. On the other hand, if studies show that they detect twice as many cancers as they induce, but that the dollar cost of the program, applied elsewhere, would result in a greater number of cancer detections, the problem becomes more complicated but still solvable. Again, if a non-radiation technique, such as ultrasonography or thermography, were shown to be as effective as x-rays in detecting cancer and to have no deleterious side effects, this would clearly be the preferred technique. In the middle ground are problems for which benefit/cost analysis may provide information essential to informed decision-making, but can provide no final answers. A hypothetical example might be that of airport x-ray inspection devices. Analysis can yield data on the possible harm, to passengers and attendants, of such devices, on the dollar costs of the program, and on the frequency and costs of airplane hijacking before and after the program was instituted; but the value judgment must be a political-ethical decision. Perhaps the most important problems are those for which there are no data--the effect of today's radiation ten generations hence, the probability of a nuclear meltdown, the future buildup of nuclear wastes--but, at best, only guesses which may be off by several orders of magnitude. Here, benefitcost analysis cannot be done in terms of dollars or other common units but, with a different focus, it can identify issues which must be recognized if rational decisions are to be made. The questions of who pays (rural people, future generations) vs. who benefits (city people, present generation); of maximum and minimum possible harm from a given course of action; of alternative courses of action, can be spelled out. This suggests a three-fold responsibility of federal regulatory agencies: 1) to promulgate regulations based on benefit/cost analysis, when such analysis can provide clear answers; 2) to provide legislators with all relevant benefit/ cost data in areas where such data alone do not resolve the question; and 3) to bring to public attention the social and ethical benefit/cost considerations, for which quantification is impossible. Benefit/cost analysis need not be undertaken if: 1) of two equally beneficial alternatives, one is clearly less costly in both health effects and money; 2) there is, in fact, no measurable alternative; or 3) society has determined either that a given risk, e.g., nuclear meltdown, is unacceptable, however slight the odds or great the benefit, or, conversely, that the benefits of nuclear power outweigh the costs, however great. Clearly, a basic question arises in any attempt to apply benefit/cost analysis to problems where data are largely conjectural or where ethical considerations are paramount: will such an analysis help by contributing some light to hitherto obscure areas, or will it hurt by focusing attention on quantifiable parameters at the expense of less easily definable but more important areas? For instance, is the attempt, which many have made, to assign a monetary value to a human life helpful in at least providing an economic baseline for analysis, or harmful in making an insoluble problem seem manageable? It seems unlikely that a benefit/cost analysis based on translation of death and disability into lost work-years, or aesthetics into real estate prices, which resulted in a ratio of one or greater would be generally convincing given the necessarily arbitrary and narrow nature of the assigned values. However, if by even such analysis the benefit/cost ratio comes to less than one, it would suggest that, in fact, the aggregate of quantifiable and unquantifiable costs must, to a greater degree than indicated by the ratio, outweigh the benefits. Nonetheless, there are cogent reasons for society to attempt health benefit/cost analysis for major technological applications involving radiation exposure of the population in comparison with feasible alternatives which involve less or no radiation exposure. Such analyses could facilitate rational and cost-effective safety and control procedures and the avoidance of health hazards and economic dislocations associated with excessive or inadequate expenditures in relation to risk. Health benefit/cost assessments, even though present data are incomplete, can provide some guidance to decisionmakers, direct attention to gaps in knowledge, indicate priorities for research and stimulate the accumulation of needed data and analysis, and contribute to public understanding of the relevant issues and problems. Although there are great benefits to be derived from the use of various radiation technologies, there are also costs. In one form, these costs arise as a risk of adverse health effects to members of the population. This risk can be stated as a probability which, when applied to the population at risk, gives the number of adversely affected persons. This adverse effect is a social cost and, in principle, can be expressed as the monetary cost of the adverse health effect. Additional private resource costs incurred in the provision of these benefits from the various radiation technologies are, of course, also included in the total social costs. These are the terms defining health benefits and health costs in this report. This report deals with health benefit/cost analysis in terms of the needs, problems and methodological approaches of the times, the concepts and parameters, ethical considerations, and regulations governing the deleterious agents involved. It illustrates application of methodology, but does not attempt a definitive analysis. This report is not concerned with the reassessment of risks from radiation exposure. The goals of this report are to frame the problems, communicate the elements of the complex technical processes and methods of