Capital, Interest, and Rent: Essays in the Theory of Distribution
Reprinted from American Economic Association, Papers and Proceedings of the Sixteenth Annual Meeting 5 (February 1904). The discussants of Fetter's paper included Thomas N. Carver, Jacob H. Hollander, Charles W. MacFarlane, Lindley M. Keasbey, W. G. Langworthy Taylor, Richard T. Ely, James Edward LeRossignol, Franklin H. Giddings, and Winthrop M. Daniels (see ibid., pp. 199-227). Fetter's reply to their criticisms is reprinted here.
1. Logical clearness and practical needs call for a reexamination and restatement of the economic concepts of rent and interest.
This proposition expresses the thought of many contemporary economic students. The thought is reflected in the recent remarkable revival of interest in this phase of economic theory. The truth of the proposition is, however, not recognized by all. Some look upon the Ricardian doctrine of rent as an eternal verity, and deem the agitators of new economic concepts to be the pernicious disturbers of theoretical calm. Some economists cling to the traditional views as some theologians cling to outgrown creeds, oppressed with the thought that if the ancient faith gives way nothing can take its place. With rock-ribbed conservatism argument is vain, but such an attitude has one considerable justification: the recent rent controversy has been almost entirely of a negative character. The period of destructive criticism has elapsed; but erroneous concepts will not be discarded until positive and practically applicable ones are put in their places.
2. The generally accepted definitions of rent and of interest are imperfect in that they mark only a small portion of the boundaries of the concepts actually employed.
Criticism of definitions should not be unreasonably exacting. It is sufficient that the definition state the essential characteristic of the concept, for it is impossible to include in a sentence all the logical and practical developments of the central thought. It is no vital fault that the statements that rent is income from natural agents, and that interest is the income from products used in production, do not tell everything about rent and interest. But the prevailing belief is that all of the essential contrasts of rent and interest so much dwelt upon for a century past, result from the one defined and simple difference as to the kind of goods yielding the income. In fact, however, the concepts of rent and interest are not developed along parallel lines, other most fundamental terms being unconsciously introduced into them. The prevailing concepts of rent and interest, therefore, have an exceedingly complex character, and what is worse for clear thinking, this complexity is concealed beneath a simple form.
The two propositions above state the negative portion of the thesis to be here maintained. Part I of this paper, given to negative criticism, is continued in propositions 3 a, b, 4, 5 a, b, c, and summarized in 6, these together forming a demonstration that the conventional rent concept contains several conflicting thoughts. Part II, consisting of propositions 7 a, b, and 8 is an examination of two possible but inexpedient ways of making the rent and interest concepts formally consistent, by developing propositions 3 and 4. Part III, the positive solution, points out in propositions 9 a, b, c, the logical and practical line of distinction to be found in propositions 5 a, b, c, when they are consistently developed, and concludes in 10 and 11 with the outline of a new theory of distribution.
3a. Since the beginning of modern economic theory, rent and interest have been defined by social marks; rent has been said to be the income of land owners, interest that of merchants, manufacturers and city men of wealth.
This distinction deserves mention, when the most recent and one of the keenest critics in this field expresses himself as follows: "It is a commonplace of historical economics that land was first given the rank of a factor in production coördinate with labor and capital for the simple reason that in England, the home of classical political economy, the landlords formed a social class distinct from the capitalists and laborers."*40 Adolph Held, probably the first to suggest this origin, states quite dogmatically, without discussion, that "the social classification appeared so sharply in England that Adam Smith accepted it without question, and accordingly distinguished the kinds of incomes without inquiring how far property in land and capital belong together."*41 However it originated, this thought of rent as a personal income of the members of a social class, persists to-day, as may be seen in many representative definitions.*42 The conscious distinguishing of the conceptions of economic and contract incomes is a recent phase of thought, as yet but slightly reflected in the formal definitions. Ownership, though frequently thus included in the definition, has not played an essential part in economic discussion because, as used, the definition became a mere truism. Goods and incomes were not classified according as they belonged to members of different social classes, but, on the contrary, social classes were distinguished according as they were receiving incomes from particular kinds of goods. The income of the landlord as a person was made up of the yield from such varied agents that to the personal mark (membership in the land-holding class), necessarily was added at once an impersonal mark (the kinds of agents yielding the income). A man was considered to be a landlord if his most important income came from land. As the thought of rent as landlord's income and as income from land never have been very sharply distinguished, we may designate this second phase of the thought as 3b.
3b. Rent, in the conventional treatment, was therefore said to be the income derived from natural agents, and interest that from produced, or artificial agents.
When this is made, as it was, the central thought of rent, that part of the income of landlords that is derived from improvements is excluded and is declared to be interest. A minor fallacy then appears in that rent is either landlord's income or income from land, as is most convenient to the immediate purpose of the writer. The principal thought in rent remains, however, that of income from the use of natural agents. The grave difficulties in the application of this thought will be later criticized (in 7b). Other ideas now to be noted were, however, from the first, associated with the original thought.
4. The characterization of rent as that income from material agents which does not enter into cost of production, and of interest as the income which does so enter, was a shifting of the central thought of the concept; what was, at first, thought to be a merely incidental peculiarity of land rent, became its essential feature, and then the center of a more general concept of rent.
If this idea did not originate with Malthus and Ricardo, it was emphasized strongly in their criticisms of Smith as the main peculiarity of land rent. The supposed peculiarity of the relation of land rent to price rested on fallacious reasoning, due to the unconscious introduction of new conditions into the concept.*43 The gradual displacement of the earlier conception of rent as income from land, by the no-cost-of-production concept, is one of the interesting chapters in the history of economic theory. First, the no-cost camel thrust only its nose into the tent, then it crowded out entirely the former occupant. To-day the no-cost concept is in large degree dominant, although the old definitions, the old arguments, and many inconsistent conclusions of the older treatment remain. Marshall's treatment of rent and quasi-rents shows the orthodox order of distributive theory dissolved into chaos by illogically conserving the older thought while developing a newer one. The quasi-rent doctrine, however, takes a long step in the right direction, for it recognizes the likeness of the yield of land and of other concrete goods.
