Wage Policy and Distribution Theory

Abstract

Wage policy is a commonly accepted part of anti-inflation policy, and almost as commonly accepted is the view that in the rules on which wage policy is based the average productivity of labour should be a key variable. The role of average productivity in determining both average and sectoral wage shifts has been examined by Lancaster (1958)2 on the basis of an assumed economy in which labour is the only scarce factor of production. Not surprisingly average productivity is of major importance in such a system, although it is shown that sectoral shifts do depend on other factors such as labour mobility and substitution elasticities in product markets. It seems of some interest to enquire whether and to what extent average productivity remains relevant if production is assumed to take place under more general conditions than with one scarce productive factor. It is the purpose of this article to examine wage policy in a system which allows for substitution of factors, and for technical change, with several scarce factors. To understand this task more fully it is instructive briefly to survey contributions to the question of productivity-geared wage policies. This is done in Section I. In Section II the problem of extending the theory to cope with factor substitution and technical change is examined, and the conclusions are discussed in Section III. It emerges that wage movements may still be related to average productivity if factor endowments change, but at the expense of knowing factor-substitution elasticities. Moreover, the lower these elasticities the greater the overall and sectoral wage change for given changes in average productivity. If, in addition, technical change is introduced, some knowledge of its extent and bias must be available if wages are to be adjusted properly. Before beginning the survey there is a question of availability of information for wage adjustment that must be mentioned. This is fundamental to any wage policy. Given a formula for wage adjustment, the data which it requires must always be at least one period late. The wage is to be fixed at the beginning of a period, and data, at best, is available only up to that time. For this reason alone only approximate success is ever possible. In order to avoid becoming involved with this problem it will be assumed that the data to be used for the policy under discussion have been correctly forecast for the coming period.