Pure and fading 'variable capital'

a immanent critique of a core concept of Marx's and marxian theory Geert Reuten * University of Amsterdam, Department of Economics <[email protected]> <http/www.feb.uva.nl/pp/greuten> draft 16 May 2007 [ISMT-15.c07 ]

Abstract Marx's concept of 'variable capital' is appropriate to the aims of Capital Is level of analysis. It is inadequately concretised in Capital II, Part Two (the turnover time of circulating capital). That text lacks the distinction between turnover time and wage time and so cannot grasp their opposite movements which turn 'variable capital' into a fading category. It also lacks the category of 'capital investment time'. The result is a defective rate of profit formula (Capital III). These flaws are traced to Marx's implicit inputoutput framework. Finally, the category of circulating capital is reconstructed, fitting both the general theoretical systematic of Capital and its counterpart in the balance sheet of enterprises. Key words: variable capital; circulating capital; turnover time; Marx's treatment of time JEL classifications: B14, B24, E11, E22

I am grateful for critical comments on earlier versions of this paper by Chris Arthur, Riccardo

Bellofiore, Patrick Murray, Tony Smith, Dirk Damsma and Murat Kotan. I thank Peter Thomas for polishing my English without implying him in my style or the final result.

Contents Introduction ..................................................................................................... 1 1. Labour-time and Marx's treatment of capital investment time .............. 1.1. Capital I: abstraction from the time aspect of capital.......................... 1.2. Time in terms of turnover time of circulating capital.......................... 1.3. Projection of the notion of variable capital into Capital III ................ 1.4. Conclusions ......................................................................................... 2 2 4 6 8

2. 'Variable capital' and the pure form for 'wage time'............................... 8 3. The turnover of capital: an analysis and further critique of Capital II, Part Two ............................................................................................. 3.1. Marx on the turnover of variable capital ........................................... 3.2. The neglect of time of capital investment .......................................... 3.3. Marx's implicit input-output framework for the turnover of CC....... 4. A concise reconstruction ........................................................................... 4.1. Circulating capital and its fading variable capital component........... 4.2. Implicit and explicit precautionary stocks......................................... 4.3. Total capital investment and the rate of profit...................................

11 11 13 16 19 19 22 23

5. A corollary for 'the' transformation problem......................................... 24 Summary and Conclusions ........................................................................... 25 References ....................................................................................................... 26

Introduction Enterprises, generally, do not 'advance' the wages equivalent to labourers. It is the other way around. Because wages are paid only after some period of performance of labour, labourers provide an implicit 'production credit' to enterprises. Although Marx was aware of this, it is nevertheless not shown in his notion of 'variable capital' as presented in his Capital. One purpose of this paper is to detect the reasons for this; another is to provide a reconstruction of this key concept in Marx's and marxian theory. Whereas this advance by labour is generally the case, its extent increases with the decrease in the production and sales time of commodities (that is, the decrease in the turnover time of circulating capital), and with the increase in the wage time (end of day, week, month). Today, for the modal enterprise at least, investment in variable capital has faded away (it has not faded away in, most importantly, construction). Marx did not foresee the extent of these changes. Nevertheless, he was aware that both of these do change, the first one (change in turnover time) being the major subject of the 200 pages long Part Two of Capital II. This enlarges the variable capital puzzle. I begin in Section 1 by briefly reviewing Marx's treatment of 'time' in the middle parts of Volumes I and II of Capital. Alongside, I outline Marx's presentation of those concepts that are the main subject of the paper. Section 2 sets out the main problematic of the paper the opposite movements of the turnover time of circulating capital and of the wage time both in empirical and in theoretical terms. Against this background, Section 3 undertakes a more substantial analysis of Capital II, Part Two (turnover time) in order to detect its theoretical shortcomings. Its main conclusion is that Marx adopts a notion of capital in which the input costs (for means of production and labour-power) coincide with the investment of capital. His focus is on the turnover time of the former (inputs) at the neglect of the time aspect of capital advanced, or capital investment. Marx wrongly believed that his abstraction from credit at that point should have a parallel in abstracting from the time aspect of capital investment. Taking account of these shortcomings, the Section 4 presents a reconstructed concept of circulating capital, one that fits not only the general theoretical systematic of Capital, but also its empirical counterpart in the assets side of the balance sheet of enterprises. The final Section 5 is a brief note on some implications for the (in)famous transformation problem. *****

I abbreviate page references to Capital. "G" refers to the German edition (MEW). "F" refers to the English FowkesFernbach translations (Penguin). "U" refers to the English Untermann translation of Volume II (Lawrence & Wishart).

1. Labour-time and Marx's treatment of capital investment time 1.1. Capital I: abstraction from the time aspect of capital In the middle part of Capital I (Parts Three to Six), Marx presents the process of production of capital. Its Chapter 9 (F: 320) opens as follows:1 'The surplus-value generated in the production process by C, the capital advanced, ... presents itself to us first as the amount by which the value of the product exceeds the value of its constituent elements. The capital C is made up of two components ...' (emphasis added) Marx then introduces the following formula for what he calls, 'capital advanced':2 C=c+v (1)

where c is constant capital, laid out on means of production, and v variable capital, expended on labour-power all in monetary dimension.3 Marx indicates (F: 321; cf. 311-13) that constant capital encompasses a sum of money for 'wear
1 2

This is Chapter 7 of the German edition (M: 226). Marx uses (in German) the symbol C for capital advanced, which the translator has taken over in

English (and that I have underlined as C). It should not be identified with the commodity shape of capital C (in the German text, with W for Ware = commodity, there is no room for this confusion).
3

For mainstream economics, components of capital are called 'variable' when they vary directly

with the rate of output (also called variable costs); components are constant when they are invariable to the rate of output (also called fixed or constant costs). For Marx and marxian economics, the criterion for 'variable' is the value augmentation or 'valorisation'. Thus 'variable capital' refers to the component of capital that is laid out on labour-power and which is theorised to be the source of value augmentation. On the other hand, the component of capital laid out on any means of production is called 'constant capital' in that in the course of production its value remains merely conserved. (Capital I, F: 317.) In other words, the value of variable capital turns into value-added; the value of the constant capital is just reproduced. For the mainstream notions referred to above, Marx would rather use the terms 'circulating capital' and 'fixed capital' (see further section 1.2)

and tear' of machinery. Thus the implication is that total capital advanced (as including the total fixed capital outlay on machinery and buildings) is larger than C. In the rest of Capital I, however, this issue moves to the background.4 Given Marx's method and given the objective for the middle part the explanation of surplus-value this is fine. On basis of the method, the blending out of this point is also appropriate for the end part of Capital I where the accumulation of capital is presented as the transformation of surplus-value into capital.5 The same might seem the case for another simplification that Marx applies, namely, for variable capital. Note that he indicates in Chapter 4 that, because wages are usually paid at the end of some period of work (end of the day or week), it is rather that, to some extent at least, 'the worker advances the use-value of his labour-power to the capitalist. ... Everywhere the worker allows credit to the capitalist.'6 In terms of the formula above, a sum of money v is nevertheless expended on wages at some point, and in Marx's view of the matter it is advanced. In fact, Marx's treatment of capital in Volume I abstracts from the time aspect of capital advanced, for both constant and variable capital. The explicit notion of time is that of labour-time and labour-time is considered in reference to the time necessary for a process of production of commodities. And of course the labour-time is not affected by the wage time (end of day or week). Nevertheless, when reading Volume I we do not know because it is implicit that Marx abstracts from the time aspect of capital advanced (or, what I call 'capital investment').7 In sum, in Capital I Marx abstracts from the time aspect of capital advanced. This is appropriate in terms of his objective of the explanation of surplus-value in the same Volume.

