Paper BELIEFS ABOUT RADIATION: SCIENTISTS, THE PUBLIC AND PUBLIC POLICY Hank C. Jenkins-Smith,* Carol L. Silva, † and Christopher Murray ‡ Abstract—Human behavioral responses to potential hazards are mediated by the beliefs held about those hazards. This holds whether the “behavior” under consideration is the provision of advice about the hazard, statements of support for policies that address the hazard, or personal behaviors in response to the hazard. This paper focuses on beliefs about radiation and the implications of those beliefs for views about radiation protection by both scientists and members of the U.S. public. We use data from a large sample of scientists, collected in 2002, and a series of surveys of the U.S. public collected in 2007. Among scientists, we focus on how beliefs about radia- tion are related to policy prescriptions for radiation protection. Among members of the lay public the focus shifts to the relationship between beliefs about radiation risks and policy preferences for nuclear energy and nuclear waste policy options. The importance of the differences and similarities in the patterns of beliefs of scientists and the lay public are discussed. Health Phys. 97(5):519 –527; 2009 Key words: radiation; radiation risk; radiation dose; National Council on Radiation Protection and Measurements INTRODUCTION THE SCIENTIFIC debate concerning the effects of low doses of radiation on human populations has direct implica- tions for quantitative assessments of the risks posed by nuclear activities, and for public policies involving those activities. Dose-response (D-R) assumptions drive much of the estimated health effects of radiation on populations from such facilities as nuclear repositories and reactors, and from the procedures of nuclear medicine. Variations in the assumed shape of the D-R curve—whether there exists a threshold of exposure below which there are no measurable health effects—fuel public debate and con- troversy over the number of deaths that might occur due to normal operations and accidents. § For these reasons the choice of an assumed D-R function has become integral to policy debates concerning nuclear technolo- gies and waste management. We focus our analysis on three questions. First, given the central technical focus on D-R relationships, how do these assumptions play a part in scientists’ judgments about the risks posed by nuclear energy and waste management? The choice of D-R assumptions plays a part in the broader policy debate, as the resulting estimates of morbidity and mortality from small expo- sures become the basis for argument and formation of more general risk perceptions among the public.** While one can conjecture that the assumption of a linear no-threshold (LNT) relationship will lead to greater perceived risks than would a sub-linear, threshold rela- tionship (SLT), we have seen no tests for this effect, or its magnitude, within scientific and technical communities. We provide such a test, using data from a mail survey collected among samples of scientists in the United States and European Union (EU) in 2002. The second question we address is how, more broadly, scientists and the public reach judgments about the risks posed by radiation in two controversial domains of public policy: nuclear energy and the disposition of spent nuclear fuel. Of course, among scientists the D-R function is only one (albeit important) element likely to play a part in assessments of risk. But scientists, like people more generally, are likely to bring to bear a fairly complex set of considerations and heuristics in arriving at judgments about nuclear risks (Rothman and Lichter 1987; Barke and Jenkins-Smith 1993; Plutzer et al. 1998; * Center for Applied Social Research, University of Oklahoma, 660 Psarrington Oval, Norman, OK 73019-0390; † Department of Political Science, University of Oklahoma, 455 West Lindsey Street, Room 205, Norman, OK 73019-2001; ‡ Center for Applied Social Research, University of Oklahoma, 2 Partners Place, 3100 Monitor, Suite 100, Norman, OK 73072. For correspondence contact: H. C. Jenkins-Smith, Center for Applied Social Research, University of Oklahoma, 660 Psarrington Oval, Norman, OK 73019-0390, or email at jenkinssmith@gmail.com. (Manuscript accepted 8 May 2009) 0017-9078/09/0 Copyright © 2009 Health Physics Society § Miller (2006) describes the implications of different radiation D-R assumptions for the number of deaths estimated to occur over time due to the Chernobyl reactor accident. ** See, for example, the use of estimated long-term deaths from a terrorist attack on the Indian Point nuclear power plant (up to 518,000 deaths from cancer among people living within 50 miles) in Lyman (2004). This report has become the basis for an information campaign describing the plant as one of the “top threats” to public health and safety on the region. 519