Attribution of Collective Causal Responsibility to Individual Actors in Stochastic Systems
26 Pages Posted: 23 Nov 2021 Last revised: 6 Mar 2023
Date Written: February 16, 2023
Abstract
We develop an attribution of collective causal responsibility to individual actors in stochastic nonlinear systems: we take an existing measure of collective causal responsibility from Vallentyne (2008) and Baumgärtner (2020), introduce four elementary axioms on the attribution of collective responsibility to individuals, and propose a measure of individual causal responsibility that is uniquely implied by these axioms. Our approach is inspired by Shapley's (1953) fundamental concept of how to divide a collective effect into individual marginal contributions. The generic setting is a set of systems with potential regime shifts where actions affect the regime-shift probabilities -- e.g. technical systems, traffic systems, financial markets, managed ecosystems, or the Earth's climate system. This paper closes several gaps in the literature on responsibility attribution: our concept is intensive, has cardinal properties and allows for multiple actors with simultaneous actions. This is relevant for implementing efficient incentive schemes and liability for managers of stochastic systems.
Keywords: attribution, causal responsibility, probability, multiple actors, regime shift, systemic risk, Shapley-value
JEL Classification: K13, K32, Q50
Suggested Citation: Suggested Citation