Industrial Concentration and Interindustry Wage Determination

Abstract

THE hypothesis that firms with market power pay higher wages than competitive industries has implications that are important to many areas of policy. If firms with market power do pay higher wages, the excessive portion of those wages would be a rather large addition to the social costs of monopoly. In addition, wages that are inconsistent with labor market characteristics and not uniformly sensitive to business cycles will hamper the implementation of macroeconomic stabilization programs. Previous studies in this area have usually tested the market power hypothesis by determining the relationship between industry concentration and industry wages.' In addition, studies focusing upon other aspects of interindustry wages, such as the effect of unions2 and plant size3 have included concentration as an explanatory variable in their models. Results of these studies have differed with respect to their findings on the importance of concentration. Some studies found concentration to be important in the wage determination process while others did not. The research reported in this article indicates that the concentrationwage relationship changes significantly over the business cycle making cross sectional studies sensitive to the year used for the analysis. This finding is consistent with the results of models developed to describe the wage determination process over time. The method of analysis used here focuses upon the concentration-wage relationship at two points in the business cycle. A cross sectional test of a model of interindustry wage determinants in manufacturing is presented for the years 1958 and 1967. The results of these tests indicate that six independent variables can explain over 72% of the variation in wages found in each of the two test years, with each of the six coefficients statistically significant in both years. However, the impact of the independent variables change considerably over the business cycle. Of particular interest are findings that support the hypotheses (1) that concentration's effect upon wages appears to change over the business cycle, thus providing support for the spillover hypothesis and (2) that the wage-concentration relationship is not linear.