ERI Discussion Paper Series No.50
A REVISED VERSION OF THE EPA MODEL OF AUSTRALIA
(Introduction)
The first version of the EPA World Economic Model was completed in 1982, making it one of the first world economic models. IT covers the G7 countries (Japan, USA, Canada, West Germany, France, Italy and the UK) plus Australia and South Korea. An Australian model has been included since 1980, and has been developed and refined by Japanese and Australian economists working at the EPA. The Australian contributions include Challen and Fitzgerald (1981), Johnston (1982), Challen (1983), and Butlin (1985)
This is a convenient time to reexamine the Australian model for two main reasons.
First, the estimation period of the EPA World Economic model has recently been updated from 1974.11984.4 to 1980.11988.4. As would be expected, this change in the estimation period has resulted in a need to modify the specifications of some equations. Major work on that process has already taken place at the EPA, and one purpose of this review is to make an Australian contribution to that revision task.
Second, the process of financial market deregulation in Australian was largely completed in the early to mid1980s with the floating of the exchange rate, the abolition of exchange controls and the freeing up of the baking system. Now that the postfinancial market deregulation era is several years old, we have a better understanding of how to model it. This is important for the EPA Australian model, because this era dominates its new estimation period.
As is always the case with such reviews, the author's own prejudices are never far in the background. This comes through in particular approach taken to equation dynamics, the stress placed on diagnostic testing, and the emphasis on the theoretical foundations of producer behaviour. Similar prejudices can be found in some recent literature and in some in some existing macroeconometric models. At the same time, the discerning reader will filter these prejudices and reach his own conclusions on which recommendations to accept, and which to reject.
This review refers to three models. The first model is the EPA Australian model and is described as the "existing model". For the discussing of simulation properties, the "existing model" refers to the final version of the model estimated for the old estimation period (1974.11984.4). For the discussion of equation estimation, we refer to various versions of equations estimated over the estimation period (1980.11988.4). As already mentioned, a key purpose of this review is to help finalise the current set of tentative equations for this new estimation period. The second model is the Murphy model of Australian discussed in Murphy (1988a, 1988b, 1989, 1990). It is referred to as the "Murphy model". It was developed by the author, and is now used regularly to produce fiveyearahead quarterly forecasts, and for policy analysis. It is used frequently in review as a source of ideas for improving the existing model. The third model is the "revised model". This collects together the various suggestions for equation revisions in a complete model. This is convenient because it enables us to assess these ideas by comparing the simulation properties of the three models.
This paper is organized as follows. The second section is concerned with the data and statistical characteristics of the existing model, and the third section considers its broad macroeconomic structure. These two sections set the scene for an equationbyequation discussion of the model in the fourth section, where equations for the revised model are developed. The remainder of the paper is concerned with evaluation the revised model relative to paper is concerned with evaluating the revised model relative to both the existing model and the Murphy model. In the fifth section the various approaches to model evaluation are introduced, and results for those approaches which focus on equationbyequation evaluation are presented. Results for systembased approaches to model evaluation are reported in the next two sections. Specifically, the sixth section covers tracking of history, and the section covers some standard counterfactual experiments. An overall evaluation of the revised model is offered in the concluding section.
Structure of the whole text

full text(PDFFormat 321 KB)

page31. Introduction

page42. Data and Statistical Characteristics

page42.1 Model Size

page42.2 Estimation Period

page42.3 Equation Dynamics

page52.4 Equaion Testing

page62.5 Dummy Variables


page63. Broad Macroeconomic Aspects

page63.1 Financial Sector

page63.2 Export Market

page73.3 Producer Behaviour


page74. Individual Equations

page74.1 Private Consumption

page94.2 Private Investment

page114.3 Import Demand

page124.4 Labour Force

page124.5 Producer Behaviour: Employment & Price Deflator for Absorption

page174.6 Wages

page174.7 Price Deflators: Private Consumption,Private Investment & Government Expenditure

page204.8 Exports of Goods & Services: Demand & Supply

page214.9 General Government Expenditure

page224.10 Interest Payment on Government Debt

page234.11 Money Demand

page244.12 Interest rate Term Structure

page254.13 Net Incom Received from Foreigners

page274.14 Net Demand for Foreign Assets


page295. Model Evaluation

page316. System Properties: Tracking of History

page347. System Properties: Counterfactual Experiments

page357.1 Fiscal Shock

page367.2 Monetary Shock

page377.3 Foreign Price Shock

page387.4 Export Price Shock


page398. Concluding Comments

page40References

Appendices

page42A. Model Listing for the Revised Model

page53B. Variable Definitions

page54C. Results for Tracking of History

page57D. Results for Counterfactual Experiments
