1 Roemer, John E., ‘Should Marxists Be Interested in Exploitation?’ Philosophy and Public Affairs 14 (1985) 30-65Google Scholar

2 The term ‘forced’ is used herein as in Elster’s work. In Making Sense of Marx (Cambridge, MA: Cambridge University Press 1985) 211-12 Elster defines ‘force’ as implying the presence of constraints that leave no room for choice; coercion is a stronger condition requiring the presence of an intentional agent. Elster states (214-15) that the most reasonable way of interpreting Marx’s claim that the worker is forced to sell his labor power is to say that although ‘the worker can survive without selling his labor power, he can do so only under conditions so bad that the only acceptable course of action is to sell his labor power.’ Thus, the worker is forced to sell his labor power when ‘(i) the offered wage rate is above the alternative and (ii) the alternative is below some critical level.’

3 See, for example, Analytical Foundations of Marxian Economic Theory (Cambridge, MA: Cambridge University Press 1981), ch. 7.

4 A General Theory of Exploitation and Class (Cambridge, MA: Harvard University Press 1982)Google Scholar

5 ‘Property Relations versus Surplus Value in Marxian Exploitation,’ Philosophy and Public Affairs 11, 4 (Fall 1982) 281-313

6 ‘Should Marxists Be Interested in Exploitation?’

7 Ibid., 30-1; 36

8 Ibid., 36

9 Ibid., 52

10 Roemer, , ‘Exploitation, Class and Property Relations,’ in After Marx, Ball, Terence and Farr, James, eds. (Cambridge: Cambridge University Press 1984), 203-4Google Scholar

11 Ibid., 204; Roemer, ‘Should Marxists Be Interested in Exploitation?’ 52

12 Ibid., 54, 65

13 Roemer claims that a kind of exploitation may occur even in a simple subsistence exchange economy populated by independent proprietors (hence no D²) holding unequal amounts of financial assets: those with above (below) average financial holdings will work less (more) than average to attain subsistence, in effect transfering surplus labor from poor to rich proprietors without creating a surplus product or a labor market (Ibid., 43-4).

14 Ibid., 44

15 ‘Should Marxists Be Interested in Exploitation?’ 47. This definition constitutes a shift from his prior work. Prior to 1985, Roemer identified alienation with hierarchical and autocratic work processes, that is, with D ; Roemer, A General Theory of Exploitation and Class, 104-5; ‘Exploitation, Class, and Property Relations,’ 197-9.

16 Ibid., 47-8

17 Ibid., 47

18 Ibid., 51

19 As I.I. Rubin has observed, social and not technical relations are the appropriate realm of social theory; that which is invariant to social forms cannot explain purely social phenomena. See Rubin, Essays in the Labor Theory of Value (Detroit: Black and Red Press 1972).

20 See, for example, Roemer (1985), 34-6.

21 The result that labor is unforced is not always established in Roemer’s work by assuming that workers can go to a ‘farm.’ However, his formal models invariably incorporate an assumption which has the same effect. For example, he builds up a model on page 113 of the General Theory of Exploitation and Class as follows: ‘An agent can engage in three types of economic activity: he can sell his labor power, he can hire the labor power of others, or he can work for himself ‘ (emphasis added). The last phrase brings in the farm. In his formal models, this assumption provides the continuity and interiority necessary for deriving the existence of an equilibrium. All of Roemer’s formal models of exploitation take the form of equilibria of this kind, which preclude unemployment. That unemployment (one aspect of forced labor) is a peripheral concern is illustrated by the fact that it is mentioned only once in the index to his 1982 opus, and is omitted from the index to his 1981 book.

