8. “Exploitation” of Labor—“Monopsony”

IS THE argument for the tolerance of the strike threat that the free market price of labor is “unfair”? Or is it alleged that the market price would otherwise be forced below the free market value? It is often difficult to tell. However, it is the second allegation that we are to consider in this chapter. Can the threat or use of the strike prevent the workers’ exploitation? Can it somehow achieve the results that a competitive labor market would bring about?

We can get the issues into focus, I think, if we envisage all labor union bargaining in which the threat of a lockout or a strike is a factor as falling into three categories: (1) a struggle over the division of the spoils of monopolistic exploitation, consumers being the exploited party, with investors and the union protecting one another in this process but quarreling about the reward (see pp. 50 and 128); or (2) as an attempt by management to neutralize exploitation of consumers, when such exploitation via the strike threat happens to harm investors also; or (3) as an attempt by managements to exploit labor for the benefit of investors by the use of monopsonistic power, while that attempt is resisted via the strike threat. It is category (3) alone which is now under consideration.

“Exploitation” of the workers means, under the rigorous definition discussed in the previous chapter, forcing the terms of wage contracts below the “natural scarcity values” which the unhindered free market would have determined, thereby creating a “contrived plenitude” of labor. Is it possible, then, for the strike threat to assist the conclusion of such wage contracts as are likely to be offered when managements are seeking to maximize profits (which means “minimize losses”) in the absence of any effective power on their part to “shut in” and hence “exploit” labor? If managements are somehow able to suppress competing demands for the services of workers to whom wage offers are made, can strike power effectively countervail such subterfuges as are used?

If a strike threat or strike has had the effect of raising the wage rate in a particular occupation from below the free market level to that level (that is, to the natural scarcity level), the situation will be characterized by the condition that no worker can be found prepared to accept less than the wage rate established for the occupation concerned. If a union insists, in these circumstances, that no interlopers shall be allowed to work for less, that is proof that, in the union’s judgment, their forcing up the wage rate was not a countervailing of monopsonistic exploitation. The test of whether any wage rate is at or below its natural scarcity level is whether any other persons are prepared to perform identical work for that or a lower wage rate. As long as any such persons exist, the price of labor is above the free market level, and if a strike threat has enforced it, the effect must have been not to nullify attempted exploitation of the workers, but the exploitation of others—investors, or displaced workers, or excluded workers, or consumers.

I have never recognized in practice the existence of the conditions here specified as tests for the antimonopsonistic effects of strike-threat pressures. No case studies that I have seen have ever left an impression of the required conditions having been fulfilled. Hence for some readers the remainder of this chapter, as well as Chapter 9, will appear redundant. They may feel that in pages 99-123 I am discussing a chimera—a merely notional circumstance which cannot be found in the real world. Such readers can usefully jump to Chapter 10. But I am trying to reach readers who will be extremely loath to accept the convictions to which I myself have been led. I propose therefore to examine contrary arguments with care. The well-entrenched idea that union power is essential to correct monopsony or oligopsony needs patient examination.

We need not discuss at length the case of monopsonistic exploitation of wage earners which occurs incidentally, through monopolistic output restraints superimposed upon a previous condition of competitive supply. If the monopolist has no power to restrict his employees’ mobility, any “exploitation” is likely in practice to be negligible; and although the situation imagined is one which needs rectification in the collective interest, the defensible remedy is “antitrust,” not recourse to private duress.

Monopsonistic exploitative power directed effectively against labor would, as we have seen, be derived from the ability of managements (independently or in collusion) to use some device or stratagem in order to retain workers under their direction although they are remunerated at less than they could otherwise command elsewhere (which means at less than the “natural scarcity” value of their services). Expressed in the terminology of the previous chapter, exploitation would require that managements could “shut in” their employees and thereby create a “contrived labor plenitude.” Such a situation implies that the labor provided is so specialized that the workers have no alternatives of comparable value and that they have been somehow tricked (a) into training themselves for that particular specialization, or to become attached to the operation in question, or (b) into chaining themselves to it by agreeing to a “lock-in” contract.¹ The crucial point is that the workers concerned would have avoided that specialization or refused, say, “lock-in” terms of remuneration if they had foreseen the possibility of the exploitation. Any effective organization to shut in labor in this way would, I feel, always be at least discernible if not conspicuous.

The theoretical possibility of monopsonistic exploitation occurring cannot be challenged of course; and evidences of actual monopsonistic influences are not lacking. Nevertheless, I am inclined to accept H. C. Simons’ 1944 judgment that “monopsony in the labor market is . . . very unsubstantial or transitory.”² His conclusion was, I think, based on realistic observation and perception. The only convincing example of continuous “pure” labor-purchasing monopsony which I have found (apart from explicit and formal “lock-in contracts,” discussed on pp. 101-102) is that of the recruitment of Africans for the South African mines, which is organized through a centralized agency.³

If monopsonistic exploitation of labor exercised on behalf of investors were practically important, however, it would almost certainly manifest itself in the form of discrimination in wage rates offered according to management’s judgment of each individual worker’s alternatives.⁴ Some workers have a wider range of skills or other valuable attributes than others and under monopsony some might be paid more or less, not in relation to their special competence in their actual employment but in relation to what they are believed to be able to command elsewhere (see pp. 65-67, 163-166).

