If these no-rent instruments "help to create " a product," it is a strange use of words which attributes such a joint product to the one factor, labor. Two things are certain: one is that Professor Clark's specific productivity is productivity in the sense already discussed, and the other is that when labor and capital coöperate to produce a result that result is their joint product. It is a verbal absurdity to ascribe it to either factor alone.

Now there is no such thing as empty handed or unaided labor.1 Some material instrument is always employed, and as long as this is true, we are left without any method of disentangling the specific product of labor. If it is "impossible" in primitive individual economy, it is a clinched and riveted impossibility in the complex productive processes of an exchange economy.

as in A

As a matter of fact there is no such thing as specific productivity. It is even theoretically inconceivable. The familiar concepts of physics find here no analogue. In physics a joint resultant may be analyzed, the familiar case of the parallelogram of forces. ship which travels south-east (nearly) for a distance of five miles may be actually carried by the wind three miles to the east, and by the tide four miles to the south. But if a man cuts down a tree it is vain to speculate as to what fraction of the work is done by the man and what fraction by the ax. Let two men plow a field with the aid of a team and a plow, the one driving and

"

1 Professor Clark is inconsistent on this point. At a number of places he tries to send laborers to work unaided" or empty handed on marginal or no-rent land. But he generally finds himself compelled to allow them a minimum of equipment, a hoe or a spade or some other simple tool. But in order to display to us the isolated product of labor he systematically belittles the importance of this equipment, and tries to make it appear negligible. Now small quantities are for some purposes mathematically negligible. But here the assumption is quite invalid. This will appear from the fact that on page 184 Professor Clark himself puts a mere hatchet into the hands of a man otherwise without tools, and then credits the hatchet with a productivity" of what is called a 500% rate of interest." Distribution of Wealth, pp. 84, 88, 89, 92, 95,

100, 160, 184.

the other holding the plow handles. It would puzzle the greatest mathematician in the land to say just how much of the plowing is done by the team, plow, and the men respectively. And if the plowing cannot be fractionally ascribed to the respective factors, then neither can the value of the plowing.

There is in fact an error in elementary mathematics in Clark's method of estimating specific productivity. If an extra laborer is added to a going concern, it is true that there may be a definite increase in the product as a result. It may even be possible to estimate this increase approximately. Moreover it may be a correct business judgment to justify the employment of the extra man on the ground that the extra product warrants the expense. But if the increase in product over what it would have been without him be regarded as his specific creation, and the same method be used to measure the specific creation of each of the coöperating factors, it is mathematically impossible for the sum of the productivities specifically attributed to the factors to be equal to the actual aggregate product. This whole is not equal to the sum of all its parts.

2

Hobson, in his criticism of the productivity theory has attempted to show this by a numerical example.

1 W. M. Acworth in his little book, Elements of Railway Economics, shows in a very interesting fashion the making of a railway rate on much the same principle. New traffic, not otherwise to be secured, may be profitably carried by fixing a rate which will a little more than cover the extra expense entailed in carrying it, even tho the new traffic does not help defray any of the general charges. But plainly this method of making a rate cannot be generalized. It can be used only when the new traffic is to be added to a "total situation which already provides for the carrying of the general charges. Now it would be incorrect to call the extra expense involved in carrying the new traffic its cost of transportation. But this would be quite analogous to ascribing the increase in product, resulting from the employment of an extra laborer, to the laborer as his specific creation. It pays to take on the new traffic, and it pays to employ the extra laborer. But the extra expense is not the cost of transportation of the new traffic; nor is the increase in the product the specific creation of the new laborer.

2 This is Professor Clark's view. And it is worth noting that J. Maurice Clark in his review of Davenport's Economics of Enterprise (Political Science Quarterly, June, 1914, p. 319) has given recent expression (in defending the productivity theory against Davenport's criticism) to the same belief.

In a primitive fishery let us say that one man fishing alone could make a catch of ten; a two-man group a catch of twenty-two; a three-man group of thirty-seven; a four-man group of sixty; a five-man group of seventy-two.1