22 ‘Should Marxists Be Interested in Exploitation?’ 41,44

23 Marx, Karl, Wage Labor and Capital, in Elliott, John E., Marx and Engels on Economics, Politics, and Society (Santa Monica: Scott Foresman 1975), 89Google Scholar. Marx, , Theories of Surplus Value, Vol. III (Moscow: Progress Publishers 1963), 49Google Scholar; Capital, Vol. I (New York: International Publishers 1967), 605; Capital, Vol. III (New York: International Publishers 1967), 795; Grundrisse (New York: Random House 1973), 700

24 ‘Basic’ exploitation, in turn, takes two forms in Marx’s exposition. Under primitive communism, use of resources is sufficiently productive to reproduce society at a subsistence level, but not high enough to generate surplus labor and a surplus product above that level. Under other, progressive, societies, including a future communism, use of resources is sufficiently productive to generate surplus labor and surplus product. It is the social and human conditions surrounding the generation and distribution of surplus labor and surplus product, however, and not the surplus-creating capabilities associated with the product use of resources per se, which is ‘basic’ to such progressive societies. See Marx, Contribution to the Critique of Political Economy (New York: International Publishers 1970), 13; Marx, ‘Critique of the Gotha Program,’ in Marx and Engels, Selected Works (New York: International Publishers 1974), 322-4.

25 Marx, Capital, Vol. I, 443

26 Marx, Capital, Vol III, 609

27 In Roemer’s model, the possibility that asset-rich workers may exploit asset-poor capitalists is presented as rendering Marx’s alternative formulation—that is, that capitalists exploit workers—redundant and/or wrong. We shall return to these claims in Section VI. But note here that because Roemer’s definitions and assumptions concerning exploitation differ significantly from Marx’s, it is not possible to move directly from Roemer’s categories to a critique of Marx. For Marx, labor, as manual and mental effort and energy, is integral-along with land, raw materials, and other forces of production-to efficacious use of resources in production (BE). But because labor is the active human element in (SHE), primarily in and through production (PE) and secondarily through asset ownership per se (SE). Corn or oil, for example, being non-human, can enter into BE, but not SHE. Capitalists, although human, are not (except for managerial labor), according to Marx, ‘direct producers.’ Therefore, although capitalists might be said to be open to subjection to a form of SHE based purely on respective wealth positions (that is, SE), they could not be subject to that form of SHE that proceeded through efficacious use of resources (PE). Thus, we may surmise (or speculate) that if a resurrected Marx agreed to use corn or some other commodity as a ‘value numeraire,’ he, too, would conclude that corn could be (BE) ‘exploited.’ But, contra Roemer, he would not therefore be impelled to withdraw the ‘claim to be interested in labor’s exploitation only because labor is exploited in the first [BE] sense’ in favor of the revised claim that labor, unlike corn, oil, et cetera, is exploited ‘in the second [SHE, SE] sense’ (Roemer, 38), because Marx, armed with careful distinctions among BE, SE, and PE, makes no such claim and has no need of such a revised claim.

28 Walrasian general equilibrium is a paradigm within economics in the sense of Imre Lakatos. Its central idea is to examine how autonomous agents with given resources, tastes and technologies, can achieve mutually consistent and socially optimal outcomes via competitive markets. By socially optimal is meant maximum feasible levels of consumption. Production in this paradigm is understood as an instantaneous process: inputs used in production are converted without any time delay into the outputs which the given technology makes feasible. This allows the economist to abstract from any disutilities associated with the actual process of production itself. In effect, the Walrasian paradigm treats production as exchange with nature. Each agent formulates supply and demand behaviors for all goods based on their relative prices. Equilibrium then exists for that vector of relative prices obtains at which the quantities of each input and output good supplied and demanded are exactly equal. There is assumed to be a coordination mechanism (the ‘auctioneer’) which varies the vector of relative prices until some vector is found for which demand supply are equal in all markets. Agents participating in this equilibrium need know only the price vector in making their decisions. The Walrasian equilibrium concept has received wide currency in economics because of its formal elegance and simplicity, but its viability as an explanation of observed economic behavior has been questioned. Within the logic of this paradigm, equilibrium is contingent on a set of stability conditions for demand and supply equations.

Critics outside this paradigm take issue with the notion that technology, tastes and endowments are are given and independent of the socioeconomic processes which the theory purports to explain; these ‘givens’ can instead be regarded as contextually defined, evolving interactively with other social phenomena. A succinct and definitive exposition of Walrasian equilibrium theory is contained in Debreu, Gerard, The Theory of Value (New York: Wiley 1959)Google Scholar. A sympathetic discussion of the methodological bases of this paradigm is found in Roy Weintraub, E., Microfoundations (Cambridge: Cambridge University Press 1980)Google Scholar.