Most of the time Marx abstracts from fixed constant capital. 'When we refer ... to constant capital

advanced for the production of value, we always mean the value of the means of production actually consumed in the process, unless the context demonstrates the reverse' (1867F: 321; cf 1867G: 227).
5

Marx's objective and proceeding in the middle part of Capital I is discussed in Reuten (2004a); Capital I, Ch.6, F: 278; German edition Ch.4, section 3, M: 188. Thus the term 'capital investment' which also refers to the total capital accumulated by

the final part is discussed in Reuten (2004b).

enterprises aims to make this time aspect explicit. Bellofiore and Realfonzo (2003) treat another problem about 'capital advanced', which is not the subject of this paper; it does not have to do with 'time', but with bank finance, an issue that Marx abstracts from until Capital III, Parts Four and Five.

1.2. Time in terms of turnover time of circulating capital Capital Is abstraction about the time aspect of capital becomes explicit in Capital II.8 The intermediate step is the introduction of the capital circuit of M C ... P ... C' M' (2)

(as well as the analogous circuits) in Part One of Volume II. The time aspect of the capital circuit (the period of time from M to M') almost inevitably leads on to the introduction of the turnover time of capital (in Chapter 5 of Part One), which is the subject of Part Two of Volume II. Key to it is the notion of circulating capital, and hence the distinction between the latter concept and total 'capital advanced' as including fixed capital.9 Fixed capital (F) is the part of total capital that is invested in those means of production that last by convention for over a year (e.g. buildings and machinery). Let the yearly depreciation of F (wear and tear) be a fraction F (e.g. =1/12 when it depreciates in 12 years time).10 Now consider that within each year a number of n production processes are carried out (we might for example have n=6 production processes, that are simply carried out subsequently, each one taking two months). Thus F is the value-fraction of F used up, and reproduced, in one year, over as many production processes as there are carried out (i.e. n).11 Then F/n is the fixed capital component of constant capital (c) in each production process, which is also replaced (on average) in each process. Its other component is raw materials and instruments used up in a process (RMI). Then Marx's notion of 'constant capital' is represented by equation (1), which, like all equations below, is expressed in monetary prices ( etc). c = F/n + RMI (tot is explained below.)
8

[per tot]

(3)

It becomes explicit when the exposition requires explicitation. Note that this procedure is akin to As indicated in footnote #4, Marx most of time abstracts from fixed capital. However, the The factor is the average of all the components of F (plant, or structures, and all the types of Although theoretically we should call an instrument of production 'fixed' when it lasts beyond

the method of systematic dialectics (see e.g. Reuten 2000: 140-52).

Chapters 8-11 (of Capital II, Part Two) are almost fully devoted to fixed capital.
10

equipment).
11

one production process, we (Marx, capitalists) adopt the convention of taking the year as a reference point. Marx also distinguishes instruments of labour that are not 'fixed' according to this convention, but that nevertheless last for several production processes within a year.

Although Marx does not exactly produce this equation, it is a fair representation of the text. The text also suggests that we have for the circulating capital (CC): CC = c + v [per tot] (4)

Marx's concept of circulating capital denotes a quantity of capital (CC) in reference to the time required for one process of production as well as the sales of commodities (e.g. 2 months together).12 Marx calls the sum of the production time (pt) and the sales time (st), the turnover time of circulating capital (tot).13 Thus CC is calibrated to this turnover time. Now moving to total capital laid out per year (TC), we would have:14 TC = F + c + v [per year] (5)

(Note that the F/n component of c is, on average, required as 'working capital' for the replacement of F. Thus F/n stands not merely for an accounting device but for actual 'replacement investment'.) Each one turnover there is a production of surplus-value of: s = s'v [per tot] (6)

where s' is the rate of surplus-value (or the rate of valorisation of variable capital) as explained in the middle part of Capital I. Hence for the surplus-value produced per year (S) we have: S = sn [per year] (7)

(Cf. Marx's formula in Capital II, F:383; explicit versions of the equations above can also be found in Capital III, Parts One to Three.) Thus it seems that the Volume I notion of 'capital advanced' is left untouched. This is especially so because most of the time (in all of Capital) Marx adopts the simplifying
12

Although equation (1) is apparently similar to equation (4), I use different symbols for

(circulating) capital, so leaving open a conceptual distinction. Nevertheless, in hindsight is seems that the notion of capital in Capital I is that of circulating capital.
13

In fact, Marx calls the time it takes to sell commodities the 'circulation time'. Hence he has

Production time + Circulation time = Turnover time. Given the concept of 'circulating capital', which relates to the turnover time, this terminology is confusing. Marx also makes a distinction between 'working time' and 'production time'. For the purposes of this paper I let these two coincide.
14

Here I abstract from the growth of CC during the year.

assumption of F=0. This simplification is justified to the extent that there is some quantitatively stable connection between F and c. There may also be theoretical reasons for this simplification. In all, or most of, Part Two of Capital II, the circulating capital advanced is framed in terms of an advance of c plus v. The focus of the presentation is on the length of the turnover time as well as on the factors contributing to a more efficient turnover (increasing turnover).15 Evidently an increasing turnover (n rises) generates, ceteris paribus, an increasing production of surplus-value per year (see equation 7). Note that the rate of surplus-value (s') proper is not Marx's concern here, as it has been analyzed in Capital I. What receives attention is the fact that given a synchronically uniform rate of surplus-value (say, per hour) between enterprises (in the same or in different branches), the yearly surplus-value produced for capitals of equal magnitude differs according to their turnover time. (Thus rates of profit diverge according to differences in turnover time analogous to the divergence of rates of profit according to different organic compositions of capital, as treated in Capital III, Part Two.) Marx discusses this matter in a Chapter entitled 'The Turnover of Variable Capital' (Capital II, Chapter 16). In fact, this chapter is not about the turnover of variable capital specifically. Rather it is about different turnover times of circulating capital (c+v), given equal c/v ratio's. The matter of the turnover of variable capital as an 'investment of capital' is in fact not discussed at all in Capital II, Part Two. Before taking this up in sections 2 and 3, I briefly show how the Capital III notion of variable capital is similar to that of the one discussed above. 1.3. Projection of the notion of variable capital into Capital III We saw in the pervious sub-section that Marx's notion of variable capital in Capital II is not essentially different from that in Capital I. This subsection briefly shows that the same notion is projected into Capital III. I offer merely a couple of examples. In Part One, Chapter 2, when introducing the rate of profit, Marx continues to treat variable capital as an advance: 'Since the capitalist can exploit labour only by advancing constant capital, and since he can valorize the constant capital only by advancing the variable, these are both one and the same in his eyes ... the actual degree of his profit is determined not in relation to his variable capital but to his total capital; ...' (Marx 1894F: 133; emphasis added)
15

See Smith (1998) for a discussion of this part in connection with 'just-in-time production'.

