This project improves our understanding of models of dynamic decisions and bounded rationality. The research will consider issues of dynamic stochastic choice by providing axiomatic characterizations of two types of stochastic dynamic choice rules. These two rules have many properties in common and coincide in static problems. The key distinction is whether the choice rule exhibits a preference for flexibility (preferring larger choice sets) or error aversion (preferring smaller sets). Another component of the research studies market interaction of agents operating under cognitive constraints such as limited attention, imperfect memory, and bounded ability to form contingent plans. The PI develops a new theory to explain how the market mechanism allocates agents' attention. The model yields a number of predictions: limited market participation, ex post consumption heterogeneity, and high volatility of asset prices as compared to the fundamentals.
The PI also studies two dimensions of temporal preferences that can be derived from a recursive utility model. The goal is to derive and analyze the premium that an agent is willing to pay for early resolution of consumption uncertainty and to examine the exact form of the aversion to autocorrelation.
The educational component of the award includes the use of a new collaborative open science approach to developing a decision theory textbook that will be made available online at no charge. The planned approach also contributes to graduate education by enlisting graduate students in textbook authorship.
The project provides foundations for models that are being used in practice, thereby contributing to our understanding of their strengths and weaknesses. It also analyzes novel dimensions of preferences and choice. These models are key to understanding many fundamental issues in economics. The project yields broader impact because adding new and better tools for analysis improves our ability to formulate and evaluate economic policy.
