The recursive approach to time inconsistency

We introduce a new class of finite horizon stochastic decision problems in which preferences change over time, and provide a proof of the existence of a recursively optimal strategy. Recursive optimization techniques dominate many areas of economic dynamics. However, in decision problems in which tastes change over time, the solution technique most commonly applied is not recursive, but rather strategic (subgame perfection). In this paper we argue in favor of the recursive approach, and we take the necessary theoretical steps to make the recursive methodology applicable.