Agent-Relative Prerogatives and Suboptimal Beneficence

The first part of Chapter 11 uses considerations of sequential choice to argue that suboptimal beneficence is impermissible. The second part shows how the prohibition on suboptimal beneficence follows from an agent-relative theory that understands permissible actions in terms of a dominance principle defined over both the agent-relative and the agent-neutral ordering. This theory incorporates agent-relative prerogatives that ensure that agents are not required to do what is impartially best, yet rules out suboptimal beneficence. The third part shows that the prohibition on suboptimal beneficence is in tension with dynamic consistency, since it leads to violations of expansion consistency condition BETA. If an agent makes use of myopic choice principles (which are purely forward-looking) or sophisticated choice principles (that make use of backwards induction), then there can be cases in which he can, by means of a sequence of permissible choices, bring about an outcome that is deemed to be impermissible from the outset. This problem is addressed by developing global choice principles that ensure dynamic consistency.