Expected prices as reference points : theory and experiments

I show theoretically that applying the model of Köszegi and Rabin (2006) to a simple purchasing decision where consumers are ex-ante uncertain about the price realisation, gives - when changing the underlying distribution of expected prices - rise to counterintuitive predictions in contrast with a "good deal model" where consumers are predicted to be disappointed (rejoice) when the realised price is perceived as being worse (better) than the other possible realisation. While the underlying ideas of both models are similar with respect to expectation-based reference points, the different results come from the concept of Personal Equilibrium in Köszegi and Rabin (2006). The experimental results show some support for the simpler good deal model for a number of different real consumption goods though the support is weaker for goods that either have a salient market price or no market price outside of the experiment.