The Pure Time-Preference Theory of Interest: An Attempt at Clarification

The Pure Time-Preference Theory of Interest: An Attempt at Clarification^*

{By Israel M. Kirzner}

For almost a century a particular theory of interest has been again and again discussed, refuted, defended, ignored, forgotten, and rediscovered; somehow it has managed to survive. This theory is the pure time-preference theory (often to be referred to in this paper as PTPT). For the most part this theory has, especially during the last half-century, languished as a basically discredited, definitely unfashionable, point of view. Yet the theory was never finally interred—nor did it even wholly expire. The theory is often described as Austrian, but, as we shall see, it is not the only and not the best known Austrian theory of interest. In recent decades a certain revival of discussion has emerged surrounding this pure time-preference theory (partly, no doubt, as a result of the modest revival of interest in Austrian economics generally). Almost invariably, contemporary economists have reacted to renewed discussions of the pure time-preference theory with utter disbelief and plain bewilderment. These critics have found the theory simply incredible; the idea that the phenomenon of interest is in no way dependent upon physical productivity is one that the critics find patently absurd; that serious thinkers should accept this absurdity, they find quite incomprehensible. The present paper does not seek to argue any superiority of PTPT over its competitors in the field of interest theories. Rather we seek to dispel the bewilderment that moderns display in regard to it. This task of clarification will turn out to involve certain “philosophical,” extra-economic issues that are of significance for economists in their own right, in several respects.

The Interest Problem

Much—perhaps all—will turn out to depend on the way in which the interest problem is formulated. For present purposes we adopt a modern formulation of the problem, but wish to emphasize that this formulation is very similar in spirit and character to classic formulations going back to Schumpeter¹ and Böhm-Bawerk.² The modern formulation we cite is that of Hausman.³ Hausman points out than an “individual's capital…enables that individual to earn interest. If the capital is invested in a machine, the sum of the rentals the machine earns over its lifetime is greater than the machine's cost. Why?” Common observation, that is, tells us that possession of a given stock of capital funds can, by judicious investment (say, in a machine), yield a continuous flow of income (annual rentals net of depreciation) without impairing the ability of the capital funds to serve indefinitely as a source of income. The problem is, how this can occur. Why is not the price of the machine (paid by the capitalist at the time he invests in the machine) bid up (by the competition of others eagerly seeking to capture the net surplus of rentals over cost) to the point where no such surplus remains? We are seeking, then, an explanation for an observed phenomenon which is, in the absence of a theory of interest, unable to be accounted for. Absent a theory of interest, no interest income ought to be forthcoming, except as a transient phenomenon; competition ought to squeeze it out of existence.

Neoclassical theory has, in a variety of versions, seen interest as obtained and paid in return for special productive services of one kind or another. J.B. Clark and F.H. Knight saw capital as providing a flow of productive services, of which interest is the irrepressible expression. Competition does not erode it; ownership of a stock of capital inevitably confers title to a corresponding income flow. (We do not discuss, in this paper, the extent to which this theory in fact addresses the interest problem as formulated above.) For “Austrian” versions⁴ of the neoclassical explanation, again, ownership of capital expresses the provision of a special productive service (“waiting”) required in order to enjoy the enhanced fruits of more lengthy (“roundabout”) processes of production. Competition cannot erode interest income: it has to be offered if potential capitalists (with positive time preference) are to be persuaded to provide the waiting (needed in order to be able to enjoy the enhanced output available through capital-using production). And, given the productivity of waiting, it pays to offer interest in order to elicit that waiting. In the celebrated “Cambridge Controversy” of some years back, this neoclassical view of interest was the only view advanced as an alternative to the neo-Ricardian perspective on interest as a surplus (to be explained, not by appeal to market exchange relations, but by “the relations between workers and capitalists including possibly their relative bargaining power”).⁵

It will be useful for us to examine more carefully the abovecited “Austrian” variant of the productivity view (actually most carefully developed by Irving Fisher) in order to point up features of the alternative Austrian theory of interest, the pure time-preference theory. The pure time-preference theory (PTPT) was developed, largely from roots in Böhm-Bawerk, by Frank Fetter⁶ in the United States, and later by Mises.⁷ As Fetter pointed out, Böhm-Bawerk's own position appears in a number of ways to be an inconsistent one, demonstrating at some points the pure time-preference view, at other points the Fisherian physical productivity view. Let us note certain features of the Fisherian “productivity of waiting” theory.

We should note, first of all, that this theory is not necessarily vulnerable to the basic Böhm-Bawerkian criticism of all productivity-of-capital theories of interest. Böhm-Bawerk had criticized such theories of interest because they ignore the essential interest problem (as formulated above). It will not do to say that the machine yields interest (in the form of a flow of rentals that is greater than the cost of the machine) because the machine is physically productive. The interest problem consisted in asking why, given this physical productivity, did not the market value of the machine rise to reflect fully the rentals it is able to generate. Of course a tree produces fruit; the interest problem consists in the dilemma posed by the apparent failure of the market price of the tree to equal the value of the total fruit output. Simple theory tells us, after all, that the value of inputs and the value of outputs tend to equality in competitive markets.

The productivity-of-waiting theory of interest escapes this criticism by arguing that besides the machine, besides the tree, yet another “input” is required in order to command the flow of rentals (or fruit). This additional input is “waiting”; once the services of waiting are properly included in the list of needed inputs, the interest dilemma evaporates. Competition indeed squeezes out all surplus above marginal productivity returns. Interest is, in this view, the marginal productivity return on a scarce factor, viz., waiting.

The Fisherian “productivity-of-waiting” theory emphatically recognizes the significance of time preference. It is the circumstance of positive time preference that renders waiting a scarce, costly factor (to which interest can be attributed as a productivity return). But the interest income that might be considered the reward received by investors to induce them to provide waiting is at the same time seen as made possible only by the productivity of waiting. (It is in this respect, particularly, that PTPT differs from the Fisher theory. For PTPT it is incorrect to see interest income as the “fruit” of anything. Rather PTPT sees interest income as a receipt that results from the pattern of prices governing intertemporal exchanges, with these prices expressing the prevalently positive time preferences of the participants.)

