David schweickart

Download 207.95 Kb.
Date conversion02.06.2016
Size207.95 Kb.

A Democratic Theory of

Economic Exploitation

Dialectically Developed


IT IS STARTLING TO realize that the concept of economic exploitation, which has been the focus of intense philosophical debate for what seems like decades now, was barely touched on in John Rawls's 1971 masterwork, A Theory of Justice, the book that ushered in the present era of Anglo-American social and political philosophy. The subject was broached just once by Rawls, and only to be dismissed as being of such secondary impor­tance as to be "out of place here."1 The concept, however, had begun to attract the attention of a generation of students and young faculty who were rediscovering Marx, to the point that it could not much longer be ignored, not even in Harvard Yard. Robert Nozick, in his famous junior-colleague, neoconservative rebuttal to the liberal Rawls, devoted a full nine pages to attacking "Marxian exploitation," concluding that "Marxian ex­ploitation is the exploitation of people's lack of understanding of economics."2

Many articles followed, most concerned with defending some version of Marx's theory against Nozick's rather crude critique. The debate changed decisively in 1982 with the publication of John Roemer's General Theory of Exploitation and Class.3 Roemer, a mathematical economist of uncommon talent, provided a brilliant technical articulation of a labor-value theory of exploitation (dispensing with the often criticized labor theory of value),

From David Schweickart, "A Democratic Theory of Economic Exploitation Dialectically Developed," in Roger S. Gottlieb, ed., Radical Philosophy (Philadelphia: Temple University Press, 1993), 101-22, with the permission of Temple University Press.


then argued that an alternative "property-relations" theory is superior, be­ing at once faithful to the basic moral intuitions of Marx, more general in scope, and less subject to counterexample.

Roemer's book and his many supporting articles spawned a minor indus­try, which I will not attempt to survey. Let me say simply that I think Roemer is right that there are serious problems with the labor-value theory of exploitation, at least as it is commonly understood. But Roemer does not take his analysis far enough—or rather, he takes it in the wrong direction. In what follows I argue that his property-relations theory of exploitation represents a dialectical advance over the labor-value theory, just as the labor theory represented a dialectical advance over the Smithian theory that pre­ceded it, but that the property theory itself needs to be superseded by a "democratic theory" of exploitation. This democratic theory now has im­portant practical relevance, immediate in the case of Eastern Europe, less immediate but quite real in the case of Western capitalism.

I set out my argument as a "dialectical" argument. Since the word "dia­lectic" is unfamiliar in many quarters and evokes active hostility in others, let me clarify what I mean. In a sense, dialectics is storytelling, that is, a form of discourse having a narrative structure. In what follows I tell a story about how, in response to certain conditions, a certain theory arose; how deficiencies became increasingly evident; how, in order to overcome these deficiencies, the theory was replaced by a new theory that did not suffer the same deficiencies; how this theory developed new deficiencies, calling for yet another theory, and so on. Although specific historical circumstances are relevant to the story I tell, a dialectic account is intended to be not a historical narrative per se, but a tracing of the "logic" of certain conceptual transformations.4 (As is noted, the actual historical sequence does not al­ways match the logical sequence.) To my way of thinking, a dialectical account allows for a richer analysis of the concept of exploitation than is provided by either a strictly historical account or an ahistorical analysis of alternative theories.

Before getting to the argument proper, let me offer a general definition of exploitation, then some preliminary remarks about theories of exploitation.

Let us take the following as a definition: X exploits Y if and only if X takes unfair advantage of Y. This definition highlights two aspects of phenomena we wish to investigate: that one agent has benefited at the expense of another agent, and that something morally wrong has occurred. To narrow the domain of our investigation, we concentrate on economic exploitation, that is, on those cases where the benefits in question are tied directly to the production and consumption of material goods and serv­ices—although, as we shall see, the logic of our exploration will push us beyond this restriction.

A definition of exploitation is one thing. A theory of exploitation is another. It is instructive to ask why one needs a theory of exploitation. It is my view that a theory of exploitation is required when the exploitation is not obvi­ous, when the situation appears to be just, yet something seems wrong. One needs a theory to get beneath appearances to an underlying reality. (The sun appears to revolve around the earth; we need a theory to explain how it could be otherwise.)

A theory of exploitation should do three things. It should clarify the nature of the injustice, that is, make clear what normative principles are being violated and, if necessary, defend those principles. It should elucidate the mechanisms and institutions through which exploitation occurs, that is, explain how certain structures generate exploitative relationships and how certain structures mask the nature of those relationships. Finally, it should present an alternative vision, that is, a model of a nonexploitative society. As we shall see, each of the theories we consider can be understood to be attempting these tasks. How well a theory succeeds will condition our evalua­tion of that theory.


Let us begin not with "Marxian exploitation," but with what might be called a theory of "feudal exploitation," a theory associated with the Enlighten­ment attack on feudal privilege. Let us think of feudalism as a social order in which serfs work their own land but may neither leave it nor sell it without their lord's permission. In addition, they are compelled either to pay a tithe to their lord or to work gratis on his land for a certain period of time each year. These payments are legitimized as payment for protection. So long as this protection really is necessary, a case might be made that there is no exploitation, but once the serfs no longer need the lord yet are compelled to submit to the feudal regulations anyway, then they are being (feudally) exploited.

