Why is it that some unions or groups of workers don't start their own business? What an easy way to give workers access to the means of production: buy machinery and rent space, and so forth, just as a private entrepreneur does. It is illuminating to consider why unions don't start new businesses, and why workers don't pool their resources to do so.

This question is of importance for what remains of Marxist economic theory. With the crumbling of the labor theory of value, the underpinning of its particular theory of exploitation dissolves. And the charm and simplicity of this theory's definition of exploitation is lost when it is realized that according to the definition there will be exploitation in any society in which investment takes place for a greater future product (perhaps because of population growth); and in any society in which those unable to work, or to work productively, are subsidized by the labor of others. But at bottom, Marxist theory explains the phenomenon of exploitation by reference to the workers not having access to the means of production. The workers have to sell their labor (labor power) to the capitalists, for they must use the means of production to produce, and cannot produce alone. A worker, or groups of them, cannot hire means of production and wait to sell the product some months later; they lack the cash reserves to obtain access to machinery or wait until later when revenue will be received from the future sale of the product now being worked on. For workers must eat in the meantime.[1] Hence (the story goes) the worker is forced to deal with the capitalist. (And the reserve army of unemployed labor makes unnecessary the capitalists' competing for workers and bidding up the price of labor.)

Note than once the rest of the theory, properly, is dropped, and it is this crucial fact of nonaccess to the means of production that underlies exploitation, it follows that in a society in which the workers are not forced to deal with the capitalist, exploitation of laborers will be absent. (We pass over the question of whether workers are forced to deal with some other, less decentralized group.) So, if there is a sector of publicly owned and controlled (what you will) means of production that is expandable so that all who wish to may work in it, then this is sufficient to eliminate the exploitation of laborers. And in particular, if in addition to this public sector there is a sector of privately owned means of production that employs wage laborers who choose to work in this sector, then these workers are not being exploited. (Perhaps they choose to work there, despite attempts to convince them to do otherwise, because they get higher wages or returns in this sector.) For they are not forced to deal with the private owners of means of production.

Let us linger for a moment upon this case. Suppose that the private sector were to expand, and the public sector became weaker and weaker. More and more workers, let us suppose, choose to work in the private sector. Wages in the private sector are greater than in the public sector, and are rising continually. Now imagine that after a period of time this weak public sector becomes completely insignificant; perhaps it disappears altogether. Will there be any concomitant change in the private sector? (Since the public sector was already small, by hypothesis, the new workers who come to the private sector will not affect wages much.) The theory of exploitation seems committed to saying that there would be some important change; which statement is very implausible. (There's no good theoretical argument for it.) If there would not be a change in the level or the upward movement of wages in the private sector, are workers in the private sector, heretofore unexploited, now being exploited? Though they don't even know that the public sector is gone, having paid scant attention to it, are they now forced to work in the private sector and to go to the private capitalist for work, and hence are they ipso facto exploited? So the theory would seem to be committed to maintaining.

Whatever may have been the truth of the nonaccess view at one time, in our society large sections of the working force now have cash reserves in personal property, and there are also large cash reserves in union pension funds. These workers can wait, and they can invest. This raises the question of why this money isn't used to establish worker-controlled factories. Why haven't radicals and social democrats urged this?

The workers may lack the entrepreneurial ability to identify promising opportunities for profitable activity, and to organize firms to respond to these opportunities. In this case, the workers can try to hire entrepreneurs and managers to start a firm for them and then turn the authority functions over to the workers (who are the owners) after one year. (Though, as Kirzner emphasizes, entrepreneurial alertness would also be needed in deciding whom to hire.) Different groups of workers would compete for entrepreneurial talent, bidding up the price for such services, while entrepreneurs with capital attempted to hire workers under traditional ownership arrangements. Let us ignore the question of what the equilibrium in this market would look like to ask why groups of workers aren't doing this now.

It's risky starting a new firm. One can't identify easily new entrepreneurial talent, and much depends on estimates of future demand and of availability of resources, on unforeseen obstacles, on chance, and so forth. Specialized investment institutions and sources of venture capital develop to run just these risks. Some persons don't want to run these risks of investing or backing new ventures, or starting ventures themselves. Capitalist society allows the separation of the bearing of these risks from other activities. The workers in the Edsel branch of the Ford Motor Company did not bear the risks of the venture, and when it lost money they did not pay back a portion of their salary. In a socialist society, either one must share in the risks of the enterprise one works in, or everybody shares in the risks of investment decisions of the central investment managers. There is no way to divest oneself of these risks or to choose to carry some such risks but not others (acquiring specialized knowledge in some areas), as one can do in a capitalist society.