analysis, and provide a basis for more informed governmental decision-making and public participation in the issues. The need to measure benefits, risks, and costs that include subjective values and are subject to pluralistic social values adds greatly to the difficulties of health benefit/cost analysis. Many investigators have made monetary estimates of the value of life, of the biological damage caused by radiation exposure, of the expenditure justified to avoid a given radiation exposure, of the value of a man-rad, or of the value of a unit of risk to a life. There has been no internally consistent, theoretically justified conceptual basis for these studies, and although there has been some consistency among the estimates based on various premises, they must be viewed with some skepticism. Whether the method is acceptable is a question for consideration along with the question of alternative approaches. Chapter III of this report presents an exposition of conventional economic benefit/cost analysis on the assumption that data are available where required. This exposition brings to light many problems in an analysis of activities in which radiation is produced. Radiation is considered in three major contexts: 1) its use in medical and dental diagnosis; 2) its use in medical therapy; and 3) its production in various stages of the nuclear fuel cycle. Each of these situations is different, and examples from all of them are utilized in illustrating basic concepts of benefit/cost analysis. In some situations, one economic strategy may so dominate all others that there may be no need for valuing health or illness in monetary terms. However, more extensive analysis is usually required, leading to the need to find a common unit, e.g., monetization, for very elusive values. Also, inequities may arise, e.g., the cost may be borne by a segment of the population which does not share equally in the benefits. Another serious distributional problem relates to intergeneration effects: oil used today for the benefit of the present population may impose increased cost on unborn generations; dangerous long-lived radionuclides produced today for a present benefit will be a cost to future generations. When standard discounting procedures are used, future generations and their welfare are largely ignored. A special problem is irreversibility. In conventional benefit-cost analysis, decisions are assumed to be reversible. Such is not the case in many aspects of benefit-cost analyses involving radiation and alternative modalities. For example, this occurs both in the case of production of longlived dangerous nuclides by nuclear reactors and in the burning of irreplaceable fossil fuels. Improbable but serious accidental events pose another problem in analysis. Melt-down of a nuclear reactor would be an example of this, and it is shown that a special type of analysis is required for this situation. It is also shown that when the costs of a new program are very high and when its benefits are uncertain, conventional benefit-cost analysis may undervalue the health and social costs. A fundamental assumption of conventional benefit-cost analysis is that numan welfare (which benefit-cost analysis seeks to maximize) may be measured in monetary or materialistic terms. This assumption has fundamental problems. The materialistic system may not be the best one; for instance, it cannot handle the situations in which the risks and benefits do not accrue to the same people. Benefit-cost analyses in the health field tend to be filled with many uncertainties and intangibles and to raise moral issues. Some people feel that such analyses should not be attempted at all since they appear to equate dollars with human lives. Other people feel that benefit-cost analyses, in spite of their obvious shortcomings, can assist governmental agencies and other public institutions in arriving at a rational allocation of resources and personnel. Irreversible processes, quality of life, risk avoidance, distributional effects, incommensurability, and ethical considerations are not adequately addressed in conventional benefit-cost analysis. Therefore, under these conditions, conventional benefit-cost analysis cannot provide an exclusive basis for decision-making. Chapter IV of this report reviews and assesses the efficacy of benefitcost analysis for purposes of decision-making in the regulatory agencies, i.e., in the context of administrative and judicial decision-making on the control of the harmful externalities of such regulated activities as energy production. The regulatory process is a relatively obscure one for most citizens. It is remote in spatial and emotional terms, is complicated and is not readily amenable to public understanding and participation. The questions about the uses of benefit-cost analysis in the regulatory process which have been raised in this report are significant in that they relate to societal capacity to protect human health and welfare now and for the future generations which will bear the risks resulting from contemporary decisions on radioactivity and other harmful substances. In the regulatory process, the only apparent alternative to conventional benefit-cost analysis has been cost-effectiveness analysis. As discussed in this report, cost-effectiveness analysis requires the articulation of objectives, the weighing of the alternative means to achieve the various articulated objectives, and the selection of the least costly approach. For regulation of nuclear energy sources of radioactivity, use of the cost-effectiveness approach would entail the establishment of societal health objectives and risk parameters (e.g., carcinogenic risks) by legislative or other institutional processes which are deemed acceptable as being socially representative. |