Similar difficulties confront judgments, based on evidence regarding business performances, of market power in product markets. Repeatedly, in the administration of the antitrust laws, the courts have wisely refused to answer the question whether the prices—or some other aspects of performance—of a combination are "unreasonable," by some test of what would be reasonable under competitive conditions, and have found violation in the mere existence of the combination. And, in so far as the courts have tended to move away from "abuse of power" and toward the existence of "power itself" as evidence of monopolizing or attempting to monopolize in cases involving large firms, the tests of market power have tended to emphasize structural rather than performance considerations. If the degree of monopoly possessed by a firm is to be estimated by comparing the prices, output, investment and profits of this firm with what these prices, output, investment or profits would be if the firm were subject to competitive restraints, two major difficulties arise. First, there is the question of standards; are the restraints to be those associated with pure competition or with some sort of "workable" competition, and, if the latter, what sort? Second, there is the problem of isolating the effect of market power on the prices, output, investment and profits, under observation, from other influences. The study of business performance has its uses in estimates of market power in conjunction with structural evidence but only in rather special situations can performance tests alone yield unambiguous findings.
So far as I can see, the same difficulties plague attempts to estimate the market power of unions by means of observations of union performance. Again there is the question of standards. Are we comparing the behavior of union wage rates with the assumed behavior of wage rates in a purely competitive labor market which is apparently what Machlup has in mind? Or is the standard of comparison the assumed behavior of wage rates in the absence of unionization, which is apparently what Reynolds and various other people consider to be appropriate? Secondly, assuming we have chosen our standard, will we find it possible to isolate statistically the influence on wage rates, or other dimensions of performance, of union power from all the other influences at work in a changing economy? A failure to establish empirically a clear connection between unionization and the terms of the wage bargain does not, to my mind, dispose of the question of labor monopoly.
In product markets, it is much easier to assemble information relevant to the market power of a firm by considering the limitations imposed by the firm's position in the market than it is by observations of the firm's performance, and I suspect this is true of the market power of unions. Needless to say, it is not at all easy, in either case, to evaluate this information. If we are to consider the area of freedom open to the union in wage negotiations as well as the limitations imposed by the external market environment, we must presumably start with the product market which defines the employers with whom, and the number of jobs with respect to which, the union will desire to negotiate. Unless all the em
ployers in a well-defined product market are included, the union's area of freedom is bound to be severely circumscribed by the product substitution of non-union for union output. Given complete control of the jobs in a well-defined product market, the union may be able to increase its market power by setting limits to the introduction of labor-saving technological changes or by increasing the number of jobs by "featherbedding" operations. There may also be opportunities of taking advantage of demand elasticities in the sale of the product by controlling or influencing output and price. The union might be said to occupy its market fully when all opportunities of improving wages, hours, and working conditions, within the unavoidable limits imposed by the elasticity of product demands and unalterable production functions lie within its control.
The union in the course of acquiring its market position may find it necessary to engage in organizing strikes and secondary boycotts; to press for closed shops; to absorb "independent businessmen-workers" into the union or drive them out of business; to insist on the employment of non-working standby crews, and do many other things designed ultimately to improve wages, hours, and working conditions. All or most of these are "well-established practices" of trade unions and Lester admonishes us, "Merely to condemn as 'monopoly' almost every well-established practice of trade-unions serves, therefore, to confuse rather than to shed light on, the significant issues." I agree that to condemn these practices as monopolistic is wrong since condemnation implies a judgment based on some public interest standard. But to analyze these practices in relation to the market power or degree of monopoly achieved or achievable by unions seems to me not only desirable but necessary. Needless to say, the conclusions of such analysis have no necessary relevance to a public interest finding of "unreasonable" power or "abuse of power." I take it for granted that all these and other union practices contribute—or are thought to contribute—to improvement of wages, hours and working conditions. Consequently I agree with Lester that there is no reason for selecting out certain of these practices, such as the closed shop or industry-wide bargaining, as monopolistic, to the exclusion of others. Certainly these particular practices may in various circumstances increase the degree of union power but so does any kind of labor organizing. The union is a monopolistic arrangement by definition and it may be reasonably assumed that a union will take such steps as it can to increase the degree of its monopoly control in order the better to perform the functions for which it was organized.