The transformation problem: Is the standard commodity a solution

Review of Review of Radical Political Economics Vol. 32, 2 (2000) 265-281 RADICAL POLITICAL ECONOMICS ELSEVIER The Transformation Problem: Is the Standard Commodity a Solution? Ajit Sinha* Department of Economics, LBS National Academy of Administration, 248179, India Mussoorie Received3 March 1998; accepted16December1998 ABSTRACT This paper attempts to closely investigate the relationship between Ricardo’s searchfor an invariable measureof value and Sraffa’s Standardcommodity on the one hand and Marx’s transformation problem and his notion of exploitation on the other. The paper arguesthat Marx’s notion of exploitation is subtly different from the Sraffian notion of exploitation as suggestedby Garegnani. And that the Standard commodity is not a solution to Marx’s transformation problem as suggestedby Eatwell. The paper is an attempt to initiate a fruitful dialog between Marxists and Sraffians on the question of labor as a unit of measure and its relation to the notion of exploitation. JEL classification: El 1; D46 Keywords: Marx; Sraffa; Transformationproblem * E-mail: ajitsinha@lbsnaa.ernet.in 0486-6134/00/$- seefront matter0 2000 URPE. All rights reserved. 266 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 1. Introduction Since Steedman’s (1977) provocative critique of Marx’s theory of value and the transformation problem, Marxists have almost exclusively identified the Sraffian position with Steedman’s, and have maintained that it is antagonistic to Marx’s theory.’ This has led to a neglect of other significant Sraffian authors who have presented more sympathetic interpretations of Marx’s theory of value and the transformation problem. The paper is an attempt to bring this “blackout” to an end. Elsewhere (Sinha 1996) I have argued that Marx’s theoretical structure can broadly be classified as “surplus approach economics”; and to that extent Sraffa’s contribution should be seen as a positive contribution to Marxism. This, however, does not mean that Sraffa’s project was entirely Marxist. Some leading Sraffian scholars, who are sympathetic to the Marxist project, have interpreted Marx’s theory of exploitation as another variant of Ricardo’s theory of distribution, and Sraffa’s Standard commodity as the solution not only to Ricardo’s problem of the “invariable measure of value” but also to the transformation problem in Marxian economics.’ In this paper I shall argue that these are incorrect inferences. The Standard commodity was designed to solve a problem which was different from the transformation problem, and its application to the transformation problem amounts to a crucial shift in Marx’s concept of necessary labor. In Marxian economics labor-time as a unit of measure derives its ’ The hostility with which Steedman’s critique was received by the Marxist camp has been an unfortunate chapter in the history of Marxist economic theory. I would not classify Steedman as anti-Marxist despite the provocation. ’ In the 1970s John Eatwell published two papers on the transformation problem (Eatwell 1974, 1975), where he argued this point. Eatwell acknowledges a great debt to Garegnani’s unpublished works for his understanding of the nature of the problem. Since then some of Garegnani’s papers have come out in print (Garegnani 1984, 199 1) that to some extent compliment Eatwell’s claim. Garegnani (1984), however, disagrees with Eatwell’s attempt to draw “a ‘direct link’ between the use of the Standard commodity and the use of the labour theory of value in Marx, ‘to expose the nature of (the) exploitation [of labour]“’ (323, f.n. 52). This criticism by Garegnani, however, is based on his interpretation of the role of the labor theory of value in Marx as an analytical one, which “leaves no room for its use in expressing exploitation (which, it seems, would only consist of showing what is already clear, that is that workers do not get the entire product).” Below I shall have occasion to criticize Garegnani’s position. On the other hand, Pier Luigi Porta (1986), a non-Sraffian, has also argued that the problem the Standard commodity is designed to solve is essentially a Marxian one. Porta’s claim, however, is of a different nature and more extreme. He argues that Sraffa’s reading of Ricardo is a reading of Marx into Ricardo. A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 267 legitimacy from the notion of exploitation which is different, in a subtle way, from the Sraffian notion of exploitation. In section 2, I elaborate on Ricardo’s problem of the “invariable measure of value” and Sraffa’s solution to this problem. In section 3, I distinguish the significance of labor-time as a unit of measure in Marx from Ricardo’s. I argue that Marx’s concept of exploitation is subtly different from the Sraffian concept proposed by Garegnani and others. In section 4, I take up Eatwell’s interpretation of Marx’s transformation problem, and argue that it is a misinterpretation. Section 5 contains some concluding remarks. 2. Ricardo’s Problem and the Standard Commodity In the Introduction to Ricardo’s Works vol. I, Sraffa (1951) tells us that Ricardo’s theory of distribution was built on one crucial “correction” of what he had called Adam Smith’s “original error.