1.—We have next, therefore, to look at the causes which lead to the maintenance of relatively separate groups of wage earners, and then at the forces which govern their relative levels of earnings.

2.—First of all let us make clear some of the characteristics of the relatively separate groups of wage earners in the United States to-day. They vary greatly both in size and in kind. They are apt, however, to be conceived as similar because of the force of logic. It is not entirely satisfactory to classify them either as horizontal groups (having reference to their position in the scale of skill, or of society) or as vertical groups (having reference to their separation by industries). For the position of certain groups may be due both to the influence of those forces which bring about horizontal divisions, and of those which bring about vertical divisions. Such, for example, is the position of a craft which requires a measure of education and training which those who are placed by circumstances at the bottom of the industrial scale cannot easily get, and which besides it is difficult to enter because of trade union regulations.

Marshall has described the situation in England in terms that roughly fit the facts in the United States also. He suggests that the different occupations may be thought of "as resembling a long flight of steps of unequal breadth, some of them being so broad as to act as landing stages." "Or even better still," he writes, "we may picture to ourselves two flights of stairs, one representing the 'hard-handed industries' and the other 'the soft-handed industries'; because the vertical division between the two is in fact as broad and as clearly marked as the horizontal between any two grades."[30] The position of any relatively separate group is usually to be accounted for only as the result of many forces, each of which has some effect upon the rest. For example, barriers of custom or on vested right may limit the field of employment for women. This would tend to establish one level of earnings for women, and a different one for men. As a result women might find it harder to get the training necessary to enable them to compete with men. And so the interaction of causes would proceed.

So much in the way of preliminary remark upon the characteristics of the relatively separate groups of wage earners in the United States to-day.

3.—Among the causes which account for the existence of these groups there are some which if they stood alone would merely modify the applicability of the idea of a general rate of wages.

Such, for example, is the fact that the wage earner's knowledge of existing opportunities for employment is limited. Considerable discrepancies of wages for the same work may arise; although the facilities for the spread of information regarding wages has greatly improved, especially in the more skilled trades. Then there are, also, various expenses of removal, both material and psychological, such as are involved in the shifting of a family from the city in which it has long been established.[31] There are, also, the handicaps and hazards attached to the learning of a new job or trade even though the new job holds out hopes of considerably better wages than the old one. All such facts as these—for but a few examples have been chosen from among many—however, are reconcilable with the theory of a general rate of wages. They are but minor qualifications of a broad general principle. Other facts challenge that theory more seriously. They really do point to the existence of relatively separate groups of wage earners, each with an economic career somewhat independently determined.

First among them must be put the inequality of natural ability possessed by individuals, and the consequent fact that the numbers who possess the inborn capacity required for certain kinds of work is relatively small. It results from this limitation of the higher forms of natural ability, that the wages received for the more skilled forms of labor may be considerably higher than for the less skilled forms without such an increase of numbers in the more skilled groups as would bring down their wages to the general level. The competition for employment on the tasks demanding skill is limited; separate groups develop. It is impossible to tell the extent to which differences in inborn capacity would lead to the formation of relatively separate groups of labor, if all the other assumptions underlying the theory of a general rate of wages were fulfilled in fact. Prof. Taussig has expressed this well. "What would be the differences in wages, and to how great an extent would groups and classes persist, if all had the same opportunities, and if choice of occupation were in so far perfectly free? Would wages then differ only so far as they might be affected by attractiveness, risk, and other causes of equalizing variations? Would coarse manual labor, for instance, then receive a reward nearly as high as any other labor, nay, conceivably (since the work is dirty and disagreeable) higher than any other? Would the soft-handed occupations lose entirely the advantages in pay which they now commonly have? The answer must depend on our view as to the limitation of natural abilities. It is clear that some gifted individuals—a few men of science and letters, inventors and engineers, business men and lawyers, physicians and surgeons—would tower above their fellows, and would obtain in a competitive society unusual rewards. But would physicians as a class secure higher rewards than mechanics as a class? They would do so only if the faculties which a capable physician must possess are found among mankind in a limited degree. And mechanics, in turn, would receive wages higher than those of day laborers only if it proved that but a limited number possessed the qualities needed. On this crucial point, to repeat, we are unable to pronounce with certainty. What are the relative effects of nature and of nurture in bringing about the phenomena of social stratification, we cannot say."[32]