As a matter of logic, the Fisher productivity-of-waiting theory deals with the interest problem (as formulated above) in impeccable manner. The only way through which the validity of the productivity-of-waiting view (at least insofar as we have presented it thus far) can be challenged is by disagreeing with the concept of “waiting” as a productive factor service. The critic may refuse to recognize that waiting (or “time”) is productive, or is regarded by prospective producers as a scarce factor. The fact that an inevitable time delay must be accepted before current efforts bear fruit, need not mean that time is a necessary ingredient in the production process; it may simply mean that this production process is a slow one, yielding a result less valuable (in terms of anticipated attractiveness) than an otherwise similar, more speedy process.

Clearly this question of whether to treat waiting as a productive ingredient must be recognized as a strictly “philosophical” question. No economic reasoning (and certainly no amount of empirical research) can reveal whether time is (or should be treated as) itself an active productive agent or only a medium permitting the flows of inputs to cumulate into the product.⁸ Moreover it is conceivable that a given observer may be prepared to recognize time as an active productive agent in some processes of production (the maturing of wine, perhaps?) while refusing to recognize it as such in other time-consuming processes of production (as, perhaps, in considering the operation of an old-fashioned, slower machine as compared with that of a faster, more modern generation of machines). In any event the relevance of the “productivity-of-waiting” theory depends entirely on this “philosophical” question. (Part of the dilemma posed by Böhm-Bawerk's own statement of the theory of interest was that, on the one hand, he explicitly refused to accept time or waiting as an independent factor to stand with labor and nature,⁹ yet seemed on the other hand to attribute interest to the physical circumstance that roundabout processes of production are more productive. If time and waiting are not themselves to be considered productive agents, no interest could emerge as a result of the productivity of time-consuming processes of production (any more than interest can be ascribed to the fruitfulness of a tree).¹⁰

The basis on which the PTPT is dissatisfied with the Fisher (productivity-of-waiting) solution to the interest problem is thus strictly a non-economic “philosophic” one—viz., a view of time (and thus of waiting) that sees it (as Böhm-Bawerk himself apparently saw it) as a neutral background medium, rather than as a positive, active ingredient in productive processes. It is this philosophic perspective that underlies the pure time-preference theory to be discussed in this paper. Given this philosophic perspective, then, we note that the interest problem has, therefore, thus far not been touched in the slightest by any productivity-of-roundaboutness considerations.

The Pure Time-Preference Theory of Interest

The alternative “Austrian” theory of interest that we wish to clarify in this paper, the PTPT, is that pursued consistently by the American Frank Fetter and by Ludwig von Mises. This theory solves the interest problem by appeal to widespread (possibly universal) positive time preference. If, in fact, people do prefer (other aspects of the situation aside) to achieve their goals sooner rather than later, then the dilemma posed by the machine and its rentals, or by the tree and its fruit, dissolves. The price paid for a tree tends systematically to fall short of the sum of its annual fruit yields because, when the tree is bought, the yields are only prospective yields. One is simply not prepared to pay $100 today in order to command $100 worth of fruit in five or ten years time, no matter how ironclad the contract for the fruit delivery may be. The prospect of $100 available in the future has less attractive power than does the prospect of $100 available immediately. The PTPT argues that this solution of the interest problem is entirely sufficient to account for the interest phenomena we observe, in all their manifestations, in the simplest consumption loan context (in the pure-exchange economy), or in the most complex of financial-industrial situations. After all, production processes do take time, hence the present price of input services must, given positive time preference, systematically and repeatedly, fall short of the nominal value yielded in the future by their marginal productivity. A portion of currently emerging output must then regularly be retained each year by the capitalist who has, sometime in the past, advanced the sums needed to pay for the input services whose output is now emerging. This retained interest income is not caused by, or made possible by, the physical productivity of anything.

Such time-preference considerations are able, then, to account completely for the phenomenon of interest. They are able, entirely without any appeal to productivity of roundaboutness, to answer the question formulated in the interest problem. On the other hand, as we have seen, the productivity of roundaboutness offered no solution at all to that problem (for those not recognizing waiting or time as a productive agent). Hence, in the view of its protagonists, this “pure” time-preference theory¹¹ is entirely adequate for its objective; no productivity considerations can possibly enter at all into the explanation offered for interest.

It is the latter contention that many theorists, from Böhm-Bawerk's time down to our own, have found simply incredible. These critics of PTPT find it unbelievable that the claim can be made that the market phenomenon of interest is never, in any way, under any circumstances, to be attributed to the productivity of capital or of roundaboutness, waiting, or time. It will be instructive to examine a recurring theme pursued by these critics in expressing their incredulity.

Sheep, Rice, and Austrian Hocus-Pocus

These critics argue that simple hypothetical examples demonstrate that, at least under certain conditions, a positive rate of interest necessarily emerges, being strictly determined by physical productivity. These examples, the critics maintain, demonstrate that the old argument with which Böhm-Bawerk had demolished the simple productivity theories (viz., the argument that competition ought to drive the market price of the productive agent to the point where it no longer yields a value surplus) cannot entirely drive productivity considerations out of the interest-theoretic picture. Whether or not we can find a logical flaw in the Böhm-Bawerkian argument, the stylized “facts” of these examples prove the argument to be specious. Productivity considerations are clearly sufficient to account for interest. It cannot, therefore, be the case that the explanation of interest must run in purely time-preference terms. The examples used by these critics vary in their details. H.G. Brown talked of fruit trees,¹² Irving Fisher talked of sheep,¹³ Knight talked of his Crusonia plant (an edible plant that grew at a fixed rate continuously),¹⁴ quite recently Samuelson constructed a rice example to drive home an essentially similar point. What these examples have in common is (1) that they postulate a given rate of physical productivity to the capital stock, a rate invariant to scale of production or level of wealth; (2) that these examples appear, at least at first glance, to escape the impact of the traditional Böhm-Bawerkian argument (against productivity theories of interest) by confining the example to the context of a single good world (in which the critics believe value productivity measures must coincide with physical productivity rates).¹⁵ Let us examine the most recent of these examples, Samuelson's rice case (which he raised in a discussion of a different doctrinal issue).

