Judicious management of uncertain risks: II. Simple rules and more intricate models for precautionary decision-making

Rational decision theory could be more fully exploited for the prudent management of uncertain-risk situations. After an integrative circumscription of the precautionary principle (PP), 10 key issues are discussed covering assessment, decision and control. In view of this, a variety of decision-theoretic considerations are explored. 'Sufficient evidence' for precautionary action is treated as a diagnostic decision in the framework of signal detection theory. Thus, an assessment about 'danger' versus 'no danger' should depend on prior probability, evidence strength and the relative seriousness of false-positive versus false-negative outcomes. From an illustrated survey of simple and more complex decision rules, it appears that 'precaution' may be variously expressed via, for example, maximin utility, minimax regret, maximin expected utility and maximising expected utility-minus-regret. Logically, serious uncertain risk (against modest benefits) may provoke a precautionary approach -- under the PP. In contrast, however, an uncertain 'great opportunity' (against modest costs) may elicit a venturous approach -- following the venture principle (VP). Thus, the PP amounts to a basic attitude rather than a normative principle, whose practical application hinges on straightforward albeit uncertain decision variables. Decision postponement, temporal discounting and risk--risk tradeoffs are summarily reviewed. General conclusions are drawn and some suggestions for policy-making and further research are proposed.