Geldtheorie und ökonometrische Großmodelle

Abstract

Monetary Theory and Large-Scale Econometric Models

In the past decade, a large number of macroeconomic, large-scale econometric models have been conceived in the U.S. A., which have had considerable influence on the analysis of economic development and the framing of economic policy. In this connection special interest has been shown of late in the question of what hypotheses on the transmission of monetary impulses to real economic magnitudes have been introduced into those models. This interest resulted above all from the fact that the forecasts made with the large-scale models were ın part rather unsatisfactory, and the economic policy strategies derived from them proved not very successful, especially unsatisfactory efficiency of the large-scale models is decisively attributable to the fact that in their construction more recent findings in monetary theory were not given sufficient consideration. The first part of the study depicts important theoretical characteristics of nine large-scale econometric models. All the models are conceived with a strong leaning towards Keynesian income-expenditure theory, i. e.,the nominal interest rates are assigned a central transmission function, while conceptions of quantity theory have been given hardly any consideration. This finds expression in some of the models primarily in the fact that the monetary base, currency in circulation, demand and time deposits are either not included in the model at all or are treated as endogenous variables, i.e., they are not assıgned any monetary function. Other models work with inadequately specified wealth calculations, or certain price levels and nominal wages of an exogenous character. In the second part the author ventilates the methodology of the applied econometric estimating procedures. In this connection he makes reference to the controversy between the proponents of the structural equations on which estimation is based in the large-scale models and the monetarists, who prefer direct estimations by way of equations in reduced form. In this context attention is drawn to the special problems which arise when it is desired to determine the relative efficiency of monetary or fiscal policy measures with the aid of one of the two methods. In the author’s view, a decisive point for the different assessment of this question is the correct determination of exogenous and endogenous variables, and adequate consideration of the important channels of action of monetary and fiscal policy. The third part shows that mispredictions and economic policy mistakes made in recent times are primarily due to the fact that faulty monetarytheory conceptions had a decisive influence on the construction of the largescale models. For example, the significance of changes in the growth of the quantity of money for the inflation rate was not recognized and the influence of inflation expectations on the level of the nominal interest rates was not taken into account. A further important aspect is the failure to distinguish between nominal and real values, and between the volume of money and the credit volume. So consideration of these problems seems absolutely necessary for the improvement of the forecasting efficiency of large-scale models.