Mr. Hobson's general criticism of the dosing method of isolating the specific product seems to me valid. But his numerical example will hardly be accepted as meeting Clark's case squarely. In the first place he cites a case of increasing returns. Clark seems always to have in mind decreasing returns. As a matter of fact the increase of product due to the employment of one more laborer, itself increases to a maximum, and then decreases (with successive increments of labor). In the second place he is dealing with technical or physical productivity, whereas Clark has " value productivity (if there be such a thing) in mind. Finally he cites a coöperative group of fishermen dividing the physical product. Clark is thinking of an employer who hires labor for wages. In the numerical illustration given, the four-man group is the most advantageous, for its product divided by the number of men gives the maximum quotient of fish per man. But if an employer is hiring the men at a wage of five fish per man, it will

pay " him to employ a fifth, or a sixth, or a seventh man. He will employ additional men until he comes to the one whose employment brings an increase in the joint product which barely exceeds the wage which has to be paid. This would make the illustration more like Professor Clark's. Now if the increments resulting from the employment of the fifth, sixth, and seventh men are 12, 7, and 5 respectively, and if we concede that the employer would hire the seventh man, tho he gains nothing by doing so, this seventh man would then be Clark's "marginal" laborer. Hobson

1 The Industrial System, 1910, p. 114.

would now say, "The total catch is 84, which gives a product of 12 fish per man." But this allows nothing for the services of the equipment, boat, tackle, etc. Hobson attributes the whole catch to the single factor, labor, which is quite as inadmissible as anything Clark does. The latter would call five fish the (" marginal ") productivity of the seventh laborer. Five would also be the "effective" productivity of each of the other laborers, making the total product of the laborers, according to Clark's reasoning, thirty-five fish. By subtraction the product which must be attributed to the equipment would be forty-nine fish. (Eighty-four less thirty-five.) Now how absurd to say that out of the total of eighty-four, thirty-five fish are produced by labor, and forty-nine by capital! The contributions of the two factors are as "hopelessly merged" and as impossible" of disentanglement as in the case of the single fisherman cited by Clark (see above, p. 155). How much of the catch is due to labor, and how much to equipment? "Not for his life" can Clark - or anybody else— tell.

The real mathematical error lies in not attributing to the coöperation of the rest of the group any part of the so-called "marginal product " of five fish. The five fish as well as the other seventy-nine are the " product of a coöperation between seven men and certain equipment. It cannot by any necromancy be fractionally ascribed. Hobson must be credited with having seen and expressed this with perfect clearness even tho the issue is not squarely joined in his numerical illustration.

But there is something" specific " in this illustration. What is it? Merely the wage,- the "reward" of "product." The men are paid what they are paid. It seems a very obvious proposition

labor, not its

that if "x" is the increase in product resulting from the addition of an increment of labor, "a," to what Davenport would call a "total situation," that the factors of this total situation play a role in coöperation1 with the increment of labor in the creation of "x." And if "x" is called the specific product of "a," we may borrow a phrase from Davenport and say that "the defect in the theory lies in the simple untruth of the assertion.' When the new situation is viewed as a whole "x" is not the fractional part of the total product logically attributable to "a." There is no such fraction and can be none.2

The writer is aware that this view will not find ready acceptance. There are in America, broadly speaking,

1 It may be that Davenport falls a trifle short of his usual skill in expressing his ideas when he says that productivity is "relative" not "specific." (The Economics of Enterprise, p. 153.) But Professor A. S. Johnson, in his very able review (Quarterly Journal of Economics, May, 1914, p. 441) should not have missed the point so completely as to say that Davenport's relative productivity" and Clark's " specific productivity" are the same. If language means anything (and Professor Clark's language is of a crystal clear lucidity), Clark means by "specific productivity" exactly what Davenport understands him to mean.

"No separate dose has any separate product." And also, "If the employment of a tenth shepherd means twenty more sheep per annum than the employment of nine, it cannot be maintained that twenty sheep form the separate product of the tenth shepherd, but only that a ten group is more productive by twenty sheep than a nine group." Hobson, The Industrial System, p. 115.

Davenport's formulation is as follows; "How then proceed to attribute to any one of the factors the increase of the proceeds due to the joint employment? So long as either glove is necessary to the worth of the pair, how tell how much either is worth? Which leg of a three-legged stool supports the stool? All that we can say is that if the stool is worth $3, one can afford to pay $3 not to be deprived of any one leg of it. So $2 may be offered to get back a lost glove out of a $2 pair. Thus it is easy enough for the entrepreneur to determine how much he can afford to pay for an item of productive goods or labor to go with his present equipment, but this is not at all to attribute to the extra item all the increase of gain which will accrue with the addition of the extra item. One buys, say a horse, to go with a wagon which otherwise would be useless. But this is not to attribute to the horse all the result from both horse and wagon. The horse would be equally useless without the wagon. In the last analysis, the entrepreneur himself could not isolate and determine a specific serviceability for gain relatively even to himself, but only that which he can afford to pay to get the thing or to refuse to keep the thing. And, as we have seen, no one of all these different sums that the entrepreneurs can respectively afford to pay or refuse has any special title to be regarded as the specific significance of the productive factor." The Economics of Enterprise, p. 147.