I'm glad this thread will continue to be active. The general questions raised/criticisms made, do not really seem new to me, which is perhaps why I attempt to go slightly into more basic issues (for example, the meaning of GDP as it relates to accumulation). And before asking whether accumulation is possible under certain conditions, what it is, and could it already have stopped occuring? However absurd that notion might be on first sight, some online fringe Marxists seem to be saying that actual companies in today's era of capitalism are loss-making, that we're dealing as it were with zombie-capitalists, which normally should have went bankrupt.
(Just as a side-note, let me imagine an individual case, where a new company is created, which operates for one year and then goes bankrupt (so has an overall negative profit rate). Nevertheless, during the one year that it operated, it produced and sold some commodities (and so contributed to a rise in the GDP figure, despite no accumulation occuring in this company).
Your clarification of Pannekoek’s views on this issue is very welcome. His review of The Accumulation of Capital is significant as, afaik, the only contemporary critique of Luxemburg’s theory from the revolutionary left of the Second International, and for this reason is of particular interest in this debate.
To recap, in the passage I quoted, what Pannekoek is challenging is the apparent absurdity of the capitalists producing simply for the sake of producing, which results in ever-growing amounts of commodities being thrown onto the market in successive years. In fact he argues the purpose of this apparently aimless activity is the self-valorisation of capital.
You are absolutely right to point out that what Pannekoek is addressing here (along with Luxemburg of course) is the abstract diagram of capitalist reproduction in Capital Volume 2, and this leads him - perfectly logically and consistently - to dismiss the criticism that it does not include the more fundamental contradictions for capital of realising surplus value or indeed the falling rate of profit. So you could well argue that his critique of Luxemburg is not the same and I did not intend to misrepresent his views. But I would argue that the phenomenon he is addressing here is real enough: overproduction – or, to be more precise; capital’s inherent tendency towards overproduction driven by the growth of productivity and the restriction of the available market, and he is, I believe, right to criticise Luxemburg for not seeing the significance of its manifestation in her search for the 'missing' effective demand.
In an earlier back-and-forth with me, Link (in post #49) stated:
Quote: I dont understand the connection to asking about increases in labour cost and/or hours worked? Generally though constant capital increases at a faster rate than variable capital as labour productivity increases.
With all of Link's stress on the importance of constant capital, let's not fall into the error of forgetting about variable capital. Variable capital must continue to grow as well, that is, the number of emplyees increase (which should be reflected in the statistics by an annual increase of the national/total labour hours worked). Otherwise, if surplus value could somehow be invested 100% solely in constant capital, that would mean these additional machines required no additional employees (to operate). Link's reference that constant capital grows faster than variable capital is a matter relating to their proportional shares, which however leaves the fact that in absolute terms the whole, including its component variable capital, increases. I continue to question how reliable GDP is as an indicator of accumulation, given eg the total employees in US doesn't signficantly increase annually (perhaps on the contrary even), while somehow GDP is going up. Link's answer would seem to be that the entire increase of GDP is almost solely due to an increase of constant capital (with near zero% increase of variable capital, which we account in actual terms of labour hours worked). If the objection (of Link) to my framing was, that to calculate the sum/value of variable capital, we musn't look at labour hours, but should just look at prices (wages), then my response was that the connection between value and price is lost.
Dman you are correct to question the reliability of GDP figures but they are a convenient indicator of what is happening in the economy. We cannot take them as too precise as they dont account for social wage nor the various types of waste production that marxists recognise. They are just an indicator.
I think you have misunderstood me completely regarding variable capital. Yes i am focussing on constant capital in relation to luxemburgs theory becuase that is precisely what she does. She says accumulation of new constant capital is absolutely impossible without extra capitalist markets yet you do not seem to address this problem.
You are correct that variable capital grows as well as constant capital and the mass of surplus value. I have never said otherwise and in fact the quote you use from me does say precisely that variable capital grows. I have consistently argued that the size of the economy and the size of world population have grown enormously over the past century but Luxemburgists have tended either ignored these empirical facts or minimised them becuase they dont fit in with Luxemburgs theory that accumulation is absolutely dependant on extra capitalist markets.
Quote: Dman you are correct to question the reliability of GDP figures but they are a convenient indicator of what is happening in the economy. We cannot take them as too precise as they dont account for social wage nor the various types of waste production that marxists recognise. They are just an indicator.