Often people who do not wish to bear risks feel entitled to rewards from those who do and win; yet these same people do not feel obligated to help out by sharing the losses of those who bear risks and lose. For example, croupiers at gambling casinos expect to be well-tipped by big winners, but they do not expect to be asked to help bear some of the losses of the losers. The case for such asymmetrical sharing is even weaker for businesses where success is not a random matter. Why do some feel they may stand back to see whose ventures turn out well (by hindsight determine who has survived the risks and run profitably) and then claim a share of the success; though they do not feel they must bear the losses if things turn out poorly, or feel that if they wish to share in the profits or the control of the enterprise, they should invest and run the risks also?

To compare how Marxist theory treats such risks, we must take a brief excursion through the theory. Marx's theory is one form of the productive resources theory of value. Such a theory holds that the value V of a thing X equals the sum total of society's productive resources embodied in X. Put in a more useful form, the ratio of the value of two things V(X)/V(Y) is equal to the ratio of the amount of productive resources embodied in them, M (resources in X)/M (resources in Y), where M is a measure of the amount. Such a theory requires a measure M whose values are determined independently of the V ratios to be explained. If we conjoin to the productive resources theory of value, the labor theory of productive resources, which holds that labor is the only productive resource, we obtain the labor theory of value. Many of the objections which have been directed toward the labor theory of value apply to any productive resources theory.

An alternative to the productive resources theory of value might say that the value of productive resources is determined by the value of the final products that arise from them (can be made from them), where the value of the final product is determined in some way other than by the value of the resources used in it. If one machine can be used to make X (and nothing else) and another can be used to make Y, then the first machine is more valuable than the second, even if each machine contains the same raw materials and took the same amount of time to make. The first machine, having a more valuable final product, will command a higher price than the second. This may give rise to the illusion that its products are more valuable because it is more valuable. But this gets things backwards. It is more valuable because its products are.

But the productive resources theory of value doesn't talk about the value of the productive resources, only about their amounts. If there were only one factor of production, and it were homogeneous, the productive resources theory at least could be noncircularly stated. But with more than one factor, or one factor of different kinds, there is a problem in setting up the measure M to get the theory stated in a noncircular way. For it must be determined how much of one productive factor is to count as equivalent to a given amount of another. One procedure would be to set up the measure by reference to the values of the final products, solving the ratio equations. But this procedure would define the measure on the basis of information about final values, and so could not be used to explain final values on the basis of information about the amount of inputs.[2] An alternative procedure would be to find some common thing that can be produced by X, and Y, in different quantities, and to use the ratio of the quantities of final product to determine the quantities of input. This avoids the circularity of looking at final values first; one begins by looking at final quantities of something, and then uses this information to determine quantities of input (to define the measure M). But even if there is a common product, it may not be what the different factors are best suited for making; and so using it to compare them may give a misleading ratio. One has to compare different factors at their individual best functions. Also, if two different things can be made by each resource, and the ratios of the amounts differ, there is the problem of which ratio is to be picked to provide the constant of proportionality between the resources.

We can illustrate these difficulties by considering Paul Sweezy's exposition of the concept of simple, undifferentiated labor time.[3] Sweezy considers how skilled and unskilled labor are to be equated and agrees that it would be circular to do so on the basis of the value of the final product, since that's what's to be explained. Sweezy then says that skill depends on two things: training and natural differences. Sweezy equates training with the number of hours spent in training, without looking to the skill of the teacher, even as crudely measured by how many hours the teacher spent in training (and how many hours his teacher did?). Sweezy suggests getting at natural differences by having two persons make the same thing, and seeing how the quantities differ, thus finding the ratio to equate them. But if skilled labor of some sort is not best viewed as a faster way of producing the same product that unskilled labor produces, but rather as a way of producing a better product, then this method of defining the measure M won't work. (In comparing Rembrandt's skill with mine, the crucial fact is not that he paints pictures faster than I do.) It would be tedious to rehearse the standard counterexamples to the labor theory of value: found natural objects (valued above the labor necessary to get them); rare goods (letters from Napoleon) that cannot be reproduced in unlimited quantities; differences in value between identical objects at different places; differences skilled labor makes; changes caused by fluctuations in supply and demand; aged objects whose producing requires much time to pass (old wines), and so on.[4]

The issues thus far mentioned concern the nature of simple undifferentiated labor time, which is to provide the unit against which all else is to be measured. We now must introduce an additional complication. For Marxist theory does not hold that the value of an object is proportional to the number of simple undifferentiated labor hours that went into its production; rather, the theory holds that the value of an object is proportional to the number of simple undifferentiated socially necessary labor hours that went into its production.[5] Why the additional requirement that the labor hours be socially necessary? Let us proceed slowly.