Samuelson's case was introduced in the course of a recent critical reconsideration of Schumpeter's zero-interest doctrine.¹⁶Schumpeter had argued (on the basis of reasoning reflecting Böhm-Bawerk's arguments against the productivity theories of interest) that in a world in circular-flow equilibrium, the rate of interest must be zero, with all output value decomposed into land, rent, and labor wages, with nothing left for any interest share. Samuelson objects that a possible technological case refutes the Schumpeterian argument. The case Samuelson identifies is that of 100 units of rice ripening into 110 units of rice during the period of one year, without the input of any labor or any scarce land. This case shows, Samuelson claims, that final value need not necessarily be wholly swept back through the market as factor payments to labor and land; apparently we have 10 units of rice (“real interest income”) that can be attributed to no factor service—only to the productivity through time of the initial rice stock. Mere ownership of rice capital confers title to a possible perpetual flow of annual rice consumption income. This annual income is clearly interest earned by the rice capitalist. This income is accounted for entirely by the physical productivity of rice. Samuelson hastens to anticipate the obvious Schumpeterian response. Schumpeter had emphasized that, with interest zero, “the greater magnitude of the forest is already imputed back in value to the saplings.” So that today's 100 units of rice already have the value of next year's 110 units: “these foreseen changes…only conserve the already calculated value of the process,” without involving any creation of new value. But Samuelson hotly denounces this response as

pure deception. Real rice is being produced net. Kuznets can measure it. You can eat 10 [units of] rice every year and still not impair your circular flow income.…No hocus-pocus of backward imputation—of forest to sapling, or rice grain to rice grain—evades the naive fact of productive interest.¹⁷

It may be instructive to note how this kind of example appears to escape the Böhm-Bawerkian critique. Naive productivity theories explain the interest on the capital sum invested in a tree by virtue of the tree's physical fecundity. Böhm-Bawerk's criticism pointed out that physical productivity does not necessarily mean value productivity. In value terms it is surely still possible, in principle, for the value of the tree to be equal to the sum of the values of all future fruit crops. The presently considered fruit tree, sheep, and rice examples seem to escape this problem. One writer has in fact claimed¹⁸ that this is the outstanding virtue of the single-good economy (he is thinking specifically of Knight's Crusonia-plant economy). Rates of productivity can be arrived at directly, since “capital stock” and “income” consist of the same physical entities. No resort need be had to calculation in “value terms,” with all its attendant pitfalls. So that these productivity examples, from sheep to Crusonia to rice, do indeed demonstrate that, even with zero rate of time preference, present rice exchanges for future rice at a rate that expresses the physical productivity of rice. If this demonstration were enough to settle the problem posed by the phenomenon of interest, the issue would indeed be closed. Time preference need have nothing to do with the emergence of interest; interest as a phenomenon, and the particular rate of interest established in the market can, it is clear, be entirely explained by physical productivity (at least in certain contexts). Yet the matter is far less simple than this.

Interest, Own-Rates of Interest, and
Intertemporal Exchange

What these examples demonstrate is that physical productivity affects (or even “determines”) the intertemporal exchange rate (the own-rate of interest) on sheep, on rice, and on Crusonia, respectively. One hundred units of 1987 rice exchange, in 1987, for 110 promised units of 1988 rice. With this trade repeated each year, the rice owner can consume 10 units of rice each year (“real interest income”) without eroding the (“capital”) base that yields this annual income. We shall attempt to show, however, that from the Fetter-Mises PTPT view, these demonstrations do nothing to advance understanding of the general phenomenon of interest; nor do they, as we shall see, demonstrate the impotence or irrelevancy of Böhm-Bawerk's refutation of simple productivity theories.

It may be useful to review the impasse we have apparently reached. On the one hand, these sheep and rice stories show that an annual consumption income drawn from the physical fecundity of a source can in principle be indefinitely enjoyed without eroding the continued existence and productiveness of that source. On the other hand, the logic of the Böhm-Bawerkian reasoning refuting productivity theories of interest has not itself been addressed—the reasoning has been denounced as hocuspocus not because of any demonstrated logical fallacy but apparently because these sheep and rice cases are supposed to serve as counterexamples showing precisely that phenomenon which the Böhm-Bawerkian reasoning purported to have proven to be impossible.

Reflection should surely convince us: (a) that the Böhm-Bawerkian reasoning must still be reckoned with—after all, no flaw in its logic has as yet been identified in these examples; (b) that these examples demonstrate the possibility of an income different from that for which Böhm-Bawerk sought an explanation—the possibility, that is, of an income which the Böhm-Bawerkian reasoning never questioned; (c) that since it was the Böhm-Bawerkian reasoning that underlay what we have called the interest problem, that interest problem has not been touched at all by these sheep and rice stories. Let us try to explain all this. We must return to our original statement of the interest problem.

The interest problem, we recall, asked how it is possible for an individual to invest capital funds in a way that yields a perpetual net income. Why does not the market bid up the price of all the “machines” (in which the individual might plan to invest his capital) so that no net annual yield remains. This question, it should be noted, did not challenge the physical possibility of a tree of infinite life producing an annual crop of fruit (or a tree of finite life producing an annual crop of fruit large enough to permit the planting, out of fruit output, of a replacement tree when the parent tree dies). The question merely asked why, in the absence of any other theory of interest, the market does not bid up the price of the tree to the point where in fact no net annual financial yield is possible from investing in trees. This is what Fetter hammered away at, the distinction between goods and values in this context: “A theory of interest must be essentially a value-theory.”¹⁹ Samuelson has proven that ownership of rice permits an indefinite stream of annual rice consumption; he has not disproved the contention that the anticipated perpetual flow of rice consumption is already fully recorded in the market valuation of the initial rice source. Samuelson does not, in fact, appear to wish to deny this contention; he appears merely to conclude that this contention constitutes a meaningless incantation which does not affect the undeniable realities of perpetual annual consumption flows—duly measured by Kuznets, or somebody.

