Chapter 3 - Simultaneous equations models

Abstract

This chapter reviews a number of identification issues around the classical workhorse of econometrics, the simultaneous equations model in various forms. The influence of the birth of macro-economics and the formulation of the first macro-econometric models in the thirties inspired research on simultaneity, leading to both the development of estimation methods for these simultaneous models, and an insight into the characterization of the identification of such models. The chapter presents the relevant theory for the conventional simultaneous equations model. The identification problem posed by simultaneous equations models becomes more complicated when there are restrictions on the covariance matrix of the disturbances. As to measurement error in regression in general, it has been known for more than a century that the method of least squares produces a systematically underestimation of the effect of a regressor when that regressor is contaminated with measurement error. Simultaneous equations models with errors in variables were introduced by Goldberger.