29 Roemer, A General Theory of Exploitation and Class, 196

30 Marx, Capital, Vol. I, 152-3, 233, 235, 264-5

31 Marx, Grundrisse, 649-52

32 Marx, Capital, Vo1. I, 639

33 David McLellan, Marx Before Marxism (New York: Harper 1970), 169

34 Marx, , Economic and Philosophical Manuscripts. In Bottomore, Tom, Karl Marx: Early Writings (New York: McGraw-Hill 1964), 163Google Scholar

35 Marx, Capital, Vol. I, 75

36 Marx, Capital, Vol. I, 570

37 Ibid., 174

38 Although grounded in DOPA and the economic necessity for most workers to sell labor power ‘voluntarily,’ the de facto servitude of labor under capitalist domination, according to Marx, rests on additional factors, namely, ‘a working-class, which by education, tradition, habit, looks upon the conditions of that [capitalist] mode of production as self-evident laws of Nature,’ a despotic capitalist organization of production which ‘breaks down all resistance,’ the constant creation of an industrial reserve army which keeps wages ‘in a rut that corresponds with the wants of capital,’ and the ‘dull compulsion of economic relations [which] completes the subjection of the labourer to the capitalist’ (Marx, Capital, Vol. I, 737).

39 Ibid., 184-6. Elsewhere, Marx makes clear, however, that the capitalist’s dominion is imperfect. Workers fight back, and capitalists must impose supervisory costs to reduce shirking (ibid., 330-2).

40 This last alienation overlaps, but is not identical with, Roemer’s notion of alienation as the sale of products embodying one’s labor. The two coincide only under a simple exchange economy of independent proprietors who labor to produce products for market exchange. Under capitalism, capitalists, no less than workers, sell products to anonymous purchasers and operate under the domination of alien market exchange processes beyond their control. But workers are subject to an additional alienation: disposition of their products are under the domination of capitalist employers. Thus, capitalists and workers are both dominated and alienated by market exchange processes; but the capitalist class through alienation confirms ‘its own power’ and experiences a ‘semblance of human existence.’ The working class feels ‘annihilated’ by alienation and experiences through it its ‘powerlessness and the reality of an inhuman existence’ (Marx-Engels, The Holy Family. Marx-Engels, Collected Works, Vol. 5 [New York: International 1975|).

41 If the worker is related to the product of his labor as an alien, powerful object, then ‘an alien, hostile, powerful and independent man is the lord of this object.’ If work activity is alienated and unfree, then ‘it is under the domination, coercion, and yoke, of another man,’ namely, ‘the capitalist (or whatever one likes to call the lord of labour)’ (Marx, Economic and Philosophical Manuscripts, 130-1).

42 Ibid., 131

43 Marx, Capital, Vol. I, 712

44 Ibid., 571

45 Marx pointedly summarizes his view of the subsidiary position of distribution in his 1875 critical remarks on the Gotha Program. Here, he states that ‘it was in general a mistake to make a fuss about so-called distribution and put the principal stress on it.’ Distribution of the means of consumption, he continues, ‘is only a consequence of the distribution of the conditions of production themselves.’ But that distribution ‘is a feature of the mode of production itself.’ Under the capitalist mode of production, the conditions of production are concentrated as a class monopoly ‘in the hands of non-workers in the form of property in capital and land, while the masses are only owners of the personal conditions of production, of labour power’ (Marx, Karl and Engels, Frederick, Selected Works [New York: International Publishers 1974], 325Google Scholar). Because of this bi-polar class distribution of property, the wage-worker is permitted to live ‘only in so far as he works for a certain time gratis for the capitalist (and hence also for the latter’s co-consumers of surplus value).’ Through its domination and exploitation, the ‘whole capitalist system of production turns on the increase of this gratis labour by extending the working day or by developing the productivity’ or intensity of labor. ‘Consequently, the system of wage labour is a system of slavery and indeed of a slavery which becomes more severe in proportion as the social productive forces of labour develop, whether the worker receives better or worse payment’ (ibid., 329).

