Collaborative Research: Evidence in Economic Models

该奖项资助经济理论方面的研究。该小组将研究获取和使用确凿的资料或证据，使有关各方能够证明他们希望作出的某些主张。什么时候才能设计社会制度，为利益相关方提供足够强大的激励，以获得足够可信的证据，从而实现有效的结果？当效率不可能实现时，由可能的最佳机构产生的效率扭曲的性质是什么？研究小组将开发用于研究这些问题的数学模型，包括使分析足够易于处理以产生明确结果的工具。更广泛地说，这项工作将促进在这一大类有趣环境中应用经济模型的研究。通过提供使分析更易于处理的工具，这项研究可能会产生新的经济制度类别，并为现有制度提供新的视角。这些工具在研究组织结构方面也很有用，比如公司和政府。该项目侧重于智能体获取噪声证据的机制设计。一般模型与广泛的“硬信息”经济模型有联系，如贝叶斯说服和测试模型。该团队将开发简化机制设计问题的工具，以便能够专注于更易于处理的机制类别。特别是，他们将制定条件，在此条件下，人们可以确定委托人在代理人提出任何索赔时要求的证据和/或委托人在给定任何类型报告时建议的证据生成行动。这种识别将大大简化最优机理的表征。研究人员还将比较最优机制与委托人无法承诺的结果。最后，它们将描述可能产生有效结果的环境以及无法实现效率时扭曲的性质。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This award funds research in economic theory. The team will study the acquisition and use of hard information or evidence which enables interested parties to prove some assertions they wish to make. When can social institutions be designed in such a way as to provide strong enough incentives to interested parties to acquire enough sufficiently credible evidence as to enable efficient outcomes? When efficiency is not possible, what is the nature of the distortions from efficiency generated by the best institutions that are possible? The research team will develop mathematical models for studying these questions, including tools for making the analysis sufficiently tractable to yield clear results. More broadly, the work will facilitate research of applied economic models in this large class of interesting environments. By providing tools for making the analysis more tractable, the research may generate new classes of economic institutions and shed new light on existing ones. The tools will also be useful in studying the structure of organizations, such as firms and governments. The project focuses on mechanism design when agents can acquire noisy evidence. The general model has connections to a wide range of economic models of "hard information", such as Bayesian persuasion and models of testing. The team will develop tools for simplifying the mechanism design problem to be able to focus on a more tractable class of mechanisms. In particular, they will develop conditions under which one can identify the evidence the principal will request given any claim by the agent and/or the evidence generation actions the principal will recommend given any type report. Together or separately, this identification will greatly simplify characterization of the optimal mechanism. The investigators will also compare optimal mechanisms to outcomes when the principal cannot commit himself. Finally, they will characterize environments in which efficient outcomes are possible and the nature of distortions when efficiency is not achievable.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.