The three facets of a theory of exploitation are evident here. The moral principle being violated is the right of an agent to dispose of her own labor and property without interference, so long as no one else's property rights are violated. The key mechanism of exploitation is the lord's control of the means of coercion, though this is supplemented by an ideology of nobility rights and natural place. The nonexploitative alternative envisaged is laissez-faire capitalism.

A major difficulty with the theory in this form is the problem of specify­ing what counts as "one's own" property. Presumably, the feudal lord would dispute the claim that the properties of his serfs are exclusively theirs. A full theory of feudal exploitation requires a theory of property rights.


Historically, this issue was settled by means of revolution. Philosophi­cally, the key figure is Locke, who argues that property rights are natural, taking precedence over the acquired, consent-based rights of political au­thorities. Locke takes pains to give an account as to how "the world, given as it was to the children of men in common," should come legitimately to be the private property of individuals "without any express compact of all the commoners."5 Locke cites the mixing of one's labor as the basis of just property acquisition, subject to the restrictions that one not accumulate so much that it spoils and that one leave "enough and as good" for others.

Rather than analyzing Locke's account, which from our vantage point seems laughably full of holes, let us look at its contemporary reincarnation: Robert Nozick's "entitlement theory," which jettisons Locke's quaint "labor mixing" idea and his spoilage restraint. Nozick correctly identifies the logi­cal core of Locke's idea: "Individuals have rights," Nozick proclaims. "So strong far reaching are these rights that they raise the question of what, if anything, the state and its officials may do."6

Nozick proceeds to argue that if property is justly acquired and freely transferred, then no political authority may touch that property, not even via taxation, without the specific consent of the property owner. (Whether or not the political authority has been constituted democratically is irrele­vant.)7 Property is justly acquired if by voluntary exchange or by claiming that which belongs to nobody else, the latter acquisition subject to the "Lockean proviso" that there be "enough and as good remaining."8

Nozick's entitlement theory has been subject to withering rebuttal. Two main lines of argument stand out. The first and most obvious notes that the theory is irrelevant to real-world concerns, since the current distribution of wealth can in no way be thought to be the product of just acquisitions. Early capi­talist accumulation derived from the theft of common land, the slave trade, the extermination of native peoples, and other assorted horrors. Just acquisi­tion and voluntary exchange had little to do with it. (Nor have they to do with much of the subsequent capital accumulation, as countless financial scandals, our own savings and loan debacle among them, make evident.)

The second line of argument is more basic to our theoretical investiga­tion. The dynamic of pure laissez-faire tends to bring about institutions and effects clearly at odds with the common good. In particular there is nothing in a (neo-)Lockean "entitlement theory" of exploitation to block the estab­lishment of monopoly power. So long as fraud and physical coercion have been avoided, monopolies cannot be condemned, no matter how much wealth they amass, no matter how wasteful, no matter how far above costs they set their prices. If producers freely agree to fix prices or if a monopoly concern buys out its competitors or drives them out of business by selling below costs for a while, no objection may be raised.

A Democratic Theory 53

Critics ask the obvious question: Why should property rights have such absolute weight? Surely it is legitimate to do what all societies in fact do: place some restrictions, in the name of the common good, on what people may do with "their" property. To this objection Lockeans have no adequate reply. The theory has hit a dialectical contradiction: A theory grounding itself in an appeal to universal human freedom justifies institutions that, if allowed to develop freely, negate the effective freedom of almost everyone.


Smithian/neoclassical theory may be understood as a response to this con­tradiction. Adam Smith initiated the great search for a theory of value. To be sure, Smith was interested in the positive question—Why does the "in­visible hand" of competition set the price of corn at x instead of at 2x?—but the positive issue is intertwined with a normative one. If, as Smith argues, there is a "natural price" for each commodity, then there is a case for intervention when certain institutions prevent the market price from settling at this natural price. Monopoly is identified as a culprit, as are certain governmental practices. Against these Smith directs his attack. The state should refrain from interfering with the workings of the free market unless intervention is necessary to keep markets free, that is, competitive, that is, nonmonopolistic.9

In the Smithian theory monopolies are exploitative, since they cause prices to deviate from their natural values. But what is this "natural value"? Smith offers the "adding-up" theory:

When the price of any commodity is neither more or less than what is sufficient to pay the rent of the land, the wages of labour, and the profits of the stock employed in raising, preparing, and bringing it to market, according to their natural rates, the commodity is then sold for what may be called its natural price.10

But this theory is unsatisfactory without cogent theories of rent, wages, and profits. These, either positive or normative, Smith fails to provide. Why should society attempt to ensure that commodities sell at their "natural" prices, when certain components of that price might be problematic? To Smith it seems obvious (and his intuitions are not wrong on this point) that natural prices are (usually) better than monopoly prices; but in "solv­ing" the value question as he does, he merely brings to the surface the deeper positive and normative questions.

Historically, the search for a more adequate theory of value passed from Smith to Ricardo and then to Marx. The resulting "labor theory of value," its roots stretching back through Ricardo and Smith to Locke, was then


abruptly challenged toward the end of the nineteenth century by the "neo­classical revolution," which introduced the paradigm that still dominates mainstream economics today.