“3 Smith (1776) in the course of his analysis of price relations, had suggested that “natural prices” of commodities are determined by adding up independently determined “natural wages,” “natural profit,” and “natural rent.” Ricardo’s contention was that the three distributional variables cannot be independent, since they must be bound by the size of the net output. With the aid of his particular theory of rent, which suggested that marginal land does not pay any rent, Ricardo was able to put the question of rent aside and concentrate on the relationship between the remaining two distributional variables -wages and profits. A straight line inverse relationship between wages and rate of profits could readily be established in a one commodity corn-profit model, where corn happens to be both input and output in the system and wages are paid, as share of net output, in corn as well. In this case, a rise in wages would mean a proportional fall in the rate of profits, which can be directly calculated since both output and input are given in the same homogeneous unit, corn. In a multiple commodity case, however, the proposition that wages and the rate of profits are inversely and proportionately related cannot be all that readily established. The reason for this is simple. The surplus ’ I follow Sraffa’s reading of Ricardo. However, Sraffa’s reading of Ricardo is one among three major competing readings. For a frontal attack on Sraffa’s “surplus approach” framework, see Hollander (1979, 1992) and also Hicks (1985), Hicks and Hollander (1977), and Casarosa (1978, 1985). This school has interpreted Ricardo’s theory to be very much in the neoclassical framework. Carvale (1985, 1991) and Carvale and Tosato (1980) have introduced an interesting element in the debate by rejecting both approaches (i.e., Sraffa’s as well as Hollander’s) on the ground that their frameworks are static, whereas Ricardo’s framework was essentially dynamic. In my opinion Carvale and Tosato are still much closer to Sraffa than Hollander. 268 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 product left in the hands of the capitalists and the total capital investments (either including or excluding real wages) would in general be two bundles of heterogeneous collections of different goods. In other words, the rate of profits cannot be calculated unless all the commodities are reduced to one homogenous unit. This led Ricardo to the question of prices as a solution to the aggregation problem.4 If one could determine the exchange ratios of all the commodities, then one commodity could be chosen as numeraire and all the others could be measured in units of the numeraire commodity. In other words, all the commodities could be measured in terms of money. By hypothesizing the labor theory of value, i.e., that exchange ratios of commodities are equal to the ratios of direct and indirect labor-time needed to produce those commodities, Ricardo attempted to root the theory of prices solely in technology-labor-time being the representation of technology-so that the role of prices can be taken out of the question of distribution, and that the same exercise, as in the corn-profit model, could be conducted in the general case of multiple commodities as well. As Ricardo in a letter to McCulloch, dated November 16, 1820, wrote: “After all, the great questions of Rent, Wages and Profits must be explained by the proportions on which the whole produce is divided between landlords, capitalists, and labourers, and which are not essentially connected with the doctrine of value” (quoted in Sraffa 1951, p. xxxiii). Ricardo, however, was well aware that the labor theory of value did not determine the correct prices of production in the case of a general competitive capitalist economy. This is because the rate of profit is calculated as return on investment per unit of time. Thus two equal amounts of investments, if measured on the basis of the labor theory of value, would give different rates of profit if their turnover time were different. Since, in general, it would be too restrictive to assume that all investments have identical turnover time, the labor theory of value must be “modified” so that all capital investments receive the same rate of profits. This in itself did not bother Ricardo much. What concerned him was that, given the “modified” values, any change in wages and consequently the rate of profits would affect the “cost” elements of all the commodities including the numeraire, or the money commodity, in all sorts of ways. This means that Ricardo’s statement that “after all, the great questions of Rent, Wages, and Profits...are not essentially connected with the doctrine of value” cannot be established, since the measurements of those divisions seem to be dependent on prices. This led Ricardo to search for an invariable measure of value, which would be a commodity with such properties that a change in the distributional 4 At an early stage of the writing of the Principles Ricardo wrote to James Mill, “I know I shall be soon stopped by the word price” (letter dated December 30, 1815, quoted in Sraffa 19.51, p. xiv). A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 269 variables in itself will not affect its “cost” conditions, and so if wages and profits are measured in this commodity their relation will become transparent, i.e., free of all the complications created by changes in prices. Though Ricardo could not discover his invariable measure of value, Sraffa’s Standard commodity has the properties Ricardo was looking for, in this particular context, in his invariable measure of value. Sraffa (1960) showed that Ricardo was wrong in attempting to first determine prices and then rate of profits, since prices and the rate of profits must be determined simultaneously. The reason prices of commodities change when the distribution of the net output is altered is because of “inequality of the proportions in which labour and means of production are employed in the various industries.” If the proportions were the same in all industries, then a rise or fall in wages would proportionately deduct from or add to the profits in all the industries without the need for a change in original prices, since prices change only to redress the discrepancy in the rate of profits that would arise if the proportions were not the same. But Sraffa showed that even though, in general, prices will change in all sorts of ways in the face of a change in the distributional variables, one can always construct a “composite commodity” from any given real system that