Union Influence and the Swedish Phillips Curve

Abstract

Almost without exception studies of the relationship between rates of wage inflation and unemployment levels have failed to specifically consider the influence of trade unions. The extent of organisation within the labour force is either ignored, or assumed to be reflected in the size of the estimated coefficients of the so-called Phillips curve. Jacobsson & Lindbeck (1969), in their study of Swedish data for 1955-67, are able to ignore the unionization problem because of their short sample period during which it could be safely assumed that labour force organisation and the structure of collective bargaining had remained relatively unchanged. The main problem with the Jacobsson-Lindbeck sample period-apart from the lack of degrees of freedom in their estimated equations-is the small degree of variation in their labour market variable, the level of unemployment. Despite the important role found for this variable-it is the only statistically significant variable in the JacobssonLindbeck wage equation-one has little confidence in an estimated Phillips curve for which the range of the unemployment variable is only from 1.1 % to 2.5 %. In this paper both the slope and position of the Phillips curve are assumed to be functions of the degree of labour force organisation (as measured by the extent of unionization). Equations are then estimated for the period 1911-72much longer than the Jacobsson-Lindbeck timespan. The model developed is aggregate in nature. No attempt is made to incorporate the ideas of, for example, Edgren, Faxen and Odhner (1969), concerning the and sheltered sectors of the economy. To incorporate such ideas would require the development of disaggregated models similar to those in the later paper of Jacobsson & Lindbeck (1971). The present model is not however, necessarily in conflict with the sectoral model. In the Edgren-Faxen-Odhner model it is the room available for wage and profit increases in the competitive sector which is determined by international price movements and domestic productivity increases. Actual wage increases in the competitive sector and hence in the economy as a whole are assumed to be the