Rational inattention: Prices and information in economics

This thesis studies rigidity of prices as an implication of agents' limited abilities to process information. It applies a framework called "rational inattention". Rationally inattentive agents have a limited information capacity. They can not process all available information, but they are free to choose what pieces of information to process, i.e. what to pay attention to.;The text is divided into four parts. In the first one, I introduce and discuss basic implications of rational inattention. The next three chapters are three independent papers: (1) Rationally Inattentive Seller: Sales and Discrete Pricing: This paper presents a model of a rationally inattentive seller responding to shocks to unit input cost. The model generates price series simultaneously exhibiting all three of the following features that can be found in the data. (i) Prices change frequently. (ii) Responses of prices to aggregate variables are delayed. (iii) Prices move back and forth between a few rigid values. Discrete pricing arises even if the unit input cost varies in a continuous range. Results of the model also agree with the evidence that reductions in price, e.g. sales, are usually short-lasting and that the highest price in a sample tends to be the most quoted price. Discrete and asymmetric pricing is a seller's optimal response to his limited information capacity Moreover, the model provides rationale for faster responses to aggregate shocks in industries with more volatile idiosyncratic shocks as well as for a steeper Philip's curve in less stable aggregate conditions. (2) Rigid Pricing and Rationally Inattentive Consumer: this paper proposes a mechanism leading to rigid pricing as an optimal strategy. It applies a framework of rational inattention to study the pricing strategies of a monopolistic seller facing a consumer with limited information capacity. Unlike in the previous paper, it is the consumer, who needs to process information about prices, while the seller is perfectly attentive. It turns out that the seller chooses to price discretely even for a continuous range of unit input costs, i.e. charges a finite set of different prices only. The price usually stays constant when unit input cost changes only a little. The seller does so to provide the consumer with easily observable prices and thus stimulate her to consume more. In the model's dynamic version, this mechanism implies that prices respond to cost shocks with a delay. (3) Discrete actions in information-constrained tracking problems, written jointly with Christopher A. Sims: in this paper we consider a restricted class of models with rationally inattentive agents and show analytically that in this class of models, for a wide class of situations where all the inputs to the problem---the distribution of exogenous randomness and the objective function---are very smooth, behavior of RI agents is nonetheless discrete. This provides some reassurance that the apparent discreteness in the computational solutions in previous work was not an anomaly, and also may give us some insight into the conditions under which RI solutions are likely to emerge as discrete. We also display classes of examples in which capacity-constrained agents will choose continuously distributed behavior, to supply qualitative guidance as to when capacity constraints are likely to lead to discrete behavior.