would not be affected by a change in distribution. Sraffa called this composite commodity “the Standard commodity.” By discovering the Standard commodity, Sraffa proved that the rate of profit can be determined irrespective of the knowledge of prices, as long as we take real wages as given and measure it in terms of the Standard commodity: “The straight line relation between the wage and the rate of profits will therefore hold in all cases, provided only that the wage is expressed in terms of the Standard product. The same rate of profits, which in the Standard ’ Ong (1983) has argued that Sraffa’s Standard commodity is not a solution to Ricardo’s problem of the “invariable measure of value.” In his interpretation, Ricardo was searching for a measure that would distinguish unambiguously any change in the price of a commodity caused by a change in the conditions of its production, particularly in the context of accumulation with given real wages and decreasing returns on land. And such a measure is impossible to find. This is because a measure of itself is dependent upon the prevailing rate a change in the “condition of production” of profits. Kurz and Salvadori (1993) have argued that Ong is wrong, because in Ricardo’s dynamic system the real wages do not remain constant (see f.n.4). They have correctly argued that Ricardo’s problem of “invariable measure of value” can be separated into two aspects: (i) the condition of invariability in the case of changes in distribution within a given technical environment; and (ii) the condition of invariability in the case of changes in technological environment. The Standard commodity is a solution to the first problem, whereas the second problem has no genera1 solution. As Harcourt (1990) has pointed out, Sraffa considered Ricardo’s sought-for invariant measure of value in the context of historical time and changes in technology to be a “will o’ the wisp.” 270 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 system is obtained as a ratio between quantities of commodities will in the actual system result from the ratios of aggregate values” (1960: 23). Thus in one stroke Sraffa established the basic proposition of what he called classical economics, which is that all the variables of production and distribution or all the aggregative variables of a competitive capitalist system are determined from outside the market (i.e., they are determined in a socio-historical context) independent of prices (see Garegnani 1984). This is also the gist of Marx’s critique of “vulgar economics” in Capital. An interesting aspect of Sraffa’s Standard system and the Standard commodity is that they not only reaffirm the basic insight of Ricardo’s corn-profit model, but also go a long way to vindicate Marx’s method, in so far as Marx insisted that the rate of profits should be calculated independently of prices. In other words, profit is a non-price phenomenon. 3. The Significance of Labor-Time in Marx Though it is quite clear that Sraffa’s Standard commodity goes a long way in vindicating certain aspects of Marx’s theoretical structure, we, nevertheless, must note that Marx’s transformation problem is not identical to Ricardo’s problem of the invariable measure of value, which the Standard commodity was specifically designed to solve. In Marx, the problem was to find a medium of translation from the valueaccountin B to price or money-accounting with given distributional variables. The Standard commodity, on the other hand, gives us a 6 The most general formulation of Marx’s price and value systems, in the tradition of Bortkiewicz (1949) and Seton (1957), can be represented as: p = (pA + pdl)(l + r) (1) il=AA+l (2) where p, 1, and 1 are the vectors of prices, direct labor-time per unit of output, and labor-values respectively; A is the matrix of input coefficients; and d and r are the real wage vector and the rate of profits respectively. The first set of equations independently solves for the rate of profits, r, and relative prices. However, these prices and the rate of profits will not coincide with the value relations. Therefore, the transformation problem is about translating the relative prices into labor-time units so that the two accounting systems form a “meaningful” relation, Within this approach, Morishima (1973) has shown that positive profit is possible if and only if there is positive surplus value in the system. In some cases of join-production some perversity could arise, as shown by Steedman (1977). In such cases, Morishima and Catephores (1978) have argued that the “necessary labor” should be calculated by the minimum labor time needed to produce at least the wage basket. See Rankin (1987) for a defence of Morishima and Catephores’s approach of linear A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 27 1 medium of translation from one price-accounting to another when the distributional variables change, given total output and technology. Even though the two problems are intricately connected, since the cause of the diversion of prices from value-ratios is the same as the changes in prices due to changes in distributional variables, they are not identical. At one place (Introduction to Works Vol. I) Sraffa indirectly alludes to the transformation problem, and urges us to keep it separate from the problem of invariable measure of value: . . .Ricardo was not interested for its own sake in the problem of why two commodities produced by the same quantities of labour are not of the same exchangeable value. He was concerned with it only in so far as thereby relative values are affected by changes in wages. The two points of view of difSerence and of change are closely linked together, yet the search for an invariable measure of value, which is so much at the centre of Ricardo’s system, arises exclusively from the second and would have no counterpart in an investigation of the first. (Sraffa, 1951: xiix, emphasis