Addressing the Diversity of Risks and Accounting for Systemic Risks: Two Proposals for Improving Clarity in Philosophical Discussions of Risk

The lack of engagement of philosophy with decisions made under conditions of risk and uncertainty has lately received increasing attention. But philosophers have devoted little thought to the development of a conceptual framework for distinguishing different types of risks. This article begins by illus- trating the need for a more nuanced conceptual framework. As the normative considerations risks give rise to are highly varied, ethicists need to distinguish between different types of risks. It then offers two ideas. First, it proposes that we understand ‘risk’ as a multi-dimensional concept, i.e. a concept that can always be evaluated in different dimensions, e.g. a qualitative, a quantitative and an epistemic dimension. While this understanding of risk underlies the current practice of risk analysis, making it explicit is of value, in particular for the phil- osophical debate. It draws attention to the diversity of risks, helps to group them sensibly and sharpens existing arguments. Second, the article introduces a novel account of the notion ‘systemic risk’. Criticising existing accounts, it proposes we understand ‘systemic risk’ as referring to a risk that endangers the function- ing of a system and originates in the system or its background noise, where a system is conceived of as a set of interdependent, interacting components that form a complex, functionally defined whole. The article identifies a number of structural features shared by systemic risks and evaluates the distinctive norma- tive considerations they give rise to. It then argues that the notion ‘systemic risk’ is particularly suited to characterise some of the most pressing risks faced in an increasingly interconnected world. While the notion ‘systemic risk’ rests on the understanding of ‘risk’ as a multi-dimensional concept, it also complements it. Its necessity shows that in addressing risks, it is sometimes helpful to move beyond a characterisation based solely on their dimensions.