Capital and Interest

THE COLLECTED WORKS OF F. A. Hayek

VOLUME XI

CAPITAL AND INTEREST

PLAN OF THE COLLECTED WORKS

Edited by Bruce Caldwell

The plan is provisional. Minor alterations may occur in titles of individual books, and additional volumes may be added.

THE COLLECTED WORKS OF F. A. Hayek

VOLUME XI

CAPITAL AND INTEREST

EDITED BY LAWRENCE H. WHITE

The University of Chicago Press

F. A. HAYEK (1899–1992), recipient of the Presidential Medal of Freedom in 1991 and cowinner of the Nobel Memorial Prize in Economics in 1974, was a pioneer in monetary theory and a leading proponent of classical liberalism in the twentieth century. He taught at the University of London, the University of Chicago, and the University of Freiburg.

LAWRENCE H. WHITE is professor of economics at George Mason University.

The University of Chicago Press, Chicago 60637

The University of Chicago Press, Ltd., London

© 2015 by The Estate of F. A. Hayek

All rights reserved. Published 2015.

Printed in the United States of America

24 23 22 21 20 19 18 17 16 15 1 2 3 4 5

ISBN-13: 978-0-226-27487-4 (cloth)

ISBN-13: 978-0-226-27490-4 (e-book)

DOI: 10.7208/chicago/9780226274904.001.0001

Professor Hayek and the Theory of Investment, Frank H. Knight. © 1935 Economic Journal. Reproduced with permission of Blackwell Publishing Ltd.

Maintaining Capital Intact, A. C. Pigou. © 1941 Economica. Reproduced with permission of Blackwell Publishing Ltd.

Library of Congress Cataloging-in-Publication Data

Hayek, Friedrich A. von (Friedrich August), 1899–1992, author.

[Works. Selections]

Capital and interest / edited by Lawrence H. White.

pages cm — (The collected works of F. A. Hayek ; volume XI)

Includes bibliographical references and index.

ISBN 978-0-226-27487-4 (cloth : alkaline paper) — ISBN 978-0-226-27490-4 (e-book) 1. Capital. 2. Investments. 3. Interest. I. White, Lawrence H. (Lawrence Henry), editor. II. Title. III. Series: Hayek, Friedrich A. von (Friedrich August), 1899–1992. Works. 1989 ; v. 11.

HB171.H4266 2015

332'.041—dc23

2014045424

♾ This paper meets the requirements of ANSI/NISO Z39.48-1992 (Permanence of Paper).

THE COLLECTED WORKS OF F. A. HAYEK

Founding Editor: W. W. Bartley III

General Editor: Bruce Caldwell

Published with the support of

The Hoover Institution on War, Revolution, and Peace Stanford University

The Cato Institute

The Earhart Foundation

The Pierre F. and Enid Goodrich Foundation

The Heritage Foundation

The Morris Foundation, Little Rock

Contents

Editor’s Introduction

One. Some Observations on the Imputation Problem (1926)

Two. On the Problem of Interest Theory (1927)

Three. Utility Analysis and Interest (1936)

Four. Capital Consumption (1932)

Five. Saving (1934)

Six. On the Relationship between Investment and Output (1934)

Seven. Professor Hayek and the Theory of Investment by Frank H. Knight (1935)

Eight. The Mythology of Capital (1936)

Nine. Technical Progress and Excess Capacity (1936)

Ten. The Maintenance of Capital (1935)

Eleven. Maintaining Capital Intact by A. C. Pigou (1941)

Twelve. Maintaining Capital Intact: A Reply (1941)

Index

Footnotes

Editor’s Introduction

Hayek on Capital and Interest before The Pure Theory of Capital

The present collection of articles spans the first fifteen years of Friedrich A. Hayek’s professional career, a period almost entirely devoted to technical work in economic theory. Regarding the difficult issues of capital and interest theory, the period can be divided into two unequal sub-periods. The first two articles represent the young Hayek’s Vienna period, during which he had been most importantly influenced by his distinguished University of Vienna teacher Friedrich von Wieser.¹ The later articles represent the more mature Hayek’s London period, during which he was teaching at the London School of Economics, and had been most importantly influenced by his post-university mentor Ludwig von Mises. The second period followed his 1931 lectures published as Prices and Production and culminated in his 1941 book The Pure Theory of Capital.² In these later articles, as in The Pure Theory of Capital, Hayek developed in his own distinctive way a perspective on capital and interest not grounded on the work of Wieser but rather on the work of Wieser’s Viennese contemporary Eugen von Böhm-Bawerk. Additional influences were the work of the Swedish economist Knut Wicksell, who had developed Böhm-Bawerk’s theory more rigorously, and the business cycle theory of Mises, who built on Böhm-Bawerk and Wicksell.