In the next citation we see not only Marx's abstraction from fixed capital in operation (this is not my worry), but we see also the variable capital component in the denumerator for the rate of profit (which is worrying).16 Thus my worry, further expanded upon in the next section, is that Marx identifies circulating capital with (the circulating part of) capital advanced and so identifies equations (1) and (4). (I replace the translator's C by C.) 'This excess [surplus-value] then stands in a certain ratio to the total capital, as expressed by the fraction s/C, where C stands for the total capital. We thus obtain the rate of profit s/C = s/(c+v), as distinct from the rate of surplusvalue s/v.' (Marx 1894F: 133) At the beginning of Chapter 3 (F: 141) and then throughout that chapter we see this repeated: s = C s c+v s'v C

p' =

(8)

where p' stands for the rate of profit and s' for the rate of surplus-value. On the next page (F: 142) Marx states that he will neglect turnover. Engels sees that there is a problem and inserts the remark that 'the formula p' = (s'v)/C is strictly correct only for a single turnover period of the variable capital, while for the annual turnover the simple rate of surplusvalue s' has to be replaced by s'n, the annual rate of surplus-value, n standing for the number of turnovers that the variable capital makes in the course of a year ...' (Engels 1894F: 142)17 Hence, as Engels (1894F: 167) sets out in his Chapter 4 of the book, for several turnovers we have:18

16 17

In section 3 we will see that there are also problems with the constant capital component. Following on from this, all of the next chapter, 'Chapter 4: The effect of the turnover on the rate I have replaced (s'n)v for (s'v)n. Formula (8a) in the English text (Penguin) contains a printing

of profit', is inserted by Engels.

error (cf. 1894G: 85).

p' =

sn (s'v)n = C C

(8a)

Engels (F: 166) also specifies C as including total fixed capital (F my symbol), and proposes TC = F + c + v, that is, in the form of a numerical example (compare equation 5 above). Similar formulas (abstracting from fixed capital and turnover time) are produced in Part Two (cf. F:254 where Marx states the two assumptions) and in Part Three (see especially its opening page and page F:324). 1.4. Conclusions Marx (and Engels) thought that the turnover time of capital has effect on the expression for the numerator of the profit rate (correct) and that, generally, capitalist development is such that the turnover time is speeded up, which has an upwards effect on the rate of profit (correct).19 However, the time aspect of capital investment also has an effect on the denumerator for the rate of profit, and especially also on the component of 'variable capital'. I now turn to this problem.

2. 'Variable capital' and the pure form for 'wage time' I discuss Marx's neglect of the time aspect of capital investment in two stages. In the current section I proceed by staying close to Marx's account of time, that is, turnover time. I merely connect turnover time with 'wage time'. In section 3 I come to investment time proper. There is something strange about Marx's notion of variable capital. As indicated in section 1.1, Marx is well aware that labourers 'to some extent' advance their labour-power to capitalists in contradistinction to capitalists 'to some extent' advancing wages to labour. On the one hand, this is an empirical matter (it makes quite a difference when wages are paid at the end of the month instead of at the end of the day; or when the average production process takes a couple of weeks instead of a couple of months). On the other hand, this is a theoretical matter of deciding what is the pure form. I take the empirical matter first. Marx probably had in mind that wages are paid at the end of the day, with the custom for it being on the edge of moving to
19

This is correct independently of other factors that might perhaps push the rate of profit down.

Engels posits this in his Chapter 4 of Capital III. Marx mentions it in Chapter 14 of the same book. However, there is no thorough interconnected treatment of the capital composition and turnover time.

at the end of the week (see Capital I, Chapter 20, e.g. F:684-5).20 Probably he had further in mind that the modal production process takes months rather than weeks (all examples and other evidence provided in Capital II, Part Two point at this). If wages are paid by the day then the formula for circulating capital provided in section 1.2 is a pretty good first approximation: CC = c + v [per tot] (4)

For each one production process we neglect the one day wages advance of labour to enterprises. If wages are paid by the week this becomes somewhat problematical as a first approximation. (If the turnover time is one month the wages advance is about v so invalidating (4).) Marx had in mind that for the modal machinal-manufacture the turnover time of circulating capital (tot) 'largely' extends the wage time (wt), which was probably correct for advanced capitalist economies in the last decades of the 19th century. In our day for capitalistically advanced economies it is especially important to distinguish between forestry, agriculture and construction on the one hand, and all other output categories on the other hand. In the USA today, for example, the last category makes up about 90% of the total output. For this category, the part of the total capital invested in wages is approaching zero. At the beginning of the 21st century, the sales time has been squeezed ('just in time production') and the production time has been and is being reduced enormously (the production of cars, e.g., is down to under a day's time). On the other hand, the wage time has made a gigantic move up, and is being moved up, to a month for most occupations. (Thus whereas in Marx's days the modal tot > wt, we have now rather a modal tot < wt.)21 In any case, a more general formula for circulating capital would be: CC = c + v [per tot; = (tot wt)/tot] (9)

I will expand on this formula in section 4. For now it suffices to say that if =1 this equation reduces to Marx's equation (4); and that empirically today for the modal enterprise <1.

20 21

German edition Chapter 18 (M: 565-6). I say modal because this is different in forestry and parts of agriculture and especially because

this is different in construction (houses, industrial buildings, ships); the latter category has an important weight in the total (for the USA today the weight in the total GDP of construction is 6.8%, that of agriculture 1.4% and that of forestry, including fishing, 0.4%).

