Alternative Measures of Labor's Share

Abstract

One of the most fundamental problems in economics is the distribution of among the factors of production. As national is the sum of the factor incomes, it is natural that most empirical studies have used national as a base. Now that increasing portions of the earned by the factors are either taxed or not distributed, gross factor shares have decreasing relevance as measures of welfare. As the various factors of production do not benefit either in direct proportion to taxes paid, or immediately from undistributed income, there is a real need for an alternative measure of share which will reflect these items. Labor's share of disposable is such a measure. Recently Frane and Klein published their data on the distributive shares of disposable income.1 While these data were originally developed with the objective of providing better bases for the derivation of separate consumption functions, they also have the effect of opening a new area of distribution analysis. This new measure of share is compared with the more traditional labor's share of national income (Chart 1).2 It is apparent from the comparison that the cyclical and secular pattern of change in share is significantly affected by the particular statistical measure employed. Labor's share of national income, L/NI, rises considerably more than share of disposable income, L/DI, during the depression, and considerably less than L/DI during the decade of the forties. Furthermore, L/DI exhibits a rather marked upward trend not in evidence in L/NI. The different behavior of these two series attests to the importance of the various tax and undistributed items. The new measure has the advantage of permitting an analysis of the effects of the various governmental and business policies. For example, it would be possible to examine the impact of a change in the personal or corporate tax structure or a change in business' profit retention policies upon the shares of disposable income. Similarly, the effects of changes in contributions for social insurance, transfer payments, etc., could be studied. Although the basic purpose of this paper is to present, rather than account