Models of Dynamic Choice

该奖项资助经济决策理论的研究。有四个不同的项目。每个人都寻求开发新的方法，将人类行为的各个方面纳入决策的正式数学模型。第一个项目建立在其他学者（Kreps-Porteous和Epstein-Zin）之前开发递归实用新型的工作基础上。该项目模拟了具有这种效用函数所表示的偏好的个人为不确定性的早期解决所支付的溢价；换句话说，个人愿意放弃多少收入来了解他未来的消费水平？该溢价可作为递归实用新型参数值选择的标准。第二个项目研究在具有标准偏好的标准竞争金融市场中运作的有界优化决策者。新理论产生了许多预测：有限的市场参与，事后消费异质性，以及与基本面相比资产价格的高波动性。方案三提供了准双曲折现的一个新的公理化。这种随时间变化的偏好模型越来越广泛地应用于行为经济学的应用中。所提出的公理化和隐含的参数测量方法对实验工作是有用的。这些方法可以推导出独立于效用函数的短期折现因子。目标是开发一种方法，消除现有方法不可避免的实验混淆。项目四研究了个体决策者的模型，这些决策者没有充分认识到他们未来的选择在多大程度上忽视了他们当前的偏好。该项目提出了一种方法来引出这种对未来的不准确信念。我们的目标是通过研究这些学习失败是如何长期持续下去来解释这种行为。动态决策模型是理解许多经济现象的关键，例如储蓄、投资、税收和增长。这项研究将改进可用于模拟这些现象的工具箱。其结果将是分析经济行为和经济政策影响的新方法。这些方法不仅对经济学家有用，而且对决策理论家和金融学者也有用。

This award funds research in economic decision theory. There are four distinct projects. Each seeks to develop new methods for incorporating aspects of human behavior into formal mathematical models of decision making.The first project builds on previous work by other scholars (Kreps-Porteous and Epstein-Zin) that developed recursive utility models. This project models the premium that an individual with preferences represented by this kind of utility function would pay for the early resolution of uncertainty; in other words, what fraction of his income would the individual be willing to give up in order to learn his future consumption levels? This premium can be used as a criterion for the choice of parameter values in recursive utility models.The second project studies boundedly optimizing decision amkers operating in otherwise standard competitive financial markets with standard preferences. The new theory yields a number of predictions: limited market participation, ex post consumption heterogeneity, and high volatility of asset prices as compared to the fundamentals.Project Three provides a novel axiomatization of quasi-hyperbolic discounting. This model of preferences over time is increasingly widely used in applications of behavioral economics. The proposed axiomatization and the implied method of parameter measurement are useful for experimental work. These methods make it possible to elicit short run discount factors independent of utility functions. The goal is to develop a method that will eliminate experimental confounds that are unavoidable with current methods.Project Four studies a model of individual decisions makers who do not fully recognize the extent to which their future choices disregard their current preferences. The project proposes a method of eliciting such inaccurate beliefs about the future. The goal is to explain such behavior by studying how these failures of learning can persist in the long run.Models of dynamic decisions are key to understanding many economic phenomena, such as savings, investment, taxation, and growth. This research will improve the toolbox available to model these phenomena. The result will be new methods to analyze the effects of economic behavior and economic policy. The methods will be useful not just to economists but also to decision theorists and financial scholars.