This is the one sense in which ceteris paribus the standard rate (“the rate for the job”) principle might be held to be defensible. A nondiscrimination rule could weaken or destroy the profitableness of monopsonistic discrimination where it might otherwise occur (in theory at least). The difficulty in practice, however, would be to distinguish the case in which any rule of uniformity would discriminate against a worker who has not only poorly paid alternatives but is also handicapped in other ways, for example, by being more expensive to employ, or by reason of being initially less efficient for the work offered.⁵

Among the other kinds of possibly exploitative action which fall under the monopsony heading and certainly do occur, are agreements or understandings not to “poach” labor—particularly labor of a special type—in order to keep down wage rates or salaries. But unless action of this type is organized openly and is accepted as desirable by all parties as a method of achieving some other agreed collective objective,⁶ it can, I guess, seldom have much success. It seems to me that as soon as any agreement to maintain a wage-rate ceiling, or otherwise to limit demand for any kind of labor, begins to have any effectiveness, all the phenomena of “labor shortage” must emerge. It will then be to the interests of both managements and individual workers, who can hardly long remain unaware of the “shortage,” to find subterfuges for getting around the agreement. I do not suggest, however, that we must necessarily leave the matter here. Some economists have alleged that collusive monopsonies to forego “pirating” have at times existed in named districts of the United States.⁷ They have charged that certain managements have agreed to recruit no worker unless the firm which last employed him has given its approval, and that “gentlemen’s agreements” not to steal one another’s employees have been arranged. But if the facts were as these economists allege, and as discernible to government authorities as they were to the economists, one wonders why antitrust officials did not at once step in. The allegations, which refer to a period of vigorous general antitrust enforcement, concern conduct of patent illegality by the “employers,” yet the supposed offenders were not prosecuted; and as their managements would presumably have denied the accusations, the evidence of the monopsony is, to say the least, far from convincing.

In so far as monopsony in the labor field does occur, however, the unfairly low wage rates due to contrived plenitude will be accompanied by wage rates in other labor markets which are set at “incidental contrived scarcity values,” that is, higher than they would otherwise have been, because some of the captive labor (which the theory assumes) is withheld from those markets. Hence although monopsonistic power may, theoretically, enable certain investors to benefit directly at the workers’ expense, we must not overlook the countervailing advantage to the workers in the markets in which labor is rendered relatively scarce, i.e., endowed with an “incidental contrived scarcity” value. Labor’s strongest case against such monopsony (if it has any importance at all) may be more its injustices to the “shut-in” workers, and its regressive consequences for workers generally as consumers, than any general redistribution of income which can be effected through it in investors’ favor.

The form of pure monopsony which is most likely to be occurring today is that due to “lock-in contracts.” A corporation may, as part of the remuneration it offers to attract and retain labor, include “employer’s contributions” to, say, a staff pension fund; or it may issue employee bonus shares, or other forms of conditionally owned capital to “loyal” personnel. Those who remain in the service of the corporation until normal retirement age (or for some other specified period) will retain these rights; but should they take up other employment before, they may stand to lose a large capital sum in accumulated benefits.

To see the problem in due perspective, let us remember that no person is forced to commit himself to such a contract. He may be held to have accepted, willingly and without duress, a condition which means that, for him to change jobs profitably in the future, the remuneration and prospects offered him elsewhere will have to be greater than his existing earnings by a sum at least equal to the accrued value of “the employer’s contribution.” If the system is abused, then the issue resembles that encountered in installment selling. The suggestion is that people can be tricked into tying themselves by contract because they do not realize the implications of their commitment. The remedy then requires careful specification of the conditions under which such contracts may be legally entered into. It seems to me that, as part of a wider plan for eliminating the strike-threat influence, there would be no harm in legislation to render void or unenforceable all “lock-in contracts” unless (a) they are genuinely providing an incentive for investment in human capital; or (b) they are an agreed means of reimbursing removal expenses advanced; or (c) they are necessary for the completion of a specific piece of work (that is, a bridge, or a round trip on an ocean vessel). Of these three (a) appears to be the most important. A contractual lock-in can be a bona fide device for conferring a “reasonable” measure of property on the investment a firm makes in imparting special skills and trade secrets to its personnel. In the absence of such a contract, those whose services have been rendered more valuable through expensive training could otherwise be “poached” by firms which have made no contribution to their training costs.

I am prepared to go further and concede that even when the “lock-in” contract possesses one or more of these attributes which render it beneficial, its duration should be limited—the limitation being no more and no less arbitrary than the limited period over which a patent monopoly is allowed to run. And it ought, I suggest, to be laid down that all “employers’ contributions” to pension funds and the like should, subject to the exceptions just mentioned, belong unconditionally to the wage earner (or salary earner). This will mean that if the worker accepts other employment, he may take with him the surrender value of any “employers’ contributions” (to pension funds or similar retirement savings plans).

In most cases firms seem to resort to these “lock-in” devices innocently enough, and often obviously beneficially for both parties as well as for the community. The aim is solely to attract and retain the workers whom managements find it profitable to hire, by what seems to them to be a simple and wholly legitimate type of offer. Nevertheless, the “lock-in contract” does differ in principle from other nonwage methods of retaining staff.⁸

A wholly different suggestion is that labor is monopsonistically exploitable, not by reason of its being locked in, or tricked into an exploitable specialization through some managerial stratagem, but by reason of some inherent immobility of labor. To judge this idea we must begin by recognizing that labor market freedom is not limited, restricted, or rendered imperfect because labor movement is not costless. Through the existence of obstacles to mobility (including man-made barriers), switching from one occupation to another, or moving from one district to another, involves costs; but only if these costs are man-made—the consequence of deliberate action by governments, by unions or by managements—can we say that the market is restrained. It may, of course, be profitable for the community, collectively or privately, to invest in the steps needed to reduce spatial mobility costs when they have a natural origin, as when a tunnel is cut through a mountain range; and any such reduction of costs in expanding, say, the area of competition, is widening the spatial range of coordination.⁹ But that process is not removing restraints. For instance, if the workers’ immobility over space is the result of his lack of capital to finance movement, relatively cheap labor in a district (whether the “plenitude” is contrived by lock-in or otherwise) will create an incentive for entrepreneurs outside the sheltered area to advance removal expenses in the form of loans or in return for appropriate “lock-in contracts” (under which an agreed minimum period of service is deemed to wipe off the debt).

