UC Berkeley

Individual decision making is the foundation for almost all economic analysis. In this dissertation, I study various choice problems in the absence of perfect information.

Related decisions are often observed in isolation without direct measurement of correlation in beliefs across state spaces or complementarity in tastes across prize spaces. In my first chapter, my coauthor David Ahn and I introduce a novel model with two decision problems with distinct states and prizes, which we call small worlds, without observation of bets that are contingent on the realization of both worlds. We characterize an appropriate version of subjective expected utility, where choices are made as if there is a joint distribution over the product of the state spaces and a joint utility index over pairs of prizes from both prize spaces. Turning to identification, the joint utility index over pairs of prizes and the marginal belief over each small world are identified, but the uniqueness of the joint distribution is more subtle. If the utility index is separable across prize spaces, then the correlation across state spaces is unidentified; but if there is any complementarity across prizes, then the joint distribution is exactly identified.

A growing literature attributes choice anomalies, such as stochastic choices, to limited attention. Most choice models with limited attention do not model attention itself. In my second chapter, I propose a choice model with gradual attention, in which a decision maker’s attention is limited due to time constraints. With richer choice data including the associated decision time, I provide an axiomatic characterization of this model. I show the model primitives such as the preference can be uniquely identified. To make the model comparable to existing choice models with limited attention, I consider its static version without time data and show identification is not always possible, which highlights the novelty of utilizing the decision time information in understanding choice behaviors. I examine aggregation issues of this model and give necessary and sufficient conditions for the existence of a representative agent under heterogeneity. In my model, the decision maker's consideration set is growing monotonically as decision time increases. As a consequence, more time leads to better choices. I test this choice pattern using experimental data from a well-known study and find supportive evidence.

In a persuasion problem, an information sender designs an experiment for an information receiver, in order to persuade the receiver to change actions. While persuasion is all about manipulating the receiver’s posterior beliefs, little is known about the effect of changes in receiver's risk attitudes on persuasion outcomes. In my third chapter, I study how the receiver’s risk attitudes affect the set of implementable outcomes in a persuasion problem. I consider a two-period consumption model, where the receiver’s future income is uncertain and he has the option to transfer his wealth intertemporally through saving. I present a local comparative statics result, showing that any marginal change in the receiver’s prudence either expands or shrinks the set of implementable persuasion outcomes. Risk aversion does not affect persuasion outcomes, except through pivoting the receiver’s optimal action under the prior belief. Other higher order risk attitudes have no direct impact on persuasion outcomes.