Robert Kurz, David Graeber’s “Bullshit Jobs” and the collapse of capitalism

by Jehu

In section 3 of the Apotheosis of Money, Robert Kurz takes on a problem that has been discussed by Marxists for almost 80 years. The discussion is important because, I believe, its conclusions will have more to say about what a post-capitalist society actually looks like than any of the issues typically raised by the Left:

“it is clear that, taken as a whole, the share of those unproductive workers (viewed from the perspective of surplus value production) who only represent social consumption, that is, “overhead costs”, is constantly increasing.” way to state this is that an ever increasing share of the total labor time of society is being wasted on unproductive labor —   an issue raised by David Graeber in his recent essay, On the Phenomenon of Bullshit Jobs.

Kurz’s essay is significant to me, because if, as I think, communism is free time and nothing else, society seems to be preparing for just this possibility. The collapse of capitalism will likely appear, as it did in the Soviet Union, as a crisis in the form of massive unemployment, which locks billions out of the labor market, making it possible for the social producers to eliminate most work in a single stroke. The preparation for a society founded on free, disposable time for the many, who can use this free time for self-activity in whatever way they want, may just be taking the form of an ever increasing quantity of labor time that is being unproductively expended by capitalist firms.

Do bullshit jobs even exist?

One of the big problems in even establishing that labor is going to go away and take capitalism with it is establishing that unproductive labor exists and can be quantified by some means. Defining the category, unproductive labor, has been really difficult — so difficult that many labor theorists have given up on the attempt after one or two stabs at it. For instance, in 2007, Chris Harman tried to address it in an essay “The rate of profit and the world today”, within the context of a much larger discussion of stagnant capitalist growth. In this essay, Harman references numerous Marxist writers who attempted to solve the problem before him:

“Moseley, Shaikh and Tonak, and Simon Mohun have all noted another feature of capitalism’s most recent development – one highlighted by Kidron back in the 1970s. This is the growing “non-productive” portion of the economy.”

Among bourgeois simpleton economists, the problem of unproductive labor is not really a problem, since they define productive labor as any labor for which someone is paid. There is, however, no commonly agreed on definition of unproductive labor among Marxists and the effort has been mostly hit or miss. The result are wildly varying estimates of the volume of labor in society that is unproductive. According to Harman:

“Fred Moseley estimates the numbers in commerce in the US grew from 8.9 million to 21 million between 1950 and 1980, and the number in finance from 1.9 million to 5.2 million, while the productive workforce only grew from 28 million to 40.3 million.  Shaikh and Tonak calculate that the share of productive out of total labour in the US fell from 57 percent to 36 percent between 1948 and 1989.  Simon Mohun has calculated that the share of “unproductive” wages and salaries in “material value added” in the US grew from 35 percent in 1964 to over 50 percent in 2000.  Kidron calculated that, using his wide definition, ‘Three fifths of the work actually undertaken in the US in 1970s was wasted from capital’s own point of view’.”

These wildly varying estimates of the unproductive labor in the economy resemble the wildly varying estimates of the debate over the law of the tendency of the rate of profit to fall; and the discussion has remained just as unsettled, since no one seems to be able to agree with anyone else. Labor theorists seem to all agree that much of labor being performed is unnecessary and could be eliminated; they just don’t agree on the little things, like what labor is actually unproductive, how much of this unproductive labor exists in the economy, and where in the economy this unproductive labor is actually being performed.

So, although they cannot tell us anything concrete about unproductive labor, labor theorist all agree it’s out there somewhere and it is a problem.

Robert Kurz made his own attempt at the problem in 1995. He, like Harman, did not make an attempt to empirically quantify the problem, but he tried to define it conceptually. What I like about the way Kurz approached the problem is that he failed spectacularly, I now believe, but moved the ball down the field in a very significant way. (When I say I now believe this, I mean, I now believe it since I recommended it to @DebbieMacAlpin this weekend.) His failure, I think, brings an end to the approach favored by Marxists for four decades now.

While rereading Kurz’s essay, it occurred to me why he ran into difficulties that led him to predict the collapse of the dollar in that essay. That difficulty results, in part, from the assumptions he made when discussing unproductive labor.

Unproductive labor and the emergence of the service economy

After WWII, a number of observers began discussing the increasing weight of the service sector in the US economy and many were talking about it as if it prefigured the emergence of a “new economy”; our “economy” was becoming “post-industrial” as it had previously become “post-feudal”. Kurz dismisses this idea by stating industrialization was not a new form of economy but a replacement of the previous mode of production, founded on peasant production, by capitalist relations of production:

“This superficial reflection, however, completely ignores the fact that the first of these changes in the reproductive structure did not by any means constitute an internal development within capitalism, but coincided with the very history of capitalism’s formation and rise. Not only were the techniques and material contents of production modified, but the basic forms of social relations were also convulsed by a long, painful and turbulent transformation.”