The Problem of Imputation

Recalling his studies in economics, psychology, and law at the University of Vienna between 1918 and 1921, a period during which Wieser returned from government service to the university, Hayek wrote: In economics it was my teacher Friedrich von Wieser who directed my interest to the intricacies of the subjective theory of value, on one particular problem of which, the theory of ‘imputation,’ I wrote during the following year and a half a doctoral dissertation while employed in a temporary government office.³ With his first degree (doctor juris) in hand, the twenty-two-year-old Hayek landed the government job in October 1921 by presenting a letter of recommendation from Wieser to one of the directors of the office, the prominent economist Ludwig von Mises, who was then forty years old. While working in the office, Hayek wrote a thesis for his second degree (doctor rerum politicarum). He then took a leave of absence to spend the period from March 1923 to May 1924 as a non-degree graduate student and research assistant at New York University. In autobiographical notes, Hayek continued the story, linking his thesis to a journal article on imputation published in 1926:

I continued to register at the university [the University of Vienna] for the degree of doctor rerum politicarum, and in the summer of 1922 started work on a thesis on the theory of imputation, for which I got my second degree in February or March 1923, just before I went to America. Though I learnt a great deal in the work and the article on Zurechnung [imputation] which later emerged from it is a fairly respectable performance, I rather hope that no copies of the thesis have survived.

. . . During the first six months in New York, I used my spare time . . . to turn my Vienna thesis on imputation of value into an article which Wieser had given me to understand might be used for the Handwörtenbuch der Staatswissenschaften but which ultimately appeared in Conrad’s Jahrbücher. On the whole I felt somewhat tired of the subjects which had chiefly occupied me in Vienna during the preceding year or so, such as the theory of subjective value or the problem of economic calculation under socialism.⁴

The absence of Mises’ name from the 1926 article calls for explanation. Hayek wrote his 1923 thesis for Wieser, on the Wieserian approach to imputation. Although he wrote it while working in a government office where Mises also worked, they were not yet closely connected. Hayek reworked the thesis at Wieser’s behest, into the shorter article included here, while away in New York. Hayek did not join Mises’ private seminar until after he had returned from the United States in May 1924. It is therefore not surprising to find that the article reflects the influence of Wieser and not that of Mises. Mises’ direct influence on Hayek could only barely have begun when the article was completed in late 1923, although that influence was evident in Hayek’s newer work by the time the article finally appeared in print in 1926.

Hayek was, however, already aware of Mises’ argument—which first appeared in a famous 1920 article—that a socialist planning board, in an economy without competitive entrepreneurial bidding for the services of factors of production, could not assign appropriate prices to those factors (labor, land, machines, materials, and so on). Mises wrote in his 1922 book Socialism: To suppose that a socialist community could substitute calculations in kind for calculations in terms of money is an illusion. In a community that does not practice exchange, calculations in kind can never cover more than consumption goods. They break down completely where goods of higher order are concerned. Once society abandons free pricing of production goods rational production becomes impossible.⁵ Hayek seems to have been challenging this argument in his 1926 article when he wrote: If one holds the view that no fully satisfactory solution to the [imputation] problem has yet been found, it cannot be excluded—as several younger authors claim—that the determinants for the prices of the factors of production exist only in an exchange economy and that for this reason an imputation of value is not possible. Yet this admission would be tantamount to abandoning subjective value theory as a satisfactory explanation for economic processes and would deprive these authors of the very basis for many of their analyses.⁶

If Mises (who was thirty years younger than Wieser) was one of the younger authors to whom Hayek referred, then the claim in the second sentence above—that to deem non-market imputation of factor values impossible is to reject subjective value theory in the explanation of economic processes—signals that Hayek was siding with the methodological position of his teacher Wieser and opposing what he took to be the methodological position of Mises.

