Can Risk Assessments Conclude That There Is No Risk? D. M. Kammen, Science, Technology & Environmental Policy (STEP) Program, Woodrow Wilson School of Public and International Affairs, Princeton University, Princeton, NJ 08544-1013; and R. L. Goble and D. B. Hattis, Center for Technology, Environment, and Development, George Perkins Marsh Institute, Clark University, 950 Main Street, Worcester, MA 01610
Consumers of a risk assessment may well care about the distinction between stating that there is "good evidence that a risk does not exist" and that there is "uncertain" or "insufficient evidence that it does exist." Contemporary risk assessments, however, commonly conclude that the risks in a situation are lower than some threshold of legal or other significance; they hardly ever conclude that there is zero risk nor do they often provide such reassurance to worried people that they decide a risk is truly negligible by their own standards. Risk assessment process has not established clear guidelines for integrating negative findings and thus for evaluating them in a consistent framework. Since risk assessments are generally only performed when some group perceives a risk, assertions that there is no risk will stir debate in a setting where negative results are not well understood. Furthermore common scientific practice is more likely to identify possibilities for risk, or to characterize possibilities as insufficiently supported by evidence than to provide evidence against the existence of some risk. Using historical examples such as the age of the earth, the fear of thermonuclear detonation of the atmosphere, and cold fusion, we consider two approaches to providing evidence that there is no cancer risk from exposures to EMFs. One approach is "impossibility theorems" which represent claims that mechanisms for cancer risk would be contrary to our understanding of physics. A second approach is to assess "research exhaustion" - whether insufficient progress in specifying effects or mechanisms given the efforts expended can be construed as evidence against an effect. We discuss the feasibility of integrating such sources of negative evidence within a risk assessment.
Work supported in part by the Public Health Institute of California, Contract #23095.
Go to . . .