Capitalism was a new mode of production which overturned the earlier mode of production and the social relations founded on that mode of production — it was, in a word, a revolution in every sense of that term. However, according to Kurz, the so-called post-industrial service economy did not represent a new mode of production, but a change within the existing one.

“In other words: the loss of social importance of the industrial “sectors” might very well be identical to a crisis and a loss of importance of the market and money, in the capitalist form as a general form of reproduction”.

Which is to say, the growing weight of the service sector — the proliferation of what David Graeber called “bullshit jobs” — in the economy may be signaling the end of capitalism.

You have to realize what Kurz was doing here and why it is so important to us today: This observation by Kurz represented a real advance in the discussion of the problem of unproductive labor. For the first time, to my knowledge, a labor theorist was placing the discussion of unproductive labor within the context of a discussion of an impending collapse of capitalism. Previous discussions of the problem had been located within the context of the so-called stagnation thesis — the idea that the rate of capitalist growth was slowing. (If you follow bourgeois simpletons regularly, you know that Larry Summers raised the stagnation thesis as a way of looking at the failure of economic policy within the present crisis.)

An example of how “bullshit jobs” has been long discussed can be seen in the previously mentioned essay by Chris Harman, where he quotes an earlier essay of his from 1982:

“State intervention to mitigate the crisis can only prolong it indefinitely. This does not mean the world economy is doomed simply to decline. An overall tendency to stagnation can still be accompanied by boomlets, with small but temporary increases in employment. Each boomlet, however, only aggravates the problems of the system as a whole and results in further general stagnation, and extreme devastation for particular parts of the system.”

While Harman employed the growing mass of unproductive labor to account for the slowing rate of growth of capitalist economies, Kurz employed it to illustrate his contention that capitalism had already entered the end phase of its shelf-life. This was an entirely new way of approaching the analysis of present day capitalism and what the slowing rate of growth of advanced economies implied.

To illustrate the problem posed by the growing mass of unproductive labor in the economy, Kurz began with a discussion of the notion of unproductive labor, which he defines initially only as the opposite of productive labor. It turns out that, from the point of view of the capitalist mode of production, all labor carried on within previous modes of production is unproductive in the capitalistic sense. In fact, says Kurz, from the point of view of capital, labor in the previous modes of production can’t even be called labor as this term is use by labor theorists today; what we call labor is specific to the capitalist mode of production and is not at all to be confused with the production of useful objects.

Moreover, according to Kurz not only is feudal labor not productive in the capitalistic sense, not all paid labor within the capitalist mode of production is productive. Following Marx, Kurz argues:

“Within the latter system, all activity carried out in exchange for money or which is situated within the context of the valorization of money is formally abstract labor. But this does not mean that it would also be so in a substantive sense. In a substantive sense, abstract labor, that is, labor whose expenditure of energy really impels capitalist reproduction, is only “productive” (capital-producing) labor which effectively creates surplus value.”

So Kurz settled on Marx’s simple definition that the only labor that can be considered productive labor under capitalism is labor that produces surplus value. So far, so good.

The difficulty identifying unproductive labor

But then he had a problem: there is a lot of labor out there that might produce a profit, but does not necessarily produce surplus value. To make an analogy, at the beginning of Capital, Marx explains that not every object with a price necessarily has value and he lists in this category such things as virgin land, a person’s honor and things of this sort. The same is true for profit: not all labor that produces a profit for the owner of a capital necessarily produces surplus value.

To illustrate his argument, he discusses labor that can be said to constitute overhead within a corporation, e.g., security or HR. The labor of a security guard is an overhead cost to a company and represents a direct loss of a portion of its surplus value. As such, the corporation will endeavor to keep overhead costs like this in check by various means. Kurz asks: but what if this work was outsourced to a security services company and that company made a profit selling security services back to the the firm and other companies. Would this security services labor now be productive labor?

Kurz says, “Well, yes and no.” Yes it would be productive for the security services company in question, since it produced a profit; but at the level of the entire economy the overhead would still exist, although it would not occur on anyone’s balance sheet.

“The tertiarization [i.e., outsourcing security services — Jehu] referred to above therefore transforms, so it would seem, unproductive into productive labor, by means of its simple formal autonomization in a separate enterprise. But things are different on the plane of the total capital, which obviously does not immediately appear on the balance sheets of the so-called economic subjects, but which can, however, be theoretically and analytically reconstructed.”

The capitalist firm can reduce the cost of security by outsourcing the overhead labor to a specialized security services firm, and the specializing firm can realize a profit by reducing the labor costs (for instance, by installing a network of security cameras to replace most of the security guards) but the overhead cost of security, if it still exists, must be subtracted from total volume of surplus value produced by the total social capital. The outsourcing firm in question escapes the problem of unproductive labor within the firm only to encounter it again on the level of the total capital.