46 Marx, Grundrisse, 247

47 The concept of the forced sale of labor power is discussed at length in G.A. Cohen, ‘The Structure of Proletarian Unfreedom,’ Philosophy and Public Affairs 12 (1983) 3-33.

48 The first and second rows of this box can also be interpreted in utility-theoretic terms. If agents who do sell their labor power maximize utility, and if other agents can maximize utility while withholding their labor power from the labor market, then the sale of labor power is unforced. Conversely, forced sales of labor power occur when at least some agents can only maximize utility by selling their labor power, but the number of jobs is less than the number of such agents.

49 Insightful discussions of the emerging economic literature on the operation of markets under incomplete information are contained in Stiglitz, Joseph E., ‘The Causes and Consequences of the Dependence of Quality on Price,’ Journal of Economic Literature 25 (1987) 1-48Google Scholar; Akerlof, George, An Economic Theorist’s Book of Tales (Cambridge: Cambridge University Press 1984)CrossRefGoogle Scholar; and Strong, Norman and Walker, Martin, Information and Capital Markets (London: Basil Black-well 1987).Google Scholar

50 Keynes, John Maynard, The General Theory of Employment, Interest, and Money (New York: Harcourt Brace 1936)Google Scholar; Knight, Frank, Risk, Uncertainty, and Profit (Boston: Houghton Mifflin 1921)Google Scholar; Marx, , Theories of Surplus Value, Vol. 2 (Moscow: Progress Publishers 1969), 504-7Google Scholar; Capital, Vol. 1 (New York: International Publishers 1967), Parts 3-5. This paragraph draws on the insightful discussion of Geoffrey Hodgson, M., ‘Production versus Allocation in Economic Theory or Marx after Keynes and Knight’ (Newcastle upon Tyne unpublished 1988)Google Scholar.

51 Stiglitz, 7

52 The result in some Neoclassical models that unemployment occurs in equilibrium does not mean they implicitly recognize intra-enterprise force (D ²). For one thing, Neoclassical economists tend not to think of indeterminate labor processes as resolved through use of force. For example, Akerlof discusses gift-giving between employers and employees and the use of social custom as means of resolving this indeterminacy. Further, as discussed above, D ² is not implied by the existence of unemployment alone. In a Neoclassical setting, agents denied employment (or credit) in one market characterized by incomplete information may have other options akin to Roemer’s farm; so the forced aspect of labor power sale may be absent. Conversely, even if D ² is one possible means of resolving this indeterminacy, positive unemployment is not necessary as a condition of its existence.

53 There is no reason to believe that the existence of one non-empty cell implies that all other cells must be empty; thus, we discuss the other three cells as alignments that may occur simultaneously (even in societies in which some production occurs under conditions enumerated in cell [1, 1]).

54 This can be accomplished, as Bowles and Gintis have argued, in a number of ways —through paying those hired a wage higher than the market clearing wage, through maintaining low benefit levels for the unemployed, etc. See Bowles, and Gintis, , Democracy and Capitalism (New York: Basic Books 1986), 76-9.Google Scholar

55 It is worth emphasizing at this juncture that the ‘separation’ between an existing capitalism and an ideal fully competitive capitalism is impossible when production is permeated with indeterminacy. Note that the analysis of an idealized capitalism requires that the Walrasian conceptions of production and competition (discussed in n. 28) are relatively accurate depictions of their real-world analogues. Ex ante indeterminate production processes sever the simple connection which is assumed to exist between inputs and outputs in the Walrasian conception of economic processes.

56 Thus, the forces identified in cell (2, 1) provide both a significant source of E and are integral to Marx’s own case (2,2). Indeed, to restrict E to the differentiation between labor and labor power, as in cell (1,2), would unnecessarily ‘commit Marx to the totally unacceptable conclusion that in a benign capitalism, in which workers labor only for the time and with the intensity agreed upon when the wage contract is struck, there would be no profit and no exploitation!’ (Robert Paul Wolff, Understanding Marx: A Reconstruction and Critique of Capital [Princeton: Princeton University Press 1984], 178).

