An investigation into the stability of the Phillips curve for the United Kingdom in the period 1950-70

Abstract

The study begins with a survey of some of the theoretical work on the wage-change equation, followed by a complementary summary of some of the empirical evidence on the determinants of wage inflation in the United Kingdom for the post war sample period. Estimates are then presented of the wage-change equation for the United Kingdom for the chosen (1950-70) sample period, using annual data, in order to explore the sensitivity of the results to the use of alternative wage- rate and unemployment series, and to the use of alternative definitions of the time rate of change variables. The main enquiry of the study centres on the use of the unemployment rate as an efficient proxy for the degree of excess demand for labour in the theory of the Phillips curve. It is possible that the observed instability of the curve may reflect the fact that the reported unemployment rate has become increasingly inefficient as such a proxy. An important source of error in the reported unemployment statistics is due to the non-registration by 'secondary' workers in the labour force. Estimates are presented of some 'adjusted' unemployment series which take into account the cyclical nature of this non-registration; and which are worked up from an analysis of participation rates in the United Kingdom. These estimates are carried over into an analysis of the 'mapping' from excess demand to unemployment where the objective is to identify and measure changes in the level of 'maladjustment' unemployment (the level associated with zero aggregate excess demand for labour). Estimates are obtained of maladjustment unemployment, and these are used to further correct the unemployment statistics for changes in 'maladjustment' unemployment. Finally this corrected unemployment series is incorporated in the aggregate wage-change equation to re-examine the statistical significance of the Phillips relation.