One can sympathize with Samuelson's impatience at verbal mysticism that seems out of touch with palpable reality. Moreover we can go a little further; Samuelson was dealing not with any theory of interest, but with Schumpeter's belief that—regardless of productivity—in the state of circular flow no interest at all would emerge. It is easy to sympathize with Samuelson's sense of conviction that his rice example has shown that productivity interest can indeed emerge in this state of affairs. Yet it must be pointed out that when we turn to a world somewhat richer in assumptions than Samuelson's rice model, it becomes clear that his example (and also the earlier fruit tree, sheep, and Crusonia stories) have in fact failed to identify interest income (in the sense in which we have identified the interest problem) as a productivity return.

One hundred units of 1987 rice are expected to ripen into 110 units of 1988 rice. Suppose that the “value” of the 100 units of 1987 rice has indeed risen to anticipate this physical growth. Then in terms of the interest problem (formulated at the outset of this paper) the perpetual annual rice consumption income so made possible does not present an example of interest. The annual flow of rice income is indeed adequately explained by productivity—more to the point, there was no “problem” that demanded explanation, at all. There is, after all, no problem constituted by the circumstance that a tree yields fruit annually.

The interest problem would begin, in the context of the rice example, only if in fact the “value” of the 100 units of 1987 rice is somehow lower than that of 110 units of 1988 rice. Then we would have the possibility of a sum of abstract capital value serving as a financial source somehow generating a flow of greater subsequent value. That would indeed appear to fly in the face of economic intuition (since competition ought—absent a theory of interest—to be expected to exclude such a phenomenon). And it is of course this interest problem that PTPT solves by reference to the general subjective preference for the achievement of goals sooner rather than later.

What the preceding asserts, then, is that what Brown, Fisher, Knight, and Samuelson have identified as interest income—and which they apparently view as the only interest income needed to be discussed—is in fact not interest income at all (from the perspective of the interest problem formulated above). On the other hand, that interest income which PTPT deals with—an income which does fit the specifications of interest as formulated above—turns out to be something the very existence of which the critics appear to deny. (We recall that Schumpeter, too, was [in this respect only!] on the side of the critics of PTPT: he believed that in the circular-flow no pure interest would be present.) Obviously the entire debate appears to have degenerated into a squabble about the meaning of terms.

From the Fisherian perspective the semantic bickering may appear even more deplorable. After all, the Fisherian view is that interest, the income received in return for providing services of “waiting,” is at the same time, the reward received by the capitalists (who provide these services) for their sacrifices (sacrifices which obtain their poignancy from the prevalence of positive time preferences). A discussion such as the above (in which the issue is made to appear whether the term interest income is to refer to the fruits of rice fecundity or to a value differential attributed to time preference) must seem doubly regrettable. Let us go back to basics.

The Interest Problem that Calls For Solution:
The Competing Versions

The attention of economists over the centuries has been attracted to the real-world, palpable, phenomenon of interest income largely due to its “surplus” character. Apparently it is possible, in the market economy, to command a steady, regular income merely by possession of a capital stock. As Böhm-Bawerk introduced the phenomenon:

Whoever is the owner of a capital sum is ordinarily able to derive from it a permanent net income…This income…arises independently of any personal act of the capitalist. It accrues to him even though he has not moved a finger in creating it…. It can be derived from any capital, no matter what be the kind of goods of which the capital consists, from naturally fruitful, as well as from barren goods, from perishable as well as from durable goods, from replaceable as well as from irreplaceable goods, from money as well as from commodities. And, finally, it flows without ever exhausting the capital from which it arises, and therefore without any necessary limit to its continuance…. And so the phenomenon of interest presents, on the whole, the remarkable picture of a lifeless thing, capital, producing an everlasting and inexhaustible supply of goods.²⁰

With this starting point for discussions of the interest phenomenon, it is easy to see how physical fecundity (while it offers a tempting ingredient for interest theorizing) came to be dismissed (in Böhm-Bawerk's chapter refuting the productivity theories). The point is that the entire discussion begins with the assertion that value productivity is a common fact of life; this fact of life was, after all, asserted to hold for barren goods and for money just as it holds for naturally fruitful goods and for commodities. It was this Böhm-Bawerkian identification of the problem as referring to value productivity that was at the heart of the formulation of the interest problem cited at the start of this paper. Now, the prospective interest theorist can choose one or the other of only two options: he can deny that this interest phenomenon in fact exists (so that there is nothing to be explained), or he must seek an explanation for it within its own framework. Schumpeter chose the first option. Others, implicitly denying the phenomenon that Böhm-Bawerk took as his starting point, focus on physical productivity as a sufficient basis for the (undeniable) market phenomenon of interest. But in so doing they have not offered a new explanation for the Böhm-Bawerkian phenomenon: they have simply denied its existence, and chosen instead to talk about something else—something easily confused with it (because both might serve plausibly as underlying basis for the surface phenomenon of market interest). PTPT theorists are then entitled to accept the existence of the Böhm-Bawerkian phenomenon and to offer their own explanation for it. From this perspective, then, the situation can be summed up as follows:

First, there is a commonly observed phenomenon of market interest. No one disputes this observation. Second, for Böhm-Bawerk this phenomenon of market interest is simply the expression of the existence of an underlying phenomenon of pure value productivity. This interpretation of market interest renders the phenomenon insoluble through reference to physical productivity, and leaves it calling out for alternative explanation. Third, PTPT accepts this Böhm-Bawerkian interpretation (and offered time-preference insights to account for the phenomenon of value productivity). Fourth, Schumpeter, while not disputing the Böhm-Bawerkian interpretation of the observed phenomenon of market interest, held that this entire phenomenon is a transient one that would disappear in the state of circular flow. Fifth, other theorists dispute Böhm-Bawerk's interpretation of the observed phenomenon of market interest. In effect these writers deny the existence of the (underlying “value productivity”) phenomenon which Böhm-Bawerk asserted to be reflected in the observed phenomenon of market interest. For these writers the observed phenomenon of market interest reflects nothing more than physical productivity; there is no “value productivity” to be explained. Sixth, finally, for modern theorists following Fisher, the entire discussion in the present section represents an incomprehensible lapse into meaningless mysticism and metaphysics. For these theorists the task of a theory of interest is not at all that of solving the problem formulated at the start of this paper (and expressive of the Böhm-Bawerkian formulation at the start of this section). Were the Fisherians to consider that (Böhm-Bawerkian) question, they would consider it solved immediately once one has identified “waiting” as a missing factor. Since waiting is a scarce, costly factor, its market value tends to equal both its marginal product and its marginal cost (in terms of foregone sooner consumption). Interest is both the reward for sacrificed earlier consumption and the fruits of the enhanced output made possible by waiting. For Fisher and the modern writers the interesting question, and thus the task of interest theory is (quite apart from possible disputes concerning the existence of the phenomenon to which the Böhm-Bawerkian formulation refers) nothing more than that of identifying the determinants of intertemporal prices. In undertaking this latter task, Fisher develops a perfectly adequate framework in which both physical productivity and time-preference considerations have their place.

We may conclude then: (a) Most of the modern bewilderment of PTPT stems, we would argue, from the Fisherian unconcern with the interest problem that was posed by Böhm-Bawerk. Most moderns, following Fisher, do not see the task of interest theory to be to account for a phenomenon that somehow (in the absence of a theory of interest) “ought not to exist.” As a result, the Fisherian approach finds the PTPT view bizarre (since their own approach finds time preference and physical productivity both valuable elements in their theory). We shall return to consider this Fisherian view of PTPT in the subsequent section. (b) Sheep, rice, and other stories introduced to demonstrate the adequacy of a pure productivity theory (and thus to refute PTPT) express a different misunderstanding of PTPT. Without recognizing it, the authors of these examples have in effect denied the very existence of the phenomenon which Böhm-Bawerk identified as that calling for explanation (i.e., the underlying phenomenon of pure value productivity). From the Böhm-Bawerkian point of view, therefore, these stylized examples deal with a phenomenon which (1) is other than the one Böhm-Bawerk is concerned with, and (2) offers no theoretical challenges (“of course” trees produce fruit!).

We have thus identified two distinct sources of modern bewilderment with PTPT. One source stems from an implicit denial of the value-productivity phenomenon seen as calling for explanation in Böhm-Bawerk's view (and in PTPT'S view). From this denial it was easy to move to assuming that the phenomenon to be explained by a theory of interest is that which would underlie the market rate of interest in the absence of the value-productivity phenomenon—viz., physical productivity. From this latter perspective, PTPT'S refusal to accord to physical productivity any role appears simply incomprehensible. The second source of modern bewilderment stems, not from any denial of the phenomenon that Böhm-Bawerk sought to explain, but from an entirely different conception of what constitutes the function of a theory of interest. Whereas for Böhm-Bawerk a theory of interest is called for to account for an otherwise inexplicable phenomenon, for Fisher a theory of interest is called for to identify the determinants of a particular market rate of exchange (viz., the intertemporal rate, which implicitly expresses the rate of interest). From this Fisherian perspective it is easy to understand how bizarre it appears for anyone to deny any role for physical productivity in a theory of interest. We turn to elaborate on this last point. The point turns out to involve an old dispute between Fisher and Böhm-Bawerk that throws considerable light on the entire issue.

The Existence of Interest vs. the Rate of Interest

“Some writers,” Fisher wrote, “have chosen, for purposes of exposition, to postulate two questions involved in the theory of the rate of interest, viz., (1) why any rate of interest exists and (2) how the rate of interest is determined.”²¹ Fisher dismisses this distinction as being unilluminating, “since to explain how the rate of interest is determined involves the question of whether the rate can or cannot be zero.”²² The purpose of the present section of this paper is (a) to present the case for the distinction criticized by Fisher—a distinction in fact made by Böhm-Bawerk, as we shall see—and (b) to show how failure to understand the rationale for the distinction has generated the widespread modern bewilderment with PTPT referred to earlier.

No better defense for Böhm-Bawerk's distinction need be found than the lucid discussion that he himself provided. Böhm-Bawerk was criticizing Fisher for not distinguishing between “originating forces” and “determining forces”:

All interest-originating causes undoubtedly are also determining factors for the actual rate. But not all rate-determining factors are also interest-creating causes…. When we inquire into the causes of a flood we certainly cannot cite the dams and reservoirs built to prevent or at least mitigate inundations. But they are a determining factor for the actual watermark of the flood…. Similarly, there are other circumstances besides the actual interest-creating causes that bring about or enhance the value advantage of present goods over future goods.²³

It will be observed that Böhm-Bawerk's distinction faithfully expresses the formulation of the interest problem that he offered (as cited in the preceding section). There is a phenomenon (the existence of interest income) which calls for explanation (just as the occurrence of floods calls for explanation). The required explanation need not necessarily invoke all those “forces” which may be relevant for the determination of the particular rate of interest prevailing in the market. It is true that a complete listing of all aspects of all the “rate-determining forces” would at the same time explain why the interest rate is other than zero. But to say that all these forces are responsible for the interest phenomenon would be highly misleading. If someone stands amazed at the flow of city traffic along one of its central avenues during morning rush hour, and asks, “Why is traffic so heavy?”, the answer to the question should presumably run in terms of the need of people to get to work. It will simply not do to invoke the traffic-light system as an explanation (even partially) for the volume of traffic (even though it is certainly the case that the size of that volume has been, in part, determined by that traffic-light system). For Fisher, it is quite clear, the “problem” of interest is simply to provide a full catalogue of the rate-determining forces. Fisher's analysis leads him, inevitably, to recognize the interaction, among these forces, of physical productivity and time preference. Modern theorists, following Fisher, are understandably bewildered by PTPT statements denying any role for physical productivity. But to endorse the PTPT denial of a role for physical productivity is not necessarily to deny that physical productivity is to be listed among the forces combining to determine intertemporal rates of exchange. The PTPT's denial refers strictly to the problem of accounting for the phenomenon of interest. For the solution of this problem, PTPT finds physical productivity to be no more relevant than traffic lights are (for accounting for the phenomenon of morning rush hour traffic) or than dams are (for Böhm-Bawerk's example of accounting for the occurrence of floods).