However, because the labor employed in a particular area or in a particular trade would be more valuable if it were not for (a) the pecuniary cost of movement to other areas, or (b) the workers’ own inertias or preferences, or (c) the impossibility of circumventing union-imposed demarcations or like restraints, that does not imply that it ought to be remunerated at what it would be worth in the absence of those conditions; or that if it is not so remunerated, it is being “exploited;” or that the strike threat can nullify the “exploitation,” Labor is only being “exploited” when the lock-in barriers are man-made; and in practice, unless fraudulent enticement is present, the only clearly discernible man-made barriers which restrict labor mobility in this manner are created by labor unions themselves or imposed in response to their political pressures. This is a glaring truth which ought not to be a matter of controversy among scholars. It is confirmed by any dispassionate observation of the institutions of the modern industrial system. In offering jobs in any area in which certain labor happens to be plentiful and cheap, and certain occupations therefore profitable, yet no managerial restraints on mobility exist, it is absurd to suspect or accuse managements of aggravating the cheapness. On the contrary, the initiatives (which lead to offers of employment to workers displaced or excluded by contrived labor scarcity) mitigate the injustices. Indeed, they raise the incomes of those employed in the cheap labor areas.¹⁰

To get our logic into perspective, it is useful to enunciate a general principle about the determinants of the wages flow and its distribution. Within any area sheltered by economic distance, by human enertias and by union-imposed restraints, in the absence of any “shut-in” of labor contrived on behalf of investors, the flow of wages wilt be highest and the distribution of the flow will be most equitable, when every wage rate is fixed at the lowest level necessary to retain or attract labor for each activity judged to be profitable. I stress that word “profitable.”¹¹ The ideal will be most closely approached (a) the more successfully any labor which may have been “underpaid” for any reason can break through natural or man-made barriers to an occupation in which its remuneration is higher, and (b) the more successfully such transfers can dissolve privileges, that is, eliminate such “overpayment” of some as may be causing others to be “underpaid.”

The term “overpayment” in this context can be defined as “remunerated higher than is compatible with ideal resource use, maximization of the wages flow and distributive justice.” Under this definition, labor will be “overpaid” when it is benefiting from some removable obstacle to equality of opportunity. The test of whether labor is “overpaid” in any undertaking is whether additional workers, technically qualified to do the work (or potentially qualified) in the judgment of managements, would find it profitable to accept employment in the undertaking at less than the wage rates ruling,¹² in the absence of legally imposed barriers, or barriers enforced through unions. Within any sheltered area there might be no “underpaid” or “overpaid” labor in relation to that area. But in relation to a wider area, if the shelter were the consequence of deliberate restrictive action, labor could be “underpaid” or “overpaid” respectively according to whether the restraints imposed were confining labor to the sheltered area or keeping the competition of interlopers out.

In the absence of deliberate restraints, wage rates will tend to rise in any expanding industry or firm by reason of inelasticity of labor supply.¹³ There is no exploitation of investors involved in such a case. Similarly, there is no exploitation of labor when, because there happens to be inelasticity of labor supply in a declining industry or occupation, the wage rates at which such workers as choose to remain in the industry¹⁴ can escape displacement are forced down. The downward wage-rate adjustments they must accept to retain employment are in no sense a consequence of any monopsonistic power. They are the kind of pricing needed to soothe the pains of recoordination.¹⁵

Consider, for instance, the rapid decline of laundry work through the competition of washing machines and laundromats. This may well have meant that many laundry employees found themselves caught in that occupation. The owners of the laundries were unable to prevent their more versatile and enterprising workers from leaving, in spite of laundrymen’s unions having been unable to force money wage-rate increases similar to those which inflation was permitting elsewhere. It seems almost absurd to suppose that monopsonistic exploitation was a factor assisting the survival of laundries in these circumstances. But lack of mobility on the part of some of their employees, especially the older ones—victims of their own inertia—may well have helped the survival of laundries. In doing so, it would have contributed to an orderly transition to a new division of labor, with a minimum disturbance of established expectations. Hence we must be careful not to attribute to abuse of monopsony the consequences of lack of enterprise on the part of many of the older laundrymen; and we must remember that the relative fall in rates of remuneration in the laundry industry did not aggravate but rather softened the harsh effects due to change confronted with very human inertias.

It is now possible to enunciate a corollary of the general principle stated above (see p. 103), a corollary which may have seemed outrageous to many readers if it had been put forward earlier. If labor is cheap partly because it is unable to move from a firm, activity, occupation or area except at prohibitive costs of movements, but entrepreneurs are in no way to blame for these costs, the susceptibility of labor to monopsonistic exploitation is not enhanced by the immobility. Let us consider the possibility under “natural monopsony,” that is, in the case in which there is no collusion and hence no visible intention to exploit, which is the usual signal for antitrust to intervene. Can it be argued that, because antitrust protection is ruled out, labor is forced to rely on strike-threat defense?

A distinction must be made between “natural monopoly” and “natural monopsony.” We get “natural monopoly” when economies of scale are so important that only one firm can operate economically in a particular activity in a particular area. Now such a concentration of economic power vis-à-vis consumers does not automatically create monopsony vis-à-vis labor and suppliers of complementary resources;¹⁶ and if it does, it does not necessarily confer the ability to exploit. “Natural monopsony” requires that there shall be one firm only that offers a particular type of employment in a sheltered area. And under such monopsony, exploitation of labor still requires some device which either lures workers into locking-in themselves to employment in that undertaking or locks them in directly.

