scientific article; zbMATH DE number 3886946

From MaRDI portal
Publication:3219650





This contribution is a revised and expanded version of a lecture originally given at the 4th World Congr. Econ. Soc., Aix-en-Provence, 1980 (see Zbl 0538.62095). The author's main point is that econometrics has, since its birth in the 1920's, overemphasized the classical statistical paradigm, and ignored the system aspects. It is better to overemphasize the latter, because there is no other way to be sure that what has been identified, is a real object and not merely an artifact created by the prejudices of the econometrician. While some earlier papers mainly have dealt with the general concept of modelling, the present paper mainly reexamines the notion of identifiability. After a redefinition of basic concepts, the author reanalyzes some well known econometric models from the system-theoretic viewpoint. Among them is the application of stochastic realization theory to the estimation of simultaneous equations. By touching the foundation of statistics, the paper raises issues far beyond the context of econometrics.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3219650)