What is most pertinent to the present purpose is that this thought of rent, as usually developed, is in its nature a compromise. The old idea and the new are entertained, together. The same old formal definition is retained; the newer distinction, brought in to modify and explain, only complicates and confuses the rent concept. Certainly none of the contemporary supporters of this view have as yet framed a definition that is more than temporizing. But even if a choice were made between these two essentially different concepts of rent (and of interest) ambiguity would not be banished, for in all the older discussion of rent and interest another distinction has been assumed whose significance usually has been quite unsuspected, but which in fact contains the key to the problem.
5a. An essentially different distinction between rent and interest is tacitly introduced into the discussion when the amount of the bearer, or source, of rent is expressed in physical terms as to quantity and quality, while the bearer, or source, of interest is expressed in the general value unit as a principal sum.
That this distinction is made a part of the conventional concepts will be recognized by all students of economic theory. Equally evident is it when once attention is called to the fact, that this is done without recognizing the changed point of view thus taken toward the two kinds of goods. The Columbus of economic theory who stood this egg on end is Professor John B. Clark. All the standard texts declare, in discussing interest, that capital consists of concrete goods, and is neither mere money nor mere abstraction, yet at the same time they speak of capital as of uniform quality and as yielding a uniform rate of income. This is said to contrast capital strikingly with land, which is measured by the acre, and differs from unit to unit. Professor Clark, in his brilliant criticism of this confused thought, has vividly pictured the varying grades of "capital goods" as he calls them, and has shown that artificial agents can be viewed in concrete form and expressed in physical terms in the same way as natural agents usually are. Most students, therefore, are ready to recognize the truth of a statement that would have been startling some years ago: the contrasts supposed to reside in the objective differences between natural agents and capital are but subjective differences due to the points of view taken by the thinker when he chooses to express the quantity of goods in different modes.
These differing modes of expressing the bearers of the two incomes involve corresponding differences in the conceptions of their maintenance and of their income. As these conceptions are but phenomenal forms of the thought expressed in 5 a, the statement of them will be numbered 5 b and 5 c.
5b. In estimating its net income, the bearer of rent is thought of as materially unimpaired by use, being preserved in identical form or in kind; the bearer of interest is thought of as maintained of undiminished value, expressed in terms of some conventional standard.
This is a contrast in point of view that is entirely unrelated with the contrast presented in the formal definition, and confusion results. The taking of different points of view is allowable; indeed, it is necessary if all aspects of any subject are to be considered. The inconsistency is in unconsciously shifting the point of view and believing that the differing natures of the objects were the cause of the differences observed. Two similar houses viewed, the one from the front and the other from the rear, appear to be very differently planned. The one blind man who got his idea of the elephant by touching the tusk is said to have argued long with the other who had caught hold of the animal's tail. Debates as hopeless as this, result from the shifting of the concepts here under discussion.
A side light on the theoretical analysis above may be given by a brief suggestion of the historical conditions in which the distinction took its rise. The rent contract, almost universally employed in the Middle Ages in transferring the temporary control of wealth, involves a legal fiction. Land, houses, cattle, whose use is delegated to the tenant, must, according to the terms of the contract usual in such cases, be returned in the same condition as when borrowed. The performance of this contract is literally and physically impossible; but by means of agreements as to repairs and replacements, the agents can be restored in equally good condition. Every rent contract for the use of agricultural land is in its terms a disproof of the idea that rent is paid alone for the original and indestructible qualities of the soil; yet the fiction of a perpetual rent-bearer deceived Ricardo and has continued to deceive. The interest contract came into use much later, as a money economy arose; hence, its employment was confined, until the last century, almost entirely to money loans and to the transfer of city wealth. This chance historical parallelism between land, rent, country and landlord on the one hand, and machines, interest, city and merchant on the other, explains many of the fallacies that beset economic thought in the first conscious attempts to analyze value.*44 The rent contract and the interest contract are not in any essential way connected with land and produced agents respectively, and the chance use of them for transferring certain kinds of goods has within the last century become less and less common. The contrasting form of contract in rent and interest (and a corresponding contrast in the mode of estimating the income bearer in economic rent and interest) was introduced into the older concepts alongside of the formally recognized characters, making the concepts complex and contradictory.
5c. Contract rent (corresponding with the thoughts in 5a and 5b) is treated by all writers as an absolute amount, not as a percentage of the income bearer; contract interest is treated as a percentage of a principal sum. A similar distinction is made in the case of economic rent and economic interest at certain moments.
The conception of economic income being more subtle than that of contractual income, is less easily grasped. Contractual income is personal, economic income is impersonal. While it was contractual rent that drew the attention of the earlier economic students, it is economic rent (using the term in a broader sense than mere land rent) that constitutes the real problem in economic theory.
Here also a word of economic history throws light on the origin and occasion of this distinction as applied to the contractual incomes. The theorists of 125 years ago found contract rent in extensive practical use. While mainly used in reference to the income of land, the word rent was taken in a much more general sense both in English and in the continental languages. Houses and machines were then rented as pianos and automobiles are now. At first the income from land was specifically distinguished as "land rent," but Ricardo's authority specialized the term "rent" in English economic theory, and, ever since, economists have struggled in vain to establish their word usage in the place of that sanctioned by many centuries. A part of every conventional discussion of rent is given to explaining that "in the economic sense" it means only the income from land considered apart from improvements.
The renting contract doubtless was the exclusive mode by which the temporary use of wealth was given and acquired in primitive communities. It certainly continued throughout the feudal period to be all but universal in the rural economies. The interest contract was an impossibility until the rise of a money economy. Money came into use first in the cities, and there also was felt most strongly the inconvenience of the renting contract. The ventures of the merchant at home and abroad required goods so various in quantity and quality, so difficult to measure exactly except in terms of value, that the borrowing of them was hardly possible except in the form first of general purchasing power, that is, under the interest contract. And it is so to-day. The differing practice was due to business convenience, not to an essential difference in the economic nature of the goods, and while in fact machines can be and are "rented," land and other natural agents are often temporarily acquired nowadays under the interest contract. As contractual incomes both rent and interest are found alternating in practice, and just because the contracts are so different in outer form, the incomes appear to have in many ways essentially different characters.
6. There are thus included in the generally accepted concepts of rent, without formal recognition, three essentially different and often conflicting thoughts:
(a) It is the income of a special social class, marked by the ownership of a special class of physical agents (the characteristics being somewhat shifting).
