Part III, Chapter XVII
INTERFERENCE TO RAISE WAGES IN PLACES AND OCCUPATIONS WHERE THEY ARE ALREADY FAIR
§ 1. IN the analysis of Chapter XIV. attention was confined to the effects of interference designed to force up wages in places and occupations where they are unfairly low to a rate equivalent to that paid for similar labour enjoying similar incidental advantages elsewhere. It is sometimes claimed, however, that wages which are low and fair, as well as those which are low and unfair, may often be forced up with advantage to the national dividend. We have now to inquire how far and in what circumstances this claim is valid.
§ 2. In his very interesting book, Wages in Practice and Theory, Mr. Rowe urges strongly that the forcing up of wage rates may stimulate employers to improve their methods of organisation and technique, not only where, as in the case described above on p. 562, exploitation is taking place, but throughout the general body of industry. The extent to which this kind of reaction occurs must depend on how far, in different industries, the best known practice is already adopted by the main body of the employers there and on the openings that there are for bettering that practice. But some favourable reaction may be looked for not infrequently. As a result of it, the marginal net product of any given quantity of labour of given capacity will be indirectly raised, and the national dividend pro tanto enhanced; just as it will be if increased wage rates lead to an improvement in the work-people's capacity in the manner described in the chapter that follows. This consideration creates some presumption in favour of a policy that would push wages upwards gradually and by small degrees, even where they are already fair and where there is no question of reactions on workpeople's capacity. In this chapter, however, reactions of this kind are left out of account, and our problem is studied on the assumption that employers' technique and methods of organisation are not appreciably affected by wage policy.
§ 3. First, as was implied in the course of Chapter XIV. § 2, fair wages at particular points may sometimes emerge as the result of a conflict and cancelling among two or more unfair elements. Thus in some place or occupation the value of the marginal net product of labour may be abnormally high because the number of workpeople there is kept down by custom or by heavy costs of entry. If this circumstance operated alone, the wage rate would be unfairly high. But it may also happen that these workpeople are out-bargained by their employers, and compelled to accept a wage less than the value of the marginal net product of their work. If this circumstance operated alone, the wage rate would be unfairly low. Conditions are possible in which these opposing tendencies exactly balance one another, so that the resultant wage stands precisely at the level of fairness. But this fairness embodies two elements of unfairness, interference with one of which would not benefit, while interference with the other would benefit, the national dividend. The interest of the dividend requires that the wage rate should not be fair, but should be put at a level that embodies the former of the two cancelling elements of unfairness. Interference, therefore, is desirable in spite of the fact that the wage is fair. It is obvious, however, that an exact cancelling of elements of unfairness that act in opposite senses is, in a high degree, improbable. Consequently, wage rates that, as wholes, are fair cannot reasonably be suspected, except on very special grounds, of embodying any element of unfairness. Hence the considerations set out in this paragraph, though they point to complications in certain sorts of unfair wages, are only of academic interest as regards fair wages.
§ 4. Secondly, in order that, in any particular place or occupation, a wage that is fair as compared with other places or occupations may also be right from the point of view of the national dividend, a certain definite condition must be fulfilled. This condition is that in places or occupations in general workpeople are receiving as wages the value of the marginal net product of their work. If they are receiving less than this, they are without the normal inducement to give as much work as the general interest demands. The pushing up of their wages to a level that equates demand price and supply price would lead to an increase in the size of the dividend more than sufficient to compensate them for their extra sacrifice of leisure. Now, when things have settled down in more or less stable conditions, the play of economic forces tends to secure that in industries in general wages do correspond to the value of the marginal net product of labour. But conditions are liable to change, on account, for example, of new mechanical discoveries, the accumulation of capital, the opening up of foreign trade, or an expansion in the supply of the substance used as money. Any one of these changes necessarily tends to raise the value (in money) of the marginal net product of labour throughout occupations generally. The old wage, therefore, though still fair, will, nevertheless, be too low. It is to the interest of the national dividend that all wages should be raised. If, however, fairness in every individual wage rate was regarded as a conclusive reason against altering it, this change could never come about. Suppose, for example, that wage rates over the whole of industry were settled by Boards of Conciliation and Arbitration, whether wholly voluntary or partly controlled by State authority; and that the principle which each of these Boards followed was that of making its own wage rate equal to that paid for similar work in other occupations. The result, in the face of changing general conditions, would be a complete impasse. In like manner, it is conceivable that workpeople might, even in stable conditions, have their wages "exploited" everywhere to exactly the same extent, and for this reason might be everywhere receiving less than the value of their marginal net product. Here again a rigid rule against interfering with fair rates would make any correction of the abuse impossible. Hence it follows that fairness in a wage rate must not be taken as a conclusive reason against interference to raise it.