You give (other) additional reasons to question the reliability of GDP, than the one I gave (ie you ignore it), and without questioning the concept as such. Btw, I earlier also raised this reason:
" Why even is a rise in GDP taken as an indicator of accumulation – shouldn't rising productivity rather translate into a price fall of the individual commodity, and then a fall in GDP? Say after some period [say 30 years] that the 90 hours of c+v can be reproduced in 45 hours, with s remaining 10 hours throughout the period, then, doesn't this mean that the gross product's value would fall from 100 to 55 hours, even though the absolute mass of surplus value stays the same, and the rate of profit even rises (from 11,1% to 22,2%)? "
Your focus on GDP can perhaps be excused because you believe by comparing its growth rate in the periods of rising and decadent capitalism, you can challenge the (ICC's) claim that "accumulation" hasn't maintained its growth rate (but slowed down). The dispute on this ground though accepts the framework of relative comparison of GDP growth rates, whereas I just even questioned its utility and growth rate absolutely (and who knows, perhaps "accumulation" is now negative).
Quote: I think you have misunderstood me completely regarding variable capital. Yes i am focussing on constant capital in relation to luxemburgs theory becuase that is precisely what she does. She says accumulation of new constant capital is absolutely impossible without extra capitalist markets yet you do not seem to address this problem.
Surplus value is incorporated in each commoditiy, not just in the additional constant capital (which includes also raw materials, besides machinery). Otherwise, one could just as randomly decide to claim that society's surplus value is incorporated solely in subsistence/consumer goods (and luxury goods I guess), while the additional constant capital contains solely the value of C+V for some reason.
Rosa mentions constant capital eg when she speaks of English railways construction in antebellum US (whose population consisted largely of simple commodity producers), or conversely the slave-produced cotton serving as constant capital in English factories, but does not make this her sole focus. I talked about the issue of capital export in my post #14.
Quote: You are correct that variable capital grows as well as constant capital and the mass of surplus value. I have never said otherwise and in fact the quote you use from me does say precisely that variable capital grows. I have consistently argued that the size of the economy and the size of world population have grown enormously over the past century but Luxemburgists have tended either ignored these empirical facts or minimised them becuase they dont fit in with Luxemburgs theory that accumulation is absolutely dependant on extra capitalist markets.
Actually, if I'm not mistaken, there is a peculiar distinction with variable capital, namely that it (ie the wage) is not paid for in advance. To "buy" additional variable capital, the capitalists would not necessarily need to pay money for it (ie need not have realised his previous surplus value), but rather "borrow" (on credit so to speak) the worker's labour (and so "borrow" the necessities of life that resulted in the value of the worker's labour power, prior to his employment).
Let's suppose for a moment that capitalists perfectly can, by themselves, realise all their constant capital; the point would still remain that they can't buy with it the new variable capital from each other (unless perhaps in the future humans are breeded, like in a clone-factory).
I do not follow, why the growth in the size of world population is relevant to the specific point I make about GDP (in the US) and annual total labour hours worked (I anticipate, perhaps you could for example opt to argue, that the strongly growing labour force in Mexico should be included in US statistics, as if it were a part of domestic US industry). The bourgeois term "globalisation", in the last decades, designates the entrance of huge swathes of workers in the world labour force, which they see as a factor that now threatens to dry up, pointing to rural China (or also bring up as a factor to explain interest rates and inflationary trends).
Let me stress my objection against reliance on GDP as an indicator of "accumulation". I think it's a basic issue (what is the relation of value/labour hours worked to GDP, if any). Here's another example:
From 2008-2018, the Chinese labor force rose from 770 to 806 million people (even began to shrink a bit after 2016). In the same period Chinese GDP almost tripled (from 4,59 to 13,89 trillion USD).
It thus seems that in this period (untill 2018) GDP grew by 200%, whereas the labour force rose perhaps by 5%. How can 5% more workers create a value (supposing that GDP does represent value) that is 200% bigger?
Am I asking a stupid question here?
Btw, I previously noted that the increase each year in the value of existing machinery, indicates that there is increasing part of value which is not used-up (ie not reproduced).
Another (better) indicator of "accumulation" (than GDP) is perhaps the mass of annual corporate profits. The evolution of US corporate profits is marked by a steady rise during the whole post-war period (see here; or just last 20 years here), but it's particularly sharp during the roughly 2001-2006 period (profits doubled then, roughly from $750 billion to 1,5 trillion).
If we hold that profits derive from/represent ("new") surplus value, my stupid question (also, and perhaps even more) for this indicator is, how does the evolution of annual profits relate to total hours worked, if at all?
Surely the US labour force/hours worked between 2001 and 2006 didn't double. But the annual mass of profits doubled.
