Dynamic Choice in an Uncertain World

不确定世界中的动态选择

• 批准号: 1258665
• 负责人: Drew Fudenberg
• 金额: $ 33.05万
• 依托单位: Harvard University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2013
• 资助国家: 美国
• 起止时间: 2013-09-01 至 2016-07-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1258665&HistoricalAwards=false
• 关键词: Dynamic; Choice; Uncertain; World

This project will help develop a better understanding of human behavior in both individual decisions and in 'games', that is, settings where the decisions of various agents can interact and each agent's payoff may depend on the actions of others. In the setting of individual decisions, the project contributes to our understanding of dynamic choice, and in particular choice under uncertainty, where uncertainty includes uncertainty about one's own preferences (for a consumer, which goods will best serve their needs; for a firm, their costs and possible revenues) as well as uncertainty about the preferences of others (e.g. the costs of other firms) and also uncertainty about the plans and intentions of other agents. The project will also extend the PI's past work on temptation and self-control to take into account habit formation, with the idea that the amount of temptation depends on what the agent has learned about his preferences in the past. In the setting of games, the project will study how and when agents manage to cooperate via reciprocation even when they only imperfectly observe the intentions of the others- as for example when a friend says they are busy or sick and thus unable to come help you. The project will also develop improved ways to predict the long-run equilibrium outcomes of a game when players are initially uncertain about their opponent's strategies and learn about them by observation, where these observations are only partial: For example a losing bidder in an auction might see the winning bid but not see the losing ones, and not observe how much the other bidders would have been willing to pay.Intellectual Merit: The work on choice under uncertainty will provide axiomatic characterizations of various sorts of stochastic choice behavior in dynamic settings, including the dynamic logit commonly used in empirical studies of consumer purchases, as the stochastic choice in these models can be interpreted as a consequence of the agent being uncertain of his or her own preferences. These theoretical foundations will shed light on when the specification is likely to be reasonable; they also may lead to alternative and equally tractable specifications that are better fits for some situations. The work on habit formation will incorporate insights from cognitive neuroscience to build more accurate models of how habits develop and how temptation distorts economics decisions. To understand reciprocal cooperation, the project will use experiments on repeated prisoner dilemma and ultimatum games when the intended actions are implemented with noise, comparing the outcomes to those when intentions as well as actions are observed, and to the outcomes in one-shot games. This will provide data to guide theories of altruism, reciprocation, and social preferences. The work on learning opponents? strategies will provide a better solution concept for many field settings and lab experiments, and thus a new tool for analyzing them.Broader Impact: The project will be strengthen ties between economic theorists and experimenters interested in behavioral economics, between economists working on decision theory and those doing estimation of demand systems, and between economists and psychologists interested in addiction, habit formation, and willpower. The grant will also provide support for graduate students doing research on related topics. Taking a longer term view, we are hopeful that the proposed research will enhance our understanding of how and when reciprocal altruism leads to cooperation; this is of fundamental importance in many branches of social science and is also a key issue in evolutionary biology.

该项目将有助于更好地理解人类在个人决策和“游戏”中的行为，也就是说，各种代理的决策可以相互作用，每个代理的回报可能取决于其他代理的行为。在个人决策的背景下，该项目有助于我们理解动态选择，特别是不确定性下的选择，其中不确定性包括对自己偏好的不确定性（对消费者而言，哪种货品最能满足其需要;对于公司来说，他们的成本和可能的收入）以及其他人偏好的不确定性（例如，其他公司的成本）以及其他代理人的计划和意图的不确定性。 该项目还将扩展PI过去在诱惑和自我控制方面的工作，以考虑习惯的形成，其想法是诱惑的数量取决于代理人过去对自己偏好的了解。在游戏的设置中，该项目将研究代理如何以及何时通过互惠合作，即使他们只是不完美地观察到其他人的意图-例如，当一个朋友说他们忙碌或生病，因此无法来帮助你。该项目还将开发改进的方法来预测游戏的长期均衡结果，当玩家最初不确定对手的策略时，并通过观察来了解它们，这些观察只是部分的：例如，拍卖中的失败投标人可能会看到获胜的出价，但看不到失败的出价，并且不会观察其他投标人愿意支付多少。智力优势：在不确定性下的选择工作将提供公理化的各种随机选择行为在动态设置，包括动态logit通常用于消费者购买的实证研究，在这些模型中的随机选择可以解释为代理人不确定自己的偏好的结果。这些理论基础将阐明规范何时可能是合理的;它们也可能导致更适合某些情况的替代和同样易于处理的规范。关于习惯形成的研究将结合认知神经科学的见解，以建立更准确的模型来研究习惯是如何形成的，以及诱惑是如何扭曲经济决策的。为了理解互惠合作，该项目将使用重复囚徒困境和最后通牒游戏的实验，当预期的行动被实施时，噪音，比较结果与观察到的意图和行动时的结果，以及一次性游戏的结果。这将为利他主义、互惠行为和社会偏好的理论提供指导数据。学习对手的工作？更广泛的影响：该项目将加强经济理论家和对行为经济学感兴趣的实验者之间的联系，加强从事决策理论研究的经济学家和从事需求系统估计的经济学家之间的联系，加强经济学家和对成瘾、习惯形成和意志力感兴趣的心理学家之间的联系。该补助金还将为研究生提供相关主题研究的支持。从长远来看，我们希望拟议中的研究将提高我们对互惠利他主义如何以及何时导致合作的理解;这在社会科学的许多分支中具有根本的重要性，也是进化生物学中的一个关键问题。