Evaluating the Benefits of Uncertainty Reduction in Environmental Health Risk Management

Abstract

Cost-benefit analysis provides a methodology for choosing the "optimal" control decision in environmental health risk management cases. It does not, however, offer guidance on how to minimize the chance that because of uncertainty regarding risks or costs, the choice made will turn out to be inferior to other feasible choices. Statistical decision theory provides one approach for addressing issues of uncertainty in environmental decision making. This paper illustrates the development and application of a computer program that uses decision theory and numerical methods to evaluate alternative strategies for making research and control decisions under uncertainty. Based on user-supplied descriptions of the cost and efficiency of control options and probabilistic estimates of population exposure and toxicity, the program determines the largest amount of money that should be spent to improve estimates of exposure and of risk and indicates the relative efficiency of concentrating research efforts on resolving each component uncertainty. The dependence of several measures of the value of information on the initial level of uncertainty is examined, as is the sensitivity of these measures to conditions of the decision problem, such as the number of available strategies and the monetary value assigned to a statistical life. Using the computer program and accompanying guidelines presented in this paper, decision-makers can improve the quality of information-gathering efforts and can move more confidently from the research phase to the implementation of control technologies.