I now take the theoretical matter. Considering its end result (i.e. regarding its final stage of development) the theory in Capital is wrong (equation (4), i.e. =1, is wrong). In order to improve on the theory, the first question to answer is about the 'pure form' for the theory (in this case the pure form for ). For Marx at least, this is always the first question to answer. Then the question is if Marx's starting point (=1) was right or wrong. Usually for difficult theoretical matters, Marx explicitly uses the term 'pure form' or 'pure theory'. In this case, however, he doesn't to my knowledge. As this matter is core to the theory in Capital generally, I infer that Marx was not aware of the problem (in section 3 I propose reasons for this). Thus the question is if the pure form is the 'advance' of labour to capital (<1), or rather the 'advance' of wages to labourers (=1). The actual or historical convention should be immaterial to the answer. First. If the advance of labour to capital were the pure form, then labour would always be paid at the end of a production process (or perhaps a full turnover of capital). Generally this would mean that we theoretically start off with wildly diverging wage times. It would mean that the labour market is highly non-transparent regarding wage levels. It would also mean that in branches with a relatively low turnover of capital for example shipbuilding or forestry only wealthy labourers could be employed. This situation would hamper the mobility of capital. Second. If the wages advance were the pure form (i.e. Marx's starting point), wages would be paid at the beginning of the production process the question is for how many days or hours in advance. They could be paid at the beginning of the production process for so many days (or for so many seconds) in advance as is cost efficient for the enterprise; this time of advance in terms of cost efficiency would tend to converge for enterprises.22 There would not be the labour market and the capital mobility distortions of the first case. This avowedly concise argumentation indicates that Marx in Capital I was (implicitly) treating the pure case, and that he was right to initially neglect his own observation about the actual advance of labour by workers in Chapter 4 (section 1.1 above). Then the theoretical stage for concretising the initial presentation (or the initial model depending on ones views on Marx's method), should have been Part Two of Capital II (turnover), and a next one perhaps Part One of Capital III (rate of profit).
22

Cash payment for e.g. one hour at the beginning of each hour, is not cost efficient for the

enterprise, so it might be cheaper to pay more hours in advance; with the current technologies, one might have rather continuous streams of payment.

10

That Marx failed to do this is a pity, but given the phenomena in his days he cannot be reproached for that. This is different for 20th century marxian theorists; 'variable capital' was such a 'hard core' issue that they did not undertake the theoretical reconstruction.23 In preparation for setting out of some lines along which the matter might be reconstructed (section 4), I first delve further into Marx's concept of the turnover time of circulation capital. In the course of this we will see that there is also another serious problem that requires theoretical reconstruction.

3. The turnover of capital: an analysis and further critique of Capital II, Part Two 3.1. Marx on the turnover of variable capital The appropriate place for Marx to deal with a concretisation of Capital Is notion of variable capital, would have been Part Two of Capital II, and especially Chapter 16: 'The Turnover of Variable Capital', or perhaps also Chapter 15: 'Effect of the Time of Turnover on the Magnitude of Advanced Capital'. Let me first indicate that Marx studies the turnover, in most of Part Two, 'on the assumption of simple circulation of money, without any regard to the credit system' (F: 261; cf. e.g. 309, 336, 340),24 the credit system being presented in Capital III, Parts Four and Five. Thus he assumes 'that the capitalist operates only with his own capital' (F: 336). I say 'most of' Part Two. In fact, this Part has the status of a research manuscript that was polished by Engels into about 200 printed pages, and it is not surprising for such a manuscript to contain passages that anticipate later stages of the systematic presentation, including the credit system. As indicated at the end of section 1.2 above, Chapter 16 is not about variable capital specificaly. What is remarkable about Chapter 15 is that Marx at one place has the crucial coupling of turnover time and wage time nearly at hand. At this point he discusses a case in which 'constant circulating capital [is] advanced for 6 weeks. Wages, on the other hand ... are paid in shorter intervals, mostly weekly. So except where the capitalist forces the worker to make particularly long advances of his labour, the capital needed for wages must be present in the money form. When
23

To be sure, I include myself. See Smith (1997) on Marxian theory in the context of Lakatos's U: 185; cf. e.g. 235, 263, 267-8.

methodology and its 'hard core' notion.

11

capital returns, therefore, one part must be kept in the form of money for payment of labour ...' (F: 340, emphasis added) Thus, he sees here (this is the only one passage I detected) that an extension of the wage time would affect the quantity of variable capital (without affecting labour-time or the sum of wages paid).25 In the way he puts it, variable capital would even be nullified (=0). However, he seems to consider this impossible. The reason can be found in a passage from much earlier on (Chapter 1): 'The normal form of advance for wages is payment in money; this process must be steadily repeated at short intervals, as the worker lives from hand to mouth.' (F: 140, cf. U: 62) Because (I infer) the level of wages does not allow for savings, wages must be paid in short intervals. (In the 20th century, however, wage increases allowed for larger intervals; often interest-bearing consumer credit bridges the larger interval: 'from bank to mouth'.) However, if I am right that Marx for this reason could not believe in an increase in 'the interval', then its peculiarity is that he was witnessing 'change', i.e. an increase of the interval from end of day to end of week payment (cf. section 2).26 Even so, another surprising issue in the context of the Chapter 15 quotation above is that the general thrust of Capital II, Part Two, is that of a decreasing period of turnover.27 So generally the turnover time should (gradually) approach the wage time. For some reason, it was beyond Marx's expectations that the two should get near to one another. These two 'blocks' (and those in combination with the abstraction from credit section 3.2), prevented Marx from concretising the concept of variable capital at this stage of the presentation.

25

It could be argued that in a passage such as on page F:388 (U:317) Marx discusses a decreased

outlay of variable capital (keeping the wage time constant apart from the passage above, it is never varied), which is true; but he really discusses a decrease in that outlay due to an increase in the turnover of all the circulating capital.
26

He witnesed an increase of the interval by 600%; the 330% increase in the 20th century (from

week to month) is a considerable slow down. I suppose that we, like Marx, cannot believe that the interval will grow further?
27

It should be added that he never frames this in terms of 'forces' or of 'tendencies'.

12

3.2. The neglect of time of capital investment Quite apart from the two 'blocks', a non-concretisation of capital in terms of investment time is not justified at this stage, that is, when we have introduced turnover time. I am not sure if Marx saw the problems of capital investment time (see section 3.3). If he did, then apparently Marx believed that his abstraction from credit (until Capital III, Part FourSix) also required the abstraction from investment time. In this subsection I show that this is a mistake and that credit and investment time are separate issues. (All this is methodologically relevant to the extent that Marx's order of presentation is also an order of relative importance in the constitution of the whole of the capitalist system.) Take the following example which considers production time only (thus I apply the simplifying assumption of instantaneous sale of ready production). Suppose that the production time is four weeks. Wages are paid at the end of each week. Hence the week 1 payment is advanced for 3 weeks, that of week 2 for 2 weeks etc., so that, in total, only part of the wage cost would be a variable capital investment. Something similar applies to those elements of circulating constant capital that enter production gradually in the course of the production process. Marx insists that in the absence of credit the enterprise would continuously need a fund of circulating capital equivalent to all payments for materials and wages (i.e. c+v). For this, however, Marx's abstraction from credit is not (sufficiently) decisive at this stage. Decisive at this stage is the degree of simultaneous production (and sales) within an enterprise, which is a pure turnover time issue that Marx neglects. If production were purely sequential (one product, or one series of products, is produced after the other) then a large part of the circulating capital would indeed lie idle. However, if the sales and purchase departments of the enterprise were not already insisting on some degree of simultaneous production (thus realising more regular purchases and sales) then the financial management would. Then, given some degree of simultaneity, the financial management will use idle capital from the one process stage to finance another.28 Table 1 and Figure 1 illustrate this (constant and variable capital have been taken together). For simplicity, the figure shows a linearly regular use of the capital advanced (this linearity is non-essential). The diagonal in the N-shape represents the gradual depletion of the CC advanced. The left hand side of the figure shows Marx's sequential cases (cf. especially Chapter 15 of Capital II). The right hand side shows two simultaneous production processes. The main

28

Today this financial management is also called 'working capital management'.