(b) It is any income having a special relation to price, namely, that "it does not enter into the cost of production."
(c) It is an income that is yielded by wealth measured physically and that is expressed as an absolute sum.
In each case rent is in contrast with interest which is (a) received by a different social class, and from a different class of agents, or (b) has a supposedly different relation to the value of products, or (c) is estimated as a percentage of a principal sum or value of wealth.*45
If the incomes from wealth are to be grouped logically and classified practically as rent and interest, the three foregoing tests must be applied to each income as it appears. It is assumed in the conventional treatment that these tests give consistent results. Unless, however, the three tests are logically related, it is incredible that the results of their application should coincide in more than a small number of cases. Indeed, every contradiction that is possible by combining these independent tests occurs at one time or another in the conventional treatment of rent. The entire collapse of the old rent doctrine has been prevented only by failure to apply the tests to all cases and in full measure. The thought is shifted as convenience suggests. Starting with the formal definition framed about the first thought, the treatment shifts to the second or third. Such a method cannot be defended as a legitimate employment of a continuity concept. Continuity does not justify the cross-logic of a three-fold or four-fold principle of classification. These is no continuity in the jump from natural agents to consumer's rent, or from landlord's income to the contract to restore in kind.
In concluding the merely negative part of this paper it should be reiterated that propositions 3, 4 and 5, summarized in 6, are to be interpreted collectively. In the foregoing argument it has not been maintained that any one of the three principal thoughts contained in these three propositions cannot be made formally logical if it is developed by itself consistently. It is, however, maintained here that when these several thoughts are employed together without a recognition of the resulting complexity, fallacious contrasts and conclusions result. Differences between rent and interest, that are assumed to arise out of the nature of the two classes of agents, are but the reflection of the changing subjective attitudes of the theorists.
7a. Formal consistency might be gained if the distinction between rent and interest were made to turn on the difference in the social classes that receive the incomes; but this is almost purposeless in economic theory.
A merely formal concept of rent might be framed about the thought of a social class. Rent might be defined as the income of wealthy men or of those moving in the best society. English conditions naturally suggested to the thinkers of a century ago the contrast of agricultural land holders and city men of wealth. But it is safe to say that no such social classification ever has been or ever could be presented that is either exact or significant enough to serve in the analysis of value. The economic theory of value is essentially an attempt to explain impersonally the origin and degree of importance of goods. The social class concept of rent thus involves a distinction not primarily economic, and one that is incapable of even a moderate degree of exactness in practical application. When, moreover, membership in a social class is tested by ownership of a particular kind of agent, the social aspect of the concept almost disappears. The connection of the thought of land rent and landlords as a class could continue only in the peculiar social conditions of England, and then it corresponded only in a broad, not in an exact way, with realities.
7b. Formal consistency is possible if the distinction between rent and interest be made to turn solely on the difference in the classes of physical agents that yield them; but this distinction is quite incapable of practical application.
The only classification of wealth that ever has been suggested for this purpose is that into natural and artificial, or unproduced and produced agents, or land and "capital." Such a classification may be required to meet two tests. It is expedient only if the two classes of agents can be practically distinguished by marks or evidences that can be taken account of in the practical world; it is logical only if it is consistently applied.
Land in an unimproved state is rare if not unknown in modern societies. As nearly every concrete thing is a bit of natural material adapted artificially to some degree to man's use, everything according to this conception should have in it elements of capital and interest, and elements of land and rent. No practicable method of deciding whether a thing is land or capital ever has been suggested, much less applied.*46 When one considers the nature of the case, it appears impossible even to conceive of such a test.
Therefore economic theory, unable to make the division between land and capital along a concrete and objective line, has been led to make it along an abstract line. Rent was said to be the income from land "considered as unimproved," or "considered apart from improvements"; while interest was that part of the income of land that was to be considered as due to improvements or to produced agents. Ricardo put it that rent is paid to the landlord for the use "of the original and indestructible qualities of the soil." Few writers that have accepted the Ricardian definition, have failed to apologize for the evident error in the phrase. Ricardo apparently meant, not that all qualities were indestructible, but that they might be spoken of as undestroyed, if annually repaired. Indeed it would be difficult to find a writer that does not, both in theoretical and practical problems, give up the impossible task of distinguishing all the value due to improvements on land. It is so much easier to wave the difficulty aside by "incorporating" or "merging" the improvements into the land. It has not been recognized that the original thought has thus been departed from, that the practical difficulty has been slurred over, and that a metaphysical division has been substituted for a concrete classification. The designating of an improved field as land or natural agents, and of an improved piece of iron as capital, becomes a purely arbitrary matter. The test is not found in immobility. Are the Suez Canal, the Hoosac Tunnel, the ploughed field, land or capital? A touch of human labor is at one time believed to convert the entire material into capital, a larger amount of labor at another time is declared merely to incorporate itself with the land and become indistinguishable from it.*47 The notion that it is a simple matter to distinguish between the yield of natural agents and that of improvements is fanciful and confusing, is responsible for many errors, including the cruder part of the single tax doctrine. The distinction doubtless more nearly approaches business realities in the case of city building sites than in that of agricultural land. It must, however, be maintained that the objective classification of land and capital as natural and artificial agents is a task that always must transcend human power of discrimination.
The vagueness of the line between natural agents and capital is increased by the fact that money and artificial agents measured as "capital" can be and are so often invested in land. Where land becomes a commonly marketed form of wealth, the classification of rent and interest according to the social class of owners becomes meaningless, and the classification accorded to kind of agents grows quite out of harmony with business usage. An attempt to meet the difficulty is seen in the more recent contrast between capital from the individual and capital from the social point of view, which is an abandonment of the distinction according to the class of agents in most of its possible applications. This complicates instead of solving the difficulty, which must be logically met.*48
8. Formal consistency may be gained if the distinction between rent and interest is made to turn on their supposed relation to cost of production.