It is useful to consider the case in which the product of the monopsonistic undertaking, which we can assume constitutes the only output of the district, is sold competitively and mainly outside the sheltered area (for example, in world markets). Under our assumptions, if the demand schedule for the end-product rises, the price of this product will rise but the “natural scarcity level” of wage rates will not rise until the demand schedule has risen sufficiently to make it profitable to recruit labor from outside the area.

When such recruitment occurs for the first time it will, of course, make it essential to meet the higher wage costs ruling in competitive labor markets outside (as well as any costs of movement inwards). If, when that situation is reached, the rule of nondiscrimination is enforced (see pp. 164 el seq), it will suddenly become necessary to raise wage rates in the sheltered industry by a large jump, to bring the remuneration of the original workers in the sheltered area into equivalence with the remuneration needed to attract the newcomers. The original workers will then experience a sudden great increase in the value of their services. This big windfall gain to labor would not mean the ending of previous exploitation. It would represent either the consequence of a switch of consumer preference toward the product, or the consequence of the growth of outputs in noncompeting fields generally (through which real demands for products as a whole would be rising). These two factors together could build up a condition at which, owing to a rise in demand for the product in question, consumers (the ultimate “employers”) are all at once offering a large addition to the wage rates of labor in the area.

It is important to notice that, up to the time when recruitment from outside has become profitable, all increased revenues due to rising demand for the output will (in the absence of the strike threat or wage fixing by law) accrue to the owners of the undertaking. This is not because of some special attribute of services rendered by assets, but because the owners of assets are the residual claimants on the value of the product. If it had been a wise division of risk-taking for labor to have accepted the residual claim (renting the site, building and machinery and paying interest needed to finance inventories and work in progress—the possibility discussed in Chapter 6) the whole of the additional revenues would have accrued to labor; for in such a case the workers are entrepreneurs.

But suppose that, irrespective of any change in demand for the product, one or more other entrepreneurs perceived that the cheap labor of the sheltered area could be still more profitably used in producing commodities or services which did not compete in markets for the monopsonists’ product but did compete for labor and locally provided materials.¹⁷ Their intervention would bid up labor’s remuneration there and reduce the original monopsonist’s yield. Exploitation would occur if the original monopsonist could insert an obstacle between the workers in the sheltered area and any potential new investors who might wish to bid for their services. In other words, if some law-conferred privilege or some tolerated abuse of the pricing system conferred power on the monopsonist to block freedom of access between factors of production, exploitation could certainly happen.

Let us ask ourselves why labor in the sheltered area might happen to be initially cheap (in relation to the earning power of similar workers in other areas). One possibility is that the workers there had agreed, quite freely, to accept employment under what amounted to lock-in conditions under a long-term contract. (We are here excluding the possibility that they were fraudulently lured into accepting such a contract.) Whenever workers conclude a long-term contract of that kind, they are acting as entrepreneurs. They may gain or lose from the commitment they enter into. Should it subsequently turn out that their contract is unfavorable, that no more implies their exploitation than the opposite would imply exploitation of investors. But the workers in the area might have been cheap for quite different reasons, such as a differential birth rate in the past which has left an exceptionally large number of persons of conventional working age in relation to the resources of the district. Or the reason might have been a decline in demand for the output of a former industry. Through all such possibilities, workers in a sheltered area might have poor employment alternatives.

Let us now assume for simplicity that the exceptional cheapness of labor in that sheltered area happened to be the critical factor in causing investors, in what was to become a natural monopsony, to risk locating their plant there. In doing so the investors in the new enterprise (or rather the managers on their behalf) closed no existing employment outlets except by offering the workers better terms. They must have raised the natural scarcity value of those employed. Moreover, an expectation that countervailing increases in labor costs would not have to follow forecast increases in demand (and rising end-product prices) may well have created an additional investors’ incentive. Indeed, that very expectation could have been mainly responsible for the enterprise being located where it was—in the sheltered, cheap labor district.¹⁸

But suppose there had been several different groups of investors competing for the cheap labor stock postulated, instead of one monopsonistic corporation, would not the remuneration offered then have had to be higher? The answer is, no—not unless we can assume (a) that prospective yields in the area were then somehow greater in the aggregate so that entrepreneurs found it profitable to bid more against one another for labor in the sheltered market; or (b) that prospective yields in the aggregate being assumed the same, one or more competitors decided to attempt “predatory buying” (see p. 122) with a view to rendering the field unprofitable for the rest, that is, by bidding up wage rates so as to corner the labor supply in order then to exploit consumers.

The natural scarcity value of local labor is determined by its alternatives, but these alternatives are at once improved if entrepreneurial interlopers, who judge prospective yields from employing the local labor to be favorable, intervene with better offers. But because the district is sheltered by economic distance, it is possible that the wage rates which the new, monopsonistic undertaking can initially offer are lower than wage rates for similar work elsewhere by an amount equal to the costs of movement. Since, however, the wage rates originally offered represent natural scarcity values, the monopsonist will only be exploiting labor if by some stratagem he can pay less than this. And to do so he must devise some shut-in arrangement which renders the alternative still less favorable to the workers.

