Learning and Economic Decisions

学习和经济决策

• 批准号: 9223320
• 负责人: Drew Fudenberg
• 金额: $ 15.32万
• 依托单位: Harvard University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1993
• 资助国家: 美国
• 起止时间: 1993-07-01 至 1995-12-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9223320&HistoricalAwards=false
• 关键词: Learning; Economic; Decisions

This research consists of three projects investigating how economic agents use their past experience to guide their current decisions. The first research project studies agents who use very simple rules of thumb in choosing between alternative technologies; these rules will incorporate the idea that agents try to learn from the experiences of their neighbors. The project will investigate the long-run implications of such "social learning," and in particular on when simple rules can lead to efficient long-run outcomes. The second project studies learning in games. Here the focus is on when repeated interaction will lead play to correspond to a Nash equilibrium or to one of its refinements. The third project will develop a model of optimal contracts between a firm's owners and its manager in a setting where the owners learn about the firm's future prospects from its current performance. The model will provide an explanation for the widely-observed practice of "income smoothing," under which the manager takes actions that increase reported income when income is low, and decrease reported income when income is high. This research is important because it tackles and helps provide answers to a few of the most important issues in the economic implications of learning.

这项研究包括三个项目，调查经济主体如何利用他们过去的经验来指导他们当前的决策。第一个研究项目研究的是使用非常简单的经验法则来选择替代技术的代理人；这些规则将包含代理试图从其邻居的经验中学习的想法。该项目将调查这种“社会学习”的长期影响，特别是简单的规则何时能产生有效的长期结果。第二个项目研究游戏中的学习。这里的重点是重复的互动何时会导致游戏符合纳什均衡或它的一个改进。第三个项目将建立一个企业所有者和管理者之间的最优契约模型，在这个模型中，所有者从企业当前的业绩中了解到企业的未来前景。该模型将为广泛观察到的“收入平滑”实践提供解释，即当收入低时，管理者采取行动增加报告收入，当收入高时，减少报告收入。这项研究很重要，因为它解决了学习的经济含义中一些最重要的问题，并帮助提供了答案。