Identification of the Long-Run and the Short-Run Structure: An Application to the ISLM Model

In this paper we discuss the problem of identification in a model with cointegration. It is pointed out that there is an identification problem for both long-run parameters and short-run parameters. The identification of the equations and the cointegrating relations is achieved by linear restrictions on the parameters and a criterion for a statistical model to be identifying is given. We also define empirical identification of an estimated structure. A switching algorithm for calculating the restricted parameters is proposed. The concepts are illustrated with an empirical analysis of the ISLM model using Australian monetary data.