Professor Dreze is a highly respected mathematical economist and econometrician. This book brings together some of his major contributions to the economic theory of decision making under uncertainty, and also several essays. These include an important essay on 'Decision theory under moral hazard and state dependent preferences' that significantly extends modern theory, and which provides rigorous foundations for subsequent chapters. Topics covered within the theory include decision theory, market allocation and prices, consumer decisions, theory of the firm, labour contracts, and public decisions.

The objective of this thesis is to improve the understanding of human behavior that goes beyond monetary rewards. In particular, it investigates social influences in individual’s decision making in situations that involve coordination, competition, and deciding for others. Further, it compares how monetary and social outcomes are perceived. The common theme of all studies is uncertainty. The first four essays study individual decisions that have uncertain consequences, be it due to the actions of others or chance. The last essay, in turn, uses the advances in research on decision making under uncertainty to predict behavior in riskless choices. The first essay, Fairness Versus Efficiency: How Procedural Fairness Concerns Affect Coordination, investigates whether preferences for fair rules undermine the efficiency of coordination mechanisms that put some individuals at a disadvantage. The results from a laboratory experiment show that the existence of coordination mechanisms, such as action recommendations, increases efficiency, even if one party is strongly disadvantaged by the mechanism. Further, it is demonstrated that while individuals’ behavior does not depend on the fairness of the coordination mechanism, their beliefs about people’s behavior do. The second essay, Dishonesty and Competition. Evidence from a stiff competition environment, explores whether and how the possibility to behave dishonestly affects the willingness to compete and who the winner is in a competition between similarly skilled individuals. We do not find differences in competition entry between competitions in which dishonesty is possible and in which it is not. However, we find that due to the heterogeneity in propensity to behave dishonestly, around 20% of winners are not the best-performing individuals. This implies that the efficient allocation of resources cannot be ensured in a stiff competition in which behavior is unmonitored. The third essay, Tracing Risky Decision Making for Oneself and Others: The Role of Intuition and Deliberation, explores how individuals make choices under risk for themselves and on behalf of other people. The findings demonstrate that while there are no differences in preferences for taking risks when deciding for oneself and for others, individuals have greater decision error when choosing for other individuals. The differences in the decision error can be partly attributed to the differences in information processing; individuals employ more deliberative cognitive processing when deciding for themselves than when deciding for others. Conducting more information processing when deciding for others is related to the reduction in decision error. The fourth essay, The Effect of Decision Fatigue on Surgeons’ Clinical Decision Making, investigates how mental depletion, caused by a long session of decision making, affects surgeon’s decision to operate. Exploiting a natural experiment, we find that surgeons are less likely to schedule an operation for patients who have appointment late during the work shift than for patients who have appointment at the beginning of the work shift. Understanding how the quality of medical decisions depends on when the patient is seen is important for achieving both efficiency and fairness in health care, where long shifts are popular. The fifth essay, Preferences for Outcome Editing in Monetary and Social Contexts, compares whether individuals use the same rules for mental representation of monetary outcomes (e.g., purchases, expenses) as for social outcomes (e.g., having nice time with friends). Outcome editing is an operation in mental accounting that determines whether individuals prefer to first combine multiple outcomes before their evaluation (integration) or evaluate each outcome separately (segregation). I find that the majority of individuals express different preferences for outcome editing in the monetary context than in the social context. Further, while the results on the editing of monetary outcomes are consistent with theoretical predictions, no existing model can explain the editing of social outcomes.

The third part concerns a class of problems in non-cooperative game theory that is subject to multiple, Pareto-ranked equilibria, where decision makers are faced with strategic uncertainty. I investigate how intervention strategies based on a Schelling-inspired tipping mechanism can efficiently alter the incentives of a small subset of individuals to trigger system-wide coordination on the Pareto-superior outcome. I then apply this framework to two examples, the provision of airline security and reward schemes in organizations.

This dissertation studies several economic questions related to decision making under uncertainty. In the first chapter, I consider the question of to what extent an external observer can infer the underlying information structure by observing an equilibrium action distribution in an incomplete information game. I investigate this issue in a general linear-quadratic-Gaussian framework. A simple class of canonical information structures is offered and proves rich enough to rationalize any possible equilibrium action distribution that can arise under an arbitrary information structure. I show that the class is parsimonious in the sense that the relevant parameters can be uniquely pinned down by an observed equilibrium outcome, up to some qualifications. This implies, for example, that the accuracy of each agent's signal about the state is identified, as measured by how much observing the signal reduces the state variance. Moreover, I show that a canonical information structure characterizes the lower bound on the amount by which each agent's signal can reduce the state variance, across all observationally equivalent information structures. The lower bound is tight, for example, when the actual information structure is uni-dimensional, or when there are no strategic interactions among agents, but in general, there is a gap since agents' strategic motives confound their private information about fundamental and strategic uncertainty.The second chapter is based on my joint work with Federico Echenique, Yuta Nakamura, Luciano Pomatto, and Jamie Vinson. We propose a model of incomplete preferences, termed twofold multiprior preferences, in which an act f is ranked above an act g only when f provides higher utility in a worst-case scenario than what g provides in a best-case scenario. The model explains the experimental phenomenon, called failures of contingent reasoning, captured through a weakening of the state-by-state monotonicity (or dominance) axiom. Our model gives rise to rich comparative statics results, which demonstrate that the two important decision criteria in the literature--subjective expected utility representation and obvious dominance--can be encapsulated as the respective extreme cases of our model. We present an application to second-price auctions and illustrate that our model can explain a wide array of anomalistic bidding behaviors caused by failures of contingent reasoning, yet it can provide sharper predictions than obvious dominance.The third chapter, coauthored with Carlo Cusumano, extends the analysis of the second chapter by studying a general class of twofold preferences that compare different acts based on two possibly different utility functions. We find that the twofold preferences are useful for analyzing economic problems for at least two reasons. From a behavioral perspective, it enables theoretical models to account for the indecisiveness of a decision maker's choice in a way depending on the scale of odds. From a modeling perspective, the twofold approach can be a unified framework for modeling various incomplete preferences that differ in additional independence-type axioms to be imposed. Our series of axiomatization results provide the characterizations of the incomplete counterparts of existing complete preferences in the literature.