Behavioral Optimization in Discrete Choice and in Competitive Markets

这个行为决策理论项目的目标是开发两个易于处理的决策模型，这些模型削弱了标准的理性要求，并适应了来自经济学、心理学和市场营销的实验证据。项目的第一部分将开发一个随机选择的新模型；决策者有一致的目标，但在追求这些目标时会犯错误的模型。将决策描述为随机有两个重要的优势：首先，它有助于衡量偏好；也就是说，随机选择模型允许计量经济学家通过选择频率来量化偏好的强度。其次，随机选择模型促进了不同经济主体之间的聚合。不同个体的行为可以作为单一模型的证据。本研究的第一部分将引入一种新的随机选择模型，即广义属性选择模型（GAR），以解决经济学、心理学和市场营销文献中发现的选择数据的关键规律。我们期望这个模型成为分析消费者行为的有用工具。约束决策者将自己的行为与环境细节相匹配的能力的行为模型，已被证明在解释系统性偏离“理性选择”标准模型方面是有效的。该项目的第二部分将介绍一种将行为因素纳入竞争性经济的一般分析方法。具体地说，本研究将分析一个充满活力的竞争经济，其中消费者的能力有限，无法根据具体情况调整消费选择。这一理论的一个应用是资产价格波动：我们希望证明，在给定的一组基本面（偏好和技术）下，行为竞争均衡中的价格波动比标准竞争均衡中的价格波动更大。本研究的最终目标是提供一个简单的模型，使研究人员能够量化和衡量金融市场中行为因素的影响，并使他们能够区分这些因素的影响与基本面的影响。

The goal of this behavioral decision theory project is to develop two tractable models of decision making that weaken standard rationality requirements and accommodate experimental evidence from economics, psychology and marketing.The first part of the project will develop a new model of random choice; a model in which decision makers have coherent objectives but make errors pursuing those objectives. Describing decision making as random has two important advantages: First, it facilitates the measurement of preferences; that is, random choice models allow the econometrician to quantify the intensity of preference via choice frequencies. Second, random choice models facilitate aggregation across different economic agents. The behavior of distinct individuals can be used as evidence for a single model.The first part of this research will introduce a new model of random choice, the generalized attribute selection model (GAR), to address key regularities of choice data found in the experiments in economics, psychology and in the marketing literature. We expect this model to become a useful tool for analyzing consumer behavior.Behavioral models that constrain the decision maker's ability to match her behavior to the details of her environment have proven effective in explaining systematic departures from standard models of "rational choice." The second part of the project will introduce a general analytical device for incorporating behavioral elements into competitive economies. Specifically, this research will analyze a dynamic competitive economy with consumers who are limited in their ability to adjust their consumption choices to their specific circumstances. One application of this theory is asset price volatility: we expect to demonstrate that with a given set of fundamentals (preferences and technology), there is more price volatility in a behavioral competitive equilibrium than in a standard competitive equilibrium.The ultimate goal of this research is to provide a simple model that permits researchers to quantify and measure the impact of behavioral factors in financial markets and enable them to distinguish the effects of these factors from the effects of fundamentals.