13

point is that we see a decrease in idle capital (from triangle JBK to triangle HGI) with the (other) costs of production remaining the same. Table 1: From sequential to simultaneous processes
one sequential process stage of production time week 1 week 2 week 3 week 4 sum costs expended 20 20 20 20 80 CC laid out 20 40 60 80 80 80+s 80+s sales two simultaneous processes each at a different stage stage of production time week 1 & 3 week 2 & 4 week 3 & 1 week 4 & 2 sum costs expended 10+10 10+10 10+10 10+10 80 CC laid out 20 40 20 40 40 40+s 80+s 40+s sales

Figure 1: From sequential to simultaneous production processes

Quantity of CC advanced

H
Quantity of CC advanced

time
Sequential process 1 (4 time periods) Sequential process 2 (4 time periods) Processes 1 split into 2 simultaneous processes (each 4 time periods)

AJ quantity of CC advanced JB gradual use of the fund during 4 periods AJB effective investment JBK idle capital BK refunding of CC through sales Cost surface ABJ = cost surface CEH+DFI

CH (and DI) HE (and IF) CHE (and DIF) HGI DI

14

Hence, credit or no credit, only part of the sum for wages payment and for material inputs payment is variable capital and constant capital.29 We now see that the equation for circulating capital (4) is problematic, quite apart from the problem of turnover time and wage time. (Taking the two problems together for the example of the Table and Figure, it requires little imagination to infer that with four simultaneous processes and wages paid at the end of the week, the category of variable capital will have faded away.) One possible factor behind Marx's mistake is that he, in this Part at least, is often careless about his terminology, and especially so with 'variable capital' (which is a problem for us, and for him, even if it is a research manuscript). Generally, Marx makes a clear distinction between capital and money. However, consider, for example, the following passage where he compares the cases of two sequential processes, one (A) with a turnover of 5 weeks and the other (B) of 50 weeks. In each case the wages are 100 per week, thus in both cases the wage expenses are 5,000 in 50 weeks. Then Marx writes about A: 'In fact, therefore, in the course of the ten five-week turnover periods a capital of 5,000 is successively spent on wages, and not one of 500 ... On the other hand, it is labour-power to the value of 5,000, and not just 500, that is successively incorporated into the production process ...' (F: 384, cf. U: 313, emphasis added) The second phrase is right; the first one confuses capital investment ('advance') with payment or costs. Again: 'Thus the same amount of variable capital is spent in fifty weeks in case B as in A [wrong]; the same amount of labour-power paid for and used [right]. But in B this has to be paid for with a capital advance equal to its entire value, 5,000. In A, however, it is paid for successively by the ever renewed money form of the replacement value that is produced every five weeks for the capital of 500 advanced for each five weeks.' (F: 386-7, cf. U: 315, emphasis added; [comments added])

29

Note that efficient use of idle funds is a concern of Marx. Much of Chapter 15 is concerned with

the matter of efficient use of idle capital during the period of sales, i.e. when the proceeds have not yet been returned. Lapavistas (2000) rightfully criticises Marx on his views about the supposed relevance of relative production and sales times for the existence of idle funds (pp 223-33). Although Lapavistas, like Marx, considers degrees of simultaneous production and sales, he does not consider simultaneous production process (such as in Figure 1) or the problem of capital investment time.

15

The point is that the idea may not be wrong (the meaning of advanced capital in the second and third sentence), but rather that the term (variable) capital takes on a loose meaning. I conclude that because (variable) 'capital' has become a mixed notion, dominated by the meaning of costs and productive expenses, Marx absolves himself from the requirement to concretise it in the face of the turnover time of circulating capital. I emphasise that there are two problems. One problem is the dynamic movement of turnover time versus that of the wage time (what I called the two blocks in section 3.1). The other problem is that once we have left Capital Is pure theory of strict wages advance (i.e. at the beginning of production), then any wage time other than the pure one means that variable capital and purchase of labour-power are no longer congruent. The same applies to circulating constant capital and purchase of circulating raw materials and instruments. 3.3. Marx's implicit input-output framework for the turnover of CC In this subsection I try to detect the reasons why these problems occurred. It will turn out that the two problems, the wage time problem (3.1) and that of the neglect of the management of circulating capital as financial capital (or money capital) (3.2) are connected. One of Marx's many great theoretical innovations is that of his schemes of reproduction (Capital II, Part Three), one aspect of which is that they provide the framework for inputoutput analysis.30 In hindsight of this part of Capital (although it is the last part written) it may well be the case that Marx, during his exposition of the earlier parts of Capital, had something like an inputoutput framework in the back of his mind. I schematically present one slice of such a framework in Figure 2. In the way I present the scheme, it is also grafted on the circuit model (Capital II, Part One, cf. equation 2 in section 1.1). Note that I schematically separate the physical and the value aspect; another major innovation of Marx is of course his unification of the two (the double character of capitalist production). The continuity (hardly shown in Figure 2) is that the output is the basis for the expanded input.

30

In economics generally (i.e. marxian, orthodox and non-marxian heterodox), the reproduction

schemes have been at the basis of the development of the theory of the business cycle, macroeconomic two-sector models, and inputoutput analysis (the latter granted Leontief a Nobel prize). On these issues, see Arthur & Reuten 1998, and on Marx's schemes themselves, Reuten 1998.

16

Figure 2: Framework of Marx's reproduction scheme

VALUE INPUT constant capital (c) variable capital (v)

PRODUCTION OF CAPITAL

VALUE OUTPUT

valorisation process

valorised capital (x)

It is my hypothesis that the greatness of a scheme like the one presented in Figure 2 (in hindsight it is simple), at the same time blocked Marx to arrive at the idea that labour-power, whilst being a necessary input, might not have a congruent counterpart in capital investment. Reconsider Marx's circulating capital (CC) concept. In fact Marx does not start (in Capital II, Part Two) with a process of production and its time but rather just with CC as a quantity, namely CC = c+v. Next, the speed of CC in time (the turnover time) determines the number of turnovers in a year (n) and so the yearly surplus-value (S=sn, eqn 7). It follows that if Marx had an adequate concept of total capital invested which he thinks he has (eqn 5 above) he thus has all the building blocks for the introduction of the rate of profit in Capital III, Part One. So what can he say about the turnover of CC? In fact, Marx focuses for the time issue on c (as including the fixed capital counterpart). Apparently this is adequate, because in the middle part of Capital I, when he provides the explanation of surplus-value from variable capital, he had treated constant capital mostly as given. Given the earlier analysis it seems that the speeding up of constant capital in process (esp. decrease in all kinds of storage time) determines the turnover time. (Casting this in somewhat different terms: Now v is treated as attached to c, whereas in Capital I, c was treated as attached to v; putting it most briefly, earlier we had the management of the speed of labour in process; now the management of the speed of means of production in process, and of the speed of produced commodities in sales-process; or, in the well known metaphorical language, earlier the management of 'living labour', now the management of 'dead labour'.) So we have:

