1. The Crucial Thesis

FOR THE sake of clarity, I shall begin by presenting my case in the simplest terms. In so doing, I run the risk of losing those readers who tend at first to disagree and then are unwilling to suspend judgment. But for the reader prepared to face apparent heresies, disclosure of my position at the outset will be the most fruitful procedure. I shall explain why I believe that coercively imposed wage rates, or more directly enforced restraints on entry to any occupation, cannot redistribute income from “capital” in general to “labor” in general.

A crucial concept must be defined first—“exploitation.” I define “exploitation” as any action taken, whether or not through discernible private coercion (collusion) or governmental coercion, which reduces the value of the property or income of another person, or prevents that value from rising as rapidly as it otherwise would, unless this effect is brought about through: (a) dissolving some privilege; or (b) substituting some cheaper method (labor-saving or capital-saving) of achieving any objective; or (c) expressing a change in consumers’ preference; or (d) democratically authorized taxation.

We shall be concerned almost entirely with the exploitation of workers by investors or of investors by workers. But, as we shall see, every restraint of the price mechanism and every restraint on consumer or entrepreneurial freedom may be held to “exploit” the community as a whole; and any such “exploitation” may be viewed as harming, more or less impartially, both investors in general and workers in general (a) as consumers and (b) as income receivers. For example, if the price of leather is forced up either through collusive action among the suppliers of hides or through a strike on the part of the workers employed in tanneries, the community as consumer will be worse off. Incomes will buy less. Whatever the effects upon the relative shares of workers and investors in the value of leather produced and sold, the economic impact upon the relative value of the shares of workers and investors in all other occupations is likely to be neutral, in the sense that both groups will be affected in roughly the same proportion.

The most vital economic decisions are “entrepreneurial” decisions: They are made within an institutional framework of custom, law and knowledge—the law requiring administration and enforcement by collective action—that is, by government. These decisions, which are being continuously made, are in every case concerned with retaining, replacing, accumulating, or decumulating the physical resources employed in various possible combinations of, in certain specific activities, together with the retaining, recruiting or displacing of labor in accordance with this process. Now, the use of capital equipment, materials and labor in any activity will not occur unless some entrepreneurial remuneration is deemed possible for every increment of resources invested in that activity. Thus a firm will not retain, replace or accumulate additional assets unless the prospective output values of each increment exceed the corresponding current and prospective input values by more than the rate of interest.

The size of a firm’s inputs—those of its assets and of the labor in which it invests—will be reduced if the predicted residue shrinks through labor costs having been raised by a strike, the threat of a strike, or a wage law. “Exploitable” resources will be neither retained, replaced nor attracted. The only investments which are “exploitable” by the strike or the strike threat, therefore, are those in which entrepreneurs have failed adequately to allow for the probable use of strike power or political action as determinants of labor costs.¹

Entrepreneurs know and make allowances for the use of the strike threat or similar power to fix wage rates. Hence in general there can be no “exploitation” of retained, replaced or accumulated capital, except on the assumption of entrepreneurial forecasting errors.

For parallel reasons, labor is unexploitable. Workers can be said to be “exploited” only if, after investments by them of time, effort and saving in specialized training, unforeseen monopsonistic² or oligopsonistic³ actions (by entrepreneurs) reduce their remuneration.

Hence, in a society where any form of entrepreneurial “exploitation” has been practiced and seems likely to be practiced again, the rate of hiring at any given wage by firms or industries will decline, possibly to the disadvantage of the specific investors.

Realistically, the probability of frequent monopsonistic exploitation of labor seems quite remote. For were it not for labor union demarcations and restrictions on entry most laborers (even highly skilled ones) would be almost as versatile as those in “put and carry” services. (There are, of course, some important exceptions.) Let us consider the extreme case of highly specialized, nonversatile labor. If we consider (a) the continuous replacement of personnel who gradually leave through competing employment openings, retirement and death, and (b) (in an expanding industry) the recruitment needed for expansion, we must recognize that the probability of the workers’ exploitation is remote. The observable labor turnover between firms suggests indeed that collusive action to reduce the price of labor has virtually never been regarded as profitable.⁴ (The necessary—quite unimportant—qualifications of this assertion will be expressed later. See Chapters 8 and 9.) But the principle is based on no inference from empirical evidence (although it does seem to be consistent with observed experience): The volume of labor which will come forward to be trained for or otherwise become attached to any occupation in which (in the light of past experience) labor appears liable to “exploitation,” will be reduced to the level at which the prospective long-term benefits are equated with the prospective benefits from training and employment in other occupations.

The reader will probably reach the conclusion (after reading Chapters 7 and 8) that where nonversatile plant and equipment are provided, the possibility of investors being exploited is greater. But even here we shall see that exploitation forecast is exploitation evaded.

