The Phillips Relation: A Theoretical Explanation

relationship have been interpreted as suggesting the degree of unemployment and idle capacity requisite for the achievement of a chosen rate of change of prices, usually fixed at zero.1 It is apparent that the Phillips curve has been absorbed rather rapidly by the profession, a tribute both to the great insight furnished in Professor Phillips' original article [5] and also to the high quality of the subsequent literature. Obviously such an important idea as this, especially with the strong policy implications that seem to follow from it, needs careful examination. There is a tendency in some circles not merely to regard the Phillips relationship as a brilliant hypothesis to help explain the rate of change of prices but rather to view it as a well-established empirical fact with obvious and direct implications for policy. This being the case it is imperative that the concept be subject to intensive scrutiny; to date, practically all of the literature has been concerned with the statistical significance of the relationship. In this article we do not directly concern ourselves with this aspect of the problem, but seek to probe rather more deeply into the theoretical underpinnings of the relationship than has hitherto been done. First of all we shall seek to establish the relationship between the Phillips curve and conventional economic analysis. Then we shall show how a generally negative relationship between the level of unemploynent and the rate of wages may be derived from orthodox dynamic market analysis. However, we shall argue that the particular form of the lThough the Phillips relationship deals with the rate of change of money wages, in policy discussions it is the rate of change of prices that is generally thought of as being important, and indeed some contemporary elementary textbooks substitute prices for wages in their discussion of the relationship (cf. as an example, Lloyd G.