Regulation Under Uncertainty: Use of the Linear No-Threshold Model in Chemical and Radiation Exposure

Abstract

This paper examines the use of the linear no-threshold (LNT) model in chemical and radiation exposure. The LNT model assumes that exposure to any level of a chemical or radiation is harmful, down to even the last molecule. Used primarily to be “public health protective,” the model has been the backbone of chemical and radiation risk regulation for many decades. Given the current state of science and risk management tools, we challenge the notion that using the LNT as the default model is public health protective. First, more and more research has uncovered dose-response relationships that reveal either a threshold or, more importantly, a hormetic response, where exposure to low doses of a hazard actually yields health benefits. Second, given these more realistic alternative dose-response models, risk management tools including risk-risk analysis and health-health analysis show that regulating down to extremely low levels can have negative health consequences when ancillary risks are considered. Risk-risk analysis focuses on how reductions in target risks can lead to increases in risk from substitute chemicals or activities. Health-health analysis explores how costs of compliance are borne in part by consumers who are forced to reduce their own private risk-mitigating activities. Overestimating risk, a common feature of the LNT model, upsets the careful balancing of risks required of risk managers.

Keywords: linear no-threshold, dose-response, health and safety regulations, benefit-cost analysis, risk

JEL Classification: D62, D81, I18