国内双重差分法的研究现状与潜在问题

发布时间：2018-02-02 12:31

  本文关键词： 双重差分 自然实验 准实验 内生性问题　出处：《数量经济技术经济研究》2015年07期 　论文类型：期刊论文

【摘要】：在国内政策效果评估领域,基于自然实验的双重差分法是应用最广的研究工具,并在近年呈现加速发展的趋势。但通过归纳双重差分法的基本应用条件与国内研究现状,本文发现部分研究出现了内生性、控制组受影响、样本异质性等问题。其中尤以内生性问题较为严重,甚至使部分研究产生了一定的缺陷,所谓的自然实验往往难以成立。本文根据实验理论与相关理论基础,对国内主要研究成果进行分类与评述,并针对所发现的问题提出改良办法。
[Abstract]:In the field of domestic policy effectiveness assessment, the dual difference method based on natural experiments is the most widely used research tool. And in recent years, the trend of accelerated development. But through the induction of the basic application conditions of the dual difference method and the current situation of domestic research, this paper found that part of the study appeared endogenous, control group was affected. Sample heterogeneity and other problems, especially the problem of endogenesis is more serious, even some of the research has some defects, the so-called natural experiments are often difficult to hold. This paper based on the experimental theory and related theoretical basis. The main research results in China are classified and reviewed, and the improvement methods are put forward for the problems found.
【作者单位】： 暨南大学产业经济研究院;中国社会科学院经济研究所;暨南大学统计学系;
【基金】：国家自然科学基金项目(71203078、71333007、41371174) 教育部哲学社会科学重大课题攻关项目(14JZD021) 国家社会科学基金项目(13CJL065) 广东产业发展粤港澳台区域合作研究中心 暨南大学“跨越计划”的资助
【分类号】：F812.42;F224
【正文快照】： 国由双重差分法的研尧现r烙脘菰谖侍狻愠铝帧，

本文编号：1484516