I welcome the forum team’s very fair summary of this discussion so far. This couldn’t have been an easy task, especially since views changed and evolved in the course of the discussion, something I think was actually a sign of healthy debate.
What can we conclude from this?
I think it confirms that it is impossible to defend the letter of Luxemburg’s theory that non-capitalist buyers are necessary for accumulation to take place and remain consistent with Marx’s analysis of capital’s contradictions.
This may seem a purely abstract question. After all, as the forum team point out, historically there has never been a capitalism without a non-capitalist environment (“even if there may not be a theoretical problem for capitalism to accumulate, on the level of the historical evolution there exists a problem...”). Agreed.
But Luxemburg’s analysis of the historical evolution of this problem is partly based on her thesis that capital needs non-capitalist buyers in order to accumulate, and unless we clearly reject this at the abstract level, it comes back to bite us when we try to understand the continuing role of no-capitalist markets in decadence, because if these are somehow necessary for accumulation to take place, and if they were insufficient for the needs of expanded production in 1914, as Luxemburg herself clearly states, how are we to coherently explain how capital has continued to accumulate for over 100 years since then? In 1914 world GDP was about $5 trillion, which by 2014, after 100 years of decadence, had grown to $105 trillion; so logically capital would need over 20 times the amount of non-capitalist areas to be able to accumulate…
The ICC’s solution to this problem appears to be to re-interpret Luxemburg’s theory to argue that since 1914 the extra-capitalist milieu has been insufficient for capital’s further ‘peaceful’ expansion, but still remains sufficient to enable it to continue to accumulate, albeit through inter-imperialist warfare:
“The ICC defends the position that the period of catastrophes started at the beginning of the 20th century, precisely from the moment insufficient extra-capitalist markets were left for capital to accumulate without leading to a qualitative sharpening of the contradictions. From that moment “the ‘push’ of capitalist production towards the rest of the world accelerated. Instead of competition between individual capitals for markets within the national framework, the emphasis was now on competition between national capitals for the remaining non-capitalist areas of the globe” (Nation or Class?)” (from the forum team summary).
In other words, Luxemburg’s central thesis is transformed into: capital needs non-capitalist buyers in order to accumulate ‘peacefully’, ie. without inter-national wars to redivide the world market.
So I think the focus has now moved on to discussing this interpretation and its implications, which suggests at least two areas:
Quote: how are we to coherently explain how capital has continued to accumulate for over 100 years since then? In 1914 world GDP was about $5 trillion, which by 2014, after 100 years of decadence, had grown to $105 trillion; so logically capital would need over 20 times the amount of non-capitalist areas to be able to accumulate
If world population around 1900 was around 1,6 billion, then, suppose (a bit randomly) only a third was engaged in work, and suppose the rate of exploitation was (perhaps absurdly) low (say a 10 hour day split into 8 hours necessary labour + 2 hour unpaid labour, ie 25%), that would mean a small percentage (resp.20%) of this total working population (of 530 million) was responsible for all the surplus produced (ie 106 million, ie 6,6% of world population).
By 2004 the world population had quadrupled (6,4 billion), ie a 300% increase. Divided into 104 years, that's an annual growth rate of 2,88% in population. We still assume only a third is working (2,13 billion). Believing GDP to be an indicator of accumulation, let's say around 1900 it was $4 trillion, and in 2004 $74 trillion, an increase by 1750% I guess, or 16,8% a year.
Thus GDP grows faster than the population.
Suppose the rate of exploitation by 2004 had stayed as low as we presupposed it in 1900 (6,6% of world population produces all the surplus, so around 423 million people in 2004, or 317 million added new "surplus-producing" workers since 1900). We would be expected here to believe (according to those who uphold GDP as a reliable indicator), under our assumptions, that just 317 million new workers are responsible for the 1750% rise of GDP (from 1900 to 2004), with all the rest of the labor force engaged in reproducing itself or simple reproduction. This strains belief, and our assumption (eg of same rate of exploitation) seems difficult to maintain (if we want to cling to GDP). Let's relax our assumptions a bit.
Suppose the original 106 million "surplus-producers" in 1900 had ideally increased at the same pace as GDP did, ie by 1750%, then = 1,961 billion people should be working on surplus in 2004 I think. Recall that I estimated that (in 2004) there are in total 2,13 billion working people. So it's narrowly possible, but results in the consequence that in 2004 maybe 90% of the labor force does surplus labour, with 10% doing the necessary labour, whereas in 1900 we posited the shares were 20% and 80%. So the assumption has become, that the rate of exploitation increased significantly (from 25% in 1900, to 900% in 2004). In 2004 10% of workers (very roughly 200 million) do, what 80% did in 1900 (424 million), thus moreover an extra 224 million workers have been "freed-up" for sole devotion to surplus working.
