Collaborative Research: Modeling Stochastic Intertemporal Preferences

合作研究：随机跨期偏好建模

• 批准号: 1919263
• 负责人: Kota Saito
• 金额: $ 19.03万
• 依托单位: California Institute of Technology
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2019
• 资助国家: 美国
• 起止时间: 2019-09-01 至 2024-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1919263&HistoricalAwards=false
• 关键词: Collaborative; Research; Modeling; Stochastic; Intertemporal

This award funds research on the economic theory of human decision-making. The first part of this project will address common everyday situations where individuals make choices repeatedly over time. In many of these cases, people's preferences over various goods may change from day to day. In other words, their preferences are stochastic. This research will propose a new theoretical model to help us understand how these stochastic preferences affect choice behavior. In particular, people's preferences to consume earlier or later would be an important factor. The second part of this project will study the welfare implications of stochastic preferences. When preferences change over time, it is unclear whether policy interventions would be beneficial and help improve individual well-being. This research will provide new tools and methodologies that will inform public policy for stochastic preferences. In general, results from this project will ultimately aid researchers in developing new insights for businesses, government agencies and other institutions that will benefit the U.S. public.This research will study the dynamics of stochastic intertemporal preferences. Existing models in the stochastic choice literature have mostly taken probabilistic choice frequencies as given without explicitly modeling repetition. The first part of this project will provide a repeated choice foundation for stochastic choice. The aim will be to formalize repetition and develop a tractable (i.e., recursive) model of stochastic intertemporal preferences. Existing standard models assume agents are expected utility maximizers and have standard (i.e. additive time-separable) intertemporal preferences. However, many experimental studies suggest that these standard models are too restrictive to accurately describe true human preferences. This research will take this into account and allow for more general intertemporal preferences such as Epstein-Zin (1989). From the perspective of inference and estimation, this will be crucial as not taking into account the intertemporal structure of the dynamic problem will lead to biased estimates of agents? preferences. The second part of this project will study dynamic consistency and stationarity in the context of stochastic choice. Research on this has been limited partly due to the difficulty of modeling such concepts for stochastic preferences. For instance, a natural extension of stationarity to stochastic choice would always be violated whenever preferences are stochastic. This is important because traditional violations of stationarity have been viewed as instances of dynamic inconsistency and grounds for policy interventions. New definitions and characterizations of dynamic consistency and stationarity for stochastic choice are needed as imprecise definitions would lead to policy interventions that may be counter-productive or welfare-decreasing for agentsThis award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

该奖项用于人类决策经济学理论的研究。这个项目的第一部分将解决个人随着时间的推移反复做出选择的常见日常情况。在许多这样的情况下，人们对各种商品的偏好可能每天都会发生变化。换句话说，他们的偏好是随机的。本研究将提出一个新的理论模型来帮助我们理解这些随机偏好是如何影响选择行为的。特别是，人们对提前或推迟消费的偏好将是一个重要因素。本项目的第二部分将研究随机偏好的福利含义。当偏好随着时间的推移而改变时，尚不清楚政策干预是否会有益并有助于改善个人福祉。这项研究将提供新的工具和方法，为公共政策提供随机偏好的信息。总体而言，这个项目的结果最终将帮助研究人员为企业、政府机构和其他机构开发新的见解，使美国公众受益。这项研究将研究随机跨期偏好的动态。现有的随机选择文献中的模型大多采用给定的概率选择频率，而没有明确地建模重复。本项目的第一部分将为随机选择提供重复选择的基础。其目的将是将重复形式化，并开发一个可处理的(即，递归的)随机跨期偏好模型。现有的标准模型假设代理是期望的效用最大化者，并且具有标准的(即附加的时间可分离的)跨时间偏好。然而，许多实验研究表明，这些标准模型过于严格，无法准确描述人类的真实偏好。这项研究将考虑到这一点，并考虑到更普遍的跨期偏好，如Epstein-Zin(1989)。从推理和估计的角度来看，这将是至关重要的，因为不考虑动态问题的跨期结构将导致对代理人的有偏见的估计？偏好。本项目的第二部分将研究随机选择背景下的动态一致性和平稳性。这方面的研究一直受到限制，部分原因是难以对随机偏好的此类概念进行建模。例如，当偏好是随机的时，平稳性到随机选择的自然扩展总是会被违反。这一点很重要，因为传统上违反平稳性的做法被视为动态不一致的例子，是政策干预的理由。需要对随机选择的动态一致性和平稳性进行新的定义和表征，因为不精确的定义将导致可能适得其反或减少代理人福利的政策干预。该奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。