The requirement that an object have utility is a necessary component of the labor theory of value, if it is to avoid certain objections. Suppose a person works on something absolutely useless that no one wants. For example, he spends his hours efficiently making a big knot; no one else can do it more quickly. Will this object be that many hours valuable? A theory should not have this consequence. Marx avoids it as follows: "Nothing can have value without being an object of utility. If a thing is useless so is the labor contained in it; the labor does not count as labor, and therefore creates no value."[6] Isn't this an ad hoc restriction? Given the rest of the theory, why does it apply? Why doesn't all efficiently done labor create value? If one has to bring in the fact that it's of use to people and actually wanted (suppose it were of use, but no one wanted it), then perhaps by looking only at wants, which have to be brought in anyway, one can get a complete theory of value.

Even with the ad hoc constraint that the object must be of some use, there remain problems. For, suppose someone works for 563 hours on something of some very slight utility. Is its value now determined by the amount of labor, yielding the consequence that it is incredibly valuable? No. "For the labor spent on them (commodities) counts effectively only insofar as it is spent in a form that is useful to others."[7] Marx goes on to say: "Whether that labor is useful for others, and its product consequently capable of satisfying the wants of others, can be proved only by the act of exchange." If we interpret Marx as saying, not that utility is a necessary condition and that (once satisfied) the amount of labor determines value, but rather that the degree of utility will determine how much (useful) labor has been expended on the object, then we have a theory very different from a labor theory of value.

We can approach this issue from another direction. Suppose that useful things are produced as efficiently as they can be, but that too many of them are produced to sell at a certain price. The price that clears the market is lower than the apparent labor values of the objects; a greater number of efficient hours went into producing them than people are willing to pay for (at a certain price per hour). Does this show that the number of average hours devoted to making an object of significant utility doesn't determine its value? Marx's reply is that if there is such overproduction so that the market doesn't clear at a particular price, then the labor was inefficiently used (less of the thing should have been made), even thought the labor itself wasn't inefficient. Hence not all of those labor hours constituted socially necessary labor time. The object does not have a value less than the socially necessary number of labor hours expended upon it, for there were fewer socially necessary labor hours expended upon it than meet the eye.

Suppose that every piece of linen in the market contains no more labor-time than is socially necessary. In spite of this, all the pieces taken as a whole may have had superfluous labor time spent upon them. If the market cannot stomach the whole quantity at the normal price of 2 shillings a yard, this proves that too great a portion of the total labor of the community has been expended in the form of weaving. The effect is the same as if each weaver had expended more labor-time upon his particular product than is socially necessary.[8]

Thus Marx holds that thus labor isn't all socially necessary. What is socially necessary, and how much of it is, will be determined by what happens on the market!![9] There is no longer any labor theory of value; the central notion of socially necessary labor time is itself defined in terms of the processes and exchange ratios of a competitive market![10]

We have returned to our earlier topic, the risks of investment and production, which we see transforms the labor theory of value into one defined in terms of the results of competitive markets. Consider now a system of payment in accordance with simple, undifferentiated, socially necessary labor hours worked. Under this system, the risks associated with a process of production are borne by each worker participating in the process. However many hours he works at whatever degree of efficiency, he will not know how many socially necessary labor hours he has worked until it is seen how many people are willing to buy the products at what price. A system of payment in accordance with the number of socially necessary labor hours worked therefore would pay some hard-working laborers almost not at all (those who worked for hula hoop manufacturers after the fad had passed, or those who worked in the Edsel plant of the Ford Motor Company), and would pay others very little. (Given the great and nonaccidental incompetence of the investment and production decisions in a socialist society, it would be very surprising if the rulers of such a society dared to pay workers explicitly in accordance with the number of "socially necessary" labor hours they work!) Such a system would compel each individual to attempt to predict the duture market for the product he works on; this would be quite inefficient and would induce those who are dubious about the future success of a product to forgo a job they can do well, even though others are confident enough of its success to risk much on it. Clearly there are advantages to a system which allows persons to shift risks they do not themselves wish to bear, and allows them to be paid a fixed amount, whatever the outcome of the risky process.[11] There are great advantages to allowing opportunities for such specialization in risk-bearing; these opportunities lead to the typical gamut of capitalist institutions.