It must be borne in mind, however, that the specially cheap labor available for the “naturally monopsonistic” undertaking (specially cheap to the extent of labor mobility costs) is initially equally available for all potential investors to bid for, before the enterprise is established. Hence the actual wage rates it is found necessary to offer (in order to attract the workers from their previous occupations) are determined, so to speak, in an environment of potential competition. Stockholders in the corporation can hardly be imagined as the only capital owners whose managements realize that there are profit advantages in bringing more productive work to any cheap labor region. If other investors—at least within the same country—refrain from intervening, it must be because they feel that the lowly paid labor of the region is not cheap enough, presumably because they cannot envisage equally productive ways (or still more productive ways) of using that labor. It follows that the “natural monopsony” we are considering arises simply because whoever gets in first with one big dominating plant, excludes others from a possible big profit gain. The actual promoters are those who happen to perceive before others that, at the estimated labor costs necessary to persuade the required number of workers to accept the new jobs, their proposition justifies the risk. Therefore, if no actual additional¹⁹ bidding for labor against the monopsonist occurs, that will be due to the fact that at the new wage rates established, further investment in the area is rendered unprofitable (possibly through the additional costs of overcoming the workers’ inertia).

In a free labor market, the services of workers everywhere are, so to speak, continuously up for auction as their existing contracts of employment (which may often be by the week or even by the day) expire. Only if some potential investors are prevented from engaging in unhampered competitive recruitment of labor in an area is there monopsonistic exploitation.

Once the monopsonistic undertaking has been established and is operating, and on the assumption that the demand schedule for the product remains unchanged, only potential or actual competition from other entrepreneurs, either offering employment outside the sheltered area (in which case their offers would have to be sufficient to compensate for the high migration costs) or inside (which would mean that new productive potentialities inside the area had come to light) can raise the natural scarcity value of the given labor stock. But there is no exploitation simply because, until that happens, the labor is relatively cheap in comparison with, say, other areas of the country concerned.

Monopsonistic exploitation cannot occur, then, unless the monopsonist can contrive a “shut-in” of the complementary services he purchases. Any measure of monopsony which may exist in the absence of some deliberately contrived “shut-in” is nonexploitative. Admittedly, the imaginary “naturally monopsonistic” condition we have considered constitutes one of the situations in which the strike threat can be used to seize income at investors’ expense without direct harm to consumers or to displaced or excluded workers. But as will be further explained in Chapter 10, the prospect of costs being raised in that manner, even when monopsony is present, must deter some investment in those forms of assets which most successfully multiply the flow of wages and income.

Strike-threat action might of course be used to lessen the costs of labor mobility. Some interunion fights may have had this effect. But unless the unions can use their organization somehow to lower or demolish man-made barriers to the better utilization of labor and of the tools with which labor cooperates, they are powerless in this respect. As I insisted above (see p. 103), in practice these barriers, and the economic injustices and wastes they cause, are created almost entirely by the very union policies which purport to be fighting for justice in distribution; and the power of the unions to erect the barriers is dependent ultimately upon their use of the strike threat. Local discrepancies in real wage rates must exist for “natural” reasons as long as areas are insulated by economic distance from other areas. But the more burdensome discrepancies are caused when spheres of employment are hedged in by the standard rate, demarcations, occupational licensing, apprenticeship rules, color bars, and the like.

As an example we can consider a type of labor “shut-out” which exists in the United States where the unions have been strong enough to secure occupational licensing. When this has happened, we find the unions opposing reciprocity in respect of qualifications (determined for the different states individually). This is quite rational, given the sheer selfishness which is regarded as ethically acceptable in labor unions because, as Rottenberg points out, “If reciprocity prevailed, people would enter through the widest door.”²⁰ But from the standpoint of the workers in any state, they are denied access to opportunities, not by any management-contrived arrangements, but by union-contrived arrangements. I have found no parallel in the whole literature of the labor movement of a man-made barrier to mobility of labor which is equally obvious;²¹ and this barrier is imposed, not in the interests of rapacious capitalists, but in the interests of labor and professional organizations.

The reader must be reminded that the circumstances I postulated on page 105 in order to illustrate “natural monopsony” were highly abstract and notional. Nothing resembling that model of one firm producing a particular product in a sheltered district and selling it almost entirely outside is, I believe, to be found anywhere in practice. True “natural monopolies” are in practice almost always undertakings in which the output cannot be rendered apart from the plant; they essentially serve therefore the areas in which they are physically situated, and this fact means that there are people in other occupations in their district earning through the production of noncompeting things,²² whom it is their (the “natural monopolies”) purpose to serve. In every realistic example of “natural monopoly” undertakings of which I can imagine, there will be highly competitive natural markets for the materials and the labor they employ. The adjective “natural” here means that competitive conditions would exist in the absence of legal or collusive restraints imposed in those markets. Normally we may expect a wide range of occupations to be competing for nearly al! kinds of labor—firstly, for juveniles to choose from on entering the labor market, and secondly, for labor of different degrees of versatility which may be attracted from other firms, activities, occupations or areas at appropriate wage rates, even if the costs of labor movement to and from other areas happen to be high.

We must be careful, then, not to exaggerate the importance in the real world of some of the circumstances we have been imagining (for purely ex-positional purposes). Thus it is doubtful whether even the largest cartels can often, say, profitably force down the price of the materials they use. They may certainly benefit from purchasing economies which are sometimes achievable through large, or guaranteed continuous orders, or through shrewd investment in a developing supply source. That is something quite different. But not only does a monster corporation or a great cartel usually express a very small part of the total demand for any raw materials, but it is to its advantage that a continuing supply shall be forthcoming. In a rather different way the same is true vis-à-vis labor.

