论文部分内容阅读
在本世纪二十年代计量经济学诞生之际,提出了二步方案,以便使计量经济学变成一门定量的科学。这二步是: (1) 将所有的经济法则以及相互之间的一些关系写成动态的数学模型(方程)。 (2) 用统计的方法与真实的数据进行比较,从而确定这些模型的一些未知系数。 这些假设从来也未受到基本的科学论证,因为对这些假设进行评论需要严格的系统理论,而在本世纪二十年代,系统理论尚未问世。 如今(1980年)系统理论已发展到一个相当高的水平,因而就有了可能着手研究计量经济学的基础了。当采用系统理论对计量经济的假设进行研究时,我们很明显地发现,上面所说的朴素的方案是有严重的及固有的缺点的。在本文中,我们提出关于这一结论的某些概念性的理由及数学事实,所以本文是一个纯粹的书面讨论。
At the birth of econometrics in the 1920s, a two-step approach was proposed to make econometrics a quantitative science. These two steps are: (1) Write all the laws of the economy and some relations between them into a dynamic mathematical model (equation). (2) Compare with the real data in a statistical way to determine some unknown coefficients of these models. These assumptions have never been scientifically substantiated, as the critique of these assumptions requires rigorous system theory, which was not yet available in the 1920s. Systematic theory has now reached a fairly high level (1980), so there is a good chance of embarking on a study of econometrics. When studying the hypothesis of econometrics using system theory, we have obviously found that the simple solutions mentioned above have serious and inherent shortcomings. In this article, we present some conceptual reasons and mathematical facts about this conclusion, so this article is a purely written discussion.