Workers' Voice mailing list
October 14, 2018
RE: Replying on labor time calculation and 21st century socialism
By Joseph Green, Detroit Workers' Voice
The issue of what economic planning under socialism would look like was discussed at one of the panels at the 50th anniversary conference of the Union of Radical Political Economists (URPE), which was held at the end of September at the University of Massachusetts-Amherst. Seongjin Jeong put forward the view that money would be replaced by labor certificates, and that planning would be done according to the single measure of the labor hour.
I wasn't at the URPE conference; what I know about it is from a report written up by the left-wing economist Michael Roberts and placed on his blog, and from Jeong's draft paper "Soviet planning and the labor-time calculation model: implications for 21st-century socialism" which Roberts linked to. (1) In his paper, Jeong considers objections to his view, and as part of this, conscientiously refers to my three-part article "Labor-money and socialist planning", which puts forward a very different view. (http://www.communistvoice.org/00LaborHour.html) (2)
My article on socialist planning centered on showing that there was no single measure that could serve as the natural unit of socialist planning, not even the labor-hour, and that the use of the labor-hour as such a measure would result in duplicating many faults of capitalism. It traced the history of the idea of labor money in the socialist movement, and the repeated failures of the attempts to use labor money. It pointed out that the labor certificate under communism, as envisioned as a possibility by Marx, was only to be used for the distribution of consumer goods and not for overall economic planning nor for how workplaces would obtain the goods they needed for their operation. My article pointed to the development of methods to plan in material terms.
This might seem a rather obscure subject, but it bears on many practical matters. For example, the rationale for using market measures for environmental goals, rather than relying mainly on regulation and planning, lies in the belief that a single unit of measure is the way to achieve economic results. The rationale for reducing every decision to a calculation of profit and loss lies in the belief in a single unit of measure. And yet in reality it won't matter that much if money denominated in dollars or other national currency was replaced by calculation in labor-hours.
Moreover, Jeong also claims that the Soviet planning agencies didn't really calculate properly or use input-output tables, and that this was a major cause of the shortages and disproportions in the Soviet economy. According to Roberts, this line of reasoning led to the view that "with the development of AI [artificial intelligence], algorithms, big data and quantum power, such planning by labour time calculation is clearly feasible. Communism will work." In my view, such views slur over the fact that the problem with the Stalinist economy wasn't simply bad choices by Stalin and his successors, nor was it bad calculation due to the lack of computing power, but that the Soviet Union under Stalinism became a state-capitalist country with a new ruling class.
Given the importance of these issues, I would like to take this occasion to reply to Jeong's article, especially as Jeong focuses on several important points of economic analysis.
Jeong holds that "The Marxian model of a communist economy, in its first phase, is characterized by 'planning based on labor-time calculation' (hereafter abbreviated as PLTC)." And he writes that "PLTC is one of the essential components of Marx's communism." (3 - but from here on references to Jeong's paper will simply give the page number)
Now, in order to use the labor-hour in this way, it can't be measured by the labor of any one person -- it has to be labor of an average intensity. This is called the abstract labor hour, not in order to denigrate it, but to distinguish it from the concrete or definite labor hour exerted by a definite person at a definite time and location. Jeong writes "...PLTC...is unthinkable if the unit of calculation is actual or individual labor-time. In the first phase of communism, where the economy of time to cope with the state of scarcity is still needed, PLTC is unavoidable, and its unit should be social." (p.82)
In this system, in order to ensure that every workplace has the necessary inputs, every product is measured in terms of the amount of the congealed labor-time embodied in it. So the PLTC means that one replaces the financial measure of the US dollar or the South Korean won with the abstract labor-hour. And just as financial calculation implies that two things with the same price are equal, the PLTC would imply that two things embodying the same amount of congealed labor-time are equal and interchangeable. (The congealed labor-time is what Jeong calls "the total labor-time embodied in products", p. 83)
Yet Jeong also admits that the PLTC has trouble with environmental issues. He says that "it is difficult for Marxian PLTC to take into consideration all the diversity of human life in an emancipated communist society or ecological issues." (p. 84)
Nor is Jeong the only advocate of the PLTC who admits that it can't deal with environmental issues. Allin Cottrell and W. Paul Cockshott are passionate advocates of using labor-time as the socialist economic measure. And yet they say that "We are not claiming that labor-time calculation would necessarily do better [than capitalism] in cases where the market fails to conserve resources. ... We have no problem with the idea that environmental considerations and labour-time accounting are not necessarily reducible to a scalar common denominator, and that the balancing of these considerations may require political judgment on which opinions can differ." (4)
I discussed this in my article on socialist planning. (5) Indeed, in my view PLTC would have trouble with other issues as well, including health care, education, and child care. There would have to be one correction to PLTC after another to allow sufficient attention to be paid to them. Oops. So much for planning with a single measure.
