Research Discussion Paper – RDP 7901 Estimation and Statistical Evaluation of an Economic Model


As this paper illustrates, econometric evaluation of macroeconometric models is difficult. This paper sets out the formal and informal procedures which are available, and applies most of them to a particular version of the RBA76 model. In an important sense, the work is a response to some of the questions asked at the 1977 Conference in Applied Economic Research. It complements similar econometric work on RBA76 done elsewhere, particularly by a group in the Australian Bureau of Statistics, and some work in progress in the Bank.

The author points out that, for the version of RBA76 examined here, various econometric criteria give conflicting answers. This is a common finding, and requires subjective decisions by the model-builder in choosing a preferred model. Lesley Moore – rightly for her purposes – has avoided making these decisions. As I read the evidence, however, the paper contributes substantially to the technical aspects of RBA76-style modelling. It provides results to be checked for later versions of the model and suggests several procedures to be used in future work.