论数理经济模型有别于计量经济模型——从关于柯布-道格拉斯生产函数的一个争论谈起

Discussion on the Difference between Mathematical Economic Models and Econometric Models: from a Controversy about the Cobb-Douglas Production Function

一些研究人员在进行计量经济分析的时候,直接把数理经济学里面的数学模型当作计量经济模型使用,并且把这类数学模型所设置的各种假定当作计量经济分析理所当然的前提。本文指出,数理经济模型和计量经济模型分别属于两门不同的学科,担负着不同的任务,不能把二者混为一谈。计量经济分析担负测算因果效应和预测两种不同的任务,在两种不同的任务下对计量经济模型的要求不同。因此强调指出,不论是测算因果效应还是预测,直接简单地把数理经济模型当作计量经济模型使用都是不妥当的。

Some researchers treat mathematical models of mathematical economics as econometric models, and use them directly, take all kinds of assumes in such mathematical model as econometric analysis premises for granted when they conduct an econometric analysis. This article points out that mathematical economic models and econometric models belong to two different disciplines, responsible for different tasks. So, they cannot be mixed up. This paper studies two different tasks in an econometric analysis, that is an estimated causal effect and prediction as well as different requirements in two different tasks on an econometric model. This article emphasizes that whether an estimated causal effect estimates or prediction, directly simply treats a mathematical economic model as the use of econometric models is not appropriate.