PTPT exponents often drive home the irrelevance of physical productivity by pointing out that physical productivity is neither necessary nor sufficient for an explanation for interest (in the sense of value productivity).²⁴ They point out how, in the absence of time preference, physical productivity, no matter how great, cannot generate value productivity. And they point out how, even in a pure exchange world without production processes of any kind, the phenomenon of value productivity could arise as a result of time-preference considerations exclusively. One critic responded to this reasoning by asking whether, based on parallel reasoning, one would conclude that the striking of matches is irrelevant to the causation of fire (since not every match that is struck produces fire, and not all matches that ignite do so as a result of having been struck). The response should surely be that the striking of matches is certainly highly relevant to the prevalence of match ignition, but the striking of matches is indeed far less relevant for other questions, such as, e.g., why ignition occurs among matches rather than among Q-tips.

Admittedly, some expositions of PTPT have sometimes unfortunately permitted it to be thought that, since only time preference accounts for the existence of interest, it follows that a change in the conditions of physical productivity would invariably leave the rate of interest unchanged. But in fact PTPT implies nothing of the kind.²⁵ It is one thing to maintain that normal daily nutritional needs are in no way responsible for the existence of hospitals; it in no way follows that the extent of such needs is not a factor helping determine the size (and conceivably even the number) of hospitals. Given the presence of sickness (upon which the existence of hospitals does depend), a host of unrelated matters may participate in the determination of hospital size. Similarly PTPT theorists do recognize that since the prevalence of positive time preference does generate the phenomenon of interest, it cannot at all be ruled out that other factors (including physical productivity) may affect the determination of its rate. For example, physical productivity may significantly affect the level of wealth, and thus the marginal rate of time preference. All PTPT insists on is that, no matter how significant a role physical productivity may play among the complete list of variables affecting the rate of interest, it is fallacious to refer to interest income (expressing value productivity) as the fruit of the physical productivity of agents of production. No matter how heavily the costs of the hospital meal service weigh in the determination of hospital size, it remains a fallacy to see a hospital as a sleep-in restaurant.²⁶

Some Remarks on Methodological Essentialism

It will be observed that our defense of PTPT against the bewilderment evinced by its various critics, amounts to a partial affirmation of what has sometimes been termed “methodological essentialism.” Several historians of thought have noticed that for Menger, economic science is a search for the reality underlying economic phenomena—for their essence (das Wesen). In a letter to Walras, Menger asks, “How can we attain to a knowledge of this essence, for example, the essence of value, the essence of land rent, the essence of entrepreneur's profit…by mathematics?”²⁷ This search for essences, reflecting a philosophical approach attributed to Aristotelian influence,²⁸ would focus, then, not on the land rent paid for a particular parcel of real estate in a particular year, but upon those essential features of land rent that would be common to all examples of the phenomenon. Similarly an essentialist approach to the interest problem as posed by Böhm-Bawerk would focus not on the list of elements which together determine specific interest rates, but on those elements upon which the interest phenomenon essentially depends, elements without which the phenomenon could in fact not exist. PTPT finds these essential elements for the interest phenomenon in time preference. Physical productivity is not such an essential element of interest; but, to repeat, to affirm all this is not to deny that market rates of interest may be related to physical productivity conditions. To assert that, absent time preference, physical productivity is unable to generate any interest (in the sense of value productivity) is not to assert that, given time preference, the intertemporal rate of exchange, even in the form of the rate of value productivity, is unaffected by changes in the physical productivity of machines or of trees. All that is being asserted is that whatever role is played by physical productivity does not permit us to say that interest (value productivity) is the fruit of productive capital. (Rush hour traffic as a phenomenon may indeed be vitally affected by the timing of traffic lights; nonetheless we understand why it is accurate to describe the phenomenon of rush hour traffic as “people getting to work,” rather than as “the result of traffic-light timing.”)

From the perspective of a modernist impatient with metaphysical discussion, some of this defense of PTPT against modernist “bewilderment” may appear wholly unhelpful; in fact it may confirm the critic's impression that it is all empty philosophizing unrelated to objective phenomena. If one sees the objective of science as being to account for empirical reality, and one sees the objective of price theory to explain the structure and levels of particular prices, then it will seem only natural to seek for an interest theory that explains particular rates of interest. All else must appear metaphysical, irrelevant, and plain confusing. So long as our theory of interest, say, in the form of Fisher's standard diagram (incorporating both intertemporal productive possibilities and time preferences), can show how the market in fact generates its intertemporal rates of exchange (with respect to given physically defined commodities identified at different dates), surely our scientific task has been accomplished—without any essentialist pursuit of some metaphysical reality underlying the explained phenomena. But this modernist impatience can be shown to be unreasonable. Such a demonstration requires that we recognize the nature of the delicate interface between science and ideology. It will prove convenient to present our discussion in the context of brief reference to the celebrated Cambridge Capital Controversy of recent decades.

Science and Ideology: The Cambridge Controversy
and PTPT

For present purposes, we compress our account of the Cambridge Controversy to its barest relevant elements. One side of the debate, representing neoclassical orthodoxy, sees the phenomena of the capitalist economy, especially the assignment of income shares (including interest income), as being phenomena to be understood within the framework of market equilibrium. Market prices (and thus interest) have to be paid if consumers are to receive that which the productive capacity of the market is able to provide, and for which the consumers are prepared to pay. Interest is rendered necessary and thus, in a sense, “justified” by efficiency (i.e., consumer sovereignty) considerations.

On the other hand, the critics in Cambridge (England) vigorously deny that interest incomes are “caused by individual exchanges as constrained by technology and the availability of factors of production.”²⁹ These critics see the distribution of income between wage earners and interest receivers as being determined by such considerations as the power balance between workers and capitalists, rather than by marginal products, consumer preferences, and factor supplies.

