Continued Development of Methods for Characterizing and Ranking Health, Safety, and Environmental RisksEPA Grant Number: R827920
Title: Continued Development of Methods for Characterizing and Ranking Health, Safety, and Environmental Risks
Investigators: Morgan, M. Granger , Dekay, Michael L. , Fischbeck, Paul S.
Institution: Carnegie Mellon University
EPA Project Officer: Hunt, Sherri
Project Period: December 15, 1999 through December 14, 2001
Project Amount: $235,504
RFA: Decision-Making and Valuation for Environmental Policy (1999) RFA Text | Recipients Lists
Research Category: Economics and Decision Sciences
In the majority of risk-ranking exercises undertaken to date, the specific methods that have been employed to perform the rankings have received little attention. In the case of many of the EPA-sponsored regional ranking exercises, this has not been a serious problem because a primary objective has been to promote dialogue among stakeholders. However, if the results of risk ranking are to be used by regulatory agencies in support of risk-management decision making, they must be based on normatively justifiable and empirically validated procedures. Under support from a previous NSF grant (SBR-95120232), researchers in the Department of Engineering and Public Policy at Carnegie Mellon have begun the development, empirical testing, and refinement of such a method for ranking health and safety risks. The research proposed here is designed to complete that work, and to extend it to include environmental risks.
Risk ranking requires both risk analysis and value-based judgments. In the Carnegie Mellon method, experts provide detailed descriptions of the risk in multi-attribute terms in the form of "risk summary sheets." Then, representative jury-like groups of laypersons perform the ranking using two different methods. The first is a "holistic" ranking procedure in which the participants carefully consider all available information on the risks and then, using procedural guidance and supporting materials developed by the investigators, make overall judgments of their relative concern. The second is a "multi-attribute" (MA) ranking procedure in which participants indicate the relative importance they attach to each of the risk attributes described in the risk summary sheets. Implicit ranks can then be inferred. These two procedures provide alternative "framings," and are used to assist participants in producing their best-considered judgment. Both individual and group rankings are elicited in order to obtain the benefits of group interaction while allowing individuals to express their personal views, and in order to assess the impact of group interaction. Although development of the method has built upon theory and empirical results from behavioral social science, a substantial amount of experimentation in a realistic setting has also been necessary. An experimental test bed has been developed which involves 22 health and safety risks in a hypothetical middle school. In the work now proposed, the investigators plan further experimental studies to complete the development of the method. Studies are planned to refine and evaluate the ranking procedures using the middle-school test bed. These studies will involve investigating the importance of several different components of the risk summary sheets; investigating the importance of focusing on the risk attributes in the individual and group ranking tasks; and investigating the credibility and usefulness of the method and the resulting rankings. To date, the method has involved only risks to health and safety, but risks to the environment are also very important in many real-world contexts. Under this proposal, the researchers plan to conduct new theoretical and empirical work to identify how to characterize ecological and other environmental risks. In addition, they plan to develop materials for health, safety, and environmental risks in a new domain, so that those risks may be compared on a common set of attributes and ranked together in a single exercise. Once development of the method is complete, the investigators will complete a set of preparatory studies that will allow a careful systematic evaluation and selection of one or more real-world application domains.
In the middle-school test bed, the researchers expect to find conclusive evidence for the usefulness of the tabular and text components of the risk summary sheets and for the usefulness of the MA procedure in the risk-ranking workshops. The credibility and usefulness of the resulting rankings is expected to depend on the quality of the risk materials, the consistency among rankings produced by different procedures, the satisfaction of the participants with the ranking process and output, and the level of agreement among different groups who ranked the same set of risks. Finally, the researchers expect to demonstrate the applicability of this method to ranking health, safety, and environmental risks in a novel test bed in a new domain.