THE ECONOMETRICS OF MACROECONOMIC MODELLING

Money and inflation

The role of money in the inflation process is an old issue in macro­economics, yet money plays no essential part in the models appearing up to and including Chapter 7 of the book. In this chapter, we explore the relevance of monetary aggregates as explanatory variables for infla­tion. First, we derive money demand functions for the Euro area and, for Norway, and investigate whether these functions can be interpreted as inverted inflation equations. Second, we make a survey of inflation models that have been used in the recent past to analyse Euro area data. Moreover, we evaluate the models’ statistical properties and make fore­cast comparisons. Finally, we make a similar evaluation and comparison of Norwegian inflation models. The P*-model, which emphasises disequi - libria of monetary aggregates as the main driving force behind inflation, plays an important part in both cases. For the case of Norway, we also test an inflation equation—derived as the reduced form of the dynamic Incomplete Competition Model—for neglected monetary effects.

8.1 Introduction

The monetarist view of inflation—that inflation is always and everywhere a monetary phenomenon (Friedman 1963, p. 17)—runs contrary to the infla­tion models we have considered in the preceding chapters. Despite the notable differences that exist between them, they all reflect the view that inflation is best understood as reflecting imbalances in product and labour markets. This view is inconsistent with a simple quantity theory of inflation, but not with having excess demand for money as a source of inflation pressure.

In Section 8.2, we review briefly some results from the theory of money demand and show that this theory forms the basis for empirically stable money demand functions for the Euro area (Section 8.3) as well as for Norway (Section 8.4). Using criteria formulated by Hendry and Ericsson (1991),

we evaluate the claim that these stable money demand functions in reality are inverted inflation equations.

In Section 8.5, we survey models of inflation which have been recently used in the literature to explain Euro-area inflation. A reduced-form inflation equation, derived from the wage-price block of the macroeconometric model area wide model (AWM) of the European Central Bank, forms a baseline for comparing competing models of inflation. The P*-model of inflation suggested in Hallman et al. (1991) is one serious contender. The P*-model specifies a direct effect from the lagged price gap, defined as the lagged price level minus the long - run equilibrium price level which is implied by a long-run quantity equation. Trecroci and Vega (2002) and Gerlach and Svensson (2003) find support for this model formulation on Euro-area data, as do Todter and Reimers (1994) on German data.[66] The above models are estimated and evaluated against each other within a common framework in Section 8.6, along with a hybrid New Key­nesian Phillips curve model and the inflation equation derived from a version of the Incomplete Competition Model (ICM).

In Section 8.7, we present a reduced-form representation of the ICM inflation model for Norway, which (much in the same way as the AWM inflation model for the Euro-area data) forms a benchmark against which we evaluate several vari­ants of the P* - model, a hybrid Phillips curve and the inverted money demand inflation equation of Section 8.4. The focus remains on monetary aggregates: in Section 8.7.4 we test the robustness of the ICM inflation model for neglected monetary effects based on a sequence of omitted variable tests. Section 8.7.7 concludes and compares the findings on the two data sets.