17

CC (=c+v) ... s .... input

CC'=x output

In fact, for Marx (in Capital II, Part Two) the c/v ratio is a technically given ratio (that is, one imported from its treatment in the last part of Capital I, i.e. Parts Seven and Eight in the English edition). More specifically, we have a technically given input ratio, with given input costs for means of production and labourpower. Therefore, and in this way, CC remains all along the same quantity.31 At the same time, given the labour process, we have also a technically determined fixed output. Notice that this 'same quantity' issue of CC thus also applies to v. This is relevant for the structure of the theory in connection with the rate of surplusvalue. For equations (7) and (6) to make sense (i.e. S = sn and s = s'v), we must assume and this is what Marx does implicitly that when the turnover time decreases (i.e. n increases), we also have the same quantity of v compressed in time (in Marx's case especially by carrying out the production process and sales process in parallel instead of subsequently). We may call this compression 'organisational'. Of course, what I call a 'technical' reduction of the time for the process would affect, ceteris paribus, the rate of surplus-value (s'). Now let us grant Marx, for the moment, that it is legitimate to treat the wage time as given (as he does implicitly). Let us also grant that is legitimate to treat especially turnover matters in abstraction from credit relations. Then there is noting wrong with this model? Again, we have previously determined technically fixed inputoutput coefficients, fixed prices, as well as wt given. Then the variable for the model is the (organisational) turnover time (and so n). On this basis, the only problem is that Marx is wrong to treat the value of the inputs as capital. The only problem is that he neglects the management of financial capital (money-capital). In the absence of financial management, the model is correct: you begin with separate fixed funds of money for each of the means of production items as well as for labour-power. The funds lie idle until they need to be expended. However, no capitalist would do this; what is more, even if some (or all) would do so, it is the logic of capital (that is what Marx wants to demonstrate) that requires efficient investment of capital. Next, if we add 'financial management' as another variable to the model, then, ceteris paribus, the financial management determines the c/v ratio as a
31

The making of a commodity, or a set of commodities, e.g. a table or a house, requires

technically fixed quantities of means of production (c?!) and labour-power (v?!).

18

capital ratio. (For the intuition of this the reader may turn back to the example of Figure 1 in section 3.2, where 'funds' are switched between processes so that we have a shift in the presumed c/v ratio.) In conclusion: the main point of this paper is about Marx's and the marxian notion of variable capital specifically. The general upshot of the current subsection is that Marx's systematic for the determination of the rate of profit (Capital III, Part One) is defective because of a defective concept of capital, i.e. one grafted on an inputoutput framework.

4. A concise reconstruction 4.1. Circulating capital and its fading variable capital component As indicated, I interpret (or reconstructively interpret) Capital I as presenting the pure form in which there is a full advance of the wage to labour (section 2). In any case, given Marx's observation about the one day or one week advance of labour by workers, he should have concretised the concept of variable capital at some stage of his presentation. The appropriate stage to do so is with the introduction of circulating capital and its turnover. Here, therefore, I start my reconstruction. The reconstruction is not actually too difficult. It is only unfortunate that we loose some of the aesthetics of Marx's presentation, because the outward appearance of the categories is different at different levels (we no longer have the c+v elegantly appearing at the level of all three volumes of Capital). Thus my starting point is the formula of CC = c + v [per tot] (4)

where, initially, c and v are elementary notions of capital (treated in Capital I). These elementary notions of capital are at the same time input notions. Each one of these notions is taken over to the new level, though each one is concretised differently. In what follows, I drop the simplifying assumption of F=0 (so "c", and its counterparts below, includes a F/n component cf. equation 3). Remember that CC is a quantity, a quantity in a discrete period of time (the time of the valorisation process, i.e. the production plus sales time).32 In fact we have two quantities.

32

This is a discrete period of time, also in the case of continuous production (the time is that from

input to output and sale of one commodity).

19

The first is an input quantity in monetary terms, which is a Circulating Commodity Flow in monetary terms (CCF), that is, the value of the means of production flow (cf) and the value of the labour-power flow (vf, read: 'value flow'): CCF = cf + vf [per tot] (10)

Analogous to equation (3) we have: cf = F/n + RMI [per tot] (11)

where I interpret F/n as the replacement flow of means of production and where RMI is the flow of the other means of production (raw materials and other instruments). Like CC, CCF is calibrated to one turnover time of the CCF value. For the output (x) of the process (the end time of the turnover) we have the value of: x = cf + vf + s [end point of tot] (12)

If we substitute these terms into Marx's Capital II, Part Two, we can read most of it as it 'is'. What Marx then describes is all kinds of management to reduce the turnover time of the CCF. Strictly speaking we have the cf/vf ratio determined through engineers technique (this is somewhat of an idealisation). Then all kinds of process managers are out to reduce the time of the total process: efficient feeding of materials (reduction of their storage time); simultaneous and continuous production (reduction in output storage time); etcetera.33 Again, all this work reduces the tot, whence the number of turnovers (n) increases, whence we have an increase in the yearly surplus-value (as prior to the reconstruction): S = sn [per year] (7)=(13)

At this level, however, equation (6) above should be 'concretised' as:

We have a theoretical successive determination. The cf/vf ratio is determined at Capital Is level.

Then given a quantity of CCF we have the cf and vf quantities. Remember (section 3.3, the full new paragraph after footnote #31) that any organisational reduction of the working period by process management is assumed not to affect the quantity of vf ; hence it is assumed to not affect the quantity of s, and to have no effect on the rate of surplus-value (the rate of surplus-value is determined at Capital I level).

20

s = s'(vf)

[per tot]

(14)

(even if this delivers no new determinants beyond those set out at Capital Is level). Now we get to the second quantity. Given the 'technically' required purchases of means of production (in the form of cf) and of labour-power (in the form of vf) and given the turnover time of the circulating commodity flow (CCF) as determined by the process management, the financial management is out to reduce the quantity of Circulating Capital (CC) required for the CCF. Very simply, we may write: CC = cc + vc [per the given tot] (15)

where cc stands for the capital invested in circulating means of production, i.e. constant circulating capital; and vc for the capital invested in the wages of labour-power, i.e. variable circulating capital. (Writing CC = c + v is bound to give rise to confusion, because we have a change of meaning in comparison with the Capital I categories of c and v.) Next we have cc and vc simply determined as follows. ('Determined' - in fact I set out instead a delineation of categories.) cc = cf [per the given tot; on average ] (16)

Remember that when the process management has done its work, CCF is fixed to a determined time (e.g. one day or three weeks or one year). Then indicates the time fraction of the tot in which capital is laid out on each of the means of production items.34
34

