时 间：2021年11月25日（周四） 10：00-11：30
地 点：中北校区理科大楼A1514
题 目：Nonlinear dynamic duration panel data model with fixed effect
主讲人：李愚昊（助理教授）
主持人：崔芮 （助理教授）
主 办：经济与管理学部统计交叉科学研究院
摘 要：This paper proposes a new framework of econometric models for analyzing the nonlinear dynamic duration panel data. The duration outcomes are specified by the Generalized Accelerated Failure Time (GAFT) model with fixed effect unobserved heterogeneity. The model is multiplicative separable, thus, a first ratio transformation is used to difference out the fixed effect. This transformation results a panel of outcomes that are free from any time-invariant individual heterogeneity. We interpret an individual's `ratio' data as realizations of a counting process that is parametrically modelled. The intensity of each counting process can be constructed by specifying the distribution of error term of the GAFT model. A minimum distance estimation method is used, the resulting estimators are root-n consistent and asymptotically normal. We illustrate our method with a work absence application.
报告人简介：Dr.Li Yuhao is an Assistant Professor at Economics and Management School, Wuhan University. He obtained his Ph.D in Economics from Universidad Carlos III de Madrid in 2019. His fields of interest include Applied Econometrics, Labor and Health Economics.