The historical dialectic in this instance does not match the conceptual dialectic, since, as we shall see, the labor theory is normatively more ad­vanced than the neoclassical theory. It is best to understand neoclassical theory as an attempt to resolve, within a Smithian framework, the difficul­ties with Smith's theory suggested above. Neoclassical theory may be un­derstood to be a sophisticated development of Smith's theory, just as neo-Lockean theory may be understood (as its advocates indeed understand it) to be a sophisticated revision of Locke. As in the case of the neo-Lockeans, the neoclassicals have different targets in mind than their forefather, but they too can be seen to grapple with the conceptual difficulties passed on to them.11

The great conceptual difficulty passed on to Smith's heirs was the prob­lem of justifying the various components of his natural price. The problem is fundamentally normative, since the free market income of each class is determined by the respective components of a commodity's natural price. If the invisible hand, unimpeded by state interference or monopoly, sets prices at their "natural" values, and if there is no political redistribution of in­come, then the income of each class, indeed, of each individual, will be determined by the "natural" rates of rent, wages, and profit. But are these "natural" rates fair rates? Calling them "natural" won't suffice as a moral argument. It won't suffice now, nor would it suffice in the late nineteenth century when the issue was re-opened. If it was untenable after Jeremy Bentham and John Stuart Mill to identify moral with natural, it was even less ten­able, in the face of a growing socialist movement, to regard the invisible hand as unproblematically just.

The intellectual problem confronting the heirs of Smith was daunting. It cannot be denied: The neoclassical solution was brilliant. The ethical ground was shifted from a dated, unconvincing appeal to nature to the more plau­sible canon of productive contribution: If you and I contribute jointly to the production of a commodity, we should divide the proceeds in propor­tion to our respective contributions.

But if the ethical appeal is to productive contribution, the theoretical difficulty seems insurmountable. Suppose we have a commodity, say a bushel of wheat, which we regard as the joint product of land, labor, and capital. How can we determine that x percent is due to the land, y percent to labor, and z percent to capital, where x + y + z = 100, apart from making the purely circular move of defining the contribution of land to be whatever the market rent happens to be, the contribution of labor to be whatever the market wage rate happens to be, and so on? How can one possibly

A Democratic Theory 55

define the "contribution" of a factor of production independently of the market mechanism in such a way as to be able to demonstrate (a) that the contributions add up to 100 percent and (b) that the free market will in fact return to each factor its respective contribution?

Amazingly enough, the neoclassical theorists met this challenge. Briefly put: The contribution of a factor is defined to be the marginal product of that factor. If, for example, it takes X acres of land worked by Y laborers using $Z worth of capital to produce a thousand bushels of wheat, then we define the contribution of each acre to be the increase in wheat that could be effected if the same workers employing the same capital worked X + 1 acres of land; we define the contribution of each worker to be the increase that could be effected if Y + 1 workers worked the same land with the same capital; and we define the contribution of capital to be the increase due to the same workers working the same land with one more unit of capital. If we make enough assumptions about substitutability of factors, diminishing returns, and so forth, then a mathematical theorem (Euler's Theorem) can be invoked to show that the sum of the contributions will be a thousand bushels. If we make enough additional assumptions to ensure perfect com­petition, it can be shown that the rent per acre set by the market will be precisely the contribution of that acre, the wage set by the market will be the contribution of the worker, and the interest set by the market will be the contribution of capital. Not only that, it can be demonstrated that monopoly power will enable a factor owner to receive more than he con­tributes—thus justifying Smith's original intuition.

We have here a theory of exploitation—historically, a rejoinder to the Marxian theory; conceptually, a dialectical development form the Lockean theory. The moral principle, that agents should receive in accordance with their productive contributions, is an advance over the problematic Lockean principle of unrestricted voluntary exchange. The exploitative mechanisms are identified: any interference with perfectly competitive pricing. A vision of nonexploitation is proposed: perfectly competitive capitalism. (This form of capitalism is something different from laissez-faire, since the government is now authorized to come between mutually consenting agents, e.g., pro­ducers engaged in price fixing, so as to promote competitive conditions.)

An amazing feat, the neoclassical solution, but unfortunately inadequate to the fundamental problem. The fundamental problem, recall, is norma­tive. Hence, the definition of "productive contribution" must not only sat­isfy conditions (a) and (b) above. The definition must also be morally relevant.

The marginal product definition fails on this count. I have treated this issue at length elsewhere.12 Suffice it to say that even if we want to say that land and capital "contribute" to the final product, we cannot say the same for the landowner or the capitalist.


The basic problem is this. In neoclassical theory, which sharply distin­guishes the functions of capitalist and landowner from those of entrepre­neur and manager, neither capitalist nor landowner engages in anything that might be called a productive activity. Their roles are purely passive. They allow land to be used by whoever will pay the highest rent; they allow capital to be used by whoever will pay the highest interest. The very conditions that must be assumed so as to prove that the free market will return to each factor its productive contribution rule out considering capi­talists or landowners as being the bearers of risk or the exercisers of entre­preneurial skill—the very factors that would seem to be necessary if property income is to be morally justified. The Smithian/neoclassical theory also founders on a dialectical contradiction.


The neoclassical attempt to refurbish Adam Smith's theory of exploitation fails, but the project of building a theory of exploitation on the canon of contribution has not been exhausted. The activities of capitalists and land­owners (at least within the neoclassical framework) may not qualify as pro­ductive contributions, but the activities of workers surely do. No matter what the theoretical framework, it must be granted that workers engage in productive activity.