For Jeong, these problems are simply shortcomings of the PLTC, which show that it won't be used at a later stage of communism. He says that such "qualifications should not be considered a rejection of PLTC. Far from being incongruent with the developed phase of communism, PLTC is indispensable to the advance of communism to its developed phase." (p. 84)
But how can PLTC be a useful economic phase in the 21st century if it doesn't deal with the pressing environmental issues? How can production be organized properly without protecting the environment, and without considering such other issues as the proper support for social programs, issues that PTLC doesn't deal with? It's a farce to talk about "21st century socialism", which is faced with global warming, rampant pollution, and other problems, and put forward a planning system which can't deal with them. And it's a farce to say one has a single unit of measure, if one has to made one correction after another in the calculations with that unit.
Jeong even says "Just replacing a market price-based coordination by a labor-time calculation is not sufficient to free the system from the law of value." (p. 83) And he goes on to say that "the Marxian communist model of PLTC is inherently contradictory, in that it tends to virtually simulate and reproduce the capitalist labor-value system." (p. 84) I strongly agree that replacing financial calculation with labor-hour calculation tends to reproduce capitalist exchange.
But I disagree that it's Marx's model. On the contrary, Marx put forward the labor theory of value to show how capitalism operates and how workers are exploited, not to show how to plan socialist production. I will come back to Marx's view about what's wrong with labor time calculation in a moment. But first let's see why Jeong insists on it as the bottom line of economic calculation.
Jeong adheres to the PLTC because he believes that rational economic planning requires "using a single unit of measure". He writes that "it is impossible to achieve macroeconomic coordination and balance without adopting a single unit of measure that enables the calculation of social averages. ... Planning without this single accounting unit is simply a contradiction in terms, tantamount to the rejection of planning altogether." (p. 81)
Using a single unit of measure means that this unit is the bottom line of calculation - everything is measured or valued in accordance with this unit. This is what is done in capitalist society with money, where the dollar reigns supreme in the US. And then the best alternative is supposedly that which gives the most profit, or which come out highest in a cost-benefit comparison. It is because we are so used to it from everyday market experience that this seems utterly natural to us. Jeong's picture of the early phase of communism preserves this use of a single measure by replacing money with labor certificates based on labor hours, but he recognizes, as we have seen, that having such a single unit "tends to virtually simulate and reproduce the capitalist labor-value system".
But what is the alternative? Is there a way to carry out economic planning without relying on a single unit as the bottom line? Well, first let's see that Marx showed that no single unit would suffice, and then see whether alternative methods of calculation and planning have ever been used in modern industrial economies.
Jeong cites a passage from Marx's "Grundrisse" to back up his claim that a single unit of economic measurement is needed, and that this unit is the labor hour. But when we look a little closer at the passage, it turns out that Marx was actually in the midst of reiterating one of his main theoretical arguments against the use of a single unit of measurement.
Marx wrote: "On the basis of communal production, the determination of time remains, of course, essential. ... Economy of time, to this all economy ultimately reduces itself. ... Thus, economy of time, along with the planned distribution of labor time among the various branches of production, remains the first economic law on the basis of the communal production. However, this is essentially different from a measurement of exchange values (labor or products) by labor time." (6)
By itself, this looks like Marx was emphasizing the use of a single unit, the labor time. The problem, however, is that Jeong left out the rest of the paragraph, which deals with one of the ways in which the treatment of labor time in communal society will differ from that of capitalist society. Marx wrote:
"The labour of individuals in the same branch of work, and the various kinds of work, are different from one another not only quantitatively but also qualitatively. What does a solely quantitative difference between things presuppose? The identity of their qualities. Hence, the quantitative measure of labours presupposes the equivalence, the identity of their quality." (emphasis as in the original)
This is a contrast of abstract and concrete labor. The abstract labor hour only measures labor quantitatively, while concrete labor has definite properties, which are important. Equating things quantitatively, or comparing them only quantitatively, means ignoring the qualitative differences between them.