It is not difficult to recognize the ideological implications that can be drawn from each of these two views. As Robert Solow (representing the neoclassical side of the debate) observed, the Cambridge School saw neoclassical theory as “an important part of an apology for private capitalism. It sounds as if capitalists are entitled to their profits.”³⁰ Indeed, Joan Robinson, leading figure in the Cambridge School, asserted very explicitly: “The unconscious preoccupation behind the neoclassical system was chiefly to raise profits to the same level of moral respectability as wages.”³¹ Clearly the Cambridge critics believe that their own theory of capital provides no such comfort and solace for the capitalist system.

It may be submitted that these asserted ideological implications of alternative capital theories are profoundly important for one's appreciation of what economic science can reveal. While Solow and others³² believe the Cambridge attack on neoclassicism expresses an anti-capitalist animus, they do not deny the responsibility to deal dispassionately with the Cambridge substantive criticisms and theory. Nor do the Cambridge critics fail to recognize that any ideological dissatisfaction with the implications of neoclassicism must yield to a dispassionate search for logical flaws in that theory. The important point is that such dispassionate scientific debate holds the key to the normative, non-scientific characterization of specific income categories. The neoclassical view of interest permits it to be seen as a productivity return (parallel to the way in which market wages are perceived). This is seen to permit the view that interest is on the “same level of moral respectability as wages.” The Cambridge theory is one from which such innate respectability for interest income does not emerge. We wish to argue briefly here that the Cambridge Capital Controversy offered an array of alternative theoretical positions that was not exhaustive. A third point of view, not represented in the discussion, but one highly relevant to the underlying ideological concerns, is in fact to be found in the PTPT.

The neoclassical side of the debate saw interest as a productivity return. Any ideological defense of capitalist interest based on this side of the debate will consist in “justifying” interest as the proceeds of enhanced productivity made possible by scarce, costly waiting. In denying this defense, the Cambridge position will argue that interest is not the “justified” proceeds of productivity. (Rather it is a share of “social surplus” somehow acquired by owners of capital.) What needs now to be pointed out is that PTPT offers an understanding of interest income that may be seen as supporting its moral “respectability,” but without seeing it as the fruits of productivity. From the PTPT perspective, the neoclassically implied defense of capitalism is flawed. We have again and again pointed out how PTPT does not recognize market interest income as constituting a productivity return; that is not what interest income is. On the other hand, PTPT very definitely sees market interest as expressing a market-determined rate of intertemporal exchange. So that PTPT provides a basis, if one chooses to use it for such a purpose, for a justification of interest (as a legitimate expression of consumer preferences) that nonetheless agrees with Cambridge critics that interest is not a productivity return.

The point of all this is that we cannot, surely, close our eyes to possible ideological implications of science. Our science may well be, perhaps, ideologically untainted and value free (or, at any rate, honest efforts in this direction may be undertaken), but human beings are, as valuing citizens, vitally interested in the character of controversial phenomena. If one is asked, “What justifies interest?”, it will simply not do to defend it as reward for productivity, if analysis shows that that is not what it is. Note that for ideological (i.e., for normative, evaluative) purposes, methodological essentialism is highly relevant. One can hardly arrive at a judgment concerning the defensibility of interest income by showing that its size depends upon physical productivity. A “theory of hospitals” that fails to identify hospitals in terms of the essentiality of their medical character is likely to be less than helpful for the purposes of normative evaluation by citizens. Citizens asked to vote to support hospitals seen as sleep-in restaurants may respond differently than when hospitals are correctly seen as institutions fighting to contain dread diseases. There is every reason for science to take note of the non-scientific purposes for which scientific results may be helpfully consulted. From this perspective, the methodologically essentialist aspects of PTPT may be considered valuable features of it, rather than as obfuscating metaphysics.

Conclusion

Our defense of PTPT against modern “bewilderment” has dealt primarily with its apparently astounding assertion that physical productivity has nothing essentially to do with the phenomenon of interest. Our discussion has: (1) made it clear that PTPT does not necessarily deny a role for physical productivity in interest rate determination; (2) emphasized that what PTPT addresses is a question that is different from that of interest rate determination; (3) identified the problem dealt with by PTPT as the interest problem addressed by Böhm-Bawerk, viz., what accounts for the phenomenon of net value-productivity (in the face of market competition that might be expected to squeeze it out of existence); (4) shown that PTPT'S refusal to recognize any physical productivity role in the explanation for the existence of interest income rests on (the admittedly arbitrary) view that time and waiting are not to be seen as productive agents; (5) recognized the methodologically essentialist aspects of PTPT and argued for their relevance, especially in the context of the Cambridge Capital Controversy, for citizens' normative understanding of interest income; (6) made it clear how numerous stylized examples (sheep, rice, Crusonia, etc.) designed to demonstrate the essential role played by physical productivity in interest income generation, in fact, concern a phenomenon quite different from that upon which the Böhm-Bawerkian discussion focused. Regardless of one's opinion of the significance of income generated by physical productivity, we have emphasized the legitimacy of distinguishing between that income and the quite different income concept identified by Böhm-Bawerk and addressed by PTPT.

The upshot of the discussion, then, is that PTPT affirms the phenomenon of pure value-productivity, that is, the phenomenon in which a source of value at a given date generates a flow of values during subsequent periods that exceeds, in total, the value of the source. PTPT accounts for this phenomenon by reference to widespread (possibly universal) preference for the earlier, rather than later, achievement of goals. Market rates of interest, and market interest income, are expressions of this underlying phenomenon of value productivity (and of its PTPT roots).

---

Reprinted from The Meaning of Ludwig von Mises: Contributions in Economics, Sociology, Epistemology, and Political Philosophy, Jeffrey M. Herbener, ed. (Auburn, Ala.: Ludwig von Mises Institute, 1993).

* This paper owes much to countless discussions with Ingo Pellengahr over 13 a two-year period. His open-minded but persistent questioning concerning troublesome aspects of PTPT helped (and compelled) the writer toward the present clarification. For Pellengahr's own perspective on the matters dealt with in this paper see references to Pellengahr.