§ 5. If it were the fact, either that workpeople were exploited over the general body of industry, or that, over the general body of industry, wages were settled by Joint Boards whose sole principle of action was to establish fairness, the results of the preceding section would be of great practical importance. There would be a wide field over which interference to raise wages that are already fair would benefit the national dividend. As a matter of fact, however, the general body of industry is not controlled by Joint Boards actuated by the single principle of fairness, and there is good reason to believe that the places and occupations in which work-people are exploited form but a small part of the whole. Hence these special reasons for interfering with wage rates on some occasions, in spite of their being "fair," are not of wide application. It must, however, be conceded that, when a large and sudden change, such as might be brought about by the issue of a large quantity of paper money, takes place in the money costs of living, money wages in general will not respond immediately, and, therefore, for a time real wages all round will tend to be too low. Hence a policy which only interferes to raise unfair wages will lose opportunities for speeding up the required adjustment. If an external authority could force wages in any industry or group of industries from that old level, which is at the moment still "fair," to the new level which will be "fair" in a short time, it would benefit the national dividend. It will, of course, be understood that, when the money cost of living undergoes a large and sudden change, the response on the side of money wages, which it is the function of the authority to expedite, is not the same in all circumstances. If a given rise in the money cost of living is a part of an equivalent rise in prices generally, caused by a banking or currency change, the proper response is a rise in money wages sufficient to make real wages equal to what they were before. If the given rise in the cost of living is a part of an equivalent rise in prices generally, caused by diminished facilities for production, such as might result from the destruction of real capital in war, the proper response is either no rise in money wages, or, at all events, a rise less than sufficient to establish the old real wage; for the real demand for labour will have fallen off. Again, if the given rise in the cost of living is the result of a real rise in the cost of producing the things on which wages are predominantly spent, unaccompanied by any similar rise in the real cost of producing other things, the proper response is, for the same reason, either no rise in money wages, or, at all events, a rise less than sufficient to establish the old real wage. These distinctions are not always perceived clearly. There is also another complication of a different kind. The "cost of living" is not a single definite thing, which necessarily changes in equal proportion for all classes of the population. It may reasonably be argued that, since better-to-do persons, accustomed to more varied purchases, have a greater power of substituting things that have risen less for things that have risen more in price than poorer persons have, the sort of price movement that took place during the war really raised the cost of living proportionately more the lower we descend in the scale of earnings. This consideration affords a defence for the practice adopted in the cost-of-living sliding scales, introduced during and immediately after the war, of making wages vary less than in proportion to the Board of Trade cost-of-living figure, except for persons in the lower wage-groups. Had wages been raised equally for all wage-earners, the relative position of better-paid workers, as against worse-paid workers, would, it might be argued, have been improved and not kept constant.
