计量经济学应用研究的可信性革命

可信性是计量经济学应用研究的重要问题,其核心在于实现经济理论、统计学、数学在实证研究中的科学结合。本文基于国际计量经济学界对可信性问题的三次大讨论取得的重要进展,厘清了计量经济学探索客观经济世界过程的本质特征;进而针对应用研究中存在的滥用和错用现象,从计量经济模型的随机性设定、经济变量之间的因果关系识别以及模型的统计适切性评价等三个方面阐述计量经济学应用研究的可靠性来源。我国计量经济学的应用研究面临进一步提高可信性的重要问题,需要全面吸收和借鉴国际计量经济学界对于可信性问题的成果,改变研究模式和教学模式。

The Credibility Revolution in Applied Econometric Research

The credibility of applied econometrics is an important issue, and its core lies in the scientific combination of economic theory, statistics with mathematics. Based on the three focused discussions among econometricians, this paper clarifies the essence of the econometric way of exploring real economic world. In China, the abuse and misuse of econometric method in empirical work has attracted wide attention. This paper elaborates three sources of the credibility: the stochastic specification, identification of causal effect among economic variables and the evaluation on statistical adequacy. To improve the credibility of the applied econometric studies, we need to draw upon the achievements of the credibility revolution and change the modes of research and teaching.