Perhaps labor is the only productive activity. Such is the claim of the labor-value theory of exploitation. This is the theory usually associated with Marx. Marx argues that labor is the source of all value and yet workers receive only a portion of that value. He argues that wage labor masks the fact that workers, collectively, are paid not the value of what they produce, but rather the value of what it takes to produce (and reproduce) them as a laboring class.13

The standard response to this argument is dismissal, on the grounds that it appeals to a discredited theory of value. It is alleged that Marx's argu­ment depends on the faulty assumption that prices in a market econ­omy are proportional to labor inputs, which, in general, they are not. It should be well known by now (although it is not) that this response is wholly inadequate. For the "Fundamental Marxian Theorem"—that posi­tive profits require a positive rate of exploitation—is independent of the assumption that prices are proportional to labor values. The mathema­tical Marxists (Morishima, Roemer, and others) have demonstrated this quite conclusively.14

There is, however, a significant lacuna in Marx's argument. If it is to be understood as (at least in part) a moral argument, then it must contain a moral premise. Marx leaves this premise unspecified.
For anyone who has read Marx's basic argument carefully, an obvious candidate comes to mind, a straightforward extension of the normative prin­ciple underlying Smithian exploitation: If a person works x hours to produce goods or services consumed by others, then she should receive in return goods and services that it took others x hours to produce.

This principle is a version of the canon of productive contribution. If several agents contribute their labor to a joint product, then they should receive a share of the product that is proportional to their contribution. This version, however, carries with it a presumption that the general canon does not. The ethical principle on which the labor-value theory of exploi­tation is based implies that labor alone is a productive activity.

There are reasons for thinking that Marx himself would balk at this ethi­cal principle, reasons that I do not pursue here.15 He would, however, as­sent to the claim that labor is the only productive activity. I think he is right to do so. Marx denies that the activities associated with the owners of the other factors of production are productive activities. He distinguishes, as do most subsequent economists, capitalist qua capitalist from managers and entrepreneurs. Once one makes this distinction, one must at least grant that the capitalist qua capitalist and landlord qua landlord contribute mini­mally to the production process, certainly much less than would be required for a justification of their incomes.

Marx not only denies the capitalist's appeal to the canon of contribu­tion; he takes issue with most of the other standard justifications of prop­erty income. He shows that capital accumulation did not come about "cleanly" nor did the working class emerge voluntarily. Marx shows that even when an individual's initial capital is based on his own labor, the normal work­ings of the system will in time return to him income far in excess of his original contribution. He also shows that appeals to the capitalist's "sacri­fice" or "abstinence" are spurious.16

I find Marx convincing on these matters. Moreover, I do not think capi­talism can be saved by appealing to risk or to "time preference"—defenses not considered by Marx, but similar in structure and spirit to the ones he does address. I do not develop these issues here, since I have done so else­where.17 I am persuaded that the labor-value theory of exploitation, suit­ably elaborated, can effectively debunk all justifications of capitalism save one: the quasi-utilitarian argument that capitalism is the best of all feasible worlds. As we shall see, justifications based on this claim are the target of a more dialectically advanced theory.

Although its negative arguments against appeals to the canon of contri­bution by any class but the (broadly defined) working class are impressive, it does not follow that the labor-value theory of exploitation is secure in its moral foundation. In formulating his powerful critique of wage labor, Marx


makes the simplifying assumption that labor is homogeneous. For purposes of critique, this assumption serves well, but when one moves from critique to affirmation (to the negation of the negation), difficulties surface. If we accept the moral principle on which the labor-value theory of exploitation is based, a nonexploitative society would be one in which each worker re­ceives the full value of her product. That is, if she contributes eight hours to society, then she should receive goods and services that took eight hours to produce.

But this implies that all labor must count the same, regardless of its qual­ity. For the only plausible mechanism for guaranteeing that the labor-value principle is satisfied throughout society is to have every worker paid at the same rate, regardless of skill and regardless of the degree of unpleasant­ness.18 Many readers (though perhaps not all) will object that such an ar­rangement would not be fair. But there exists an even more compelling objection: Such an economy is not viable. Not only would such radical egalitarianism require that moral incentives (or coercion) completely re­place economic incentives, but it would also require a degree of central planning that cannot be made to function efficiently. (The theoretical ar­guments long advanced concerning the authoritarianism and inefficiency of central planning have been dramatically confirmed by the recent collapse of the Soviet model.)19

Although much more could be said about the labor-value theory of ex­ploitation, I think I have said enough to establish the thesis of this sec­tion: The labor-value theory of exploitation, although an advance over the Smithian/neoclassical theory, is unsatisfactory. Its negative moment is sound: Neither capitalists qua capitalists nor landowners qua landowners engage in productive activity. But the ethical principle on which it is based is untenable, above all because the positive vision of a nonexploitative econ­omy deriving from this principle is not only morally problematic; it is economically unfeasible. Since a meaningful theory of exploitation must be rooted in the real world, not only as a critical exposé of real justice, but as a positive exposition of a viable alternative, a better theory of exploitation is demanded.