Even the most careful empirical studies throw little certain light on the problem. For instance, J. W. Garbarino²³ and H. M. Levinson²⁴ have found that what they call “noncompetitive” industries (assuming that the concentration of output into relatively few firms implies less than normal competitiveness) have raised wage-rates by more than the average. Unfortunately, the data they present are no proof of the absence of monopsony; for the differential rate of wage-rate increases may obviously have been the consequence of strike-threat pressures possibly facilitated by joint monopoly. M. W. Reder, who has criticized these contributions on other grounds, suggests that “large firms are more dilatory about correcting overpayment . . .” (when wage rates get out of line) “than correcting underpayment”; and he finds that “large and profitable firms do tend to pay more at any one time than could be explained by the competitive hypothesis.” He finds further that “high concentration ratios . . . measured by (say) the percentage of the industry’s employment concentrated in the four or eight largest firms” are associated with “high wages at a given moment of time.”²⁵ Such findings tend, superficially at any rate, to contradict the notion that monopsonistic exploitation of the workers is correlated with scale of operations or ownership. But using data presented by G. Warren Nutter, for the extent of monopoly, Reder shows that the figures indicate “a slight (negligible) tendency for a decrease in monopoly to accompany an increase in wages. . .,”²⁶ and he refers to similar conclusions reached for the Canadian economy by D. Schwartzman. This is exactly what we should expect in the light of the analysis presented in this and the previous chapter. When exploitation of consumers by, say, manufacturers is weakened, their demand schedules for labor and other productive services will rise and, except under the wholly unrealistic assumptions we made on the preceding pages, the increased bidding for labor will raise wage rates in the industry.

To assess the practical importance of the monopsony argument for acquiescence in the strike-threat system, we should remind ourselves that in the United States and Britain this system originally emerged in spheres in which the entrepreneurial undertakings were small, and where the relations between separate undertakings could only be described as highly competitive. Obviously, then, at that epoch, resort to the private use of coercive power by the unions can hardly have been a countervailing response to monopsony. Moreover, even today a very large area of union operation is in fields in which the typical undertaking is small relative to the market, and where competition in the sale of output is not subject to any obvious restraints. Consider, for instance, building, printing, textiles, footwear, transport and mining in the United States. Antitrust has kept these industries competitive (superficially considered at any rate); and certainly one finds nothing resembling cartels or explicit price agreements, for output agreements and any other apparatus of collusion are illegal. Hence in such instances there can hardly be any question of the unions, through the use of the strike threat, having been able to act as an antidote to monopsony. Yet unions are powerful in these industries (for instance, the Teamsters). It follows that any valid argument for permitting the strike-threat influence in wage determination has to be developed so as to cover circumstances in which neither monopsony nor joint monopoly (to which I refer in the following paragraph) can be alleged. “Countervailing power” is clearly irrelevant.

Empirical evidence suggests that it is to the advantage, not the disadvantage, of a union that the undertaking with which it is “bargaining” shall be organized monopolistically. This is because its power to exploit the consumer can be magnified through joint monopoly (see pp. 72-73). But it is true also that strike-threat action on a union’s part is essential for any actual exercise of this joint power. For example, if a union is confronted with an association of undertakings which arranges the pricing of output, only the use of strike-threat power can insure its members a share in the spoils. There is no reason why this should occur automatically unless the associated managements regard collusion with the union as a means of perpetuating their ability to exploit consumers. Managements have, on occasion, taken the initiative and offered “sweetheart contracts” to raise the remuneration of their workers in return for union protection against interlopers. And there are other possibilities. For instance, a union’s initiative might force what had been competing undertakings to adopt collusion in order themselves to share some part of the spoils; or, as it would probably appear to the managements, in order the better to off-load increased costs on the public. In all of these cases, it is the public which is exploited; yet it is a form of exploitation which would be impossible if the strike-threat power were absent.

There is one further possibility. Given the legality of strike-threat pressures, managements may have found an incentive to rely upon “reasonableness” (so-called) in competition for labor; for increased reliance upon an “oligopsonistic” situation may be expected to appear in that light. It seems as though union pressures cause competing undertakings to act less competitively (that is, having a reduced incentive to substitute the least-cost method of achieving outputs for the community’s benefit). And in the extreme case the effect may be, as Albert A. Rees has pointed out, “to create effective cartels in the product market.”²⁷ Rees suggests that this may be the position in the building trades and the local service industries in the United States, but I feel that he is thinking rather of the cases discussed in the paragraph above.

In a strike-free system, an incentive to oligopsonistic “understandings” about wage rates might well arise among managements; but a strong counterincentive to bid against the rest for underpriced labor would remain. Competition could hardly be effectively restrained, one feels, except under some explicit agreement. Managers are like the rest of us, reluctant to pay more or to accept less. But that will not prevent some from anticipating future labor scarcity and, possibly through the offer of fringe benefits and stress on superior prospects, outbidding the rest. The withheld bidding of firms which hold off vigorous recruitment (via the offer of higher wage rates) until their managements feel certain that a real labor scarcity is developing, is likely to become active bidding through the fear that rivals may get in first. And this likelihood will be increased if the unions begin to act as entrepreneurs, which I am about to suggest should be their chief function when the strike-threat system is abandoned. The oligopsony possibility will be discussed in the next chapter. I shall try to show that it is unimportant.

I have been stressing the apparently rare occurrence, the inherent instability, and the ephemeral nature of labor-purchasing monopsony in the absence of conspicious “shut-in” power. I have shown the improbability of more than negligible abuse, but I cannot deny the possibility of serious exploitation in particular cases. Where managements are in a position to restrain labor mobility, “shut-in” arrangements may be discernible. There are, however, three general methods of insuring that such theoretically conceivable abuse from monopsonistic exploitation shall be avoided. Among these methods, the use of the strike threat is not included.

