Essays in Decision Theory
SodaPDF-merged-Merging Result.pdf (877.7Kb)
Access StatusFull text of the requested work is not available in DASH at this time ("dark deposit"). For more information on dark deposits, see our FAQ.
MetadataShow full item record
CitationRidout, Sarah. 2021. Essays in Decision Theory. Doctoral dissertation, Harvard University Graduate School of Arts and Sciences.
AbstractThis dissertation comprises three independent chapters that span topics in decision theory, including ethical decision-making, learning from others, and anticipatory emotions.
In Chapter 1, I model decision-making constrained by ethics, duty, law, or other virtues or principles. In addition to a true preference, the decision maker (DM) is characterized by a set of preferences that he considers justifiable. In each choice setting, the DM maximizes his true preference over the subset of alternatives that maximize at least one of the justifiable preferences. The justification model unites a broad class of empirical work on distributional preferences, discrimination, corruption, philanthropy, and other domains. I provide simple axiomatic characterizations of several variants of the justification model as well as practical tools for identifying true preferences and justifications from choice behavior. I show that identification is partial in general, but full identification can be achieved by moving to between-subject data and imposing some additional structure on true preferences and justifications. Moving to between-subject data also eliminates the consistency motives that may arise in within-subject experiments. I extend the between-subject justification model to information choice and relate its predictions to the “moral wiggle room” literature.
In Chapter 2, I model a decision maker who cares about the relative frequencies with which different items are selected by peers. The decision maker 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.
In chapter 3, I model a decision maker who fantasizes or worries about future risks by reweighing probabilities in an optimistic or pessimistic manner. The degree of optimism or pessimism toward a given lottery may depend on the lottery itself, its context, and the period in which it takes place. The representation resembles a standard recursive expected utility model, but with distorted probabilities instead of physical probabilities. The model is well suited to applications because it (a) has a flexible domain (infinite tree of continuous and/or discrete distributions on a bounded or unbounded interval), (b) admits simple comparative statics, and (c) can generate tractable parametric distorted distributions.
Citable link to this pagehttps://nrs.harvard.edu/URN-3:HUL.INSTREPOS:37368182
- FAS Theses and Dissertations