Acceptable Risk Under Moral Uncertainty

Abstract

Decisions regarding acceptable risk and risk imposition routinely confront both descriptive and moral uncertainty, yet conventional risk ethics has largely overlooked uncertainty regarding which moral framework is correct. Standard decision-theoretic models for moral uncertainty, particularly Maximizing Expected Choiceworthiness (MEC), prove unviable for evaluating risk. MEC relies on problematic assumptions of intertheoretical comparability and produces decisions that lack weak justifiability—the capacity to explain and defend choices in terms of coherent, underlying moral reasons. Simpler decision rules, including following a single favored theory, pairwise comparisons, or maximin criteria, similarly fail due to insensitivity to credences, vulnerability to preference cycles, or the de facto dismissal of moral uncertainty. These procedural failures stem from treating ethical theories as mutually exclusive systems aggregated only after factoring in descriptive uncertainty. In practice, moral uncertainty introduces independent moral reasons, affects problem framing, and interacts directly with descriptive uncertainty from the outset. Resolving questions of acceptable risk under moral uncertainty therefore requires replacing rigid decision-theoretic aggregation with deliberative, coherentist methods, such as reflective equilibrium and hypothetical retrospection, which treat ethical theories as complementary and preserve the substantive justifiability necessary for imposing risks on others. – AI-generated abstract.