Referential revealed preference theory

Reference-dependent choice behavior implies behavioral anomalies such as the so-called attraction effect, status quo bias, and endowment effect. This paper builds a new theory of revealed preference capturing preferences that depend on a reference point. The first main contribution of this work is a decomposition of the Weak Axiom of Revealed Preference (WARP) into three independent axioms. Referential revealed preference theory is then, naturally, constructed by only removing the WARP-rationales that are inconsistent with the data. This minimal deviation, in addition to explaining the mentioned behavioral anomalies, preserves the predictive power of the classical theory to the extent possible. Therefore, all sensible results are endogenously derived from the model. These results include: formation of references and reference preferences, the connection of the latter concept to the classical revealed preference, the nature of referential effects, and the characterization of choice. Interestingly, the notion of sequential rationalizability arises, endogenously, as a result in the referential revealed preference theory.