Revealed preferences in intertemporal decision making

This paper connects two tracks of previous research: revealed preference theory and intertemporal decision making. It demonstrates that, in contrast to static decision frameworks, in dynamic settings the Weak Axiom of Revealed Preference is not sufficient for a perfect correspondence between an agent's observed choices and the unobservable preferences revealed. Necessary and sufficient conditions for a perfect correspondence are derived. In addition, conditions for an imperfect correspondence are analyzed and implications for applied work are suggested. The paper's findings represent a necessary first step towards a deeper appreciation for the relationship between observed dynamic choices and revealed dynamic preferences.