§ 6. We have now to consider a much broader claim. This is that, in any occupation where wages are low, whether or not they are fair (relatively to the degree of efficiency among the workers engaged in them), they ought to be raised far enough to yield a decent subsistence to the average worker; a "decent subsistence" for the average man and the average woman respectively being interpreted in the light of the fact that the former has, and the latter has not, to support a family. Some approach to a recognition of this claim was made in a modification introduced into the British Trades Boards Act of 1918, as against the original Act of 1909. Whereas under the original Act a condition precedent to the establishment of a Trade Board in an industry is that wages there are exceptionally low, under the later Act it is sufficient that the trade be unorganised and, therefore, likely to be in receipt of wages unduly low. Elsewhere the claim has been accepted explicitly. Thus the South Australian Industrial Arbitration Act 1912 provides that "the Court shall not have power to order or prescribe wages which do not secure to the employees affected a living wage. 'Living wage' means a sum sufficient for the normal and reasonable needs of the average employee living in the locality where the work under consideration is done or is being done." In the Western Australian Act it is laid down that "no minimum rate of wages or other remuneration shall be prescribed, which is not sufficient to enable the average worker to whom it applies to live in reasonable comfort, having regard to any domestic obligations to which such average worker would be ordinarily subject." In the New South Wales Act of 1918 these general rules are given a statistical interpretation. The Board of Trade is ordered, after public inquiry into the cost of living, to declare from year to year what shall be the living wages respectively of male and female adult employees (other than those who are abnormally inefficient) in the State or any defined area thereof, and no industrial agreement shall be entered into, and no award made, for wages lower than such living wages. In the Arbitration Court of the Australian Commonwealth Mr. Justice Higgins proceeded (as regards unskilled workers) on the same principles, seeking to establish basic wages dependent on the cost of living for a normal man presumed to be the head of a family of five persons and for a normal woman presumed to support herself alone. In most of the American States, which have minimum wage laws for women, the minimum wage is defined as one sufficient to supply the necessary cost of living and to maintain the employee in health.
§ 7. In popular discussion the issue which these proposals raise is sometimes blurred by a confused idea that a living wage implies, for workmen of normal capacity in any industry which enjoys it, a "living income." This, of course, is not so. A living wage, as ordinarily conceived, for a man, is a wage that will enable the person who receives it, if he has an average family to maintain and if he has average good fortune in the matter of sickness, to earn an income sufficient for a good life. But a rate of wages that will achieve this end in these conditions will not achieve it for a man with a family in excess of the average or subjected to an unusual amount of sickness. Nor can the "living wage" take account of the fact that some workpeople need to support parents who are past work as well as their own children, or of the further fact that the wives of some workpeople contribute nothing towards the family income, while those of others contribute largely. Moreover, a wage for the breadwinner, which would provide a "living" for his family at one stage of its growth, would be quite inadequate at another stage. This consideration is very important. Dr. Bowley, for example, on the basis of his investigations into the condition of the poor at Reading in 1912, estimated that the minimum expenditure necessary for the attainment of a reasonable standard of living "at marriage would be 16s. weekly and would rise gradually to about 25s. in five years and 28s. in ten years, provided that there were four children all surviving. It would remain at 28s. for another five years, and then fall back to 16s. as the children became self-supporting." For women workers the connection between living wage and living income is even more remote, in view of the great differences between the positions of women mainly supported by their husbands, self-supporting single women, and women who are themselves the principal bread-winners of a family. These various considerations taken together make it plain that the enforcement in any industry of a living wage, in any plausible sense of that term, would go a very little way towards ensuring a "living income" even to those workpeople who regularly received it. Our natural desire to ensure in every industry a living income is thus not really relevant to the "living wage." The policy of forcing up wage rates in industries employing workpeople of such a low grade that fair rates, as defined in Chapter XIV., are less than living rates—however we may choose to define these—has, therefore, to be considered separately on its own merits.