John Roemer proposes such a theory. Its general structure is this. Class S is exploited by the complementary class S' whenever

  1. There is a feasible alternative state in which coalition S would be better off than in its present situation;

  2. Under this alternative coalition S' would be worse off than at present;

  1. Coalition S' is in a relationship of dominance to coalition S. This dominance enables it to prevent coalition S from realizing the alter­native.20

Some clarifications are in order. Roemer designates his general theory of exploitation a "property-relations" theory. Within this general theory he specifies successive forms of exploitation: feudal, capitalist, and socialist. All these forms satisfy the three conditions above. What distinguishes them are the different ways in which the alternatives are formulated—the differ­ent "withdrawal conditions." One of the merits Roemer claims for his theory is its generality. I do not concern myself with this claim here. Here I analyze as the "property-relations theory of exploitation" what Roemer calls "capi­talist exploitation." The alternative that characterizes "capitalist exploita­tion" is for S to withdraw from the larger society with its per capita share of the alienable assets—natural resources and capital goods.

Roemer is explicit in his assertion of the superiority of the property-rela­tions theory to the labor-value theory (called by him the "unequal exchange" theory).21 If we look back to our critique of the labor theory, his verdict seems on the mark. The property-relations theory does not seem to entail an undesirable or unfeasible alternative. Built into its definition is a feasi­ble alternative freely chosen. The difficulties we encountered with hetero­geneous labor do not arise. If class S would be better off—whatever the qualitative composition of its labor contributions—by withdrawing with its per capita share of society's alienable assets, but is prevented from doing so by S' (who would be made worse off), then S is exploited.

Roemer's theory pushes beyond the labor-value theory not by revision but, in true dialectical fashion, by revolution—placing property, rather than labor, at the center of the exploitation question. Roemer agrees with the general Marxian condemnation of capitalism. He disagrees, however, that the key to the exploitative nature of capitalism is wage labor. The key, he argues, is unequal property ownership.

Roemer advances some strong, sophisticated arguments in support of his highly controversial thesis. He demonstrates that it is possible to have labor exploitation (i.e., agents consuming commodities that embody less labor than they contribute) without wage labor. He shows, for ex­ample, that if we have a society where initial property endowments are unequal, it makes no difference whether those with more capital hire laborers or simply loan out their capital. Exactly the same agents will exploit and be exploited.22

He pushes the argument further. Suppose we have two islands equally endowed with labor, each with full knowledge of available technologies, but differentially endowed with initial capital. Suppose there is neither labor


migration between these islands nor any capital flow, only free trade of commodities. Roemer demonstrates that if each island apportions its labor to maximize its consumption, then unequal exchange, that is, labor exploi­tation, will occur.23

How do these arguments bear on the issue with which we are concerned, namely, the adequacy of the property-relations theory of exploitation? Cen­tral to our investigation is Roemer's claim that his property-relations theory is more adequate normatively than the labor-value theory. Unlike the labor-value theory, his theory does not require that all agents be paid at the same rate. For Roemer, the morality of the unequal reward thus comes down to the legitimacy of one's initial wealth. That is to say, if the inequality of initial assets is just, then the unequal consumption is just. If the initial distribution is not just, then neither are the resulting entitlements.

Roemer's idea here is not unattractive. It would seem to be in better accord than the labor theory with certain of our considered moral judg­ments. For example, if two islands develop their resources autonomously, one more effectively than the other, we do not want to say that voluntary trade between them is exploitative simply because the commodities of the more developed island embody less labor than those for which they are exchanged. If such trade is deemed exploitative, then the advanced island could avoid exploiting the other only by refusing to trade—even when both parties would benefit by the trade—or by disadvantaging itself. Surely the former course makes no sense. Surely justice (as opposed to benevolence) cannot require the latter.

I think Roemer is right that the implications of the labor-value theory of exploitation sometimes conflict with our sensible intuitions. His theory avoids many of the difficulties of the earlier theory. But this theory lands him in a trap—if you will, in a dialectical contradiction. To avoid the implausibly egalitarian implications of the labor-value theory (that all labor should be rewarded equally), he posits a theory whose implications turn out to be—even more egalitarian.

For consider the normative principle on which the property-rights theory is based. Recall that a class is exploited if three conditions obtain: It would be better off withdrawing from the larger society with its per capita share of society's alienable assets; the complementary class would be worse off; the complementary class blocks the withdrawal.

What is the ethical principle operative here? The conditions that S be better off and S' worse off function not normatively but to indicate that S' is benefiting at the expense of S. The blocking condition inserts causality, so that an agent may be deemed responsible. That leaves us with the nor­mative principle itself, which must be: the right of any individual or group to a per capita share of the alienable assets of society.

This is a strongly egalitarian principle. It is stronger than the famous Rawlsian difference principle, which it closely resembles. According to Rawls, "all social primary goods ... are to be distributed equally unless an unequal distribution of any or all of these goods is to the advantage of the least favored."24 The property-theory principle implies that inequalities are justi­fied only if no subgroup (not merely the least favored) would be better off with an equal distribution.

We notice that something rather odd has transpired. In avoiding the excessive egalitarianism of the labor-value principle, we find ourselves with a principle that seems even more egalitarian. If we balked at assigning equal value to all labor, regardless of quality or circumstance, why should we not balk at an alleged equal entitlement to all alienable assets? If I work harder than you or longer than you or more skillfully than you in creating material assets, why are you entitled to a per capita share?

