ABSTRACT
Tha natural rate hypothesis has played a key role in the successful monetarist counter-revolution. Friedman (1968, 1975) and Phelps (1968) used the concept to argue that Keynesian demand management policies could not permanently alter the aggregate level of employment. The subsequent incorporation of the rational expectations mechanism in new classical macroeconomic models centred on the natural rate hypothesis (Lucas, 1972; Sargent, 1973, 1976; Sargent and Wallace, 1975; Barro, 1977, 1978) led to the more extreme proposition that systematic macroeconomic policies have no effect on the level of employment (and output) even in the short run. (See Tobin, 1980, for a critical review.)
These very strong conclusions resulted, quite naturally, in a number of investigations designed to test the validity of the natural rate hypothesis and corollary propositions. One class of test involves a time series methodology proposed by Granger (1969) and extended by Sims (1972, 1980) designed to extract empirically ‘causal’ relationships between key macroeconomic variables. These tests have been applied to a host of macroeconomic issues including, amongst others, the relationship between the money supply and inflation (Sargent and Wallace, 1973), the money supply and nominal income (Sims, 1972; Williams, Goodhart and Gowland, 1976; Feige and Pearce, 1979), wages and prices (Mehra, 1977), and employment and real wages (Neftci, 1978; Sargent, 1978; Kirkpatrick, 1981, 1982; Geary and Kennan, 1982).
