Forecasting Versus Policy Analysis with the ORANI Model

ORANI is a detailed general equilibrium model of the Australian economy. It has been applied many times by economists in universities, government departments and business in analyses of the effects on industries, occupational groups and regions of changes in policy variables (e.g., taxes and subsidies) and in other aspects of the economic environment (e.g., world commodity prices). These applications have been comparative static, i.e, they have been concerned with questions of how different the economy would be with and without the changes under investigation. They have not been concerned with forecasting the future state of the economy. More recently we have experimented with the model for forecasting. In a pilot exercise (Dixon, 1986), ORANI was used to forecast growth rates of industry outputs in Australia for the period 1985-1990. This revealed some problems inherent in the use of computable general equilibrium models for forecasting. In section 2, we describe the difference between comparative static analysis and forecasting, in general terms, and with specific reference to the ORANI model. We also discuss some computational difficulties which arise in applying ORANI to forecasting. in section 3 we present some numerical examples to supplement the theoretical material in section 2. Finally, section 4 contains some brief concluding comments on the strengths and limitations of ORANI as a forecasting device.