Labour Unions and the Wage Structure: A General Equilibrium Approach

Abstract

In recent years a number of studies have been published which tend to show that wage differentials (in particular the college/high school wage differential) are greater than justified by the corresponding differences in productivity (see for example Eckaus (1964), Berg (1971), Taubman and Wales (1973) and Thurow (1975)). The evidence is tenuous, and has been vigorously controverted (e.g. Scoville (1966)). However the major resistance seems so have been on theoretical grounds. These studies directly contradict the predictions of Marshall's (1931, Bk. 6, Chs. 3-6) model of relative wages in a competitive economy based on comparative advantage, and this is taken as a-priori grounds for rejecting them (e.g. Layard and Psacharopoulos (1974)). This article is an attempt to explain the above-cited observations by a theoretical analysis of the long-run impact of labour unions on the structure of relative wages. It is generally agreed that in the short run labour unions can affect the real wages of their members. But in the long run, when the labour market has come into equilibrium, a worker of given characteristics should make the same amount whether or not he is employed on a unionized job. All the employees who originally helped organize the union (and who benefited therefrom by getting more than their market wage) will have retired, and the new employees will be the best the employer can hire for the wage offered.' But it does not follow that unions as a whole have not affected the wage structure (the relationship between the quality of a worker and his relative wage). Rather I shall show that in the long run unions increase the relative wages earned by all high quality workers (whether or not union members), and decrease the relative wages earned by low quality workers. However I shall not consider the impact of unionism on labour's share of national income, and will therefore not advance any opinion concerning the impact of unions on the real (as opposed to relative) wages of any quality of labour. The mechanism is that in the long run an employer, faced with having to pay a higher wage than he would voluntarily, will (ceteris paribus) upgrade his employment standards and attempt to hire the highest quality labour he can get for that wage (at least in the United States where most employers control the quality of labour they hire). Aggregated over many employers and unions, this represents a substantial shift in demand towards higher-quality labour. But (ignoring education, etc. for now) there are no more highquality workers for them to hire. One's natural economic intuition is that the wages of high quality labour will be bid up relative to the wages of low quality labour until the demand for high quality labour is back in balance with the supply. A good analogy is the effect of sulphur emission controls on the premium paid for lowsulphur oils. In response to the emission controls, each user attempts to buy lower-sulphur oil than previously. Of course in aggregate they can't upgrade the quality of the oils used