(1) To bring collusive monopsonistic exploitation of labor explicitly within the scope of antitrust (or other legislation aimed at eliminating socially indefensible use of private power to contrive scarcities and plenitudes);

(2) To forbid “lock-in” contracts with labor except under the conditions specified above (101-102):

(3) To encourage labor unions to take up cases of monopsonistic abuse in the purchasing of labor and:

(a) to act entrepreneurially on their members’ behalf, firstly, as an employment agency, and secondly to finance or otherwise assist the transfer of “underpaid” workers to the better-paid jobs available, the potential existence of which alone can justify the term “underpaid”;

(b) to initiate antitrust (or similar) action against what they believe to be deliberate collusive monopsonies, or “shut-in” devices, including nonpoaching agreements;

(c) to initiate similar proceedings against firms which are held to be exploiting consumers, and thereby incidentally acting against the interests of the union’s members as wage earners;²⁸

(d) to initiate private proceedings against firms alleged to be using unjustified “lock-in” devices.

But in cases (b) and (d) the unions surely would have to establish initially the absence of potential interlopers. That is, they would have to demonstrate that no outside workers were prepared to accept less than the existing, allegedly monopsonistic ally determined wage rates being paid in the firm or firms challenged. (See pp. 98-99 above.)

The typical laborer or artisan is said to be badly informed about alternatives. It is contended that he usually begins to try to become informed about the market only when he is laid off. There seems to be considerable scope, therefore, for the sort of expert guidance recommended under 3 (a); and if the unions can play on the divergences of interest which exist under monopsony or oligopsony, and are also able to disclose employment opportunities outside of the area, they can reduce or wipe out the prospective profitableness of any attempt at monopsonistic exploitation.

I have treated respectfully the notion of managements using monopsonistic power to force down wage rates only because I do not expect all readers to share my judgment that such a possibility is of negligible practical importance. But I have shown also that in so far as evidence of monopsonistic abuse is forthcoming, recourse to the courts and not recourse to strike-threat power is the defensible remedy. “Antitrust” or its equivalent is the answer, not private coercion. Admittedly, if justice is not effectively and promptly obtainable through the courts in labor disputes, men may be expected to take private steps to protect themselves against invasion of their rights. In the sort of issues with which we are here concerned, it is exceptionally desirable that “the law’s delays” should be minimized. But the cost of achieving reform in this direction—possibly the establishment of special courts—would be negligible in comparison to the benefits of eliminating the injustices which are inevitable when wage rates are determined through any form of warfare.

To sum up. Demonstrable monopsonistic exploitation seems to be remarkably rare; yet the remedy we are asked to tolerate is a general and universal right to strike. Moreover, if the right to strike is defended because it can be used defensively, the fact remains that it can equally well be used aggressively. “Those who understand the economic, social and political implications of monopoly power,” says Fritz Machlup, “must deplore the lack of imagination and intelligence of a society which in the name of ‘equalization’ embarked on a policy of combatting occasional monopsony by creating more monopolies. The hope that they may neatly offset one another is plainly naive.”²⁹

There remains one aspect of the topic which requires mention. I am convinced that neither managements nor sole proprietors do typically act in the rapacious spirit which nearly all the textbooks in the labor economics field manage to imply. This is a conclusion reached after a long academic life devoted to the study of business administration in which I have relied on the recorded experience that constitutes the literature of this subject. The late Sumner Slichter (on the whole an apologist for the strike-threat system) admitted that managements tend to appear generous when it is at all possible and to resist demands for wage-rate increases or to call for cuts, only when things are going badly.³⁰ But when an undertaking is prospering in the sense that the business is expanding, managements are often forced to pay a wage premium to get the additional labor they require, even in what is regarded as a competitive labor market. And when they have had to accept a union-demanded standard rate, there still seems to have been a tendency toward what has been called “wage drift,” that is, payment of more than the standard rate when there is “excess demand.”³¹ What Slichter may really have been observing was managerial appeasement—the unwillingness to fight on consumers’ behalf, especially when confronted with expanding demands or when inflation is relied upon to validate any concessions.

Of course, salaried managements, as distinct from proprietors, have no right to be generous at stockholders’ expense. When they appear generous in wage negotiations, I judge it to be due to their trying to show that the offers they make (on behalf of the residual claimants) are highly favorable to the workers they wish to attract or to those they want to persuade not to strike. But the popular and propaganda-perpetuated stereotypes of avaricious investors and unscrupulous managements anxious to carve out careers by exploiting the workers, and held in check only by union power, are grotesque caricatures.

“I think the robbery of labor by capital is a humbug,” concluded Mr. Justice Oliver Wendell Holmes as he approached the end of his famous career.³² This chapter has tried to show that, even it it is not “humbug,” it is a delusion.

NOTES

¹ A “lock-in contract” is one in which the employee binds himself for a period, or under the terms of which he is subject to some penalty if he leaves before the expiry of a stipulated period, for example, the confiscation of deferred pay, profit-sharing rights or pension rights.

² H. C. Simons, Economic Policy for a Free Society (Chicago: University of Chicago Press, 1948), p. 129.

³ Whether there is indeed any effective monopsonistic exploitation in this case is, however, far from certain; for it is difficult to see anything resembling a “shut-in.” The number of South African workers in the mines has been declining (because better-paid or more attractive work is available elsewhere in the Republic) and more and more foreign Africans (migrants) are being employed.

⁴ Similar discrimination by monopolists against consumers, based on estimates of the urgency of their need for the product, is normally impracticable, although sometimes contracts can be entered into prior to investment in fixed resources under which some consumers or users of products agree to discrimination against them under long-term contracts—but in their own interests (see pp. 163-166).

