Discussion Papers in

The value of selecting the best forecasting model as the basis for empirical economic policy ana-lysis is questioned. When no model coincides with the data generation process, non-causal statist-ical devices may provide the best available forecasts: examples from recent work include intercept corrections and differenced-data VARs. However, the resulting models need have no policy implic-ations. A ‘paradox ’ may result if their forecasts induce policy changes which can be used to improve the statistical forecast. This suggests correcting statistical forecasts by using the econometric model’s estimate of the ‘scenario ’ change. An application to UK consumers expenditure illustrates the ana-lysis.