It is always a scientific service to carry to its extreme possibilities any abstract distinction, for thus only can be made apparent its merits and defects. In the gradual enlargement of the no-cost-of-production notion of land rent (noted in proposition 4) until it becomes the essential thought in the rent concept, the view of Mr. John A. Hobson represents nearly this ultimate development.*49 Moved by the desire to find a basis in the theory of rent for a juster system of distribution and of taxation, he reexamines the problem and arrives at the conclusion that "the law of rent, in its extreme application, is valid for each factor." A fund is required as well to keep land and labor, as to keep capital in repair, above which sum, he thinks, the differential expenses of production "whether they be rent, interest, or wages, will not enter into the market price of the supply." While he thus narrows the conception of rent in some ways, he widens it greatly in others. He retains, though after modification, the notion of a no-cost-factor, and broadens it greatly. He stops just short of rejecting the whole distinction between land and capital as unproduced and produced agents. As a result of this and other recent criticisms, a doctrine of general rent, or of quasi-rent, is the dominant idea regarding rent to-day in many minds.*50 As a negative criticism Hobson's essay has the highest merits, demonstrating, as it does, how illusive are many of the supposed peculiarities of the various incomes in the older treatment of distribution. His idea of cost and "no-cost" factors is moreover closely in touch with realities, for cost in his discussion is a very concrete thing, representing the repair and replacement fund needed for each factor. Moreover, there is for the theory of social legislation much suggestiveness in the idea of the surplus feature in each income that is above "cost," and therefore amenable to taxation. For all this, Hobson's treatment does not yield a satisfactory solution of the problem of the rent concept, notably because rent is left quite unadjusted, and unrelated to, the interest concept. Though Hobson, in concluding, expresses the hope that he has laid the basis for a "sound theory of distribution," he recognizes the complexity of his concept and the difficulty of its application.*51
The distributive system presented by Dr. C. W. Macfarlane*52 is, however, a further step into abstraction. That writer, believing that any given factor may, at a given moment, have various relations to price, reaches the somewhat bewildering conclusion that land (which "includes all natural forces except labor") and entrepreneur's service, each may yield both rent and profit; capital may yield rent, profit and interest; and labor may yield rent, profit and gain. Whether and how far any income is thus to be named depends on whether it is "price-determined" or "price-determining," a transcendental inquiry as difficult to apply as the small boy's method of catching birds by salting their tails. As the conception that some incomes bear a peculiar relation to price grows out of fallacious reasoning, no logically sound classification of incomes can be based upon it.*53 But if it were sound, it still would be the extreme of abstraction, confined to the most subtle and probably useless economic speculation. Even if such a no-cost-of-production concept of rent could be made formally logical it still would lack expediency for a theory of distribution.
9a. Consistency must be gained by substituting for the older futile distinctions, that between the wealth aspect and the capital aspect of material goods.
Neither the physical classification of agents, nor the metaphysical classification of abstract types of income, affords an answer to the theoretical and practical problem of rent and interest; but in the consistent development of the third important thought contained in the old and confused rent concept, the desired solution is found.*54 Rent and interest, until recently, have been looked upon as corresponding respectively to two different factors of production. In recent criticism the idea of correspondence or parallelism between each factor and its income has been abandoned, but the two material factors (natural and artificial) are still retained. A better positive theory must clear up the confusion as to the differing nature of these factors. Present in the thought of the older economists, along with the distinction between natural and artificial agents, and coloring their conclusions, has been the distinction here suggested. Durable goods were sometimes thought of as yielding uses (the wealth aspect), but land was the only important class of agents that was regularly so viewed. Durable goods were sometimes thought of as saleable at their present worth (the capital aspect), but only produced agents, the materials and instruments of manufacture, were usually so viewed. Both classes of agents can be looked at consistently from either point of view, can be considered either as bearer of rent, or as discounted sum of rents, either as wealth or as capital. It is in the confusion of these contrasts that most of the old opposition between income from land and income from artificially produced agents was found. This fog is lifted when the sources of rent and of interest cease to be considered as physically distinct and objectively differing kinds of goods, and are seen to be simply the same body of income yielders, differently viewed, calculated and expressed for theoretical and practical purposes.
9b. Corresponding with the distinction between the wealth aspect and the capital aspect of material goods, are the differing thoughts as to the maintenance of the factors.
In the earlier industrial stages when exchange is rare and money but little known, it is inevitable that the uses, or rents, of durable agents should be primarily thought of. In estimating the uses, allowance must first be made for keeping the agents in physical repair. This calculation is necessary not only in making the rent contract, but in conducting the individual economy, if net income is to recur. As was shown above, the supposed durability of land and of its qualities for which rent is paid, is largely an illusion due to ignoring its constant repair. The preserving of the rent-bearer in identical form or in kind is essential to the concept of a perpetual rent.
As the money economy displaces the barter economy, and the thought moves from the valuable present rent to the present saleable value of the rent-bearer, the capital sum of value is thought of as kept intact before a net income from it is estimated. This is a primary condition of the contractual money loan, requiring the repayment of a principal sum apart from interest and this becomes the leading type of modern business calculations.
The blunder of the older economics in connecting land and rent with the one mode of calculation, and artificial agents with the other mode, has been noted above in proposition 5. Not only is it possible to view both aspects of use-bearers consistently, but clear theory and sound business practice require that this be done.
9c. As a necessary result of the distinction between the wealth and the capital aspects of agents, and of the thoughts as to the maintenance of the factors, rent must be expressed as an absolute amount, and interest as a percentage of a principal sum.
This is stated mainly for formal completeness, but it emphasizes the retention of a feature of the older treatment whose significance was unsuspected. In fact the expression of interest as a percentage marks interest as the form of income most connected with mobile and saleable agents, it makes of interest a "marginal" factor in price, a fact so much emphasized in the older treatment, it connects interest peculiarly with the element of time, as so many writers have felt it should be. Yet the percentual form of expressing interest is impossible when the income bearer is measured by physical norms, it is practically inevitable when the income bearer is expressed as a capital sum.
10. The rent and interest concepts, when looked upon as successive steps in the analysis of value, instead of as coordinate shares dividing between them the income from material agents, are made consistent internally, mutually, and with the foregoing conceptions of wealth and of capital.*55
It was suggested in proposition 5 that the treatment of land rent as an absolute amount, and of interest on produced goods as a percentage of their value, grew out of prevailing practice in the contracts for the use of wealth. Either mode of expressing income may be logical if consistently employed, and if divorced from the confusing prejudice that the difference is due to the different nature of the factors yielding the two incomes. This error recognized, economic theory must abandon the old distinction as to the differing factors. What is left in place of the old rent concept? All that was best in it, freed of error: rent is the usufruct attributable to any material agent. The uses of material agents considered apart from the using up of the agents, are in this view always and only rents. This is a logical thought, a useful one and one applicable to practical problems.