⁵ Such a worker would be harmed by the enforcement of so-called “equal pay for equal work;” for it would disallow him the effective right of bargaining. Any acceptable nondiscrimination rule would have to guard against all the possibilities which, in practice, make “the rate for the job” the most effective discriminatory device that has ever been invented (as I shall be explaining in Chapter 12).

⁶ For example, in British professional football, maximum wage rates were (until recently) enforced as a means of ensuring that the wealthier clubs did not attract all the best players and thereby destroy spectator interest in the game by reason of contests becoming too one-sided. This is an interesting example of a general problem known as “externalities.”

⁷ For example, see A. Myers and W. R. Mac Laurin, The Movement of Factory Workers (1943), quoted by Fritz Machlup, The Political Economy of Monopoly (Baltimore: Johns Hopkins Press, 1952), pp. 353-4.

⁸ For instance, offering better terms than competitors in respect of provision of special “fringe benefits,” an attractive place of work, playing fields, sports clubs, and other “extramural” facilities and amenities.

⁹ In practice, when such natural barriers are broken down, the unions tend to substitute man-made barriers for them. For example, Phelps Brown, referring to “the improvement of communications exposing once sheltered local markets to the vicissitudes of wider competition, . . .” says that this phenomenon “heightened the need felt for a union.” E. H. Phelps Brown, The Economics of Labor (New Haven: Yale University Press, 1962), pp. 40-41.

¹⁰Incidentally, we cannot assume that undertakings which are viable because labor is plentiful and cheap where they are located, are able to earn a larger proportion of the value of output for investors than can be gained for them in similar activities in other parts.

¹¹ The word “profitable” is justified here only because monopsonistic exploitation on the part of managements is assumed (in this paragraph) not to exist.

¹² In other words, wage rates raised by any sort of compulsion other than the social discipline of the market, cause the labor they remunerate to be “overpaid” in the light of the definition given.

¹³ An apparent exception is discussed on p. 106.

¹⁴ The workers’ lack of versatility, ignorance, lack of enterprise, or simple preference—or the pecuniary costs of mobility—may determine such a choice.

¹⁵ If demand for the product is declining, ceteris paribus it will be to the interests of both investors and employed that wage rates shall fall relatively to wage rates in general until the more versatile of the workers have found better openings elsewhere, which may happen rapidly or slowly. In that way, not only may the burden of lay-off be mitigated, but the contribution of the industry to the source of demands in general will be maximized (although it will be declining).

¹⁶ The labor market in question may be sheltered (in that all entrepreneurs enjoy exceptionally plentiful and cheap labor) but competitive (in that there is a wide range of employments).

¹⁷ The entry of competing capital to supply the same kind of product is of course ruled out by the assumption of natural monopsony.

¹⁸ Actually, under the simplified situation I have envisaged for exposition purposes, it could be to the advantage of the workers in a sheltered area to be discriminated against by the monopsonist during a transitional period. Such discrimination might cause expansion of profitable outputs to occur sooner, as demand for the product was rising. Recruitment from outside at wage rates considerably higher than those ruling internally could make it possible thereby to raise the remuneration of existing employees, although not to the point at which equivalence of remuneration with the lucky newcomers had been established. If, however, anything distantly resembling this situation ever came into being (which I myself can hardly imagine) in a nonstrike regime, it could be a union function to agree to the suspension of any antidiscriminatory rule and allow its existing members, for their own benefit, to be paid less than newcomers (see pp. 163-167).

¹⁹ “Additional” to the bidding offered in the previous employments from which the recruits are attracted.

²⁰ Simon Rottenberg, “The Economics of Occupational Licensing,” Aspects of Labor Economics: A Conference of the Universities—National Bureau Committee for Economic Research (Princeton; Princeton University Press, 1962), p. 19.

²¹ I do not say “equally burdensome.” The most effective barrier to spatial or occupational labor mobility is enforcement of “the rate for the job” (see Chapter 12, passim).

²² “Noncompeting” in respect of outputs, not in respect of resources used in production.

²³ J. W. Garbarino, Quarterly Journal of Economics, May 1950, pp. 299-300.

²⁴ H. M. Levinson, Study Paper 4, Joint Economic Committee, U. S. Congress 1960, pp. 2-5.

²⁵ M. W. Reder, op. cit., pp. 285-6. Actually, there is some reason to believe that large firms shrewdly recognize the advantages of outbidding competitors and purchasing the cream of the labor supply. They may obtain thereby what Reder has called a “richer skill mix.” It is a way of purchasing labor’s inputs more cheaply, and a particular economy of scale. Another suggestion mentioned by Reder is that large undertakings “over-pay” for prestige reasons. But prestige contributes to “goodwill”; and “goodwill” is an item in a firm’s stock of assets.

²⁶ Ibid., pp. 286-287.

²⁷ Albert A. Rees, The Economics of Trade Unions (Chicago: University of Chicago Press, 1962), p. 84.

²⁸ If a union can prove that the consumer is being compelled to pay too much for a certain output, for the very reasons which are keeping the earnings of those who make it unjustly low, it will have an iron-clad case; and, incidentally, it will be fighting the case also for better remunerated opportunities for diverted capital resources.

²⁹ Fritz Machlup, Wage Determination and the Economics of Liberalism (Washington: U. S. Chamber of Commerce, 1947), p. 56.

³⁰ Sumner H. Slichter, “Notes on the Structure of Wages,” Review of Economics and Statistics, February 1950, pp. 81-91.

³¹ The term “excess demand” is a misleading way of describing the situation when there are more vacancies than unemployed in any field at current wage rates.

³² In a letter to Harold Laski, quoted by Helmut Schoeck, Envy—A Theory of Social Behavior (New York: Harcourt, Brace and World, 1970).