In fact, the property-relations theory of exploitation contains within it­self a contradiction. On the one hand, it does not want to commit itself to equality of outcome. If we begin with equal endowments but employ them differently (more or less skillfully, more or less intensely, more or less wisely), then unequal outcomes should not be deemed exploitative. On the other hand, these unequal outcomes form the basis for the "initial endowments" of the next period—which are supposed to be equal.

Roemer, in his latest work, senses the strong egalitarian position toward which the logic of his theory leads. He bites the bullet in the penultimate chapter of Free to Lose, wherein he constructs a highly original proof of a rather strange theorem:

On the class of economic environments X, there is only one economic constitution F that satisfies the five axioms Pareto Optimality, Land Monotonicity, Technological Monotonicity, Limited Self-Ownership, and Protection of the Infirm. That unique constitution assigns in any envi­ronment the Pareto optimal allocation of corn and labor that equalizes the utility levels of the agents.25

This is not the occasion to analyze this theorem. Suffice it to note its import. The only "economic constitution" (i.e., principle of allocation of resources) that satisfies a set of constraints that Roemer thinks anyone of an anticapitalist bent should find nonproblematic is an allocation that "equal­izes the utility levels of the agents" (i.e., makes everyone equally happy). This is egalitarianism with a vengeance. A more egalitarian prescription is scarcely imaginable.

This is too strong an egalitarian commitment. Such a commitment con­flicts with many of our considered intuitions. More important, it conflicts with the nonutopian impulse of the Marxian socialist tradition, an impulse


that is particularly important today when the issue of feasible socialism is being hotly debated. We have here another curious dialectical reversal. A strength of the property-relations theory over and against the labor theory is its insistence that a feasible alternative exist before exploitation can be charged. This alternative, however, turns out to be wholly abstract. If we try to think concretely about alternatives, we find ourselves running up against an excessive restriction. To avoid the charge of capitalist exploita­tion, the feasible alternative must be such that no group can make itself better off by withdrawing with its per capita share of the alienable assets—regardless of how much or how little it contributed to the formation of those assets. This restriction rules out not only capitalism, but any form of feasible socialism as well—in particular, market socialism. To be sure, the issue of market socialism is controversial, but a theory of exploitation that decides the issue a priori is surely defective.26

What is called for is a theory of exploitation that is more adequate to real-world socialism and more firmly grounded normatively. What is needed is a theory that combines elements of the labor-value theory with elements of the property-relations theory in a synthesis that exceeds the sum of its parts. Precisely such a theory, I submit, is the following.


The democratic theory of exploitation rests on a principle that is not a principle: Individuals should, in some sense, legislate the rules that bind them. This principle is too vague to be a real principle. The principles that underlay all the previous theories of exploitation were sharp and clear: property rights, quantitative contribution, equality of wages, a per capita share of alienable assets. This principle is more a heuristic than a well-defined norm. It guides our thinking but does not command.

It is also self-reflexive, in a way that the other principles are not. How do we determine in what sense individuals should legislate their rules? We do it democratically. Individuals must, in some sense, democratically decide on the economic rules and goals and institutions that will constrain their subsequent behavior; they must decide the appropriate spheres for various sorts of democratic mechanisms.

So long as we focus on the normative principle itself, the democratic theory seems hopelessly vague. But the crucial dialectical innovation of this theory is to shift the starting point from the normative principle to the alternative vision. We begin not with an abstract principle, but with a far more concrete project: the specification of a model of an economically fea­sible, morally desirable alternative to capitalism.27 In all the previous theo­ries, the model of a nonexploitative economy could be derived more or less

A Democratic Theory 63

rigorously from the normative foundation. Not so with democratic exploita­tion. The model itself must be constructed not via logical deduction from first principles, but by drawing on theoretical, historical, and empirical re­search, all the while guided by the democratic heuristic.

This project, although not widely known, is in fact well advanced. In­creasingly, since about 1970, various radical thinkers and activists have been investigating the theory and practice of economic democracy.28 The central concept here is workplace democracy. Why, it began to be asked, do workers not elect their bosses? If ordinary citizens in Western democracies are com­petent enough to elect their mayors, governors, presidents, and a host of other political officials, are they not competent enough to elect their workplace managers?

As it turns out, the answer to this empirical question is unambiguous: They are. As it turns out, there are large (and growing) numbers of experi­ments—participation schemes, cooperative enterprises, networks of coop­erative enterprises—that prove beyond a reasonable doubt that workplace democracy is efficient, at least as efficient as and often more so than ordi­nary capitalist management.29 Democracy, it turns out, has positive eco­nomic as well as political and moral significance.

Starting from the basic, well-grounded premise that workplace democracy works, the task then becomes to determine what sorts of structures are nec­essary to link a system of democratic enterprises into an optimally viable, desirable democratic economy. This question, of course, is less amenable to direct empirical investigation and hence is more speculative. Still, I think we are now in position to state with considerable confidence what the basic institutions of economic democracy should be.30

Obviously, this is not the place for more than the briefest sketch, but let me at least provide that.31 A feasible, desirable, fully democratic economy must be an appropriate synthesis of three elements: workplace democracy, a free market for economic goods and services, and social control of investment. Workers should control their enterprises, one-person, one-vote. In larger enterprises this entails electing a workers' council that appoints the man­agement. Enterprises should sell to one another and to consumers freely, with supply and demand (for the most part) determining prices. (Econ­omic democracy is a market economy—though not a capitalist economy.) What the invisible hand of supply and demand should not determine is the nature and rate of new investment. For a large variety of compelling rea­sons—ethical, ecological, economical—a better mechanism is needed. Al­though there is no consensus on specific details, it is clear that national, state, and local legislatures must play an active role in generating the in­vestment fund (via enterprise taxation) and in setting general priorities for its dispensation.