1 Joseph A. Schumpeter, The Theory of Economic Development (Cambridge, Mass.:Harvard University Press, 1934).

2 Eugen von Böhm-Bawerk, Capital and Interest (Spring Mills, Penn.: Libertarian Press, 1959 [1889]). Contains translations of three volumes: vol. 1, originally published in 1884; vol. 2, originally published in 1889; vol. 3, originally published in 1921.

3 Daniel M. Hausman, Capital, Profits and Prices (New York: Columbia University Press, 1981), p. 3.

4 As we shall see, there were two “Austrian” theories of interest, both deriving from Böhm-Bawerk, see F.A. Hayek, The Pure Theory of Capital (London: Routledge and Kegan Paul, 1941), app. 1. One of these, which (unlike Hayek) we identify with Fisher (and describe as neoclassical), is that discussed here in the text. The second is the PTPT, the subject of this paper.

5 Hausman, Capital, Profits and Prices, p. 167.

6 Frank A. Fetter, “The ‘Roundabout Process’ in the Interest Theory,” Quarterly Journal of Economics 17 (November, 1902); reprinted in Frank Fetter, Capital, Interest, and Rent: Essays in the Theory of Distribution, Murray N. Rothbard, ed. (Kansas City: Sheed Andrews and McMeel, 1977). Also, Frank Fetter, “Capitalization versus Productivity: Rejoinder,” American Economic Review (December, 1914).

7 Ludwig von Mises, Human Action: A Treatise on Economics (New Haven, Conn.: Yale University Press, 1949).

8 On this see T. Haavelmo, A Study in the Theory of Investment (Chicago: University of Chicago Press, 1960), p. 47; Israel M. Kirzner, An Essay on Capital (New York: Augustus M. Kelley, 1966), p. 97.

9 Böhm-Bawerk, Capital and Interest, pp. 97-98.

10 Fetter, “The ‘Roundabout Process’ in the Interest Theory.”

11 See Ingo Pellengahr, “Austrians Versus Austrians I: A Subjectivist View of Interest,” in Studies in Austrian Capital Theory, Investment and Time, M. Faber, ed. (Berlin, Heidelberg, and New York: Springer-Verlag, 1986), pp. 10-11, for several senses in which the adjective “pure” may be understood in the present context. Also, Pellengahr, “Austrians Versus Austrians II: Functionalist Versus Essentialist Theories of Interest,” in Studies in Austrian Capital Theory, Investment and Time, M. Faber, ed. (Berlin, Heidelberg, and New York: Springer-Verlag, 1986).

12 H.G. Brown, “The Discount Versus the Cost-of-Production Theory of Capital Valuation,” American Economic Review (June, 1914).

13 Irving Fisher, The Theory of Interest (New York: Macmillan, 1930), p. 193.

14 Frank H. Knight, “Diminishing Returns from Investment,” Journal of Political Economy (March, 1944): 52.

15 Brown was dismayed at having been thought by Fetter to have failed to recognize the problem of value productivity. He believed himself to have successfully avoided this pitfall by his example. See H.G. Brown, Economic Science and the Common Welfare, 3rd ed. (Columbia, Mo.: Lucas Brothers, 1926), p. 125, n. 13.

16 Paul A. Samuelson, “Schumpeter as an Economic Theorist,” in Schumpeterian Economics, H. Frisch, ed. (New York: Praeger, 1981).

17 Samuelson, “Schumpeter as an Economic Theorist,” p. 23.

18 Donald Dewey, Modern Capital Theory (New York and London: Columbia University Press, 1965), p. 80.

19 Fetter, “Capitalization versus Productivity: Rejoinder,” 257.

20 Böhm-Bawerk, Capital and Interest, p. 1.

21 Fisher, The Theory of Interest, pp. 13f. and 474 (where Böhm-Bawerk is identified, without specific citation, as having argued for this distinction).

22 Ibid. See also J.W. Conard, An Introduction to the Theory of Interest (Berkeley and Los Angeles: University of California Press, 1959), pp. 13-14.

23 Böhm-Bawerk, Capital and Interest, p. 192.

24 Charles W. Baird, Prices and Markets, Intermediate Microeconomics, 2nd ed. (St. Paul, Minn.: West Publishing, 1982), pp. 303f.

25 Fetter was quite explicit on this point, see “Capitalization versus Productivity: Rejoinder,” 247.

26 Economists have frequently argued (e.g., F.A. Hayek, “Time Preference and Productivity: A Reconsideration,” Economica, n.s., 12 [1945]: 22-25) that whether time preference or physical productivity is to be considered the more important explanatory variable for interest, depends on which of them is expressed, in the standard Fisher diagram, by a curve having greater convexity. This makes good sense in regard to the determinants of the interest rate. It may be highly relevant to know whether the interest rate is more sensitive to a given marginal change in time preferences, than to a similar marginal change in physical productivity. But for the explanation of the existence of interest, these comparisons are hardly relevant. Even if hospital size were somehow more sensitive to changing nutritional standards than to change in the incidence of disease, the raison d'etre of the hospital remains unaffected.

27 This letter is cited in Terence W. Hutchison, A Review of Economic Doctrines, 1870-1929 (Oxford: Clarendon Press, 1953), p. 148.

28 See Emil Kauder, A History of Marginal Utility Theory (Princeton, N.J.: Princeton University Press, 1965), p. 97; Samuel Bostaph, “The Methodological Debate Between Carl Menger and the German Historicists,” Atlantic Economic Journal (September, 1978): 11.

29 See Hausman, Capital, Profits and Prices, p. 167.

30 Robert Solow, “Cambridge and the Real World,” Times Literary Supplement (March 14, 1975).

31 Joan Robinson, Economic Philosophy (Chicago: Aldine, 1962), p. 58.

32 Mark Blaug, The Cambridge Revolution: Success or Failure? A Critical Analysis of Cambridge Theories of Value and Distribution (London: Institute of Economic Affairs,1974).