MACROECONOMIC FORECASTING AND MODELING

The authors discuss methods of macroeconomic forecasting, and explain why the focus of the paper is on an assessment of, and ways of improving, forecasting based on econometric systems. The role of leading indicators in forecasting has recently undergone a resurgence of interest, and is considered in Section II of the paper. The authors establish a framework for analyzing forecasts, and derive a taxonomy of sources of forecast error. In Section V, a research program into how econometrics might help is discussed. A number of factors are emphasized: the parameterisation; two aspects of parsimony in forecasting; and strategies for robustifying forecasts against structural change. All have in common the aim of mitigating the impact on forecasts of non-constancy. Section VI considers the implications of the arguments for policy analysis, and for the role of macro-economic forecasting in policy analysis. Section VII concludes. -from Authors