Opportunity Cost:

Marshall's Criticism of Jevons

Harry Gunnison Brown

[Reprinted from the American Economic Review,
Vol. 21, No. 3 (September, 1931), pp. 498-500]

The theory of "opportunity" cost has come into such general acceptance that it receives a degree of lip service even from economists who do not clearly understand it. But there are still a good many who argue that the rent of land does not form part of the cost of production in a sense analogous to that in which wages and interest do.

A few stress the fact that much land is so unequally adapted to different uses as to make any alternative to its best use of little or no significance. And it never seems to occur to them that some labor, also (as well as some capital), is so unequally adapted to different lines as to make any alternative occupation of little or no significance. When a piece of land has alternative uses, the choice among them is just as significant, so far as supply is concerned, as is the choice among different occupations of a worker who has alternative lines of work among which he may choose.

Another difficulty -- and this seems to have confused Marshall -- arises from the notion of a no-rent margin of land use, extensive or intensive. The whole price of goods produced on this margin is said to resolve itself into wages and interest. The price of any specific kind of goods competitively produced is said to be fixed at this margin, and rent is said to be a result of the price so fixed.

But those who so formulate the relationship appear to be innocent of any suspicion that their formulation can as justifiably be turned completely around. For it is equally possible to picture to ourselves a no-wages margin where the entire price is resolvable into interest on capital and rent of land. If this fact is made clear, surely the reader will be able to see that we can picture, also, a no-interest margin where the entire price is resolvable into wages and rent.

Perhaps someone will demur at this point, contending that while there may be marginal land which yields no rent, there is no marginal labor which yields no wages. Let such a person remember, however, that for over a hundred years economists have laid stress on the intensive margin of land utilization. The "final" unit or "dose" of labor and of capital to an intensively used, rent-yielding piece of land adds nothing to the rent but merely produces wages and interest. And the same principle can be applied to labor. To be sure, there may not be any laborers who will work for nothing. But there are intensive uses of labor which yield no wages. To use a given piece of land more intensively without adding to rent is merely to apply more labor and capital to this piece of land when the added product so resulting is all absorbed in wages and interest. Likewise, to use a given amount of labor more intensively without adding to wages is merely to apply more land and capital to this amount of labor when the added product so resulting is all absorbed in rent and interest. The labor is paid no more. It is better provided with land and capital and so its product is larger. But the increase of the product all goes to pay rent and interest. The labor is used more intensively in a sense analogous to that in which land is said to be used more intensively when more land and capital are applied to it. The laborers do not work any harder. In regard, therefore, to intensive use as here defined, there is a part of the product of industry that includes no wages (and similarly, of course, a part that includes no interest), just as truly as there is a part that includes no rent.

All this may be said to be commonplace. Yet apparently some economists who can see the basic principle involved, when they are explaining interest, wages and rent by the marginal productivity analysis, can be entirely oblivious of it when they are explaining commodity price. That is why they assume that price always includes wages and interest although admittedly the price of a part of the product contains no rent!

It is this error -- for surely no economist who really sees the entire picture will deny that it is an error-which invalidates Marshall's criticism of Jevons on rent and the expenses of production. Marshall, in an extended footnote on rent and price, concludes as follows:[1]

Jevons asks (Preface to Theory of Political Economy, p. liv):

"If land which has been yielding 2 pounds per acre rent, as pasture, be ploughed up and used for raising wheat, must not the 2 pounds per acre be debited against the expenses of production of wheat?"

The answer is in the negative. For there is no connection between this particular sum of 2 pounds and the expenses of production of that wheat which only just pays its way. What should be said is:

"When land capable of being used for producing one commodity is used for producing another, the price of the first is raised by the consequent limitation of its field of production. The price of the second will be the expenses of production (wages and profits) of that part of it which only just pays its way, that which is produced on the margin of cultivation. And if for the purposes of any particular argument we take together the whole expenses of the production on that land, and divide these among the whole of the commodity produced, then the rent which we ought to count in is not that which the land would pay if used for producing the first commodity, but that which it does pay when used for producing the second."

In the light of the analysis presented at the beginning of our discussion, Marshall could with equal cogency -- no more and no less -- have criticized some one who argued that wages receivable in an alternative line should be debited against any product, and in the same words, thus:

"The answer is in the negative. For there is no connection between this particular sum of 2 pounds (wages) and the expenses of production of that wheat which only just pays its way (contributing nothing to wages)."

Similarly with the next sentence. Why might not Marshall as well have said:

"What should be said is: 'When labor (or capital) capable of being used for producing one commodity is used for producing another, the price of the first is raised by the consequent limitation of its field of production"?

And as to the next sentence after this one, might not Marshall as well have said:

"The price of the second will be the expenses of production (rent and 'profits') of that part of it which only, just pays its way, that which is produced on the no-w ages margin of labor production"?

And so we might go on to the end.

There is no intention here to go into an analysis of opportunity cost or a defense of it. Competent economists have analyzed and expounded it and, properly interpreted, it needs no defense. The present purpose is merely to show that such an attack as Marshall levels against the opportunity-cost theory as applied to rent has neither less nor more validity against the opportunity-cost theory as applied to wages or interest. The opportunity-cost theory either is not or it is a formulation which helps us better to understand the operation of the market and the forces lying back of it. In either case it has the same significance for land rent as for the interest ("profits") of capital and the wages of labor.

FOOTNOTES AND REFERENCES

1. Principles of Economics, 8th ed., London (Macmillan), 1920, p. 437.