When to rent has thus been assigned all current incomes from material agents there is no place for the old concept of interest as the yield of produced agents. But rents accrue at different points of time and vary in value accordingly. Present uses and future uses differ. A more or less durable agent represents a series of rents. The capital value of a good is the sum of its prospective rents and uses, discounted at a rate that reflects the prevailing premium on the present. Capitalization, thus viewed, is logically a later stage of the problem of value than is rent; and interest first appears in connection with capitalization. As the market expression of the all-pervasive premium of present over future, interest may appear in connection with any gratifications, whether they be yielded by natural or by produced, by material or by human, by durable or by perishable agents. There is not a writer from Ricardo to the present time by whom this universal application of interest is not vaguely recognized; there probably is not one by whom its application is not more or less inconsistently restricted.*56
11. The propositions above imply the need of a radical restatement of the theory of distribution, and suggest its essential outlines.
The prevailing theory of distribution rests upon the idea of three (more often lately, four) objectively differing factors, to which correspond three (or four) different kinds of income. Some later, more subtle, attempts to restate the theory have left it far from realities and quite unusable. Another solution may be found by combining into a logical system the three typical modes in which goods appeal to wants. First, goods appeal directly, as want-gratifiers immediately available. Here is required a theory of wants and enjoyable goods, and the technical analysis of marginal utility. The mental process here examined is chronologically the first stage of evaluation in the history both of the individual and of the race. Secondly, goods appear as more or less durable, and may be made comparable by being considered, through repairs, to be lasting use-bearers, yielding in a given short period a group of uses. Here is the place for the theory of rents. This is chronologically the second stage of evaluation, when durable goods are thought of and expressed in terms of their usufructs. Thirdly, whenever two nonsynchronous gratifications, rents or series of rents, are exchanged, they must be discounted to their present worth to be made comparable. Here is required a theory of capitalization, that is of economic interest. This is historically as well as logically the latest stage of evaluation, characteristic of a developed money economy and of a "capitalistic" era. These three phases must be observed in every complete analysis of value. They are in some respects analogous to the three dimensions in geometry. The older economic theories were curiously crude caricatures of such an analysis. The cost-of-production theory of the exchange value of commodities, (assumed to be the whole theory of value) roughly corresponded only with the first. The old theory of land rent caught a fragmentary view of the second. The old theory of interest on a narrowly conceived class of "capital," was an ineffective attempt to express the third. The theory of value in the present conception proceeds from the simple to the complex, from the immediate to the distant gratification, from the goods directly in contact with the senses, to those whose utility is indirect and only in expectation. While the negative criticism of the past three decades has wrecked the old distributive theory, many admirable positive contributions, widely diverse in character, converge to the solution here presented.
All taking part in this discussion have shown their belief that economic theorizing is worth while, and that theories both good and bad are affected by, and in turn affect, practical life. In accordance with this view, the leading proposition of the opening paper that the conventional concepts of rent and interest are illogical and inconsistent, has a corollary that these concepts are unfitted to explain the problems of the business world, and that another conception must be adopted.
To the frank and friendly criticisms offered in this debate, I shall reply as brevity permits. Those taking part in the discussion may be arranged in a continuity classification (the validity of which I fully admit) from those who for regard of traditional theories would overlook a lack of logic, to those who for regard of logic are willing to adopt new theories. The conservatives are far from harmonious in their beliefs, and by mutual cancellation they have left for consideration only a residuum of argument.*57
The prime contention of the first part of my opening paper is not, as it was assumed by Professor Hollander to be, "the historical relativity of the traditional theories."*58 That thought is a minor one, and the brief historical paragraphs were given merely as "side lights" on the origin of the errors.*59 It would be an easy task to defend and strengthen these historical references had any one of the speakers sought to controvert them at any specific point. Even the critic who first waived the whole opening paper aside as "conjectural history" gave to the historical suggestions "conditional assent."*60
Prejudgment has, I fear, caused more than one of my critics to shut his eyes to the repeatedly avowed purpose of the paper, which was to show that the traditional concepts are internally inconsistent, illogical, containing several conflicting thoughts, and that they were thus defective even in the days of Ricardo. In recognizing that some practical issues in Ricardo's time served to obscure this lack of logic, the paper had, to be sure, a suggestion of historical relativity. It is admitted by all the speakers that of recent years the emphasis on the various thoughts of these concepts has been shifted; and some would believe that this shift has cured the infirmities in logic. On the contrary I maintain that it has aggravated them. Thus, changes in industry and changes of thought have combined to enhance the difficulties inherent from the first in the older concepts.
Professor Carver has dissented generally from the negative part of the opening paper, regretting the attempt "to show that there is no basis for the scientific distinction."*61 He would explain the confusion by declaring that there are two clearly distinguishable concepts, the popular and the scientific, which at times contradict and overlap each other. As none of my critics attempted a specific disproof of this portion of my opening argument I may limit myself here to a reassertion that the socalled "scientific concept" is inconsistent in itself, that no writer has employed it without shifting thought and untenable conclusions. It is for the reader to determine whether I have not shown that the so-called "practical concept" has been confused with the so-called "scientific concept" in economists' minds. If this is true it follows that some of the supposed contrasts between rent and interest are but the reflection of the unconscious shifting in the subjective attitude of the thinker.
A test is thus afforded for any revision of the concepts; no valid contrast can be drawn between the concepts of rent and interest where there is an unconscious change from one to another of the three conceptions that have been noted. A shifting eclecticism becomes impossible when these different thoughts are clearly recognized. My critics, however, avoid a clear-cut decision, and uphold conceptions uniting two or more discordant elements. It is not easy, therefore, to say on just what ground they take their stand. They defend in the main the attempt to distinguish between land and artificial agents objectively, but their reasons are largely drawn from supposed differences in the relation of the income to price, and yet according to their own statements this distinction is not coextensive with that of the two objective classes of agents. Moreover, their arguments involve a use of the third distinction,*62 which they are endeavoring to overthrow.