If all of cf would have to be paid for at the beginning of the production process, even if they

enter e.g. halfway, then =1. (This is Marx's simplifying assumption, cf. section 3.1.) Next, <1 to the extent that payments are normally made later (because means of production enter later, or because of normal purchase conditions). Note that 'normal' purchase conditions would also apply to the payment for sales. Theoretically, therefore, one may reckon with delivery dates at the input and output side (determined by technique and process management). If all means of production items entered equally distributed over the turnover period, then =. Until a few decades ago financial management was a complicated job because often a given 'working capital' would have to be fitted to the sets of outgoing and incoming payments so that a minimum of the working capital would lie idle with, at the same time, a minimum risk for illiquidity preventing purchase of means of production. Today enterprises usually have flexible credit lines with

21

For the variable circulating capital (vc) we have:

vc = vf

[per the given tot] (17) 35 [0 1, depending on contingent tot/wt]

Similarly to , indicates the time fraction of the tot in which capital is laid out on wages, thus here the wage time (wt) comes in. In the history of capitalism, it has generally been the case that labour is advanced to enterprises, whence < 1 (i.e. for any tot; this thus includes the case of tot > wt.) The extent to which < 1 depends on (further) wage bargaining as determining the wt, as well as on the tot.36 Taking (15)(17) together we have for circulating capital: CC = cf + vf [per the given tot] (18)

This completes the reconstruction of circulating capital and its turnover (Capital II, Part Two) cf. equations (10)-(18). Before integrating this in total capital invested (section 4.3), the next subsection briefly discusses the notion of stocks that is implied in the reconstructed circulating capital. 4.2. Implicit and explicit precautionary stocks In this subsection I briefly expand on stocks. Like Marx in Capital II, Part Two, I consider average circumstances (specifically: no overheating or recession). Reconsider the circulating commodity flow (CCF) with its means of production flow (cf) and its labour-power value flow (vf), the tot of which is the outcome of process management. CCF = cf + vf [per tot] (10)

From this we derived the quantity of circulating capital as the outcome of financial management:
banks, which takes away this complicated fitting work for working capital. (There remains the fitting work for new fixed means of production items in connection with short or long term loans.)
35

More specifically, as we saw before, = (tot wt)/tot. See the next footnote on the restriction 0 For tot < wt we would formally have < 0. However, this makes no economic sense since CC

.
36

cannot be smaller than cf (see equation 18 below). Generally therefore, 0 1. (=0 for tot=wt, with wages being paid at the end.)

22

CC = cf + vf

[per the given tot]

(18)

Management of the input stock should reckon with a possible stagnation of the input supply, whence production would stagnate, and therefore there should be a precautionary input stock. Similarly, most enterprises would not want to lose clients because they cannot meet demand, and therefore they build up a precautionary output stock. This matter is one of the terrains where process management and financial management may have conflicting objectives. We may take the precautionary input stock to be a fraction of cf.37 Staying close to Marx's framework, any output stock should be calculated not in terms of an unsold output as including a non realised surplus-value, but rather as without the latter. (To the extent that we have an 'iron' stable output stock, this makes sense.) Then we may take the precautionary output stock to be a fraction of cf+vf. There is nothing wrong with this if we indeed understand that part of the input and output flows take on a stock character. This matter is relevant when we move to total capital investment. 4.3. Total capital investment and the rate of profit As indicated, the middle parts of each of Volumes I and II of Capital (the production of surplus-value and the turnover time of circulating capital) provide the building blocks for the presentation of the rate of profit in Volume III, Part One. I now briefly consider the latter in face of the reconstruction above. Earlier (section 1.1) we had: c = F/n + RMI TC = F + c + v [per tot] [per year] (3) (5)

Upon the reconstruction we have a total capital investment (K) of: K = F + cf + vf [per year] (19)

Then we have for r, the rate of 'operating surplus' over capital invested:

37

This is a simplification depending on the particular production process and esp. the

materialisation of replacement of means of production and of other instruments. Of course we might introduce a range of different fractions for the range of cf components.

23

r=

S = K

s'(vf)n F + cf + vf

(20)

This 'capital invested' should have its empirical equivalent in the assets side of the end of year balance sheet of enterprises. I suppose that the fixed capital component requires no further clarification (it may be called 'plant', 'structures', 'equipment', 'buildings' etc.).38 The balance sheet equivalents of CC = cf + vf are: 1. RMI stock; 2. work in progress (including wages only if tot > wt]; 3. output stock; (cf. the phases of Marx's circuit scheme equation 2). Given the precautionary stocks (section 4.2), the components 1 and 3 are one major component. For the rest, the balance sheet components would depend on whether production is purely sequential (which is rare) or continuous. In the pure sequential case, the proportions of 1-3 would depend on the stage of the turnover process at which the balance sheet is made up. For continuous production we would have 1-3 mixed. However, the work in progress component is most often reduced to its materials component ("stocks"), whence we find no component for variable circulating capital (eqn 17) on the balance sheet. Of course, we cannot expect otherwise because for most enterprises there is no such thing as variable capital (because their tot<wt). However, for sectors such as 'construction' and 'forestry' the tot>wt.

5. A corollary for 'the' transformation problem The critique and reconstruction set out above has consequences for the appraisal of the (unfinished) drafts of Capital III, especially Parts One and Two Part Two being the one where the (in)famous transformation problem is posited. I have argued in a previous paper (Reuten 2007) that, quite apart from the problems treated in the current paper, the positing of that transformation is methodologically illegitimate to the extent that quantities are transposed from the one level of abstraction to the other. (More precisely, it is methodologically illegitimate from the point of view of a plausible interpretation of Marx's method in Capital I, the final draft of which was written after that of the Capital III
38

Of course the valuation of these assets is highly problematical but that is not the subject of this

paper.

24

research manuscript that was edited by Engels).39 The current paper in fact highlights this methodological point. One cannot just algebraically manipulate between quantities derived at the level of pure theory and those derived at more concrete levels (in terms of the current paper: it makes no sense to transpose the Capital I notion of capital (c+v) to the level of concretion of the rate of profit, i.e. the level where time of capital investment (cf+vf) plays a role, or at least should play a role).40 This is not to say that there may be no transformation problem, rather that the framework for positing the problem as well as its possible solution was ill developed in the 20th century.

Summary and Conclusions In the middle part of Capital I Marx's focus for 'time' is the labour-time required for a production process (1.1).41 Consequently it is appropriate to treat implicitly at that point the 'pure form' of wages being advanced at the beginning of the production period (2). Implicitly then, Capital I adopts a notion of capital in which the input costs (for means of production and labour-power) coincides with the investment of capital (3.23.3). In fact, Marx here abstracts from the time aspect of capital advanced, or capital investment. All this is appropriate in face of his objective of the explanation of surplus-value in the middle part of Capital I (1.1). The middle part of Capital II, on the turnover time of circulating capital, would have been the appropriate place to make all these abstractions and simplifications explicit so as to further concretise the earlier notion of capital. In fact, Marx mainly focuses on the turnover time of constant capital, leaving its quantity and the quantity of variable capital unaffected (3.3). We have seen that two issues thus remain undeveloped. One is the distinction between, and the movements of, the turnover time and the wage time. The opposite movements of these two turns 'variable capital' into a fading category, at least for the modal enterprise (2; 3.1). The other issue that remains undeveloped is the financial management of the investment of capital; this defect results in the continued treatment of input costs and capital investment as quantitatively congruent (3.2

39

Perhaps, as I set out in the paper referred to, it is not illegitimate if we take the point of view of

the Capital III manuscript (written 1864-65) and forget about Capital I (text for the first edition written in 1866-67).
40 41

As including fixed capital this is about F+c+v versus F+cf+vf. Numbers in brackets in these conclusions refer to the sections above.

