A Sparsity-Based Model of Bounded Rationality

摘要: This paper proposes a model in which the decision maker builds an optimally simplified representation of world is "sparse," i.e., uses few parameters that are non-zero. Sparsity formulated so as to lead well-behaved, convex maximization problems. The agent's choice features quadratic proxy for benefits thinking and linear formulation costs thinking. agent then picks optimal action given his world. yields tractable procedure, embeds traditional rational particular case, can be used analyzing classic economic questions under bounded rationality. For instance, studies how boundedly agents select consumption bundle while paying imperfect attention prices, frictionless firms set prices response. leads novel mechanism price rigidity. also examine intertemporal problems portfolio with understanding returns.