57 Marx, Capital, Vol. III, 244

58 B-owles, , ‘The Production Process in a Competitive Economy: Walrasian, Neo-Hobbesian, and Marxian Models,’ American Economic Review 75, 1 (1985), 16-36Google Scholar; Bowles, and Gintis, , ‘Structure and Practice in the Labor Theory of Value,’ Review of Radical Political Economics 12, 4 (Winter 1981), 1-16Google Scholar; ‘Contested Exchange: Political Economy and Modern Economic Theory,’ American Economic Review 78, 2 (May 1988) 145-50. Also see Devine, James and Reich, Michael, ‘The Microeconomics of Conflict and Hierarchy in Capitalist Production,’ Review of Radical Political Economics 12, 4 (Winter 1981) 27-45Google Scholar. We hasten to add that these authors all view positive unemployment as a systemic feature of capitalism, and as such view the economy’s position as cell (2,2), not cell (1,2). We have discussed these authors’ work in the context of this cell because they emphasize the importance of the use of force within the labor process in their characterizations of capitalist production.

59 An analogue is found in John Maynard Keynes’ theory of effective demand and employment, in The General Theory of Employment, Interest, and Money (New York: Harcourt Brace 1936). Keynes wanted to show that, ‘even taking the assumption of neo-classical analysis [the atomistically competitive firm], he could produce non-neoclassical results’ (Chick, Victoria, Macroeconomics After Keynes: A Reconsideration of the General Theory [Cambridge: MIT Press 1983], 25Google Scholar). Yet he rejected certain aspects of ‘perfection’ in the workings of market forces-especially certainty and perfect foresight —which he considered so discordant with capitalist reality as to seriously distort the theoretical argument. See Elliott, John E., ‘Keynes on the Efficacy of Wage Cuts as Anti-Depressionary Strategy’ (Research in the History of Economic Thought and Methodology, Samuels, Warren J., ed. [Greenwich, CT: JAI Press 1988], vol. 6)Google Scholar.

60 Consider the following thought experiment. For any economic process in the second column (such as bank credit decisions), ask the question: if someone could make money by solving the indeterminacy, would that indeterminacy have occurred in the first place?

61 Marx, Capital, Vol. I, 714-15

62 Ibid., 342

63 The option of moving from factory to farm becomes increasingly difficult for an ever-larger fraction of the labor force, and thus is at best a transient element in the historical evolution of capitalism. The reproduction of wage laborer as wage laborer (under conditions of unemployment), which Marx considers ‘that indispensable requisite’ for ‘the social dependence of the labourer on the capitalist’ is thus increasingly secured. If, on the other hand, such acceptable alternatives as small landed or artisate proprietorships are readily available to workers (as in the appropriation of public lands by settlers in a ‘free colony’), the ‘constant transformation’ of wage-labourers into independent proprietors creates substantial impediments to the regular reproduction of wage labor. Moreover, the exit of large numbers of workers from the capitalist sector reacts ‘perversely’ on the operation of the labor market: ‘Not only does the degree of exploitation of the wage-labourer remain indecently low. The wage-labourer loses into the bargain, along with the relation of dependence, also the sentiment of dependence on the abstemious capitalist’ (ibid., 768-70).

64 Roemer, ‘Should Marxists Be Interested in Exploitation?’ 33,59

65 See, for example, Cohen, G.A., ‘Freedom, Justice, and Capitalism,’ New Left Review 126 (1981) 3-16Google Scholar; Elster, Jon, ‘Exploitation, Freedom and Justice,’ Nomos 26 (1983) 277-304Google Scholar; Elster, Making Sense of Marx, 216-33.

66 See Elliott, John E., ‘Justice and Freedom in Marx’s Moral Critique of Capitalism,’ Samuels, Warren J., ed., Research in The History of Economic Thought and Methodology (Greenwich: JAI Press 1987), Vol. 5, 1-49Google Scholar, and sources cited therein.