25

3.3). In effect, this results in a defective denumerator for the rate of profit (for Capital III, Parts OneTwo; cf. 5). It is not in fact too difficult to provide a reconstructed concept of capital investment linked to the - now quantitatively incongruent - input costs (4). Given the technically determined inputs of production (theoretically 'imported' from their Capital I analysis), we have a quantitatively given circulating commodity flow (CCF = cf + vf). The process management is out to reduce the time of this flow (i.e. to reduce the turnover time, in accordance with Marx's Capital II, Part Two analysis). Then given this time, the financial management is out to reduce the quantity of circulating capital required for this time of the commodity flow, ending up with an investment of fractions of these flows (CC = cf + vf). The upshot is that the ratios of the thus concretised circulating constant capital and variable capital (cc/vc) are co-determined by financial management (4). For all enterprises, the investment in variable capital (wages) is smaller than the input costs of labour-power (<1) thus everywhere labour provides an implicit production credit to enterprises. Moreover, for the large majority of enterprises today, the investment of variable capital (cc = vf) is fading away.

References
any date after a sign | means 'date of composition'; ~ means 'probably'

Arthur, Christopher J. & Geert Reuten (eds 1998), The Circulation of Capital: Essays on Volume II of Marx's 'Capital', London/New York, Macmillan. Bellofiore, Riccardo & Roberto Fineschi (eds 2007), Re-reading Marx New Perspectives after the Critical Edition, London/New York, Palgrave Macmillan (forthcoming). Bellofiore, Riccardo & Riccardo Realfonzo (2003), Money as finance and money as universal equivalent: re-reading Marxian monetary theory, in Rochon & Rossi (eds 2003), 198-218. Bellofiore, Riccardo & Nicola Taylor (eds 2004), The Constitution of Capital; Essays on Volume I of Marx's 'Capital', Basingstoke/New York, Palgrave Macmillan. Engels, Friedrich (1894), Chapter 4 (of Capital III): The effect of the turnover on the rate of profit' (see Marx 1894). Lapavitsas, Costas (2000), On Marx's analysis of money hoarding in the turnover of capital, Review of Political Economy, 12:2 (219-35).

26

Marx, Karl (1867G; 4th edn 1890 | 1866-67), Das Kapital, Kritik der Politischen konomie, Band I, Der Produktionsproze des Kapitals, MEW 23, Berlin, Dietz Verlag. (1867F; 4th edn 1890 | 1866-67), idem, Engl.transl. Ben Fowkes (1976), Capital, A Critique of Political Economy, Volume I, Harmondsworth, Penguin.42 (1885G; 2nd edn 1893 |~1870-1878), ed. F. Engels, Das Kapital, Kritik der Politischen Okonomie, Band II, Der Zirkulationsprozess des Kapitals, MEW 24, Berlin, Dietz Verlag. (1885U; 2nd edn 1893 |~1870-1878), idem, Engl.transl. Ernest Untermann (1907) (& I. Lasker), Capital, A Critique of Political Economy, Vol II, The Process of Circulation of Capital, London, Lawrence & Wishart 1974. (1885F; 2nd edn 1893 |~1870-1878), idem, Engl.transl. David Fernbach (1978), Capital, A Critique of Political Economy, Vol II, Harmondsworth, Penguin. (1894G |~1864-65), ed. F. Engels, Das Kapital, Kritik der Politischen konomie, Band III, Der Gesamtprozesz der kapitalistischen Produktion, MEW 25, Berlin, Dietz Verlag. (1894F |~1864-65), idem, Engl.transl. David Fernbach (1981), Capital, A Critique of Political Economy, Vol III, Harmondsworth, Penguin.43 Moseley, Fred & Martha Campbell (eds 1997), New Investigations of Marx's Method, Atlantic Highlands (New Jersey), Humanities Press. Reuten, Geert (1998), The status of Marxs reproduction schemes: conventional or dialectical logic?, in Arthur & Reuten (eds, 1998), 187-229 (2000), The Interconnection of Systematic Dialectics and Historical Materialism, Historical Materialism, nr.7, 137-165. (2004a). Productive force and the degree of intensity of labour; Marx's concepts and formalizations in the middle part of Capital I, in Bellofiore & Taylor (eds, 2004), 117-45. (2004b). The inner mechanism of the accumulation of capital: the acceleration triple; A methodological appraisal of Part Seven Marx's Capital I, in Bellofiore & Nicola Taylor (eds 2004), 274-98. (2007). Marx's general rate of profit transformation: methodological and theoretical obstacles; an appraisal based on the 1864-65 manuscript for Das Kapital III, in Bellofiore & Fineschi (eds 2007), forthcoming.

42 43

First Engl. transl. S. Moore & E.Aveling, 1886. First Engl.transl. Ernest Untermann 1909.

27

Rochon, Louis-Philippe and Sergio Rossi (eds 2003), Modern Theories of Money; The Nature and Role of Money in Capitalist Economies, Cheltenham/Northampton: Edward Elgar. Smith, Tony (1997), Marx's theory of social forms and Lakatos's Methodology of Scientific Research Programs, in Moseley & Campbell (eds 1997): 176198. (1998), The CapitalConsumer relation in Lean Production: the continued relevance of Volume Two of Capital, in Arthur & Reuten (eds 1998): 6794. [10,600]

28

Abstract Marx's and the marxian notion of 'variable capital' is assessed in face of the fact that enterprises, generally, do not 'advance' the wages equivalent to labourers. Rather, because wages are paid only after some period of performance of labour, labourers provide an implicit 'production credit' to enterprises. This is not only a 20th century phenomenon; Marx was well aware of this phenomenon in his days. It is argued that whereas the notion of 'variable capital' is appropriate at the level of analysis of Marx's Capital I, it is not adequately concretised in Capital II, Part Two (the turnover time of circulating capital). Two issues in particular remain undeveloped. One is the distinction between, and the movements of, the turnover time (the production and sales time) and the wage time (end of day, week or month). The opposite movements of these two turns 'variable capital' into a fading category, at least for the modal enterprise. The other issue remaining undeveloped is the financial management of the investment of capital; this defect results in the treatment of input costs and capital investment as quantitatively congruent. This again results in a defective denumerator for the rate of profit (Capital III, Parts One-Two). Taking account of the objections raised, the final part of the paper undertakes a reconstruction of Marx's category of circulating capital, one that fits not only the general theoretical systematic of Capital, but also its empirical counterpart in the assets side of the balance sheet of enterprises. [247]

29