Constructive Decision Theory

Contemporary approaches to decision making describe a decision problem by sets of states and outcomes, and a rich set of acts: functions from states to outcomes over which the decision maker (DM) has preferences. Real problems do not come so equipped. It is often unclear what the state and outcome spaces would be. We present an alternative foundation for decision making, in which the primitive objects of choice are syntactic programs. We show that if the DM's preference relation on objects of choice satisfies appropriate axioms, then we can find states, outcomes, and an embedding of the programs into Savage acts such that preferences can be represented by EU in the Savage framework. A modeler can test for SEU behavior without having access to the subjective states and outcomes. We illustrate the power of our approach by showing that it can represent DMs who are subject to framing effects.