Identifying changing taste from demand data via golden eggs.

The paper provides choice theoretic foundations for impulsive behavior and changing taste. Similar to other studies in the literature, revelation is possible owing to self-limitation. However, in contrast to the usual unrealistic assumption that self-limitation opportunities are exogenous and observable, we follow Laibson's classical paper by allowing for self-limitation through investing only in illiquid assets. This approach allows us to identify impulsive behavior and changing taste using only standard demand data. In addition, we provide a refinement for an individual's welfare that is based on libertarian considerations. The refinement uniquely identifies the revealed impulsiveness in our model, and ultimately leads to a notion of comparative impulsiveness. Finally, we relate the comparative impulsiveness in our model to existing measures of (ir)rationality. [ABSTRACT FROM AUTHOR]