Having stated this broad thesis, it is essential to stress what it does not imply: labor costs imposed by duress immediately to change the property rights of individuals or groups. The strike threat works much like a gun threat. In theory it can, therefore, redistribute property as a street robber can, to the extent to which people carry thievable assets. I am not questioning anything as obvious as that. Investors who have not foreseen the use of the strike weapon resemble the traveler who carries with him large sums of money. I shall show, in a similar way, that labor can be subject to monopsonistic exploitation by managements if any of the workers allow themselves to be “shut-in,” that is, prevented from moving to better remunerated opportunities. When this happens, the workers’ property in themselves is in some degree seized. What I shall call “the shut-in” is a form of partial slavery, unless it is simply a mutually advantageous contractual commitment, freely entered into and uninfluenced by fraud, to sell certain services for a certain price for a certain period.

The problems to be considered here fall within this theoretical framework. Wage-rate increases enforced through the strike threat benefit those remaining employed at the enhanced labor costs. But, on whom does the burden then fall? On the specific investors? On investors in general? On displaced or excluded workers? On consumers? What is the incidence of the burden? This problem is like that which economists discuss under the heading of the incidence of taxation. To express the issue in abstract terms (which means in the simplest terms), any one party to the productive process can exploit one or more of the other parties only in a measure determined by the “elasticities of supply”⁵ of the different productive services rendered by the people or by the assets employed. In this connection I propose to draw attention to four vital realities: (1) In the absence of man-made barriers to mobility, noticed above (p. 4), there is a wide range of alternative uses for a large proportion of workers and assets,⁶ a fact which implies long-term elasticities of supply. (2) Assets are often substitutable for labor (a consideration which is usually taken into account under the heading of “elasticity of substitution”). (3) If growing large scale recourse to the strike threat, accompanied by growing hostility to inflation, does not cause a disastrous cumulative decline in real income (depression without deflation), it must eventually force labor somehow to become the residual claimant on the value of the product, in order to make profitable the replacement of the complementary assets labor requires, let alone permit any growth in the stock of such assets in response to society’s saving preference.⁷ (4) Substitution of the consumption process for the saving process may contribute to the elasticity of supply of assets as such. (See p. 145)

Subject to these considerations, property may be taken from investors for the benefit of workers, either by way of the strike threat or governmental power. But how effective in practice can such transfer attempts be? Answering this question demands remembering that when everyone expects property transfers of this kind to be attempted, the long-term elasticities of supply of complementary factors must be multiplied.

I shall argue that while taxation can have limited effects in bringing about property and income transfers from rich to poor, the strike threat cannot. Forcing up the price of labor in different firms, occupations or industries does not effect an income redistribution from investors in general to workers in general. Similarly, the forcing down of the price of labor in any field by threats of managements to order a work stoppage is equally bound to fail to enrich investors in general (at the expense of workers in general). The workers’ “disadvantage in bargaining,” their “inequality of bargaining power,” the “perishability” of their labor, and so forth are relevant, I hold, only if we can assume that recruits to a trade can be somehow tricked by falsely represented prospects into specializing in that trade.

The only practically important case of monopsony (to be discussed in Chapter 8) involves some clear “shut-in” power. Yet even the Webbs, in two massive studies⁸ (both special pleading for the union movement), have presented no evidence of “employers,” in collusion or singly, ever deliberately and fraudently enticing employees into specialized occupations, with a view eventually to reducing their remuneration unfairly.

However, insofar as any firm or industry can attract and retain more workers by such methods, there will still be no redistribution of income caused in favor of “employers” as a whole or against the interests of “labor” as a whole. For other “employers” will then be forced to pay more for the labor for which they bid. If we assume that one set of occupations engrosses more labor than is just, we must assume that the rest are left with less than is just.

For identical reasons, if the entrepreneurs who have provided specialized resources in any particular field are, through wrong predictions, “exploited” by the use of strike power, other industries will have been robbed of that part of the capital which, given the reduced profits, ought not to have been invested in the “exploited” field. Ceteris paribus yields to capital elsewhere must be correspondingly higher.

The effect of wage rates determined under labor union pressure is, I shall insist, to distort society’s production structure, while it causes no redistribution whatsoever in favor of the poorer classes as such. The only income transfers that the use of strike power can effect are (1) in favor of those employed in one occupation at the expense of those in others⁹, or (2) in favor of workers as such when entrepreneurs generally have failed to forecast the extent to which, as investors, they will be subjected to duress-imposed costs.

