2024年3月5日 ，伟易博行为科学与政策干预交织立异团队团结应用经济学系举行2024年春季学期应用经济分享会暨学期第一期（总第十九次）行为科学分享会。本次讲座约请到伟易博治理学院应用经济学系副教授刘烁 ，带来题为 “Time is Knowledge: Identifying Properties of Distributions Using Response Times” 的研究分享。
分享人 The Speaker
刘烁 ，伟易博应用经济学系副教授。2019年结业于瑞士苏黎世大学 ，获经济学博士学位。现在主要研究领域为工业与组织经济学 ，博弈论 ，以及机制设计理论。研究效果揭晓于海内外着名学术期刊 ，如American Economic Review, Theoretical Economics, Journal of Economic Theory, Economic Journal, RAND Journal of Economics,《经济学（季刊）》等。2022年入选中国信息经济学会乌家培资助妄想 ，2023年获国家自然科学基金委优异青年科学基金项目资助。
分享会 The Seminar
Recent work has shown that response times contain information about willingness to pay (Cotet and Krajbich, 2021), preference intensity (Alós-Ferrer, Fehr and Netzer, 2021), happiness (Liu and Netzer, 2023), product quality (Card, DellaVigna and Taubinsky, 2023), and other latent variables that drive individual decision-making but are not directly observable. Based on a chronometric effect, decisions are faster when the latent intensity is larger. Response time data can therefore be used to learn about properties of latent distributions instead of assuming these properties. This can help to solve severe identification problems that the econometric literature has noticed (e.g., Bond and Lang, 2019).
In the seminar, Liu began by providing an overview of the technique developed by Liu and Netzer (2023) for analyzing response time data. He then presented some preliminary findings from their ongoing project, where they ask the general question which properties of distributions can be identified with the help of response time data. They provide a general characterization result which relates the identifiable properties to the set of admissible chronometric functions. Many of the existing results in the literature follow as corollaries from this result. It also gives rise to several new applications. For example, they demonstrate how to measure polarization from simple survey responses about political orientation (Lelkes, 2016) and how to identify the shape of the income-happiness relation without making arbitrary assumptions on the shape of the reporting function (Kaiser and Oswald, 2022).