This confusion may be seen in Professor Hollander's contention that the critics of the traditional distinction overlook "the composite character of the law of diminishing returns." He says that the characteristic that suffices to "differentiate land from capital as a productive good" is its diminishing efficiency in extensive cultivation; "while capital is available.....in identical homogeneous quality with respect to extensive use." Observe the reasoning by which this conclusion is reached.*63 The assumption, however, that any particular enterpriser, in enlarging his business, is forced to take up poorer land, surely is not warranted. Except in the rare case that the particular enterpriser had been using the one best piece of land, he can hire more land as good as he has, or even better, if he cares to pay the prevailing rental, just as he can hire more and better machines. The thought evidently shifts to the old dynamic and social conception of the growing scarcity of land with increasing population, and from the particular entrepreneur to the personified total population.
There is another shift, for while the physical conception of land is retained, and it is thought of in terms of acres, the particular produced goods called capital, are thought of in terms of a value unit. This creates the illusion that the differential return is peculiar to land, and that the value units of capital are of homogeneous quality. The varying yields of land are looked at in a way that makes them necessarily appear as differentials, and the varying yields of other agents are by reason of the mode of their capital expression, converted from differential incomes into homogeneous capitalized sums. What is this capital but the incomes (or I should call them rents) of productive goods, capitalized at the prevailing rate of interest? A given rent thus corresponds to one unit of capital, a double rent to two units of homogeneous capital, and a free good, or rentless unit, to no capital at all. This capitalization of rents is possible in the case of land also, the price of land being the sum of the anticipated future rents, discounted to their present worth; and the enterpriser can purchase x dollars' worth of land as easily as x dollars' worth of machines, and the units are just as homogeneous in one case as in the other. In fact, both kinds of agents frequently are bought as value units. The word "amount" in the contrast between an amount of land and an amount of capital begs the whole question, for in one case it means units measured by area and differing in yield, in the other it means the homogeneous value expression of differing units. It is impossible to escape these errors if the analysis insisted upon in the opening paper is overlooked.
Professor Carver has maintained*64 that there are abundant reasons for distinguishing between the income from land and the income from produced goods, in that interest as a personal income is a necessity to insure waiting, and thus is a condition of efficient production. This is retaining the traditional conception of the distinction between the objective classes of goods, while repudiating the traditional reasoning, and while broadening the conception of rent to any surplus or unearned income. The idea of surplus is generally very vague, but under the application of any suggested surplus-test the concept of rent would extend to numberless incomes and fractions of incomes not derived from land, and would fail to include numberless incomes and fractions of incomes that are derived from land in any usable sense of that term. Replies that, to my mind, are conclusive on the principle here involved were given in the course of the discussion.*65 It follows from this surplus conception that any portion of the income derived from produced goods that would have been saved if the rate of interest had been lower, is rent, not interest; and that any natural element of fertility in land that would have been used up except for the factor of waiting, would thereafter yield interest, not rent. Adopting for the moment the terminology of the critic, his challenge may be accepted; the proposition that "men must receive interest as a personal income to induce them [i.e., the marginal abstainers] to wait" and that "interest as a personal income is necessary to secure efficient production," not only can be but must be paralleled by like propositions concerning rent. Men must receive land rent as a personal income to induce them to bring the marginal land into cultivation and to maintain undiminished the supply of productive qualities. Thus land rent is necessary to secure efficient production continuously from land. The margin in question is not a hair line, it is in practice a zone of wide extent. This fact is the basis of private property in land as broadly and surely as the other fact is the justification of interest. We are not concerned here with the ethical question, but in each of the two cases a social policy is based on the need of maintaining the marginal units of supply, a policy which always appears unjustified when attention is directed only to the surplus cases.*66 It is in conflict with all experience to assume that the actual supply of land would be kept up to its efficiency if rent did not go to some personal agent who made himself responsible for the repairs, the restoration of fertility, and the waiting for the future involved in refraining from "Raubbau," the immediate exploitation of the land. (In some cases, it is true, this agent may be a group of men acting collectively through government, as in the case of any form of public ownership.)
As Marshall says: "The greater part of the soil in old countries....has in it a large element of capital. Man can turn a barren into a very fertile soil."*67 To deny first that the supply of land either as extension or fertility has any marginal relation to sacrifice, or is within man's control, and then when this is shown to be an error, to assert that such land is not land, but capital, and that the income from it is not rent, but interest—this is the approved mode of showing the exceptional character of rent. Are the terms land and rent thus to be refined away from any relation to the real things about which the economist begins to reason, and of which the practical world thinks whenever those terms are used?*68
Professor Taylor admits that my thesis is valid when confined to static conditions, but he adheres to "the relation-to-cost" concept in discussing dynamic conditions.*69 In his very suggestive remarks he has not revealed his thought fully enough to make clear the ground of his reasoning, but it would appear to be essentially the one just examined. While Taylor and Carver differ in some points, they agree in others, both alike rejecting the static reasoning on which Hollander bases his conclusion.
Dr. MacFarlane also holds a relation-to-cost concept of rent, but most of his discussion is given to a negative criticism of my position. His own views, though known to many readers, were not developed in this symposium. Some points will be noted below.
The attacks on my positive proposals refer in part to their supposed implications and consequences, in part to the advisability of the terms suggested.
1. Professor Taylor objects*70 that the first of the three stages in the analysis of value is not fundamental and precedent to the others, but is co-extensive with them. This criticism probably proceeds from a misunderstanding of the briefly expressed proposition. Not all goods, but only immediately enjoyable goods were said to present the first problem in the analysis of value. The second problem, that of the value of usufructs, and the third, that of the value of future uses, are, as my critic suggests, but developed phases of the general problem of value.
2. Professor Taylor believes that in criticizing Marshall's attempt to trace a continuity between rent and interest, I have denied the validity of reasoning by continuities.*71 It is not to a true continuity concept that I object, but to a pseudo-continuity concept. As the thought passes along the series from rent as an income yielded by one kind of concrete goods, to interest as the income yielded by another kind, there is unconsciously introduced a new contrast. The value expression of capital and the percentage expression of interest are equally applicable to the rent end of the series, and it is an error to assume that they are applicable only at the other end. My suggestion is to apply consistently each distinction in turn.