When a wage rate is raised so as to price some part of potential output higher than consumers are prepared to pay, the wage gain is partly at the expense of workers who would otherwise have found their most remunerative employment in that trade; partly, of course, it is at the expense of consumers in general; but hardly ever (and I shall be developing this point at some length) is it at the expense of those who provide complementary resources—i.e., the assets which, in general, multiply the yield to effort. It is consumers who ultimately pay wages; and when the market value of output of any kind is forced (whether by the right to strike or through legal enactment) above the level which the free market would have determined, the effect is, in general, actually to harm the poorer classes disproportionately. This “regressive”¹⁰ consequence is aggravated because the process keeps (in the long run) a large segment of the work force in low-productivity and low-paying jobs; or (in the short run) forces workers into short-time jobs and (encouraged by unemployment compensation) into idleness. Hence the effect of the strike-threat system upon the distribution of the wages flow is to render it less equitable.

Through the consequences of the strike-threat system upon the composition of the assets-stock, and the nature of the employment outlets available, the flow of output as a whole and hence aggregate real income will be reduced. And, because all must admit that it is highly improbable that any substantial redistribution of the shrunken real income in favor of labor has ever been thereby effected,¹¹ obviously the system has all along been reducing the flow of real wages and the average of real wage rates.

In the “classical” theory of wages, as it had evolved at Cambridge in the pre-Keynesian era (by which I mean before publication of J. M. Keynes’s General Theory of Employment, Interest and Money in 1936), the issues which I have discussed in this chapter were virtually ignored. On the points which concern society most seriously, exposition was hopelessly contradictory for this reason. An inherent part of Alfred Marshall’s imposing synthesis of the “orthodoxy” of his age was the marginal productivity theory of wage-rate determination. This was clear, for instance, when he criticized Cliffe Leslie who (in attempting to justify strike-threat actions)¹² had, Marshall showed, failed to understand why competition tended to establish equivalence of net advantageousness in labor’s earnings. Yet other passages in Marshall’s writings appear to me to have been quite inconsistent with the insight he showed in his reference to Leslie.

This criticism applies, I suggest, particularly to Marshall’s discussion of what has been called the “range of indeterminateness” under bilateral monopoly;¹³ for the circumstances imagined relate to the problem of income distribution, I maintain, only under the assumption of wrong predictions. And I find that his contemporaries and successors who have relied upon similar kinds of reasoning have never stated this assumption—either explicitly or implicitly.

Marshall does recognize, through his notion of “derived demand,” that consumers ultimately employ all the resources used. But does not his analysis treat only the particular case? For one thing, it shows that consumers are the more exploitable the greater the inelasticity of demand’¹⁴ for the output happens to be. For another, it indicates that the suppliers of fixed and circulating capital who have failed to anticipate and discount typical trade-union practices (the vital qualification which Marshall does not specifically make) are more exploitable (a) the fewer the alternative uses there happen to be for the assets they have provided, (b) the smaller the proportion of labor cost to the total cost of the output, and (c) the fewer the opportunities of replacing existing employees by others (for example, strikers by blacklegs) or by labor-economizing machinery or organization.

I contend that Marshall’s analysis does not explicitly recognize that the power of the strike threat to raise wage rates (in particular industries) depends also upon the degree to which majorities within unions are indifferent about injustices caused to minorities. Concern for the interests of union members who may be displaced into inferior employments (and possibly forced into temporary unemployment) through duress-imposed costs may curb the extravagance of strike-threat demands. But Marshall does perceive that unions will try not to drive too many firms into insolvency. Yet he does not expressly refer to the condition that the consumer will be more exploitable the more effectively the firms supplying any output can act collusively in raising prices to consumers and rely upon the unions to protect them from the competition of nonunion labor.

Hence the possibilities covered by Marshall fail to show how manipulation of the strike threat can influence the division of the value of the product of industry between “employers” and “employed.” For in the economic system as a whole typical strike-threat activities are expected at least to continue, if not to extend, in scope. They are therefore allowed for in every business decision, and this is the overriding consideration in almost every context.

Marshall’s lack of rigor on the labor issue raises a question of great sociological interest. In his Economics of Industry he seems to imply (without clearly referring to the strike threat) that workers as a whole gain through their unions. He says that their power “to sustain high wages depends chiefly on the influence they exert on the character of the workmen themselves. . . .” If this means that the unions increase personal efficiency so that the market value of the workmen is higher, the question is how the unions manage to do so. Exploiting the consumer and excluded workers could, of course, take away, in Marshall’s words, “that want and fear of hunger which depressed the physique and moral character of the working class,” on the part of the exploiters; but as it would further depress those exploited, it is difficult to see how the “working class” as a whole could benefit. He goes on to say, “Unions have been at once a chief product and a chief cause of this constant elevation of the standard of life: where that standard is high, unions have sprung up naturally; where unions have been strong, the standard of life has generally risen.”¹⁵ That unions have been a product, of which more has in fact been acquired when the standard of living generally has been rising, is beyond question. But overcoats and bicycles and cars and television sets have also been products of which more has been acquired as standards of living have risen. Hence it is quite another matter to claim that living standards generally have risen because the unions have been strong, or because people generally have more overcoats or bicycles. Marshall refers also to the unions compelling employers to treat the worker “as an equal with something to sell that they (the employers) wanted to buy.”¹⁶ Of course union officials who are allowed to use the strike threat will be treated courteously by managements, as will their tax assessors. But will management’s fears of a union’s powers enable its members to raise their earnings without exploiting people poorer than themselves? Marshall did not face this sort of question with frankness; nor, in my judgment, have most subsequent economists.

