An argument against resolute choice

It is central to the argument of Chapter 11 that the sophisticated chooser will in certain circumstances select a plan that is dominated with respect to sure outcomes by another plan. Specifically, situations will arise in which an entrepreneur will have a motive for offering his services for a (modest) fee, and the sophisticated agent will be disposed to accept such an offer, even though he could do better, regardless of the turn of events, by simply serving (without a fee) as his own agent. Those who are inclined to regard SEP as a plausible rationality condition on dynamic choice and, hence, who think in terms of a sophisticated rather than a resolute approach will take issue with this claim. They will do so on the grounds that the plan called for on the resolute approach – indeed, the resolute approach itself – is simply not feasible. On this view, the agent at a tree-continuation point will always choose in a manner that is consistent with de novo choice. Agents can make resolutions, to be sure, but they cannot carry them out. Those who set out to be resolute will end up behaving just like myopic choosers. But if resolutions cannot be effectively implemented, then the charge that the sophisticated approach is pragmatically inferior collapses. Only feasible plans can be of any relevance for determining the pragmatic respectability of sophisticated choice.