Research

Selected papers

Distinguishing Preferences from Justifications [paper] (R&R at Social Choice and Welfare, November 2023)

In a justification model, a decision maker identifies the subset of alternatives that maximize at least one member of a set of acceptable preferences (“justifications”), then maximizes his preference over that subset. This paper characterizes the justification model and uses that characterization to show that the model exhibits significantly stronger identification properties than related models present in the literature. All information about the decision maker's preferences and justifications turns out to be embedded in two simple choice patterns. These patterns can be used to construct a unique focal representation, which minimizes the constraints the model places on the decision maker.

Justification in Heterogeneous Populations [paper] (R&R at Journal of Mathematical Economics, December 2023)

Motivated by the concern that decision makers with unjustifiable preferences may exhibit consistency concerns, this paper extends the justification model of Ridout (2023) to stochastic choice. It provides tools for detecting justification that do not rely on individual-level inconsistencies, and it suggests strategies for discouraging unjustifiable choices in a heterogeneous population.

A Dynamic Model of Peer Effects on Choice (2020) [paper] (R&R at Economic Theory, March 2024)

I model a decision maker (DM) who cares about the relative frequencies with which different items are selected by peers. The DM has a Bayesian prior on these relative frequencies, which he updates as he collects more data. The model provides a unified framework for interpreting recent empirical evidence on peer effects in varied domains. I provide an axiomatic characterization of the model and show how different specializations of the representation generate different responses to peer information. I then extend the model to distinguish peer effects caused by inference about the best alternative from peer effects caused by concerns about relative status.

A Theory of Ex Post Rationalization (with Erik Eyster and Shengwu Li) [paper] (presented at Behavioral Economics Annual Meeting 2022)

People rationalize their past choices, even those that were mistakes in hindsight.  We propose a formal theory of this behavior.  The theory predicts sunk-cost effects, as well as "unsunk-benefit" effects. Its model primitives are identified by choice behavior, and it yields tractable comparative statics.

Work in progress

Trust and Respect [slides, results] (with Steve Hamilton)

We study communication or delegation between a long-lived expert and a sequence of short-lived laypeople. There are two expert types, only one of which the laypeople are willing to consult, and the laypeople monitor the expert’s reputation. In contrast to existing results on “political correctness” and “bad reputation,” we find desirable equilibria in which good types communicate truthfully and bad types are disciplined to act almost exactly like good types. These desirable results are achieved by allowing the expert’s payoff from each interaction to depend on his reputation. Costs of reputation only appear in the very long run after beliefs converge. While some good types achieve maximal payoffs, others are unable to separate themselves from bad types and are punished with very low payoffs, although communication continues unaffected.

Utility Happens (with Matthew Rabin and Tomasz Strzalecki)

We seek to understand models in which utility is experienced by agents in the literal sense of the word. In the expected utility model the agent can be risk averse, but there is no sense in which that aversion is experienced by the agent: whether she has to live with the risk for a long or short time is immaterial. A large body of work relaxes expected utility. Among them are (1) models such as Kreps and Porteus which relaxes time-separability, and (2) the many models of non-expected utility, such as Prospect Theory, which relax separability over states. Another line of work is (3) a la Caplin and Leahey where the agent’s utility is defined directly on beliefs. By a series of examples and axiomatic characterizations, we show how these three classes of preferences relate to each other. We introduce the notions of anxiety and excitement and show how their definitions within the model correspond to natural violations of dynamic consistency observable in the lab or field.