Exposure Gone “Wild”: A Call for Rational Exposure Scenarios

In my experience, most makers of “risk-based” decisions (regulators, responsible parties, risk managers) really do not know much about the nuts and bolts of what passes for a risk assessment. They know there is some threshold (i.e., criteria, guideline, standard, benchmark, etc.) they should not allow to be crossed. They know a story (i.e., an exposure scenario) for how it might be crossed. And they know that being below it is good, while being beyond it is bad (if not for the public’s health then possibly for their career). With this level of understanding, their decision is simply to do what’s necessary to stay below that threshold. Doing so might be costly and burdensome, and it might mean giving up certain benefits (Ginsberg and Toal 2009). But decision-makers are willing to bear those costs, and lose those benefits, because they (or their constituencies) really believe that if they do not, a bad outcome will happen. Decision-makers prefer an alternative that maximizes the chance for good outcomes and minimizes that for bad outcomes. But what if the chance of a bad outcome was miniscule at the threshold and did not get much larger until well beyond it? What if decision-makers were making costly, benefit-denying decisions in the mistaken belief they were averting the seeming certainty of a bad outcome (Stone and Hope 2010)? Decision-makers with no or biased knowledge of a bad outcome’s chances may choose an alternative that sacrifices benefits, and incurs large political and monetary costs, to avoid an outcome that has no real chance of ever materializing. Would unbiased knowledge of the chance for a bad outcome influence their decision, particularly if it seemed really small and the benefits they were going to lose were really important to their constituencies? If so, why do they not have such knowledge? There are undoubtedly several reasons for this, but two stand out as particularly influential: (1) the threshold method itself and (2) the uncritical use of “upper bound” exposure scenarios. In the United States, much of what passes for “risk” assessment in the regulatory community is hazard assessment, where you are either over (bad outcome) or under (good outcome) some threshold and where chance (probability) is not part of the