added) Given that the final version of the Introduction was written between 1948-50 (see General Preface to Works Vol. I), and that the properties of the Standard commodity were developed in the thirties or early forties (see Preface to PCMC), it is quite clear that the above statement of Sraffa is made with full awareness of the properties of the Standard commodity. Below I shall argue that the rationale for using labor-time as the unit of measurement in Marx’s project is different from what it is in Ricardo’s project. In Ricardo the labor theory of value plays a strictly technical role. As Garegnani (1991) has argued, the basic flaw in Adam Smith’s argument about determining prices by adding up the independently determined “natural rate of profit” on top of the “cost” may be quite easily shown today with the aid of the simultaneous equation method; in the absence of this tool, however, which was the case with both Ricardo and Marx, it is not easy to show the constraint binding on the rate of profits in a multiple commodity case. The labor theory of value was the best one could do to show the constraint. Garegnani attributes the same reason to Marx for using the labor theory of value. In this case, the question of a relation between two accounting systems does not arise. In other words, Garegnani’s contention is that in the absence of a Sraffa-type simultaneous equation method, Marx had no option but to derive the rate of profits from the labor-value magnitudes to programming method in this context. Furthermore, Shaikh (1973) has shown that the rate of profit is a monotonic increasing function of the rate of surplus value. 2’72 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 arrive at prices. Therefore, there is no “transformation problem.” One significant implication of this argument is that now the rationale for labor-values is redundant in Marx’s economic theory.’ Garegnani’s suggestion that Marx’s reason for using labor-values was “technical” leads him to interpret Marx’s notion of exploitation in purely non-labor terms: “It appears, rather, that [Marx] viewed the exploitation of labor as a fact established by overall explanation of economic phenomena where profits emerge as a share which, in spite of the apparent freedom of the wage contract, is similar to that of a feudal landlord, in that both have their only foundation in social institutions preventing the labourer from appropriating the entire product” (1991: 108, emphasis added).’ Elsewhere (see f.n. 7) I have argued that Marx’s division of the total direct labor-time into necessary and surplus labor-time draws a direct parallel with the feudal system where the peasant serf is self-sufficient but is still forced to work for the landlord. The fundamental difference between Garegnani’s understanding of Marx’s notion of exploitation and mine is that Garegnani defines exploitation as “preventing the workers from appropriating the entire net product,” whereas I maintain that Marx’s notion of exploitation means making the workers work for somebody else, gratis. The two notions of exploitation are not the same. Garegnani’s notion of exploitation becomes operative after production is over. In our definition the notion of exploitation is operative contemporaneously with production. It is true that control over others’ labor would result in control over their fruits of labor, but not viceversa. When we speak of appropriation of the fruits of labor, we are speaking of a juridical or moral category-a right to appropriate. However, when we speak of control over others’ labor, we are speaking of control over the production process, i.e. how and what to produce. Moreover, it amounts to a lack of freedom on the part of the worker. Life is measured in time, and control over others’ time is control over their lives. Marx talks about increases in the intensity of work and in the length of the working day in terms of life of the workers being sucked out more rapidly by capital: “The consumption of labourpower by capital is so rapid that the worker has already more or less completely lived himself out when he is only half-wa!. through his life.. .” (Capital I: 795). This aspect of Marx’s notion of exploitation cannot be captured by the RicardianBraffian approach. The Sraffian notion of exploitation suits the example of sharecropping, where the ” This has been the main bone of contention in the debate between Steedman (1975, 1977, 1981) and the “Marxists,” and there is no need to go over that debate here. My position on this debate can be found in Sinha (1990, 1991, 1996). a See Steedman (1981) and Hodgson (1982, 1991) for a similar notion of exploitation. And, as we shall see in this paper, Eatwell’s solution to the transformation problem also hangs on such a notion of exploitation. A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 273 landlord appropriates a share of the total output without playing any role in the production process. Capitalist exploitation, however, is the result of capitalists’ direct control, and workers’ lack of control, over the production process. The reason why the paradigmatic case of Marx’s notion of exploitation is the relation between the feudal lord and the serf is that in this relation the expenditure of surplus-labor as well as the cause of it are quite obvious and do not need analysis. As far as labour rent goes, the most simple and primitive form of rent, this much is clear. Here rent is the original form of surplus-value and coincides with it. But it needs no further analysis here that surplus-value coincides with the unpaid labour of others, since this still exists in its visible, palpable form, the labour of the direct producer for himself being separate both in time and space from his work for the landlord, with the latter appearing directly in the brutal form of forced labour for a third party. (Capital III: 928) Marx’s point is that in the capitalist system