[9月20日]计量、金融和大数据分析workshop

发布日期:2019-09-18 09:40    来源:

  【题目】: Identification and Estimation of Peer Effects in Latent Networks

  【主讲人】:Wenzheng Gao (高文政)

  Assistant Professor,Nankai University

  【时间】:2019年9月20日(周五)9:00-10:30

  【地点】:北大经济学院107室

  【摘要】:This paper studies the identification and estimation of peer effects when network links are latent because of measurement errors. We first derive conditions under which the conditional distributions of covariates based on latent network structure can be identified. Second, we show that the peer effects can be identified by adopting these covariates as instrumental variables. Based on the identification method, a three-stage semiparametric estimator is proposed and applied to study the peer effects on the financial policy of firms in United States. Our empirical results show significant peer effects and contextual effects among firms. Moreover, the peer effects become insignificant when measurement errors in network links are ignored.

  【主讲人简介】:

  Wenzheng Gao (高文政)

  Ph.D., Texas A&M University 2018

  Assistant Professor,School of Economics,Nankai University

  Research Interests: Microeconometrics, Industrial Organization, Applied Microeconomics