The problem for analysis, however, is this: an overhead cost to the total surplus produced in the economy in the form of security services now appears as a very profitable venture for a capitalist firm specializing in security services. And this apparently profitable paid labor appears to observers as equally “necessary” as the industrial paid labor that produces a chicken in every pot and a car in every driveway. The very way capitalist relations of production are structured makes it nearly impossible to distinguish between productive and unproductive labor. It can, at its extreme, make entirely wasteful labor appear productive simply because this wasteful labor produces a profit for some firm. (Think of this in terms of such very profitable defense department firms as Raytheon.)

Capitalism as a system of continuous circulation of values

So the question Kurz had to answer was this: Is there a non-arbitrary (objective) way to separate productive and unproductive labor? How could we get away from subjective determinations of whether labor was productive or unproductive. This had been the obstacle encountered by previous analyses of labor theorists and explained why the estimates of unproductive labor by various theorists varied so wildly. To avoid the mistake of previous efforts, Kurz suggested we had to think of capitalism as a complete system of circulation:

“What, then, is the decisive criterion for conceptually determining, on the plane of the total capital (that is, after eliminating the distortion that typically accompanies the point of view of the particular capital), whether or not labor is productive?  … A definition of productive labor, with reference to the mediation process of capitalist reproduction as a whole, can only be presented in the last instance in terms of the theory of circulation. In other words: in terms of the theory of circulation, only that labor whose products (as well as its reproduction costs) return to the capital accumulation process is productive; that is, labor whose consumption is recovered again in expanded reproduction. Only this kind of consumption is “productive consumption”, not only directly but also with reference to reproduction.”

To really understand the problem of unproductive labor in a capitalist economy, says Kurz, we have to think not only in terms of the production of values, but also about the extent to which these newly produced values find their way back into productive employment for reproduction of capital on a still larger scale. The circulation of the value productively consumed does not end with its immediate consumption, but reappears within capitalist circulation as additional productive labor.

Kurz hits the wall in his analysis

And this is where Kurz begins to go off the rails in his analysis.  — he assumed the problem was with labor theory itself, not the methods and assumptions of these labor theorists:

“At first glance, it seems hard to imagine how this distinction could be argued in an analytically transparent way, without falling prey to arbitrary assumptions. In this respect, Marx’s theory does not have any instruments at its disposal capable of an unequivocal assessment; hence, the Marxist debate concerning “productive and unproductive labor”, which was hardly articulated as a whole, never reached a conclusion.”

You can tell Kurz ran into a brick wall at this point as well and he admits it:

“Conceiving productive labor in terms of circulation theory may seem strange to those who, influenced by positivism, want strict definitions…”

Well, yeah. Sort of.

We would like to be able to quantify the exact mass of unproductive labor in the economy hence the need for “strict definitions”.  The entire notion of communism as non-labor time for the vast majority hinges on the definition and empirical measurement of unproductive labor in the economy. This problem is not a positivist mistake made by some labor theorists but a problem of their assumptions. Conceptually it is okay to say unproductive labor is labor whose product does not reenter the production process — and this is what Kurz has told us — but he has not told us how much labor qualifies as unproductive, nor how we might measure it objectively.

I can tell Kurz has run into a wall in his analysis because, at this point in his essay, he then begins to do what all the Marxists before him did: that is, Kurz begins to enumerate examples of particular forms of labor he thinks may be unproductive. Kurz is correct to state labor theory has no tool to decide whether, for instance, the concrete labor of a doctor is productive or unproductive. A doctor treating a steelworker is being employed differently than a doctor treating a soldier to return him to the battlefield.  But the problem Kurz is trying to clarify in his essay is not the problem of particular forms of concrete labor — he is concerned with abstract, homogenous, value producing labor.

Commodity money and the problem of bullshit jobs

How can we tell when the concrete labor expended by social producers no longer produces value? In Marx’s theory, the value of a commodity produced by such abstract homogenous labor can only be measured by another commodity — a money commodity. There is, in fact, no connection between the concrete labor expended in the production of a commodity and the abstract homogenous value producing quality of that labor within labor theory. There is no way to look at concrete labor and decide, by any characteristic of this concrete labor, whether it produces value; which is why it is impossible to tell whether a capital is productive or unproductive of surplus value simply by looking at its balance sheet.

Labor that might be considered clearly productive, like a steelworker, turns out to be unproductive when the steel is used in a nuclear sub. A construction worker performs the same labor no matter that he is putting up sheetrock in a new home or in a new facility for the NSA. In each case we are not talking about the specific useful labor performed by the workers, but the abstract socially necessary labor congealed in the product of their labor.

The quantity of socially necessary labor time in any commodity cannot be directly observed; it can only be indirectly measured in the form of exchange value. Thus, when Kurz began to enumerate the various types of concrete labor that he felt might be unproductive based on his criteria, this was a sign he had failed in his effort — he was reduced, by his own argument, to subjectively identify various forms of concrete labor like his predecessors, although his definition of productive labor was premised on abstract homogenous labor.