This is a major theme which he returned to repeatedly in the "Grundrisse" and, for that matter, in "Capital". Clearly Marx's idea was that future society will need to take account of the qualitative differences, something which no single measure, neither money nor measurement by the abstract labor hour, can accomplish. One can't deal seriously with Marx's standpoint about economic planning without taking this into account. At the same time, what's true or not doesn't depend simply on textual analysis. Marx's standpoint and writings deserve close study, but what he said about qualitative differences is correct, not because he said it, but because it corresponds to economic reality.
For example, if labor time were the single unit of measure, then one hour of a carpenter's labor would equal one hour of a welders' labor. Marx is saying no, they are qualitatively different. For example, if we need so many carpenters, we cannot replace them by so many welders. The carpenter and the welder might be paid the same, but if we are to ensure that construction can take place, we need to keep track of carpenters and welders separately.
The same thing holds when one compares two different products or material goods. One labor hour's worth of one material good is not qualitatively equal to one labor hour's worth of a different good. For example, economic planning needs to keep track separately of such things as how much food there is, and how much steel. Food that took one hour of labor time to produce (counting both direct and indirect labor) isn't the same thing as, say, steel that took one hour to produce. With respect to the abstract labor hour, one hour = one hour = one hour. With respect to money, one dollar = one dollar = one dollar. But with respect to rational economic planning, they're different. We can't eat steel. We can't use food to build bridges and buildings. So the amount of food and steel available has to be calculated separately. Thus, effective planning for a project cannot leave it at there are supplies available worth, say, a million labor hours, but has to break down what is needed into different categories. So, while Jeong claims that economic calculation absolutely requires using a "single unit of measure", the opposite is true.
In the quote from "Grundrisse", Marx was saying that economic calculation is always necessary, even under communism (we shall see later on that Jeong doesn't believe this is so in later-phase communism), but it cannot be done with the single measure.
As we have seen, Jeong himself recognizes that PLTC has difficulty accounting for qualitative diversity and environmental issues. But he doesn't realize why this is so. Effective calculation must take account of the abstract labor hour, but it must also deal with other factors of economic importance. It must not reduce matters to a single factor.
Still, thanks to living in a capitalist society and having to buy and sell things all the time, it might seem to be only common sense that calculation requires a single unit of measure. This might seem particularly so if, as is probably only too common, one wasn't taught the spirit of mathematics in school, but only one mechanical rule of calculation after another. But during the last century methods were developed to plan in physical terms. They were used by governments of vastly different types, not just the Soviet Union but to some extent by Western capitalist countries and also some newly-independent countries. This type of planning didn't just take into consideration the overall size of the economy, or the Gross Domestic Product. Nor did it simply list the different types of material goods that existed. It paid attention to the "intersectoral flows" that connected one branch of the economy with the other. Steel, for example, wouldn't be measured simply by how much it cost, nor by the total labor-hours needed to produce it. Instead steel would be measured by the resources that were needed from each other sector of the economy for the production of steel, while the other sectors of the economy were measured, in part, by how much steel they needed for their production.
This new type of calculation was first put forward in the Soviet Union in the 1920s, and was called "material balances". The method used to adjust these material balances was mathematically crude, but was an important new departure in planning. It seems to have helped inspire the development of input-output tables in Western capitalist countries, which were also taken up to some extent in the Soviet Union. In parts one and two of my article on the "labor hour and socialist planning" I explain the basic idea behind these methods in more detail, and I also show that Marx foreshadowed these methods in volume II of "Capital" where he wrote equations connecting how different sectors of a capitalist economy interact with each other.
These methods of physical planning were not communist planning, and as mentioned, were used by governments of all types. But they did show that it is possible to calculate using multiple units of measure, and that it was absolutely necessary to do so to achieve realistic results. That's why even diehard capitalist governments make a certain use of them, albeit only as an occasional supplement to overall financial planning. So these methods did not achieve, or in most cases even aim at, a fully planned economy, but they showed that some physical planning was needed even for very limited goals.