Next among the facts which account for the existence of relatively separate groups of wage earners are those which are usually summed up under the phrase inequality of opportunity. Equality of opportunity in the way of education and training, and in the way of healthy and strengthening environment would have to be assured before the theory of a general rate of wages could possibly apply. This equality of opportunity is not realized in the United States to-day.

The United States has been the scene of continuous and heavy immigration. The mass of this immigration entered into the field of unskilled labor. The great majority of these workers because of the partly unavoidable handicap of their strangeness, and their ignorance of American life, and because of their poor education, did not have equal chances with the older inhabitants to rise in the industrial scale. They could not possibly make the same use of the common opportunities—even if their natural ability were on a par with those of the older inhabitants. Furthermore, the rapid growth of our great cities and the accompanying social changes, the growth in the size of the average industrial enterprise, and the progress of standardization have all lessened equality of opportunity. The chances of the children born in the lowest industrial groups to discover and fairly test their natural abilities have declined in relation to the chances of the children more fortunately born. These conditions have certainly checked the working out of those forces on which the theory of a general rate of wages rests.

Thirdly, there is the fact that certain forms of work on which youthful labor is employed, give no preparation and training for the further stages of life and work; and these blind alley employments are filled by children born in the lowest industrial groups.

Then there are the barriers of different kinds to free movement throughout all parts of the field of employment. There are the barriers of sex which have added to the crowding of certain occupations and industrial grades. There are the barriers of race and religion, which have affected the flow of labor between different industries. Lastly, there is the barrier of color, which has prevented the negroes from developing their natural ability. These barriers may be well justified, in part or in whole, by other considerations. That question need not be considered here. But they certainly contribute to the formation of relatively separate groups of wage earners, with different levels of earnings.

4.—The existence and activities of labor unions are still another factor in the formation of relatively separate groups. In many cases labor organization tends to follow closely the lines of separation or unity established by the other causes of group separation or unity. There is often a tendency for a single union to include within its limits the whole of a group within which all the conditions underlying the idea of a general rate of wages are well fulfilled; or for various unions to merge or act together, if these conditions are well fulfilled between them. G. D. H. Cole has given a case in point. "Clearly the ease with which an industrial union can come into being depends upon the sharpness of the distinction between the skilled and unskilled in the industry concerned. Thus in the mining and textile industries, as we have already noted, there is no very sharp distinction between the two classes of workers. In mining, the boy who enters the pit has every chance of passing before many years have gone by into the ranks of the coal getters, who form the skilled section of the mining community. There is no sharp division or cleavage of interest between the main sections of the mining community. Promotion runs easily from one grade to another, and therefore, it is easier to realize a form of combination in which all the various sections are grouped together in a single industrial organization."[33]...

This tendency, however, has not been perfectly realized by any means. It often happens that the scope of a labor union will coincide with the underlying facts of unity at one time, but not permanently. The limits of particular trade unions have sometimes been set by an accident of time or place; by some episode in union history. The internal politics of the union movement has been the decisive factor in still other instances. Furthermore, industrial conditions are constantly changing and creating new lines of group separation or unity, which may vary from the lines of the existing labor unions.

Labor organization affects the formation of relatively separate groups of wage earners both directly and indirectly. First as to its direct influence. A labor union is a combination of a number of individuals, formed with the intention of advancing the material welfare of the group and for such wider purposes as the group may agree upon. The chief peaceful method of unionism is collective bargaining; its chief combative method is the strike. Labor unionism is a factor in the formation of relatively separate groups of wage earners, because each autonomous, or practically autonomous, trade union is a point of pressure upon the distributive mechanism. Each trade union strives to turn the balance of distribution in its own direction. This it does in a variety of ways.