To sum up. When the owners of assets or the suppliers of labor anticipate the possibility or likelihood of “exploitation,” as they will if society permits attempted “exploitation,” they will be unexploitable. Neither the providers of assets nor the providers of effort and skill are exploitable by one another (a) unless the former fail to predict and allow for the full cost consequences of future strike threats when they choose their investments, or (b) unless the latter fail to predict the wage-rate consequences of lockout threats or monopsonistic action by the hirers of labor when choosing and preparing for specialized employment. To me it seems unchallengeable that, because during the past half century or more the strike-threat influence has obviously been increasing, investors must on the whole have predicted the cost implications and hence have been virtually unexploitable. I use the word “virtually” because whether they have overestimated or underestimated the cost effects of strike power is difficult to judge. But if my reasoning is valid, the major consequences of society’s tolerance of the strike-threat system must have been simply a slowing down of the rate of increase in aggregate income, to the disadvantage of both participants, and with no discernible change in the proportion in which income is shared between “capital” and “labor”.¹⁷

If this argument is acceptable, the policy implications are far-reaching. The strike-threat system must be recognized as intolerable in a civilized age’, for it can be observed to be blocking the path to the maximization of the wages flow, the achievement of optimal equality and equity in the distribution of that flow, and the designing of an institutional framework to ensure security of employment without inflation.

I could end this book here. The simple argument I have presented ought to be sufficient to convince economists and policymakers. Unfortunately, I can anticipate a host of objections, some seriously advanced, other casuistic or obscurantist. Accordingly, the remaining chapters are devoted to meeting such objections as I have been able to find in academic literature devoted to the subject.

NOTES

¹ Investors in assets retained may already have been exploited. At the reduced value of such assets, their retention implies the prediction of no further exploitation.

² “Monopsony” (adjective, “monopsonistic”) means “monopolistic buying.”

³ The awful word, “oligopsony” (adjective, “oligopsonistic”), may be defined at this stage as “tacit monopsony” to distinguish it from “collusive monopsony.”

⁴ Collusive action among firms to resist “the strike in detail” (the “whipsaw”) is another matter. I deal with it on p. 47.

⁵ “Elasticity of supply” refers to the relative facility with which productive services of men or of assets can transfer to other uses without a material decline in their value. The smaller any such loss of value in alternative uses, the greater will be the elasticity of supply.

⁶ Some economists may at first be inclined to challenge this assertion. It will be discussed vigorously in Chapter 10, passim.

⁷ That is, as is to be explained later, the workers will be forced to hire or rent the fixed assets they need and to pay interest on the circulating capital because the owners of assets will only make them available on those terms.

⁸ Sidney and Beatrice Webb, The History of Trade Unionism and Industrial Democracy (London: Longmans Green and Co., 1920).

⁹ Because, as we are about to see, when duress-imposed labor costs in any activity reduce the number of workers who can be profitably employed in it, the number of workers who must compete for employment in other activities is increased, while as consumers all other workers will be disadvantaged.

¹⁰ A tax is said to be “regressive” when the proportion of the tax to the taxpayer’s income is greater the smaller his income. Thus import and excise duties and sales taxes are obviously regressive.

¹¹ I use the work “substantial” because in this context I am relying upon empirical evidence (see chapter 16). In the light of the general case argued in this chapter and the rest of the book, the word “substantial” could be omitted.

¹² Cliffe Leslie was one of a group of writers on wage questions (of whom the others were Thornton, Longe and Fleeming Jenkins) who had tried to show how union initiatives could enable a redistribution of income in labor’s favor. They had a considerable influence on John Stuart Mill during the last years of his life, when he was contemplating entry, and after his entry, into politics. I have discussed their contributions in my Theory of Collective Bargaining (Glencoe. Ill.: Free Press, 1954).

¹³ “Bilateral monopoly” means, in this context, “monopsony” (see above footnotes) among the purchasers of labor and monopoly (a union) among the suppliers of labor. Under such conditions there is no market determination of the price of labor.

¹⁴ The demand for a thing is said to be “inelastic” when a change in its price will have little influence on the amount of it that will be purchased.

¹⁵ Alfred Marshall, Economics of Industry (London: Macmillan, Ltd., 1909), p. 389.

¹⁶ Ibid., pp. 388-9.

¹⁷Within the category “labor” there must have been a consequential regressive redistribution (see footnote 10 and below pp. 168, et. seq.).