this basic relation is hidden. The separation of the means of production from the workers and a social division of labor conceal the basic relation of exploitation. Not only does the category of profit hide the notions of surplus-labor and surplus value, but also the category of wages conceals the notion of surplus-labor. As Marx writes: The wage form thus extinguishes every trace of the division of the working day into necessary labour and surplus-labour, into paid labour and unpaid labour. All labour appears as paid labour. Under the corvee system it is different. There the labour of the serf for himself, and his compulsory labour for the lord of the land, are demarcated very clearly both in space and time. (Capital I: 680) Marx’s fundamental difference from Ricardo on the question of labor-time as a unit of measure is situated here. It was his critique of Ricardo’s theory of wages that had led him to proclaim: “The result the analysis led to, therefore, was not a resolution of the problem as it emerged at the beginning, but a complete change in the terms of the problem” (Capital I: 677-79).’ In Ricardo the theory of wages refers and not per hour of labor: to wages per worker (and “his family”) “The natural price of labour, therefore, depends on the price of the ’ I have borrowed this quotation from Althusser and Balibar (1970: 23), who had In the reference from Marx given above the quoted Marx from the French translation. wording is slightly different. 274 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 food, necessaries, and conveniences required for the support of the labourer and his family” (Ricardo 1821: 93, emphasis added). This implies that the translation of the given wages into wage rate per hour of labor must necessarily assume that the length of the working day is also given from outside. This, for Marx, was the fundamental problem with Ricardo’s problematic. Marx develops his critique of Ricardo by showing that Ricardo treats real wages as the cost of (re)producing the laborer (labor-power) as well as the labor input used in the production process. He points out that labor and the laborer (labor-power) are not the same thing. Though wages represent the cost of production of labor-power, there is nothing in the determination of wages that tells us about the amount of labor input that could be drawn out of it. In other words, it was not legitimate for Ricardo to leave the question of the length of the working day out of the scope of the theory (ibid. 677ff; TSV II: ch. XV.B). Thus, contrary to Garegnani’s (1991) position, it seems that in Marx’s case the choice of labor-time as the unit of measurement of economic variables was not based on his helpless situation of having no better way of showing the constraint binding on profits, but because the central aspect of his theory-the notion of exploitative relation-was defined in terms of labor-time, i.e. as a relation between necessary and surplus labor. As Marx remarked: . . .it is clear that though the existence of surplus-Zabour presupposes that the productivity of labour has reached a certain level, the mere possibility of this surpluslabour (i.e., the existence of that necessary minimum productivity of labour), does not in itself make it a reality. For this to occur, the labourer must first be compelled to work in excess of the (necessary) time, and this compulsion is exerted by capital. This is missing in Ricardo’s work, and therefore also the whole struggle over the regulation of the normal workingday. (TSV II: 406) 4. The Transformation Problem and the Standard Commodity In Marx’s judgment Ricardo’s preoccupation with variations in prices due to changes in real wages laid unnecessary emphasis on a secondary problem. Commenting on Ricardo’s discussion on an invariable measure of value, Marx wrote: “This section VI On an Invariable Measure of Value deals with the “measure of value” but contains nothing important. The connection between value, its immanent measure-i.e., labour time-and the necessity for an external measure of the values of commodities is not understood or A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 275 even raised as a problem” (TSV II: 202). So for Marx it was Ricardo’s weakness that he failed to distinguish between “value” and “valueform.” Ricardo’s “values” are simply the one form-i.e. the moneyform-in which the value, the substance, appears, thus Ricardo’s inability to relate the substantial relations of the capitalist economy to its appearances. In Capital ZZZ,Marx devotes a short chapter (chapter XI) exclusively to Ricardo’s problem on Ricardo’s own terms, and concludes by saying: “This is a very secondary question compared with the other important points which have been dealt with in this part” (306). It should be noted that Marx was well aware of the fact that his distinction between value and value-form did not solve the problem of “average commodity” as such: “It is quite possible, accordingly, for the cost price to diverge from the value sum of the elements of which this component of the price of production is composed, even in the case of commodities that are produced by capitals of average composition. Let us assume that the average composition is 80~ + 20~. It is possible now that, for the actual individual capitals that are composed in this way, the 80~ may be greater or less than the value of c, the constant capital, since this c is composed of commodities, whose prices of production are different from their values.. .” (Capital III: 309). Thus, it is safe to conclude that Marx did not think that he had solved Ricardo’s problem, but simply considered it to be a “secondary” or a minor problem. This notwithstanding, Eatwell (1975) argues that Sraffa’s Standard commodity is also a solution to Marx’s transformation problem. His argument rests on the claim that Marx defines the “necessary labor time [. . .] in two ways: (i) as the value of ‘the