ABSTRACT
It has already been shown that macro-econometric modelling methodologies permit a range of theoretical specifications to co-exist. This chapter develops this theme in two main ways. First, it examines the statistical properties of macro economic models in order to give some insight into the reasons for this interpretative flexibility. Second, it shows why neither econometric testing nor forecast performance appear to be decisive in proving or disproving an economic theory. 1 It therefore shows why it is difficult for both economists and policy makers to choose between economic models.
