THE IDENTIFICATION PROBLEM IN SYSTEMS NONLINEAR IN THE VARIABLES

作者: Bryan W. Brown

DOI: 10.2307/1912253

关键词:

摘要: This paper examines the identifiability of coefficients a single equation in simultaneous model which is nonlinear only variables. The concept this motivated and developed using closely related observational equivalence. framework then utilized to develop necessary sufficient conditions for when disturbances are required be independent exogenous approach recommended by Fisher shown yield but not identifiability. For several relatively common special cases found simplify familiar rank condition linear model. THE SIMULTANEOUS EQUATION MODEL that variables has enjoyed widespread application economics. Such models parameters typically seem as well, viewing equation. In many nonlinearity arises due endogenous entering different forms equations (logged unlogged form, example). macroeconometric includes real, nominal, price variables, nonlinearly related.2 Whatever reason it important determine under such can identified.

参考文章(3)
Franklin M. Fisher, IDENTIFIABILITY CRITERIA IN NONLINEAR SYSTEMS Econometrica. ,vol. 29, pp. 574- ,(1961) , 10.2307/1911805
Thomas J. Rothenberg, IDENTIFICATION IN PARAMETRIC MODELS Econometrica. ,vol. 39, pp. 577- 591 ,(1971) , 10.2307/1913267
Roger Bowden, THE THEORY OF PARAMETRIC IDENTIFICATION Econometrica. ,vol. 41, pp. 1069- 1074 ,(1973) , 10.2307/1914036