INTRODUCTION

Any attempt to evaluate Friedman's monetary theory must first come to grips with a number of apparent anomalies in his position. To begin with, there is the obvious difficulty that Friedman has never explicitly presented a complete theoretical framework or model. This feature of Friedman's work is generally accepted (Shaw, 1983:423). His theoretical framework appears to have wandered through numerous phases beginning and ending with an attempt to impose the properties of the traditional quantity theory of money on to a system of unbacked fiat money and credit. The story begins in traditional quantity theory fashion with support for 100 per cent reserve banking (Friedman, 1953c), and ends on the same note with a call for an overhaul of the Federal Reserve system that will render control of the monetary base effective (Friedman, 1985). But in between these expressions of support for the traditional quantity theory Friedman engaged in several theoretical detours, notably: (i) a restatement of the quantity theory in the Cambridge tradition as a theory of the demand for money (Friedman, 1956, 1959); (ii) a reinterpretation of the IS–LM framework to include permanent income and Fisher's distinction between real and nominal interest rates (Friedman, 1974); (iii) the analysis of the natural rate of unemployment and the vertical long-period Phillips curve (Friedman, 1975); and (iv) the outline of a dynamic analysis designed to explain the overshooting of the final equilibrium position by nominal income and prices (Friedman, 1974; Friedman and Schwartz, 1982:ch. 2; Taylor, 1976).