RDP 1977-05: Modelling Monetary Disequilibrium 1. Introduction

In recent years the proposition that “money matters” – for output and prices in the closed economy, and for domestic variables and the balance of payments in the small open economy – has received growing support. One area of the relevant literature is that dealing with the effects of monetary disequilibrium, defined as the gap between the demand for and supply of money.^[1] The view that monetary disequilibrium has a major role in economic dynamics is based on the premise that money is a buffer stock for the private sector.

This paper examines the various channels for the effects of monetary disequilibrium – on consumption, prices, wages and capital flows – which have been detected in previous studies. These channels are examined in the consistent framework provided by the quarterly RBA76 Model of the Australian economy. The paper also discusses some evidence for a direct influence of excess money on the demand for labor and for business fixed capital. The existence of these channels, which have not been discussed in the earlier literature, is a logical consequence of the view that money is a buffer stock for firms as well as households.

Section 2 examines the arguments for monetary disequilibrium having a special role in economic dynamics. This section also sets out the channels for monetary disequilibrium which are examined in this paper and discusses briefly the related problem of estimating the demand function for money. Section 3 presents two alternative models: the first has only the more widely recognised consumption channel for the influence of excess money, while the second has no monetary disequilibrium effects. The properties of these models are compared with those of the model discussed in Section 2, using several tests including a comparison of the within sample dynamic simulation performance of the models. The response of each model to a sustained increase in real government spending is also examined.

The comparison of the models with and without the channels for the effects of monetary disequilibrium is of particular interest since the model which excludes this effect could be characterized as a simple version of many neo-Keynesian macroeconomic models. In this model, the effects of monetary and budgetary policy operate through changes in interest rates, and changes in government spending and tax rates, and the normal relative price and interest rate effects in goods and asset markets are represented.^[1]

In all three of the models considered in Section 2 and 3 the quantity of money is determined resiaually in the short run, in line with the view that money is a buffer stock. In Section 4 the results of estimating models in which the quantity of money is determined by the traditional convergent first order adjustment process are presented, to provide evidence on the relative ability of the two approaches to deal with the sizeable monetary fluctuations of recent years. Section 5 sums up.

It is worth noting at the outset that, for several guestions addressed in the paper, the estimation evidence is not fully consistent with the answers suggested by the examination of the simulation properties of alternative models. This illustrates the difficulties in the evaluation of econometric models which are discussed for example by Dhrymes et al (1972), and in view of the somewhat different answers given by different criteria, the reader will need to form his own judgements on some issues.

Footnotes

Much of this evidence is presented in the forthcoming volume on the work of the International Monetary Research Program. [1]

For some the major conceptual weakness in this version of the model is its omission of standard wealth effects. It should be noted, however, that the stock of business fixed capital plays an important role in the investment function, and that inventories of goods also play a role in the model. [2]