The assumption here still remained though, that the rate of exploitation was a relatively low 25%, initially in 1900. If, on the other hand, which perhaps is also more realistic, it was modestly higher in 1900, perhaps eg 4 hours surplus/6 hours necessary ie 66,6%, then, 176 million (ie 33% of 530 million) were "surplus-producers", which to keep pace with the +1750% in GDP = would require 3,26 billion workers I think, which in any case already exceeds the existing total of workers in 2004 (2,13 billion, so the reality falls over a billion short for the model to hold). The pool of necessary workers in 1900 in this case also was already smaller, and even if they are nearly all freed-up, insufficient to be able to bridge this gap.
Perhaps to posit a kind of mathematical model/law, my hunch is, that attempts to explain/account for the discrepancy between annual GDP growth (here 16,8%) and population growth (here 2,88%) by way of invoking increased exploitation, encounter a certain limit, namely the initially posited rate of exploitation can't be higher than a certain percentage.
Going backward from my assumptions: 2,13 billion decreased by 1750%= 115 million, which is 21,7% share of workers of 530 million in 1900, or 2,17 hours in the 10 hour day, so, the rate of exploitation must be posited as below 27,7% as a starting point in the model for 1900. Otherwise the discrepancy between GDP and population growth can't be explained.
Then either we reject GDP (and others suchlike, possibly also the mass of profit) as a reliable indicator for accumulation, or perhaps we start doubting accumulation is actually happening/possible.
Another issue is the reliability of inflation adjustments (of eg GDP). If inflation is underestimated by say 1% each year, over eg 115 years (1900-2015), prices would be 3,14 times above their undistorted level. In other words, to correct such a faulty existing inflation adjustment procedure, the absolute GDP figure in 2015 would have to be divided by 3,14.
Incidentally, I take these figures (1% annual underestimate, spanning 1900-2015), because I heard an investment analyst claim that gold over this time had a "real" (ie inflation-adjusted) rate of return of "just" 1% per a year. I haven't looked up what inflation-adjustment they used, but it seems to indicate that an ounce of gold in 1900 (with compound interest) buys 3,14 times as many goods in 2015. This could mean, for example; that the value of goods has declined 3,14 times as rapidly as gold's value, or just that the paper-price of gold is 3,14 times overpriced (ie a speculative bubble, like exists in wine bottles or art works). But if neither of those two options are the case (for gold is also used an an industrial material), then, if prices are to reflect actual value, the inflation-adjustment which is relied upon here (and I guess the same everywhere) is wrong by 3,14 times.
But regadless of gold, the point is that an error in annual inflation calculation can lead to seriously distorted figures. With new inventions, machinery etc, we can't know what their price evolution is, since they weren't even invented in 1900.
The historical peak of the rate of surplus value and the “new normal” of the Chinese Economy: A Political Economy Analysis (professor Hao Qia, 2018). This study provides (see on page 7) the estimates, that the rate of explotiation in China in 2008 was 255% (after which it stagnated), while in the US in 2008 it was 350%.
Surplus Value Production and Realization in Marxian Theory Applications to the U.S., 1987-2015 (professor Jonathan F Cogliano, 2017). This study (see page 14) estimates the US rate of exploitation by 2015 hovered around 400%.
In my crude model above, I had assumed a starting rate of exploitation (worldwide) of 25% in 1900, which should reach 900% in 2004 (if GDP were to be a reliable indicator). From the studies I linked here, it seems fair to assume, that it was surely not 900% in 2004 worldwide.
Let's be very generous though and assume it was in reality as high as 500% worldwide in 2004 (ie a 10 hour day split into about 8,35 surplus labour hours + 1,65 hours necessary labour), in order to go backward and calculate how much it would be in 1900 (if GDP were a reliable indicator of accumulation), instead of our previous procedure of just positing (a randomly low) 25% rate in 1900.
In 2004 we assumed 2,13 billion workers, so of which 1,78 billion (83,5%) create surplus value; decreased by 1750% gets to 96,2 million surplus workers in 1900, which is 18,15% of total work force then (530 million), so (1,815 sruplus hours divided by 8,185 necessary hours) the rate of exploitation had to be about as low as 22,2% in 1900 worldwide.