Regarding the Determinants of Union Wage Policy

Abstract

M. W. REDER offered his recent and interesting paper, Theory of Union Wage Policy, ' as a contribution to the debate between two schools of thought on the question of the of wage policy. One school, that characterized as emphasizing determinism, is represented in the paper by the writings of Prof. John T. Dunlop. The other, that characterized as emphasizing is represented in the paper by the writings of Prof. Arthur M. Ross. This paper may likewise be viewed as a contribution to the debate over the of wage policy. The issue with which we shall be primarily concerned is this: How could one empirically test the rival hypotheses stating that, typically, the important of wage are (i) political considerations, or (2) economic considerations? Reder, in commenting upon this aspect of the problem, says: Each argues that his statements apply to 'most cases' or to a class of significant cases, etc. Unfortunately, all the empirical evidence that is available consists of examples illustrative of one contention or another; but what is needed is a census . . . of wage decisions, so that some quantitative estimates might be made of the number and importance of the cases conforming to the various competing theories. 2 Unfortunately, Reder does not go on to tell us in what terms such a census should be drawn. That is, how, precisely, we would determine to which theory a given case conformed. In our opinion, the answer to this question is not so patent that it need not be discussed. It is our contention that no satisfactory empirical test can be devised unless careful attention is paid to the following two points: First, the concepts union wage policy and determinants (of wage policy) have as their empirical reference some behavior on the part of individuals. Unless the behavioral content of these concepts be specified unequivocally, we cannot know the nature of the empirical evidence that would be required to test the competing hypotheses.3 Second, the concept determinants (of wage policy) is susceptible to two widely different interpretations.4 On the one hand, it may refer to motivation. On the other hand, it may refer to certain aspects of the union's economic behavior other than motivation. The evidence necessary to test the competing hypotheses will depend upon which interpretation the investigator is putting upon the concept determinants. In what follows, we will attempt, in this order, to: (i) present the schematic paradigm in terms of which any particular behavioral model of wage should be cast; (2) elucidate the two interpretations of the concept determinants (of wage policy); (3) propose one particular behavioral model of the concept wage policy. Vis-'a-vis this model, will be understood in the nonmotivation sense. In the attempt to define a satisfactory model, a major difficulty arises from the consideration that the union's wage is probably determined by tactical as well as substantive considerations. A way around this difficulty is proposed. Finally (4), we shall consider, briefly, the analytical