Learning and Information

学习与信息

• 批准号: 9730181
• 负责人: Drew Fudenberg
• 金额: $ 22.48万
• 依托单位: Harvard University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 1998
• 资助国家: 美国
• 起止时间: 1998-07-01 至 2002-09-30
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=9730181&HistoricalAwards=false
• 关键词: Learning; Information

This grant supports the continuation of several research projects. One project studies new issues in monopoly and oligopoly pricing that arise as firms come to have more information about the purchasing habits of consumers. Previous work by the PI studied the use of `upgrade discounts` by monopoly sellers of software, and `buybacks` of old goods by sellers who want to prop up the price of a new Improved version. The proposed work would study attempts by firms to `poach` the current customers of their competitors by offering them special discounts or other inducements to switch. The project will analyze equilibrium poaching both with short term contracts and when long term contracts are practical; one question will be the overall impact of this sort of price discrimination on consumer welfare. A second project has developed equilibrium concepts for extensive form games, based on the idea that players may persistently make mistaken forecasts of how opponents would respond to actions that have never been tried. Previous work studied the case where players have no information about their opponents' payoff functions, and so are unable to make deductions about how the opponents would respond to untried actions. The proposed work will study the way that prior information about opponents' payoffs can restrict the set of equilibrium outcomes. A key issue in this development is what sort of prior information about payoffs should be considered.

这笔赠款支持几个研究项目的继续。其中一个项目研究垄断和寡头垄断定价中出现的新问题，这些问题是由于公司越来越了解消费者的购买习惯。PI以前的工作研究了垄断软件销售商使用“升级折扣”，以及希望支撑新改进版本价格的销售商对旧产品的“回购”。拟议的工作将研究公司试图通过提供特别折扣或其他诱惑来“挖走”其竞争对手的现有客户。该项目将分析短期合同和长期合同可行时的均衡偷猎;一个问题是这种价格歧视对消费者福利的总体影响。 第二个项目为广泛形式的游戏开发了均衡概念，其基础是玩家可能会持续错误地预测对手对从未尝试过的行动的反应。以前的工作研究的情况下，球员没有关于他们的对手的支付功能的信息，因此无法作出推论对手将如何应对未经尝试的行动。拟议的工作将研究的方式，对手的回报先验信息可以限制一组均衡结果。在这一发展的一个关键问题是什么样的先验信息的回报应予以考虑。