One may conceptualize these briefly described institutions as a synthesis of three different forms of democracy. In a real if limited sense the market mechanism is a democratic institution. Individuals vote their preferences with their purchases. The productive apparatus responds to consumer de­mand. Workplace democracy is democracy of a different sort, democracy on the model of the ancient Greek polis or the early town meetings in this country, that is, direct democracy. The arena is small enough for all indi­viduals so inclined to voice their views and try to persuade others; the issues are near at hand and the results often swiftly perceived. By contrast, social control of investment is necessarily a mediated, representative de­mocracy. The need for coherence and consistency in investment planning and for balancing costs, benefits, entrepreneurial incentives, regional versus national interests, and the like precludes the direct, popular formulation of investment priorities, though not, of course, popular input. Economic de­mocracy is thus a synthesis of individualist, direct, and representative de­mocracy, giving individuals the opportunity for democratic participation as consumers, as workers, and as citizens.

It is clear on reflection that the democratic theory of exploitation is a synthesis and sublation of the previous two theories (see table 1 for a sum­mary of the dialectical argument). The labor-value theory of exploitation is right that neither capitalists nor landowners are necessary for a healthy economy. (Quite the contrary!) It is right too in its condemnation of wage labor. The property-relations theory is right to focus attention on property rights, for these rights, as currently constituted, block the emergence of a democratic economy. They serve as the fundamental bulwark of capitalism. (Why can't workers elect their bosses? Why must democ­racy stop at the factory gate? Because the factory does not belong to them. Why can we not set priorities democratically as to the development of our economy? Because investment decisions are made by the "private" sector, i.e., by the owners of the means of production. Why can't we change things? Because money dominates politics, and that money—most of it—derives from property.)

Both the labor-value theory and the property-relations theory, however, are too narrowly normative. That is to say, each centers on a normative principle, sharply defined, and this renders each theory inadequate in the face of what is perhaps now the fundamental question for radical theorists: Is there a viable alternative to capitalism? Democratic theory preserves the valid insights of the earlier theories but confronts this question head on—and in my view, successfully.

A Democratic Theory 65 Table 1. Summary of the Dialectical Argument

Theory of











Just acquisition/

The state




Free exchange





Capitalists don't





Labor value


Equal exchange

of labor

Eqaul access to

Wage labor


Equal wages

Equal assets






alienable assets

Legislative self 

binding rules







I do not mean to suggest that the democratic theory of exploitation is the final form of exploitation theory. History has not come to an end. The democratic theory is itself unfinished. The reader cannot fail to notice a great many gaps in my argument, and large areas only roughly sketched.

But even when the gaps are closed and details provided (as I am confi­dent they can be), the theory here presented remains radically incomplete. For there is a huge area of exploitation, also the focus of much attention by radical theorists and activists, that this account has left invisible: namely, sex/gender exploitation. This is not the place to venture even a sketch of what the next stage of exploitation theory will look like, but I am per­suaded that it will be a dialectical advance that infuses the democratic theory just presented with feminist theory so as to clarify, critique, and mark an alternative not only to sex/gender exploitation itself, but to a range of other problems that feminist theory has shown to be linked to that exploitation: social violence, militarism, carelessness (in all its meanings), the despoilation of our planet, and so on.

Theoretically and practically, there is much to be done.


  1. John Rawls, A Theory of Justice (Cambridge: Harvard University Press, 1971), 310.

  2. Robert Nozick, Anarchy, State, and Utopia (New York: Basic Books, 1974), 262.

  3. John Roemer, A General Theory of Exploitation and Class (Cambridge: Harvard University Press, 1982).


  1. A model for what I am doing is Marx's account early in Capital of the develop­ment of the form of value from the elementary form to the extended form to the general form to the money form (Capital, [New York: International Pub­lishers, 1967], 1:48-70).

  2. See John Locke, Second Treatise of Government, Samuel Moore and Edward Aveling (Indianapolis: Hackett, 1980), 18-30.

  3. Nozick, Anarchy, State, and Utopia, ix.

  4. It may be noted that although logically Nozick is arguing against "feudal ex­ploitation," the real concern has shifted enormously since Locke. Locke articu­lates a defense of bourgeois property against its usurpation by aristocratic authorities. Neo-Lockeans such as Nozick are concerned to defend bourgeois property against the threat of democratic redistribution. It is not the king that Nozickeans fear but popular majorities.

  5. Nozick, Anarchy, State, and Utopia, 150-82. Nozick pays much attention to the Lockean proviso, hoping to give it current applicability. Since my cri­tique of Nozick centers elsewhere, we need not analyze this (vain, in my view) attempt.

  6. Actually, Smith has little faith that law can fend off directly the deep tendency toward monopoly. "People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices. It is impossible to prevent such meetings, by any law which either could be executed, or would be consistent with liberty and justice." He does believe, however, that certain public works, above all roads and canals, can enhance competition (Adam Smith, The Wealth of Nations [Middlesex: Penguin, 1970], 232-33, 251).