sum of money v expended upon the labor power, ’ in effect as the share of wages in the value of output, and (ii) as ‘the value of [the] means of subsistence,’ that is, as the value of the commodities comprising the real wage” (Eatwell 1975: 550). The second definition is consistent with our objective measure discussed in the previous section. Eatwell, however, prefers the first definition. As long as the labor theory of value prevails, the two measures would be identical; however, in the case in prices diverging from labor-value ratios, which is the case of the transformation problem, the two measures will diverge. In this case, Eatwell suggests that the rate of exploitation could be defined as the profit-wage ratio measured in money terms. Such a rate of exploitation, however, will have a serious shortcoming. The “money” rate of exploitation will be dependent on the composition of the net output. That is, even though technology, money wages, and the total labor-time spent in the production process remain the same, just the change in the allocation of labor would change the rate of exploitation.” This, for I0 This point scholars, Dumenil is crucial in understanding the transformation problem. Some (1984), Lipietz (1982), Foley (1982), have argued that the value of 2’76 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 Eatwell, was at the root of the transformation problem. But thanks to Sraffa, this problem can now be solved by using the Standard commodity as the money commodity, i.e., money wages being expressed in terms of the Standard commodity. Sraffa’s analysis of the Standard system and the Standard commodity has shown that a direct relation between wages and profits of any actual system is given by r = R(1 - w*), where r is the rate of profits, R is the maximum rate of profit, and w * is the “money wage,” when the Standard commodity is chosen as the “money commodity” or the numeraire. This prompts Eatwell to conclude: “In other words, if the rate of exploitation is defined as 1 minus the proportion of total labor embodied in the ‘money’ wage, this is equivalent to the rate of exploitation in the production of an ‘average’ commodity, the standard commodity, and may thereby be directly related to the rate of profit in the system as a whole. So the rate of exploitation e=(l-W*)/W* [for a given technology] is unambiguously related to the rate of profit r = R{e/(l + e)}” (Eatwell 1975: 555).” variable capital and surplus-value should be calculated by the “value of money,” where the “value of money” is defined as the total direct labor-time spent in a “year” divided by the total money price of the net output. This makes the rate of exploitation not only dependent on prices of production, but also dependent on the composition of the net output. For example, a change in the pattern of capitalists’ consumption would change the rate of exploitation in the system, everything else remaining the same. This would be quite foreign to Marx’s notion of exploitation, which is defined at the level of production, and must be independent of the composition of the net output. Moreover, this approach destroys the concept of commodity-value as made up of C , v, and s , since the two elements v and I are measured differently than C For a detailed critique of this approach see Sinha (1997). ” Roncaglia (1978: 79) has criticized Eatwell on the ground that for Marx the money-commodity was a historical product, and therefore the Standard commodity cannot be substituted for Marx’s money-commodity. This, in my opinion, is not a serious blow to Eatwell’s position, since all Eatwell is claiming is that the Standard commpdity can be used as a theoretical device to draw a direct link between the wageprofit ratio and the rate of exploitation. The more serious problem with Eatwell’s position, as I shall argue, is that his definition of exploitation is not a correct representation of Marx’s notion of exploitation. KLUZ and Salvadori (1987: 875, f.n. 6) have also criticized Eatwell’s use of the Standard commodity as a “solution” to the transformation problem. They argue that “the main mistake” in Eatwell stems from not taking into account the fact that the ratio of total profits to total wages (measured in terms of the Standard commodity) of the real system would generally differ from the ratio of the two aggregates in the Standard system. This again seems to be an off the target criticism. Eatwell nowhere is making such claim that in his system Marx’s condition that total profits is equal to total surplus value would hold. All he is saying is that a rigorous relation between the rare of A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 277 Before discussing Eatwell’s reasons for choosing the first definition of necessary labor in the context of the transformation problem, let me point out that he has not provided any serious evidence to show that this was Marx’s position as well. It was customary for Marx to translate the labor-value measures in terms of the money commodity on the assumption that the labor theory of value holds. This was solely to simplify his presentation to the reader. It would be incorrect to think that Marx defined the “necessary labor-time” in terms of money wages paid to the workers. It would in fact be inconsistent with the definition of value, since “necessary labor” is the value of laborpower, and “the value of labor-power is determined, as in the case of every other commodity, by the labor-time necessary for the production, and consequently also the reproduction, of this specific article” (Capital I: 274, emphasis added). Measuring the value of labor-power by the value of the money-wages would be equivalent to measuring the value of a commodity with the value of money given for it. The value of a commodity, however, is not determined by the value of money given for it, but by the labor-time it takes to produce the commodity. Thus, the value of