§ 8. An argument by which it is widely supported is as follows: "An industry, which uses up the human capital without replacing it, is not self-supporting and does positive harm to the community. When, therefore, a woman is partially maintained by some other source, such as by a father, husband, etc., the industry which employs her is being subsidised from these other sources to the extent by which her wages fall short of proper maintanance." In other words, to allow the low wages, which make this subsidisation necessary, to continue, is to allow a process to go on, by which productive power and, consequently, the national dividend of the future, are steadily eaten away. This argument is invalid. It depends upon an ambiguity in the phrase "use up." If the setting to work of people at some industry wears out and destroys productive powers, which, had they not been set to work in that industry, would be available to augment the national dividend, then the destruction of this productive power ought strictly to be debited against that industry. Its social net product falls short of its private net product to that extent. But there is no general presumption that an industry employing low-grade workers and paying them a wage equivalent to what they could obtain elsewhere is using up human capital in this sense. For, if it did not employ them, they would either be employed in another industry at the same wage or they would not be employed at all; and in either event there is no reason to suppose that their productive powers would be worn out any less soon. The industry, therefore, only uses up their productive powers in the sense of using or employing them, not in the sense of wearing them out. Hence there is no difference between the (marginal) social and the (marginal) private net product of the work done in it, and no damage is caused to the national dividend by its continuance. Nor is this conclusion affected by the fact—when it is a fact—that the workers engaged in the industry are "subsidised" from other sources; because, if they were not engaged in this industry, they would still have to be "subsidised" to at least as great an extent. It is true, as was argued at length in Chapter XIV., that, if an occupation or section of an occupation is maintaining itself in being by its power to pay to workpeople of a given grade of capacity less wages than such workpeople could, and, apart from the existence of that occupation or section of occupation, would earn elsewhere, then the continuance in existence of that occupation or section involves a waste of the resources of the community. Here there is true parasitism. The essence of it, however, lies in the fact that workers are paid less than they could and would earn elsewhere. When this is not happening there is no parasitism, even though workers are being paid much less than is required to maintain them in independent self-support. The thesis that industries which pay less than "fair wages" ought to be forbidden by law to do this, even though such prohibition involves their destruction, is quite different from, and lends no support to, the thesis that industries which pay less than a "living wage" to workpeople who are in fact worth, for all purposes, less than a living wage, ought to be subjected to a similar prohibition. This common argument, therefore, breaks down, and our problem must be studied without reference to it.
§ 9. We suppose that a particular industry is employing low-grade workers, and that these workers are being paid a wage, which, in view of their comparative inefficiency, is fair relatively to that paid in other industries; and we suppose further that equality of efficiency wages in this industry and elsewhere is accompanied both by equality in the values of marginal net products there and elsewhere and also by a general equality between wages and these values. In these conditions, if the wage rate in our industry is forced up, a strong inducement is offered to employers to distribute inferior workpeople away from the occupations in which they are specially privileged, leaving these occupations to be occupied exclusively by more capable men. For example, the establishment of different (real) trade union rates of wages in different towns is met, in the main, by the abler workpeople gravitating to the towns with higher real wages, just as the establishment of the "dockers' tanner" in 1889 was met by the substitution, in part, of strong immigrants from the country for the weaker men among the old dockers. If this class of reaction occurs, it will not really happen to any substantial extent that the workers aimed at are paid higher wages than before. All that will be accomplished will be a redistribution of workers of different grades between different occupations. Consequently, no significant effect, either favourable or adverse, will be produced upon the size of the national dividend. The attempt at interference in favour of particular workpeople will, in fact, be parried by evasion. Let us suppose, however, that for some reason this reshuffling of workpeople is impracticable. Then, unless, of course, the demand for labour is perfectly inelastic, it must happen that some labour is ejected from employment in the industry where wages have been raised, with the result that, according to circumstances, it is either not employed at all or is employed elsewhere in conditions such that the value of the net product of most of it is less than it was. That this must be so follows directly from the analysis sketched out in Chapter XIV. The inference is that, employers' technique being given, apart from reactions on workpeople's capacity, to force up wages to a "living" standard in an industry where the fair wage is less than a living wage must injure the national dividend.
§ 10. At first sight it is natural to suppose that the damage done will be roughly proportionate to the number of occupations to which the interference extends; so that the dividend will be reduced three times as much if a living wage is enforced in all of three similar industries, where the fair wage is less than a living wage, as it would be if only one of them was touched. This, however, is an underestimate. When labour—and, consequently, capital also—is driven out of one occupation, it moves, in the ordinary course, into others; and, though it produces there less than it was producing before, it still produces a good deal. If only a little labour and capital are thus sent seeking a new home, and there is a large field open to them, the new contribution of each unit will be worth very nearly as much as the old contribution. But, if, with a given field open to them, more labour and capital are sent to seek work, they will have to be pushed into less highly valued uses, and the new contribution made by each unit will be less. Hence a doubling or trebling of the amount of interference, and, consequently, of the quantity of labour and capital turned loose, will involve more than a doubling or treblin of the damage done to the dividend. This consideration must be borne in mind when any inference is attempted from Australian experience as to the probable effect of forcing up the wages of low-paid industries in this country. For "industrialism is relatively simple in form and limited in extent in the Australian colonies. Agriculture is the chief occupation, and this, being untouched by the arbitration laws, is a vent for any labour or capital driven out of the industries." In this country the proportionate part played by agriculture is enormously smaller, and, therefore, the vent available, if a wide-reaching policy of forcing up industrial wages were attempted, would be far less extensive. Moreover, it is practically certain that in the United Kingdom this policy could not be applied to industry without being applied to agriculture also. The danger to the national dividend would, therefore, be very much greater than Australian experience suggests.