Marx attempts to answer the following Kantian-type question: how are profits possible?[12] How can there be profits if everything gets its full value, if no cheating goes on? The answer for Marx lies in the unique character of labor power; its value is the cost of producing it (the labor that goes into it), yet it itself is capable of producing more value than it has. (This is true of machines as well.) Putting a certain amount of labor L into making a human organism produces something capable of expending an amount of labor greater than L. Because individuals lack the resources to wait for the return from the sale of the products of their labor (see above), they cannot gather these benefits of their own capacities and are forced to deal with the capitalists. In view of these difficulties with Marxist economic theory, one would expect Marxists to study carefully alternative theories of the existence of profit, including those formulated by "bourgeois" economists. Though I have concentrated here on issues about risk and uncertainty, I should also mention innovation (Schumpeter) and, very importantly, the alertness to and search for new opportunities for arbitrage (broadly conceived) which others have not yet noticed.[13] An alternative explanatory theory, if adequate, presumably would remove much of the scientific motivation underlying Marxist economic theory; one might be left with the view that Marxian exploitation is the exploitation of people's lack of understanding of economics.

Notes

1. Where did the means of production come from? Who earlier forwent current consumption then in order to gain or produce them? Who now forgoes current consumption in paying wages and factor prices and thus gets returns only after the finished product is sold? Whose enterpreneurial alertness operated throughout?

2. However if given the values of some final products (with great latitude about which ones would serve) the ratio equations could be used to specify the measure M and that could be used to yield the values for the other final products, then the theory would have some content.

3. The Theory of Capitalist Development (New York: Monthly Review Press, 1956). See also R. L. Meek, Studies in the Labour Theory of Value (London: Lawrence & Wishart, 1958), pp. 168-173.

4. See Eugene von Böhm-Bawerk, Capital and Interest, vol. 1 (South Holland, Ill.: Libertarian Press, 1959, chap. 12; and his Karl Marx and the Close of His System (Clifton, N.J.: Augustus M. Kelley, 1949).

5. "The labor time socially necessary is that required to produce an article under the normal conditions of production, and with the average degree of skill and intensity of labor prevalent at the time in a given society." Karl Marx, Capital, vol. 1 (New York; Modern Library, n.d.), p. 46. Note that we also want to explain why normal conditions of production are as they are, and why a particular skill and intensity of labor is used on that particular product. For it is not the average degree of skill prevalent in a society that is relevant. Most persons may be more skilled at making the product yet might have something even more important to do, leaving only those of less than average skill at work on i. What is relevant would have to be the skill of those who actually work at making the product. One wants a theory also to explain what determines which persons of varying skills work at making a particular product. I mention these questions, of course, because they can be answered by an alternative theory.

6. Capital, Part 1, Chapter 1, Section 1, page 48.

7. Marx, Capital, Vol. 1, Chapter 2, pp. 97-98.

8. Marx, Capital, p. 120.. Why "stomach"?

9. Compare Ernest Mandel, Marxist Economic Theory, vol. 1 (New York: Monthly Review Press, 1969), p. 161. "It is precisely through competition that it is discovered whether the amount of labor embodied in a commodity constitutes a socially necessary amount or not… When the supply of a certain commodity exceeds the demand for it, that means that more human labor has been spent altogether on producing this commodity than was socially necessary at the given period… When, however, supply is less than demand, that means that less human labor has been expended on producing the commodity in question than was socially necessary."

10. Compare the discussion of this issue in Meek, Studies in the Labor Theory of Value, pp. 178-179.

11. Such risks could not be insured against for every project. There will be different estimates of these risks; and once having insured against them there will be less incentive to act fully to bring about the favorable alternative. So an insurer would have to watch over or monitor one's activities to avoid what is termed the "moral hazard." See Kenneth Arrow, Essays in the Theory of Risk-Bearing (Chicago: Markham, 1971). Alchian and Demsetz, American Economic Review (1972), pp. 777-795, discuss monitoring activities; they arrive at the subject through considering problems about estimating marginal product in joint activities through monitoring input, rather than through considerations about risk and insurance.

12. See the detailed discussion of his theory in Marc Blaug, Economic Theory in Retrospect (Homewood, Ill.: Irwin, 1962), pp. 207-271.

13. See Israel Kirzner, Competition and Entrepreneurship (Chicago: University of Chicago Press, 1973).

Go back