labor-power should consistently be measured by the labor-time it takes to produce the consumption of the workers, which (re)produces the labor-power. In the passage quoted above from Capital III: 309, where Marx explains why the value-price divergence would also apply to the average composition of capital industry, he goes on to add: “The 20 v can similarly diverge from its value, if the spending of wages on consumption involves commodities whose prices of production are different from their values, The workers must work for a greater or lesser amount of time in order to buy back these commodities (to replace them) and must therefore peqform more or less necessary labour than would be needed if the prices of production of their necessary means of subsistence did coincide with their values” (Capital III: 309, emphasis added). Here we have a clear statement from Marx defining “necessary labor” as an objective measure, the labor-time required to replace the workers’ means of subsistence, and not a “monetary” measure as share of the value of net output. Eatwell’s interpretation of Marx’s rate of exploitation is incorrect. Now, let us briefly discuss the reasons provided in favor of using the “money wages” as given rather than the “real wages.” Eatwell provides three reasons: (1) it is easier to conduct an analysis of the effect of changes in wages on the rate of profits in terms of money wages rather than the real wages, since it is particularly difficult to represent changes in real wages as scalar multiples; (2) the real world possibility of change or difference in the composition of real wages creates difficulty in interpreting the rate of exploitation (1975); and (3) profits and the rate of exploitation the Standard commodity. (as defined by him) can be established by the use of 278 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 it is convenient to take the money wages as given, because it is independent of a theory of wages (1974). The first point is irrelevant to the transformation problem, because it is not concerned with the issue of the changes in real wages and their effect on the rate of profits, as we have argued in this paper. The second point is correct, but it should be noted that Marx’s analysis is of a highly aggregative nature. And if all the wage goods could be aggregated into a wage good sector, Eatwell’s problem will vanish. In any case, however, Marx’s contention was that though composition of wage goods differ from culture to culture, in one given culture at a particular time the life style of the working class is usually quite uniform (Marx 1976). For minor differences from worker to worker such as dietary habits, etc., one could easily assume these commodities to have equal organic composition of capital, and treat them as one commodity for theoretical purposes (Morishima 1973). The third point goes to the very heart of the problem. It was important for Marx to have a theory of wages, so that the surplus labor-time could be independently determined prior to the determination of prices. Marx held that the life style of the working class is determined by long-term socio-historical forces (Marx 1976). Sraffa (1960: 33) showed ambivalence about taking the real wages as determined from outside the price system mainly because his analysis was concerned with the effects of variation in wages. It should be noted, however, that in Marx’s analysis the value of labor-power is determined on the basis of the value needed to produce the working class. The value of labor power is determined on the basis of the life-cycle of the workers and their families. The real wage rate is then derived by taking into account the length of the working day as well as the working life and life expectancy of an average worker, the average intensity of work, etc. All these variables are determined in the historical class struggle. Marx’s point is not that an individual worker cannot save at any given point in time. His basic point is that for the working class as a whole there cannot be persistent positive saving, i.e., the working class cannot go on bequeathing property from one generation to another. If the working class as a whole could go on saving persistently, then, in the long run, they would not remain the propertyless proletariats, and the whole basis of the capitalist mode of production would collapse. Thus, the assumption that workers consume all the wages is a sound assumption in Marx’s theoretical system. 5. Conclusion In this paper we have refrained from coming up with “yet another solution” to the transformation problem. Instead, we have argued that, contrary to Garegnani’s position, Marx’s use of labor-time as a unit of A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 279 measure is based on his notion of exploitation, which can only be understood in terms of division of labor-time into necessary and surplus labor-time. The Sraffian reformulation of the notion of exploitation as a ratio of profit to wages falls short of representing the full meaning of Marx’s notion of exploitation. Although it is true that Marx’s economics is “surplus approach” economics, and that Sraffa as well as Eatwell, Garegnani and other Sraffians have contributed greatly in rehabilitating Marxian economics, it would be incorrect to suggest that Sraffa’s Standard commodity is a solution to Marx’s transformation problem, as Eatwell has done. This conclusion we have drawn from a close reading of both Marx and Sraffa, where it is clear that both the authors did not consider Ricardo’s problem of the “invariable measure of value” as the same as the “transformation problem.” Moreover, contrary to Eatwell’s suggestion, we have argued that it is highly problematic to use the money wages as given, and that it is much more meaningful to maintain Marx’s practice of taking the real wages as given in his theoretical framework. Acknowledgments Many thanks to Riccardo Bellofiore, Geoff Harcourt, and John King for valuable suggestions on earlier drafts of this paper. I also wish to thank the three referees of this journal, David Andrews, Doug Koritz, and Gary Mongiovi, for their valuable suggestions and criticisms. It goes without saying that none of them is responsible for all the shortcomings of this paper. References Althusser, L. and E. Balibar. 