辽宁大学经济学前沿高端讲座第五十讲    

   Identification and estimation of treatment effects using instrumental variable without exclusion restriction   主讲人：江硕助理教授（厦门大学王亚南经济研究院）  

主持人：杨程助理教授（辽宁大学李安民经济研究院）  

嘉宾介绍：何真瑜助理教授（辽宁大学李安民经济研究院）  

时间：2024年5月17日（周五）  10：00-11：30（北京时间）  

地点：辽宁大学崇山校区五洲园一楼会议室 

线上地址：腾讯会议：573-887-734  

语言：中文/英文  

摘要：In this paper, we study the identification and estimation of treatment effects using instrumental variable that does not satisfy the exclusion restriction. In the absence of exclusion restriction, we show that a new type of treatment effect called the “local average treatment effect on the encouraged outcome” is point identified and it coincides with LATE when the exclusion restriction actually hold. Testable restrictions for the exclusion restriction are derived. When the effects are homogeneous, we propose different causal reduced form characterizations of the treatment effect as population regression coefficient and propose several two-step semiparametric estimators that are root-N consistent and asymptotically normal. Monte Carlo simulations demonstrates the finite sample efficacy of our estimators and tests.  

主讲人简介：

江硕，厦门大学王亚南经济研究院助理教授，美国华盛顿大学经济学博士。主要研究领域为微观计量经济学和实证产业组织。主持教育部人文社科青年基金一项。