Jeong holds that I champion the way the Soviet Union prepared
material balances as against input-output tables. But in my article, I
wrote that the "Western input-output methods...are essentially a
variant of the method of material balances. Input-output methods start
... from the point of working with balances of goods regarded as
qualitatively different. ... Western economic authors like to make
various fine distinctions between material balances, input-output
methods, and linear programming. ... [But these distinctions] don't
affect the issue raised in this article. What is important, is that all
these methods use a multitude of separate balances of qualitatively
distinct goods." (7) So I
sometimes grouped all these methods, including the "Soviet version of
material balances", under the general term "material balances". I
stressed that the various forms of material balances were not
socialist planning, and all that they had "in common with the future
economic calculation of a classless society, is that they both have to
keep track of society's production in material terms." (8)
Jeong seeks to discredit the idea of material balances because it means abandoning the single unit. He promotes input-output tables, because he believes this means using a single measure.
To do so, he writes "It is ... not correct to regard material balances as a specifically communist method of resource allocation. Even in capitalism, variants of material balances have been frequently used to allocate some strategically crucial goods when the monetary economy does not work, like during wartime." (p. 81) I agree strongly with him on this, and I repeatedly made the same point, right up to the use of the same example of wartime planning, in my article on socialist planning.
But the same points also apply to input-output tables and linear programming. They are only technical planning tools, while socialism is based on whether the working class owns and directs the economy. It's the social structure of a country, the class regulations, that define socialism, not the planning tools. And socialist planning involves planning from below as well as above. Without a socialist structure, no matter how much planning is attempted, only a limited amount can be achieved.
For that matter, in my article on socialist planning I pointed out that material balances and input-output tables aren't even the last word in physical planning. For example, these methods depend on certain assumptions about the economy, which are only approximately true (such as that twice as much production always requires exactly twice as much resources). Environmental issues could be included in the calculations but generally aren't. And so forth. These methods are but a step in the direction of realistic physical planning. They have in common with future communist planning that they make use of not one, but many separate natural units.
Jeong writes in praise of input-output tables that "input-output tables can be compiled in terms of physical natural units as well as monetary or labor-time ones". (p. 81) And this is true. But what Jeong misses is that when input-output tables are written in terms of many physical natural units, then they are being used as material balances, whereas otherwise they might not be. There is a subtlety here. The difference isn't simply whether the table has entries written in financial terms. It's whether the table attempts to combine categories which are qualitatively different into a single combined or aggregate category (for example, combining the quantity of food and that of steel into a combined category which embraces both). This is explained in more detail in my article on socialist planning, which also refers to how this distinction was drawn by the late Wassily Leontief, the bourgeois economist who was the father of Western input-output analysis. (9)
The point here is that when input-output tables represent material planning, they are an example of calculations that go beyond using a single unit of measure. When, instead, they use a single unit of measure, whether the dollar or the labor-hour, to combine entries with qualitatively different properties, they have what Leontief very politely called "a faint but unmistakable air of unreality". (10)
Jeong insists that the PLTC is indispensable in the early stage of communism, but says it will vanish in developed communism. That is how he reconciles his recognition that using the abstract labor-hour as the single unit will be similar to "the capitalist labor-value system" with his picture of communism. (p. 84) He holds that PLTC will lead to economic advances that will lay the basis for the advanced phase of communism that will do without the PLTC.
But wait a minute! If economic calculation is supposed to require a single unit of calculation, namely the abstract labor hour of the PLTC, how can economic calculation be carried out in advanced communism without the PLTC? Apparently Jeong believes that there will be no economic calculation in advanced communism. This is utterly astonishing, but that is where his argument leads.
Moreover, he writes that "labor will ... be transformed into activities", and that "the essence of Marxian communism is not the domination of labor but its abolition." (p. 84) True, Marx envisioned in his "Critique of the Gotha Program" that labor will change from something workers have to do into "life's prime want". But Jeong goes further, and describes this as the abolition of labor. And, of course, if labor is abolished, then it follows that labor-time vanishes too. But according to Jeong's system, with the lack of that single unit of measurement, the abstract labor-hour, "macroeconomic coordination and balance" is impossible to achieve, and there would be "the rejection of planning altogether."
Now, Marx too thought that labor certificates would vanish as communism developed. Jeong writes about "needs-based distribution" and pays attention to how from the start of communism some of the social product will be distributed freely. (pp. 83-84) This is in accordance with Marx. But Jeong then identifies how much economic planning takes place with how many labor certificates are circulating. This means to forget about the planning that is necessary in order to produce the goods. Marx, on the contrary, distinguished between how goods are distributed, and whether there had to be economic planning.
