THE EXTENT OF BENEFIT-INDUCED UNEMPLOYMENT IN GREAT BRITAIN: SOME NEW EVIDENCE 1

Abstract

THLE question of explaining the shift in the relationship between the level of unemployment and vacancies has been set in terms of a dichotomy. Some (Gujarati; Maki and Spindler) claim that it was due to shift in the labour supply function. Others (Taylor; Knight and Wilson) claim it was due to a change in the demand for labour function. The papers by Maki and Spindler, and by Knight and Wilson give some evidence in support of each hypothesis. Both hypotheses have also a measure of theoretical plausibility, based on changes in legislation, notably the Redundancy Payments Act (1965), the National Insurance Act (1966), and the introduction of Selective Employment Tax. These altered the incentives of both suppliers and demanders of labour. This comment is concerned with the paper by Maki and Spindler, which appeared in the November 1975 issue of this Journal. It was stimulated by dissatisfaction on theoretical grounds with the way Maki and Spindler had specified their equation.2 Their equation implies that the unemployment they believe to be induced by the payment of benefits can be simply added on to frictional, structural, and cyclical components of unemployment. The latter is approximated by a measure of GNP adjusted for trend, both current and lagged. There are two objections to this. First it is not clear how the estimating equation, which is presumably a reduced form derived from supply and demand functions, is derived from the theoretical framework laid down by Grubel and Maki. Secondly, a literal reading of their estimated equation is that the more generous unemployment benefits are the greater will GNP be for a given level of unemployment. This seems to imply that generous unemployment benefits make the economy more 'efficient' in terms of output per head. Indeed, it might well be read as evidence of the labourdishoarding hypothesis. Thus it is not really a convincing test of the 'labour-supply' argument.