It may by its wage demands test out the nature of the demand for the products of its labor. It strives to force the price of these products up to the point which seems to promise the greatest wage income for the group. It may by its pressure on the employer bring about a revision of productive methods. It seeks by its strength to secure that portion of the product which, in its view, goes to the strongest contender for it. Unions, indeed, sometimes strive to restrict the flow of labor into their craft or industry by deliberate regulation or silent obstruction. Such instances are less important than formerly in all probability. On occasion unions may even play a part in determining the field of employment for their members. Thus G. D. H. Cole points out that in England the trade unions do not recognize "differences between skilled and less skilled workers as demarcation disputes, and do not recognize the right of unskilled workers to raise such cases against skilled unions. In fact, the skilled unions virtually claim the right to do such work as they think fit, and so far as they can enforce their claim, to exclude the less skilled where they think fit."[34] Again unionism may indirectly through its wage policy cause a slowing up of recruiting of new men into the craft or industry. In short, by every means at its command, a union strives to assert the importance of its group as against other interests. Thus, in respect to the activities just described, unionism must be included among the influences which lead to the formation and maintenance of relatively separate groups of wage earners.

On the other hand, trade unionism in many indirect ways tends to have an effect in the opposite direction. By a constant adherence to certain broad policies, the trade union movement may contribute much to a realization of the conditions on which the idea of a general rate of wages is based. Such, for example, is the emphasis played by the trade union movement upon free and compulsory education, and the raising of the age of entry into industry. Such, also, is its advocacy of social legislation which is aimed to give more nearly equal opportunity to the lowest grades of industrial workers. Or, to take a third example, such is the result of the aid given by the skilled trade unions to the unskilled workers in their efforts to organize. Unionism works against the formation of relatively separate groups of wage earners to the extent that its activities contribute towards the achievement of equality of opportunity for all wage earners, and to the extent that the strong groups come to the assistance of the weaker.

5.—The main cause of the formation of relatively separate groups of wage earners, with different, though closely related levels of earnings have now been considered. As a result of these influences, it must be concluded that the determination of the wage level of each of the various groups of wage earners is a sufficiently independent process to make it necessary to account for it as such. The various groups of wage earners have relatively separate economic careers so to speak. The economic fortune of each group is not settled merely as part of one general process, though the economic fortunes of all are intimately connected. The wage situation is not to be explained as consisting of one basic level of wages with a series of equalizing differences; but rather as consisting of a series of wage levels, all of which are governed to a considerable extent by the same forces or conditions.[35]

6.—We can now pass on the final question which confronts us. How are the differences between the level of earnings of the relatively separate groups of wage earners determined?

The factors which determine the relative levels of earnings of each of the different groups may be put into two sets. First, those factors in regard to which each group stands alone and separate. Second, those which arise out of the dealings between the several groups.

"The relative plenty or scarcity of the different kinds of labor" falls in the first set. It will be remembered that this was among the three forces which, earlier in the book, were stated to be among the most constant and important in the determination of wages. The processes by and through which the facts of relative plenty or scarcity work out their effect in the distributive result have already been examined. If the numbers in any group of wage earners are high relative to the uses in which the employment of the members of that group results in a considerable addition to the product of market values, the wages of that group will be low, and vice versa. The need of the productive system for any kind of labor, relative to the supply available to fill that need is an important factor in determining the reward paid for that labor.

Furthermore, the statements in regard to the interactions to which the action of the factor of relative plenty or scarcity was subject, apply with equal force to the problem under discussion. Every human quality plays its part in the actual processes and negotiations by which the wages of the various groups of wage earners are settled. The outcome depends on many forces, some stable, some shifting and difficult to trace. Among those forces labor unionism, as the assertion of group economic power, holds a significant place.