After all, where do the necessities and luxuries of life come from? From where do all the good things come that are distributed among people, and how does one ensure that production is carried out in a way that protects the environment? This requires planning. Marx stressed repeatedly that economic planning wouldn't vanish in communism, but simply take on another form. Recall that Jeong himself, when arguing for the single economic measure, cites a passage in which Marx says that "the planned distribution of labour time among the various branches of production remains the first economic law" of communist production. Indeed, Marx wrote that such planning "becomes law, to an even higher degree" than existed under capitalism. Nowhere does Marx suggest or even hint that this planning will diminish as communism develops.
It is not Marx, but Jeong, who first assumes that the economic planning must be carried out in a way similar to that of "the capitalist labor-value system", and then concludes that therefore, to avoid the capitalist labor-value system, there must be an end to labor and calculation itself.
While I disagree with many things that Jeong puts forward, I think
he has done a service by putting them forward and looking into the
contradictions in them. If I have time, I will take up the issue of
what went wrong with the Soviet economy in a continuation of this
(1) Roberts wasn't at the conference either, and says that his report is "solely based on some of the papers presented that I have obtained and also from some of the comments on the sessions by participants." His report on the 50th URPE conference can be found at https://thenextrecession.wordpress.com/2018/10/02/50-years-of-radical-political-economy/ under the title "50 years of radical political economy".
Jeong's draft article, "Soviet planning and the labor-time calculation model: implications for 21st century socialism", can be found at https://thenextrecession.files.wordpress.com/2018/10/sovietplanningltc_seongjin_urpe20180928.pdf. It is contained in "Varieties of Alternative Economic Systems", pp. 71-87.
(2) The table of contents of the entire three-part article with the overall title "Labor-money and socialist planning" and links to all three parts can be found at http://www.communistvoice.org/00LaborHour.html. The article opposes the idea of labor money and advocates that the Marxist labor theory of value does *not* mean that the labor-hour is the natural unit of socialist calculation. Jeong's criticism of my article occurs on p. 81 of his article.
(3) From Jeong's draft paper, p. 71, the parenthetical comment is Jeong's.
quote from Cottrell and Cockshott is from "Calculation,
Complexity and Planning: The Socialist Calculation Debate Once again",
in Review of Political Economy, vol. 5, no. 1, 1993, Section II.3.1. It
is also available at
(5) See "The environment and things of zero labor content" in part 2 of "Labor money and socialist planning", http://www.communistvoice.org/26cLaborHour2.html.
(6) "Grundrisse: Foundations of the Critique of Political Economy", Translated with a Foreword by Martin Nicolaus, Pelican Marx Library (Penguin Classics), The Chapter on Money, p. 173. The passage is cited by Jeong on p. 71. I discuss this passage in some detail in the section "Concrete and abstract labor" of part 3 of my article on socialist planning.
(7) See the section "Soviet material balances and Western input-output methods" in "Labor-money and socialist planning, pt. 2".http://www.communistvoice.org/26cLaborHour2.html.
Strictly speaking, the use of input-output tables is similar to the method of material balances only when those input-output tables keep track exclusively of material goods and the intersectoral connections between them. Input-output tables can also be used for ordinary financial planning, and are widely so used by bourgeois economists and even international capitalist agencies. This is developed further in the section "Input-output tables may or may not be material planning".
(8) See the section "One, two, three, many natural units (the method of material balances)" in "Labor-money and socialist planning, pt. 1" http://www.communistvoice.org/25cLaborHour.html.
(9) See the section "Soviet material balances and Western input-output methods" in "Labor-money and socialist planning, pt. 2". It should be noted that the use of "shadow prices" in linear programming shows the departure from physical planning. This doesn't mean linear programming is useless, but that it can't serve as the bottom line of realistic calculation.
(10) See Leontief, "Essays in Economics", ch 2 and 4.
In the D/SWV list item for
October 3, 2018,
* In the excerpts from "brics-from-below reader", under Chapter 2,
in the fourth paragraph, "the [1015-1916] crisis" should be "the
* In the notice about the CVO web outage, the dates of the outage
should have been Sept. 21-26, not July 21-26. <>
Back to main
how to order CV,
Posted on October 16, 2018
Some typos have been corrected.