THE ECONOMETRICS OF MACROECONOMIC MODELLING
Summary of the findings—Norway vs. Euro area
The overall conclusion from the comparisons of inflation models for the Norwegian economy is that monetary measures do not play an important part in explaining and/or predicting Norwegian inflation. The preferred specifications of money demand do not include inflation as a significant explanatory
variable and hence the money demand equation cannot be interpreted as an inverted inflation equation. An attempt to model an inflation equation as an inverted money demand function shows clear signs of mis-specification and the MdInv model is demonstrated to be inferior to all other competitors based on in-sample evaluations as well as in forecasting (Figure 8.19). Also the P*-model, which embody several aggregates which monetarist theorists predict would explain inflation, fails to do so. Only when we augment the P*-model with the equilibrium-correction term for broad money, ecmmdt-1, does the model (P* - enhanced) appear to perform adequately in explaining and predicting the inflation series.
It is shown elsewhere in this book that the ICM gives a data congruent representation of Norwegian wage-price formation. In this chapter, it transpires that a reduced form representation of the ICM seems to perform better than the rival models it is compared to based on in-sample evaluations as well as forecasting. Moreover, there are no signs of the neglected monetary effects in the reduced form ICM inflation equation. In conclusion, the support for the ICM inflation model is much stronger in the case of the small open economy Norway, than in the case of a large aggregated economy, as is the Euro area. The AWM reduced form inflation equation emerges as the strongest contender amongst the Euro-area inflation equations and the enhanced P*-model is almost equally good.[91]
This page intentionally left blank
