ABSTRACT
This chapter is concerned with the problem of identification in the simultaneous equation model with purely random coefficients. One approach to identification is to impose zero restrictions, that is equating some elements of the matrix and to be zero. This approach is based on the fact that each structural equation of the complete system represents some hypothesis about the behaviour of certain groups of individuals or variables. It is rather obvious that all variables of the complete system will not necessarily be represented in each equation. To give the real flavour of a simultaneous equation model with random coefficients the chapter presents a version of Klein's model I. Only the variables in Klein's model which reflect consumer behaviour would appear in a consumption function and others would appear with zero coefficients. Another approach to identification is through the imposition of restrictions on the covariance matrix.