3. Dr. MacFarlane declares*72 that the outcome of my proposal is the obliteration of all distinctions between rent, interest, profit and wages. This conclusion, drawn from my statement that "interest may appear in connection with any gratification," is due to the failure to apprehend how and how far the proposed conception differs from the one apparently taken in this discussion*73 by the critic himself, that each kind of income corresponds to a particular kind of income bearer. The proposal is to look upon interest in all cases (as it is now in many cases) as being that particular phase of value connected with differences in the time of accrual of incomes. Recognizing that a day's work to-day is worth more than one next year, does not identify interest and wages. Wages payable at different points of time vary in value as do rents at different points of time, and the comparison of each series is expressed by the interest rate.
4. Dr. MacFarlane objects*74 further that the proposed view of capital identifies the capitalized value of monopoly surplus with capital in general. True it does; there is no other logical way.*75 It is not quite clear what monopoly means as the critic uses the term, but any source of income that is continuing and foreseeable can be capitalized and sold, and thus becomes homogeneous with the value of the continued control of other sources of income. When from any cause income ceases, the capitalization collapses, monopoly or no monopoly. The puzzle as to whether the $5,000 or the $12,000 are to be called interest, is merely a confusion of the problems of economic income and contract interest.
5. Dr. MacFarlane says*76 that I have tried to identify land and capital by a mere arithmetic device that does not touch the substantial differences. I would reply that because an arithmetic device has been inconsistently applied in the traditional theory, illusive contrasts not existing objectively, have been created. I dissent from Professor Carver's opinion that it is merely a question of terminology in dispute,*77 and I agree with Dr. MacFarlane that there is involved more than a question of definition.*78 The arithmetic device is significant at least in a negative criticism of the supposed contrast between rent as a differential and interest as a homogeneous income; it serves to show the fallacy in the old view as to the special relation of rent to entrepreneur's cost of production; and it sets in a clear light the error in the traditional contrast between the value expression of "capital" and the concrete expression of land. This proof of the substantial unity and continuity of the body of income yielding wealth has been suggestively styled by Professor Taylor*79 in a phrase drawn from chemistry, "allotropism." One group of elements has been mistaken under differing conditions for two elements, (the condition in this case being the subjective attitude of the thinker). Take away the fallacious contrast, apply the arithmetic device consistently, and the objective classes of "natural agents" and "capital goods" are seen to be merged into one body of wealth, presenting three value aspects: gratifications, usufructs, expectations. But identifying the substance does not identify the allotropic states; coal is not diamond, though both are allotropic states of carbon; and no more is rent the same as interest. Like most analogies, however, this one is not perfect, and may become misleading. But this has brought us to another question deserving special answer.
6. It is taken for granted that my proposition is to treat rent and interest as identical. Several of the speakers have assumed that the idea of the paper was that of John B. Clark, and thereupon they have criticized his views, not mine. My indebtedness (shared in common with all contemporary students) to the inspiration of this ablest of theoretical economists, should not impose on him any responsibility for the theory of distribution here presented. The prepossessions of some of the speakers make it difficult for them to see the full import of a denial of the parallelism between the two incomes, rent and interest on the one hand, and the two objective classes of goods, land and capital on the other. They therefore attribute to me conclusions deduced from premises of their own supplying. This is seen in the assumption that a denial of the conventional contrasts between valuable natural agents and (conventional) capital is a denial of the difference between rent and interest. It is consistent with my views to speak, as Professor Daniels does, of the identification (or merging), of the classes of wealth composing "land and capital" (in the conventional sense); but this is not an identification, as others consider it to be, of rent and interest. Having made this point as clear as I could in the limited space allotted, I can merely re-assert that this lack of parallelism is of the very essence of the contention in the opening paper.
7. Finally, it is said*80 that if the old concepts are to be rejected, it is better to devise new terms than to adapt old ones having misleading associations. To this view must be conceded a large measure of validity. Regarding the term rent there is less difficulty, as the broad meaning here suggested not only has strong historical support, but in many languages, including our own, is grounded so deeply in popular usage that no economic authority has been able to uproot it. There is needed only an elimination of inconsistent thoughts from the concept and the retention of one of the ideas that always has been present in it. Regarding interest the decision is more difficult. Only yesterday economists talked of "the theory of profits" when they meant what is now called "the theory of interest." The term interest, until recently, was used almost if not quite exclusively, as meaning the income from a money loan. This is a contractual, not an economic income, and as such is not a genus coördinate with economic rent, rather it is species of the genus contractual rent. Is it not significant that even in the classical treatment interest as an accruing or realized income expressed as a percentage never appears except as the result of a contract?
The essence of the so-called problem of interest, according to the view in the opening paper is not fundamentally contractual interest, but capitalization. The problem logically following that of rent is not that of analyzing a coordinate income, for rent absorbs all the incomes accruing from material agents at any moment of time; but it is that of the value-calculation on future incomes. The title of the opening paper might perhaps better be: "The relations between rent and capitalization." That, however, would have misled the reader approaching it with the older conceptions in mind. Either "the theory of discount" or "the theory of capitalization" would be a more appropriate term for this part of the problem than is the theory of interest, and possibly some still better term can be found. The final use of terms is a matter of social convention; but when the real nature of the problem is understood, and then the fitting terms are suggested, they will not long fail of acceptance, as the example of the rapid change in the usage of the word profits gives reason to hope.
Whatever other impression may be left by this discussion I trust it will be that I have contended for a merely verbal change. On the contrary I have outlined, whatever be its defects, a radically new conception of the whole theoretical analysis of distribution. Doubtless this session has been most profitably spent in considering the more negative phases of the subject; but the scant attention that has been given to the yet more important positive outcome of the study may leave an impression of negation and verbal criticism that is misleading.
I welcome the able, forcible and somewhat unexpected support that has been given to my thesis in this discussion by the advocates of a realistic theory.*81 Opinion on this subject is unquestionably in process of change. Even the more conservative speakers in this session have made concessions that would have been startling a few years ago. The immediate result of such a friendly interchange of views as this has been, may be to strengthen each in his own opinion; but in the end the result must be to help us all towards the right solution of these difficult and important problems in the realm of abstract economic theory.