1975. Reading Capital. London: New Left Books. Bortkiewicz, A. von. (1907) 1949. On the Correction of Marx’s Fundamental Theoretical Construction in the Third Volume of Capital. In Karl Marx and the Close of his System. P. Sweezy (ed.). New York, N.Y.: Augustus M. Kelly. Caravale, G.A. 1985. Diminishing Returns and Accumulation in Ricardo. In The Legacy of Ricardo. G.A. Caravale (ed.). Oxford: Basil Blackwell. 1991. Marx’s Views on Ricardo-A Critical Evaluation. In x.x and Modern Economic Analysis Vol. II. G.A. Caravale (ed.). Aldershot: Edward Elgar. Caravale, G.A. and D. Tosato. 1980. Ricardo and the Theory of Value, Distribution and Growth. London: Routledge and Keegan Paul. 280 A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 Casarosa, C. 1978. A New Foundation of the Ricardian System. Oxford Economic Papers 30: 38-63. 1985. The “New View” of the Ricardian Theory of Distribution and Economic Growth. In The Legacy of Ricardo. G.A. Caravale (ed.). Oxford: Basil Blackwell. Dumenil, G. 1983-84. Beyond the Transformation Riddle: A Labor Theory of Value. Science and Society 47 (4): 427-50. Eatwell, J. 1974. Controversies in the Theory of Surplus Value: Old and New. Science and Society 38 (3): 281-303. 1975. Mr Sraffa’s Standard Commodity and the Rate of Exploitation. Quarterly Journal of Economics 89 (4): 543-55. Foley, D.K. 1982. The Value of Money, The Value of Labor Power, and the Marxian Transformation Problem. Review of Radical Political Economics 14 (2): 37-47. Garegnani, P. 1984. Value and Distribution in the Classical Economists and Marx. Oxford Economic Papers 36 (3): 291-325. 1991. The Labour Theory of Value: “Detour” or Technical Advance? In Marx and Modern Economic Analysis, Volume I. G. Caravale (ed.). Aldershot: Edward Elgar. Harcourt, G.C. 1990. On the Contribution of Joan Robinson and Piero Sraffa to Economic Theory. In Political Economy in the Twentieth Century. Maxine Berg (ed.). London: Philip Allen. Hicks, J. 1985. Sraffa and Ricardo: A Critical View. In The Legacy of Ricardo. G.A. Caravale (ed.). Oxford: Basil Blackwell. Hicks, J. and S. Hollander. 1977. Mr Ricardo and the Moderns. Quarterly Journal of Economics 91: 351-69. Hodgson, G.M. 1982. Capitalism, Value and Exploitation: A Radical Theory. Oxford: Martin Robertson. 1991. After Marx and Sraffa: Essays in Political Economy. London: Macmillion. Hollander, S. 1979. The Economics of David Ricardo. Toronto: University of Toronto Press. 1992. Classical Economics. Toronto: Toronto University Press. Kurz, H.D. and N. Salvadori. 1987. Burmeister on Sraffa and the Labor Theory of Value: A Comment. Journal of Political Economy 95 (4): 870-81. 1993. The “Standard Commodity” and Ricardo’s Search for anInvariable Measure of Value.” In The Dynamics of the Wealth of Nations. M. Baranzini and G.C. Harcourt (eds.). New York, N.Y.: St. Martin’s Press. Lipietz, A. 1982. The So-Calied “Transformation Problem” Revisited. Journal of Economic Theory 26: 59-88. Marx, K. (1867) 1977. Capital, Vol. I. New York, N.Y.: Vintage. (1885) 1981a. Capital, Vol. II. New York, N.Y.: Vintage. -1 (1894) 1981b. Capital, Vol. III. New York, N.Y.: Vintage. A. Sinha / Review of Radical Political Economics 32, 2 (2000) 265-281 28 1 (1861-63) 1968. Theories of Surplus Value, part II. Moscow: Tgress Publishers. (1865) 1976. Value, Price, and Profit. New York, N.Y.: Inter-. national Publishers. Morishima, M. 1973. Marx’s Economics: A Dual Theory of Value and Growth. Cambridge: Cambridge University Press. Morishima, M. and G. Catephores. 1978. Value Exploitation and Growth. London: McGraw Hill Ltd. Ong, Nai-Pew. 1983. Ricardo’s Invariable Measure of Value and Sraffa’s “Standard Commodity.” History of Political Economy 15: 207-27. Porta, P.L. 1986. Understanding the Significance of Piero Sraffa’s Standard Commodity: A Note on the Marxian Notion of Surplus. History of Political Economy 18 (3): 443-54. Rankin, S. 1987. Exploitation and the Labour Theory of Value: A Neo-Marxian Reply. Capital and Class 32. Ricardo, D. 1821. On the Principles of Political Economy, and Taxation. Cambridge: Cambridge University Press [1951]. Roncaglia, A. 1978. Sraffa and the Theory of Prices. Chichester: John Wiley and Sons. Seton, F. 1957. The “Transformation Problem.” Review of Economic Studies 24 (3): 149-60. Shaikh, A. 1973. Theories of Value and Theories of Distribution. Ph.D dissertation: Columbia University. Sinha, A. 1990. The Concept of Value in Marx: A Reinterpretation. In Research in Political Economy Vol. 12. Paul Zarembka (ed.). Pp. 41-79. Greenwich: JAI Press Inc. 1991. The Concept of Value in Marx’s Economic Writings: A State University of New York at -??t/aue. Ph.D dissertation: Buffalo. 1996. A Critique of Part One of Capital Volume One: The Value Controversy Revisited. In Research in Political Economy Vol. 15. Paul Zarembka and Ajit Sinha (eds.). Pp. 191-218. Greenwich: JAI Press Inc. 1997. The Transformation Problem: A Critique of the “New Solution.” Review of Radical Political Economics 29 (3): 51-58. Smith, A. (1776) 1930. An Inquiry into the Nature and Causes of the Wealth of Nations. London: Methuen and Co. Ltd. Sraffa, P. 1951. Introduction to The Works and Correspondence of David Ricardo, Vol. I. Cambridge: Cambridge University Press. 1960. Production of Commodities by Means of Commodities. -bridge: Cambridge University Press. Steedman, I. 1975. Value, Price, and Profit. New Left Review 90 (March-April): 71-80. 1977. Marx after Sraffa. London: NLB. -1 1981. Ricardo, Marx, Sraffa. In The Value Controversy. I. Steedman (ed.). London: New Left Books.