RDP 8302: Economic Forecasts and their Assessment Economic Forecasts and their Assessment

I.J. Macfarlane and J.R. Hawkins

August 1983

This paper attempts to do two things. First, it looks at the question of whether economic forecasts are accurate. Second, it looks at whether it is possible to distinguish between “good” and “bad” forecasters.[1] The points raised are illustrated by examples drawn from the economic literature on the subject, from some years of experience in economic forecasting and from a sample of twelve Australian forecasters over five years.

Footnote

Several important questions about economic forecasting are not addressed in this paper. The first is whether econometric models give more or less accurate forecasts than those based on judgement. For discussion of this point see Zarnowitz (1972), Christ (1975), Jonson and Norton (1980) and McNees (1981). The second question omitted is whether structural econometric models give better or worse forecasts than those based on sophisticated auto regressive methods (such as ARIMA processes). For a discussion of this issue see Cooper (1972), Nelson (1972) Christ (1975), Hirsch et. al. (1974), McCarthy et. al. (1974) and McNees (1981). [1]