NOTES [addendum by Murray N. Rothbard—Econlib Ed.]
The argument in this paper, forced into excessive brevity at many points, should be interpreted in connection with other essays by the writer, published from time to time in the past three years. Arranged, as nearly as their special nature permits, in a logical series, they are as follows:
1. The next decade of economic theory, Publ. of Amer. Econ. Asso., 3d ser., vol. 2, no. I, p. 236-246 (read Dec. 29, 1900). Points out the relative and temporary nature of the old concepts of rent and of capital, and suggests the general direction that may be taken in their restatement [see above, pp. 74-83].
2. The passing of the old rent concept, Quar. Jour. Econ., vol. 15, pp. 416-455 (May, 1901). A detailed criticism, purely negative, of Marshall's doctrine of quasi-rent, as typical of the prevailing unsettled condition of thought on this subject [see pp. 318-354].
3. Recent discussion of the capital concept, Quar. Jour. Econ., vol. 15, pp. 1-45 (Nov., 1900). A review of the contributions of Clark, Irving Fisher and Böhm-Bawerk to this subject, criticizing especially the last named in his distinction between social and individual capital, between consumption and production goods, between natural and produced agents; concluding with a positive statement of a concept of capital, as distinguished from wealth [see above, pp. 33-73].
4. The "roundabout process" in the interest theory, Quar. Jour. Econ., vol. 17, pp. 163-180 (Nov. 1902). A criticism of Böhm-Bawerk's "Positive theory," showing that his retention of a defective capital concept is the cause of his retaining (inconsistently) a productivity theory of interest; concluding with a suggestion of the true relation of productivity to a theory of interest. The present paper unites, and develops somewhat, the various arguments in this series of articles [see above, pp. 172-191].
Notes for this chapter
A. S. Johnson, Rent in modern economic theory, p. 19, Publ. Amer. Econ. Asso., 3d ser., vol. 3, no. 4. Probably most students would not consider this explanation a commonplace and would even deny that it truly states the principal cause of the distinction in question. The author quoted makes it the main thesis of his book that the difference between land rent and interest, though thus originally observed as a merely transitory historical fact, remains of permanent significance.
Zwei Bucher zur socialan Geschichte England's, p. 160.
Note for example, Ricardo, 1817: "that portion of the produce of the earth, which is paid to the landlord," etc.; F. A. Walker, 1887: "the remuneration received by the landowning class," etc.; Marshall, 1890: "the income derived from the ownership of land," etc.; Bullock, 1897: "the return that is secured by the owner," etc.
This question is dealt with more fully in "The passing of the old rent concept," especially pp. 333-350.
This thought was stated with a somewhat different emphasis in "The next decade of economic theory," pp. 80-81.
Still other distinctions find partial recognition in current economics. See "The passing of the old rent concept," 325-332, for a discussion of space extension and of time in this connection.
The ablest attempt to face this difficulty formally, that of Böhm-Bawerk, in his "Positive theory," pp. 55-56, is quite unsuccessful. A criticism of his argument is given in "Recent discussion of the capital concept," pp. 57-65.
This idea as held by Böhm-Bawerk is more fully criticized in "Recent discussion of the capital concept," p. 63.
The "land concept of rent" in the somewhat complex form as held by Marshall, is criticized in "The passing of the old rent concept," pp. 320-325.
"The law of the three rents," article in Quarterly Journal of Economics, vol. 5, p. 263; restated in his "Economics of distribution," 1900. Likewise in vol. 5, p. 289, appeared John B. Clark's remarkable paper on "Distribution as determined by a law of rent."
The change in the rent concept is reviewed in "The next decade of economic theory," pp. 78-79.
"The law of the three rents," pp. 287-8.
Value and distribution, 1899.
The mistaken origin of the no-cost concept is shown in "The passing of the old rent concept," especially pp. 345-350.
This solution was implied in the capital concept presented in "Recent discussion of the capital concept," pp. 65-70.
This conception was briefly suggested in concluding the criticism of Böhm-Bawerk: "The 'roundabout process' in the interest theory," pp. 185-188.
The broader conception of interest was presented in "Recent discussion of the capital concept," pp. 33-73, especially pp. 49-57.
Brevity compels me to confide these closing comments to the criticisms adverse to the opening paper.
Hollander, "Discussion, A.E.A." Proceedings 5 (February 1904): 204.
See above, pp. 197-198.
See Hollander, pp. 204, 205.
See Carver "Discussion, A.E.A." Proceedings 5 (February 1904): 205.
See above, p. 205.
See Hollander, passage beginning "No entrepreneur" and ending "only in inferior efficiency," p. 208.
See Carver, pp. 200-201.
See Daniels, p. 226. Dr. Whitaker's remarks to the same effect unfortunately were not obtainable for this report.
This applies also in answer to the remarks of Professor Ely.
Marshall, Principles of economics, 4th edition, p. 224. He does not draw the conclusion, however, that is here suggested as necessary.
The interesting facts cited by Professor LeRossignol, p. 224, seem to me to illustrate, not to disprove, the view I have taken, which is far from a denial of the "surplus return" to the investor in land, or in other wealth, in a new country.
See Taylor, "Discussion, A.E.A." Proceedings 5 (February 1904): 221.
See Taylor, p. 218
See Taylor, p. 220.
MacFarlane, "Discussion, A.E.A." Proceedings 5 (February 1904): 215.
As is well known to students of economic theory Dr. MacFarlane has in his work "Value and distribution," obliterated the distinctions between the objective classes of agents yielding rents, and other incomes, more fully than has any other writer.
See MacFarlane, pp. 213-14.
See Professor Gidding's reply, "Discussion, A.E.A." Proceedings 5 (February 1904): 226.
See MacFarlane, p. 212.
See Carver, pp. 203-4.
See MacFarlane, p. 214.
See Taylor, p. 219.
Hollander, p. 209.
Daniels, Giddings, Marburg, Whitaker. Unfortunately no report was secured of Mr. Marburg's brief and pointed remarks or of Dr. Whitaker's subtle discussion. Professor Keasbey's attitude toward the question is favorable to the opening paper as against its critics, but his point of view is original, and his treatment in several ways not consistent with the views I have expressed.
Part 2, Essay 4. Interest Theories, Old and New
End of Notes
Return to top