Robustness and Uncertainty in Markets and Mechanism Design

理解不确定性下的决策是经济学中最重要的问题之一。长期以来，经济学家和决策理论家一直试图为实体如何感知和应对不确定性的现实模型建立严格的基础。冯·诺伊曼-摩根斯特恩、安斯科姆-奥曼和萨维奇的预期效用理论在过去50多年里一直是不确定性理论和应用工作的主导范式。然而，最近，经济学家、决策理论家和心理学家的重要工作已经开始解决预期实用新型的严重缺陷。许多研究表明，决策者经常和系统地违反预期效用公理。例子包括在阿莱悖论中对小概率事件的误解，以及在埃尔斯伯格悖论中未能形成独特的主观先验并以概率复杂的方式使用它。本项目研究不确定性对市场和资源配置机制设计问题的影响。该项目主要关注围绕模糊模型的问题，该模型捕获决策者对决策问题中的不确定性作出反应的设置，但不能对其环境中的随机性形成精确的概率评估。该项目研究决策理论问题，涉及抽象选择问题中这种偏好的表示，以及对市场和机制设计问题的模糊性的影响。这个项目的智力优点是对不确定性对重要资源分配机制的影响有了更深入的理解。该项目的更广泛影响包括对创新时间和结构的见解，无限视界模型中令人满意的行为的出现，以及使用简单机制，如英国拍卖。

Understanding decision making under uncertainty is one of the most important problems in economics. Economists and decision theorists have long sought to develop rigorous foundations for realistic models of how agents perceive and respond to uncertainty. The expected utility theory of von Neumann-Morgenstern, Anscombe-Aumann and Savage has been the dominant paradigm for both theoretical and applied work incorporating uncertainty for over fifty years. More recently, however, important work by economists, decision theorists and psychologists has begun to address serious shortcomings in the expected utility model. Many studies suggest that decision makers routinely and systematically violate expected utility axioms. Examples include misinterpreting small probability events in the case of the Allais paradox, and failing to form a unique subjective prior and use it in a probabilistically sophisticated way in the case of Ellsberg's paradox. This project studies the implications of uncertainty on markets and mechanism design problems for allocating resources. The project focuses mainly on problems surrounding models of ambiguity, which capture settings in which decision makers respond to uncertainty in decision problems but do not form precise probabilistic assessments of randomness in their environments. The project studies decision-theoretic questions involving the representation of such preferences in abstract choice problems as well as the implications of ambiguity on markets and in mechanism design problems. The INTELLECTUAL MERIT of this project is a deeper understanding of the effects of uncertainty on important resource allocation mechanisms. The BROADER IMPACTS of this project include insights into the timing and structure of innovations, the emergence of satisficing behavior in infinite horizon models, and the use of simple mechanisms such as English auctions.