In one respect, indeed, the previous analysis does not apply accurately to the question of different, though closely related wage levels. It is probable that the opportunities for the substitution of one type or group of labor for another type or group are more extensive and numerous than the opportunities for the substitution of one agent of production for another. And this fact limits the differences of wage levels that may arise between different kinds or groups of labor. For substitution of one type or group of labor for another is one of the ways in which changes in the relative plenty or scarcity of the different types or groups are brought about.

So much for the first set of forces—those in regard to which each group stands alone. The second set—those which arise out of the relationships between the various groups—remains for consideration.

Among these is the influence of customary wage relationships upon the course of wage movements within an industry, and to a lesser extent throughout industry. Because of the existence of vague customary relationships, wage movements affecting some groups or classes of labor are likely to stimulate similar movements among other groups; though it is plain that the efforts of different groups may not meet with equal success. This is well exemplified in the case of railway labor, of which Mr. Stockett has written, "Indeed there is every likelihood that the existence of a powerfully organized and highly paid group of labor in any industry—such as the engineers and conductors in railway transportation—far from being detrimental, may in the long run, be beneficial to the interests of the unorganized and low paid workmen. There is a tendency among the employees to keep a close watch on the wages paid to other groups of their fellow workmen, and the differential between their wage and that of some other grade of employment is jealously guarded. Thus on the railways, wage increases usually advance in cycles, an advance to engineers being followed at a close interval by an equivalent advance to firemen, conductors and trainmen. Existing differentials are more jealously maintained among the train service employees than among other railway workers, but that the latter do aim to maintain their relative level below the skilled groups is evidenced by the reference in arbitration proceedings to the advances made by the train service employees and by their claims to proportionate advances. Thus an increase in the wages of a highly paid group of employees, on account of this tendency to maintain existing differentials tends to put in motion a cycle of wage advances extending to all grades of labor."[36] Public opinion and public agencies of wage settlement have in the past been inclined to give support to the idea of the maintenance of customary relationships, even when the justification was flimsy.

Far more important is the factor of mutual aid between groups. For example, in pursuance of some general object skilled groups of labor have given support to minimum wage legislation for unskilled female labor; or again, such instances as the occurrence after the panic of 1907, when various organized groups of wage earners made common cause to resist wage reductions even for unskilled and unorganized labor. Such mutual aid plays its part in determining the wage levels of the different groups of wage earners.

This concludes the explanation of the forces which govern the relative wage levels of the separate groups or classes of labor. The actually existing differences of earnings between different groups of labor can only be explained by the combined influence of all the forces discussed.

7.—Differences in the levels of earnings of various groups of wage earners have been called "differentials." An effort has been made to explain their causes. Several practical conclusions, in regard to them, may be deduced from the preceding discussion.

Firstly, that these differentials (which may be measured by the differences between the average earnings of various occupations) result from, and in that sense represent, a large variety of actual forces; some of which can only be changed slowly and with much effort, as, for example, the relative plenty of the lowest grades of labor. As complete a knowledge as is obtainable of the various forces which produce these differentials is absolutely necessary to any project of wage regulation.

Secondly, although they represent a large variety of actual forces, it is misleading to apply such adjectives as "normal" or "natural" to them. For such adjectives inevitably suggest that the condition to which they are applied corresponds to a set of facts from which divergence can be only temporary, and is probably accidental. That, however, is not true in regard to the wage differentials which exist at any given time.

Thus, and thirdly, in any project of wage regulation, existing wage differentials can neither be accepted nor rejected blindly. A policy of wage settlement for industrial peace need not be based upon the acceptance and maintenance of all existing differentials. On the other hand, whatever revisions are undertaken should rest upon a knowledge of the forces which have established existing differentials. The policy of the South Australian Industrial Court, as expressed by its President, would seem to be a practical application of this view. To quote from one of his decisions: "PreËxisting or customary marginal differences are followed by this court as a prima facie rule, but the rule is only prima facie, and is subject to revision in the light of argument and evidence."[37]