§ 11. Up to this point we have tacitly assumed that the old-established practice of disregarding conjugal and family estate in fixing wages will be maintained. Of recent years, however, considerable attention has been paid to proposals for regulating wages on a family basis. These proposals involve a departure from the ideal of fair wages even larger than is involved in the policy of "a living wage" as described above. For they require that different men of equal quality shall be paid wages that differ in accordance with the number of children that they have. During the Great War the wages paid to soldiers (pay and maintenance allowances being counted together) were, in effect, regulated on this principle; and the war bonuses on account of the increase in the cost of living, that were paid to members of the police force, were also on a family basis. It has long been understood that a public authority, if it chooses, is free to adopt a plan of this kind for the payment of its own servants, but it has been felt that no such arrangement could be introduced into the general body of industry without causing whatever unemployment there is at any time to be concentrated upon men with large families. In the last few years, however, there has been a very marked movement in several countries towards what has come to be called the "family wage system." In Germany, owing to the serious fall in real wages, a man with a family cannot live on the ordinary rate. In consequence of this, "the system of paying extra allowances in respect of family is now very widespread. An analysis, from this point of view, of the terms of current collective agreements was recently undertaken by the German Ministry of Labour. It shows that, while there are scarcely any trades in which the family-wage principle is not applied to some extent, there are a number of important industries in which it is universally recognised, amongst the latter being coal-mining, mechanical engineering, textile and paper and cardboard manufacture." In France the system first began to assume importance in 1916, and has since developed so far that, by 1923, some 2½ million workers were affected by it. In separate industries "compensation funds" are established, to which the separate employers contribute in proportion to their wages bill, and from which the whole of the family allowances are paid. By this device, an employer makes the same total payment in respect of each worker employed by him whatever the size of his family, and there is, therefore, no inducement to anybody to engage single men in preference to married men. In Belgium, Holland and Austria this system has also had some vogue. It is easy to see that, on plans of this type, a given aggregate wage payment would yield a bigger direct return in social benefit than it would have done under a living-wage rule of the ordinary type. But the family wage, in any form, is open to a serious objection from another side. It necessarily implies the taxing of bachelors in order to provide a bounty for men with large families—a sort of bounty on parenthood. I do not now raise the question whether this is, in principle, a good or a bad thing; a question in connection with which it is relevant to observe that those bachelors who afterwards marry get compensation in later life for the damage they have suffered in youth. Plainly, however, different bachelors of equal earning power ought to be taxed equally. But, if the family-wage plan is adopted, those of them who happen to work in industries where the proportion of bachelors, as compared with family men, is small will be hit very much more severely than their confrères in industries where the proportion of bachelors is large. If taxation is to be raised for the purpose of giving bounties to the heads of large families, it would seem better to raise it through the ordinary machinery of taxation, rather than in this hidden and uneven way. This consideration is applicable to a family-wage system applied universally by State decree to all industries. As against such a system confined to particular industries there is the further objection that it must tend to fill those industries with married men with large families, and to leave there no bachelors to be taxed in their interest. In 1927 New South Wales passed a Family Endowment Act, which provides for a payment of 5s. per week to the mother for each child in families with incomes of less than the official living wage for a man and wife without children, plus £13 a year in respect of each child, and raises the necessary revenue by contributions from employers in all industries proportionate to their wages bill.