  1. Ibid., 157

  1. I observed that the neo-Lockeans are more concerned to check democratic "ex­cesses" than monarchical ones. I would venture that the "monopoly" of most concern to neoclassicals is not that of enterprises but (at least until very re­cently) that of labor. See Milton Friedman's treatment of business monopolies and labor unions in Capitalism and Freedom (Chicago: University of Chicago Press, 1962), ch. 8.

  1. See David Schweickart, Capitalism or Worker Control? An Ethical and Economic Appraisal (New York: Praeger, 1980), 5-19.

  2. Marx, Capital, esp. 146-98.

  3. Michio Morishima, Marx's Economics: A Dual Theory of Value and Growth (Cam­bridge: Cambridge University Press, 1973); John Roemer, Analytic Foundations of Marxian Economic Theory (Cambridge: Cambridge University Press, 1981).

  4. Marx is far too subtle and dialectical a thinker to base his case against capital­ism on a single normative principle. But to set out the structure of Marx's argument in its full complexity is beyond the scope of this chapter.

  5. See Marx, Capital, 713-61, 579-88, and 591-99, for the various arguments.

  6. See my Against Capitalism (Cambridge: Cambridge University Press, 1993), ch. 1.

  7. Students of Marx might protest that in Capital Marx allows that skilled labor count as a multiple of unskilled labor (44). But with Marx there is intent on constructing a (quasi-)positive theory, not a normative one. In a market economy, prices will not reflect labor values unless skilled labor counts for more than simple labor. Whether it should count for more (in a nonexploitative society) is a separate matter, one to which the labor-value theory of exploitation (as I have formulated it) gives a definite, if problematic, answer.

  1. It should not be thought that only procapitalists have made such arguments, nor should one identify the Soviet model with socialism. Many radical thinkers have argued, and for a long time, that a desirable, feasible socialism should be structured very differently from the Soviet model. This issue is taken up more fully below.

  1. Roemer, General Theory, 194-95.

  1. For a concise presentation of his case, see John Roemer, "What Is Exploita­tion?" Philosophy and Public Affairs (Winter 1989): 90-97. He has, however, had some "Second Thoughts on Property Relations and Exploitation," Cana­dian Journal of Philosophy, supp. vol. 15:257-66.

  1. Cf. Roemer, General Theory, 87-95; see also Roemer, Free to Lose: An Introduc­tion to Marxist Economic Philosophy (Cambridge: Harvard University Press, 1988), 90-98.

  1. See John Roemer, "Unequal Exchange, Labor Migration, and International Capital Flows: A Theoretical Synthesis," in P. Desai, ed., Marxism, Central Planning, and the Soviet Economy: Economic Essays in Honor of Alexander Erlich (Cam­bridge: MIT Press, 1983), 34-62. See also Roemer, Free to Lose, 99-103.

  2. Rawls, A Theory of Justice, 303.

  3. Roemer, Free to Lose, 167.

  4. It is worth noting that Roemer has recently shifted his attention from exploi­tation to precisely the question of market socialism. See his "The Morality and Efficiency of Market Socialism," Ethics 102 (April 1992): 448-64. See also Pranab Bardhan and John Roemer, "Market Socialism: A Case for Rejuvenation," Jour­nal of Economic Perspectives 6 (Summer 1992): 101-16. In Roemer's view no market socialism is "just," but designing a feasible alternative to capitalism is an appropriate and important task for a socialist theorist today.

  5. Normative considerations-the democratic principle noted above and others as well-guide the construction of the model and enter explicitly into its defense. But we do not begin with abstract principles.

  6. The literature is now too vast to cite here. The most important early theorist was Cornell economist Yaroslav Vanek, whose technical treatise, The General Theory of Labor-Managed Market Economics (Ithaca: Cornell University Press, 1970), and popular companion piece, The Participatory Economy (Ithaca: Cornell University Press, 1971), set the terms for almost all subsequent analysis. An important recent contribution from a more philosophical perspective is Carol Gould, Rethinking Democracy: Freedom and Social Cooperation in Politics, Economy, and Society (Cambridge: Cambridge University Press, 1988). My own Against Capitalism contains an extended bibliography.

  7. For a recent analysis of some forty-three separate studies, see David Levine and Laura D'Andrea Tyson, "Participation, Productivity, and the Firm's Environ­ment," in Alan Blinder, ed., Paying for Productivity: A Look at the Evidence (Washington, D.C.: Brookings Institute, 1990), 183-244.

  8. Setting out such a model should not be thought to contradict the reflexivity of the democratic principle. The institutions to be set out all give scope to demo­cratic decision making, and they are all open with respect to many details. Moreover, they are merely being advocated here, not legislated-but with the confidence that they would be readily embraced by the vast majority, if thor­oughly and fairly debated. (Joyce Rothschild and Allen Whitt cite a 1984 study that finds "the social basis of support for workplace democracy to include nearly all segments of society, especially blue-collar workers and professional/technical


workers. The only class that does not support the redistribution of power in the workplace are managers and owners." The Cooperative Workplace: Potentials and Dilemmas of Organizational Democracy and Participation [Cambridge: Cambridge University Press, 1986], 186).

31. For many more details and an exhaustive defense, see my Against Capitalism.

The database is protected by copyright ©essaydocs.org 2016
send message

    Main page