Such opposition may seem inconsistent with Hayek’s well-known later arguments supporting Mises’ critique of socialism. In the 1945 essay The Use of Knowledge in Society, Hayek pointedly rejected Joseph Schumpeter’s proposition "that consumers in evaluating (‘demanding’) consumers’ goods ipso facto also evaluate the means of production which enter into the production of these goods. Hayek argued that to impute appropriate factor values, an observer would need to know not only all consumer demand curves for outputs, but also all producer supply and demand curves for inputs for all potential combinations of prices, which amounts to needing to know every potentially relevant bit of information about production techniques and local circumstances held by any of the myriad producers or potential producers in the market. He then took Schumpeter (and by implication other economists who thought market socialism" feasible) to task for conflating an abstract model with reality:

To assume all the knowledge to be given to a single mind in the same manner in which we assume it to be given to us as the explaining economists is to assume the problem away and to disregard everything that is important and significant in the real world. . . . I am far from denying that in our system equilibrium analysis has a useful function to perform. But . . . it is high time that we remember that it does not deal with the social process at all and that it is no more than a useful preliminary to the study of the main problem.⁷

The 1926 article does not in fact conflict with this 1945 statement regarding the infeasibility of socialist central planning, however, because in the earlier article Hayek was not talking about socialism. Its subject was not the feasibility of correct pricing by a real-world central planner, but rather imputation by the explaining economist who has by construction all the knowledge necessary to solve a model that he himself builds. Hayek’s 1926 claim is that an analyst who has stipulated the tastes, technologies, and resources of a model economy might in principle impute prices to its factors of production while abstracting from decentralized market exchange. In other words, in principle the economist might solve a general-equilibrium (GE) model for input prices, simultaneously with solving for consumer prices and for all input and output quantities. That Hayek was proposing such an approach to the factor-pricing problem is apparent in his remark toward the end of the article that under Walras’ leadership, the mathematical school of economics has already tackled successfully a similar set of tasks.⁸

In a pure exchange economy without production, given a set of consumer goods endowments, the consumers’ subjective values are the sole cause of exchange and the sole determinant of equilibrium prices. In the simplest case of two consumers and two goods, the explaining economist can, to put matters in standard terms, draw a two-dimensional Edgeworth-Bowley box and thereby illustrate the simultaneous determination of the equilibrium range of mutually beneficial prices and quantities traded (or the single equilibrium price and associated quantities traded, if we assume that perfectly competitive behavior prevails in the bilateral exchange). In the case of a pure exchange economy of many consumers and many goods, an economist who knew all preferences and endowments could likewise determine equilibrium prices and quantities traded. Hayek in his 1926 article proposed that economists could extend the subjective value theory of the pure-exchange-economy model to a production-economy model although, as he noted, altogether different conditions apply to factors of production, and thus their prices require a separate explanation from consumer goods prices.⁹ In 1926 it remained to be rigorously established that a GE model of a production economy with variable factor proportions could indeed be solved for the prices of the factors of production.

Mises and Hayek differed from their opponents in the socialist calculation debate of the 1930s and 1940s over how much relevance the solution of a GE model would have for guiding factor pricing in a real-world non-market economy. Their opponents, the market socialists led by H. D. Dickinson, Oskar Lange, and Abba Lerner, offered the solving of a simultaneous-equation GE model, together with trial-and-error adjustment of prices toward equilibrium, as practical techniques for real-world planning.¹⁰ Hayek in 1926 made it clear that he was not asserting the feasibility of socialist planning in a complex real-world economy, because complete information about endowments and tastes could not be taken for granted in the real world. He cautioned: Each solution of the imputation problem must take into account the totality of all the complementary goods used in an economic system and all the needs for whose satisfaction products are employed. This necessity may render impossible the practical application of imputation to any comprehensive economic system.¹¹ In 1945 he emphasized the additional necessity for taking into account the totality of relevant possibilities for productive transformation of goods, and—in place of 1926’s cautious may—insisted against Lange and Lerner that the inescapable dispersion of knowledge does indeed render impossible the practical application of the GE-model-solving approach to directing any complex real-world national economy.

The 1926 article does, however, take a different view from the 1945 article with regard to the explanatory sufficiency of equilibrium theory. The two sides of the socialist calculation debate differed on how much relevance the GE approach has for understanding the factor pricing of a real-world market economy. In 1926 Hayek in Wieserian fashion offered subjective value theory as a satisfactory explanation for economic processes. But in 1945, Hayek cautioned that a GE model does not deal with the social process at all.

If GE theory is insufficient for understanding the market process, then it needs to be supplemented (or partly replaced) by a theory of price and quantity adjustment through entrepreneurial competition or something like it. Hayek would contribute to such a theory in his later essays The Meaning of Competition (1946) and Competition as a Discovery Procedure (1968). Israel Kirzner, in Competition and Entrepreneurship (1973) and later works, has further developed the Mises-Hayek entrepreneur-driven approach to the market process. Franklin Fisher has been the most important developer of a neoclassical mathematical analysis of equilibrating price adjustment.¹²

Hayek continued to consider Wieser’s omniscient-planner framework analytically useful for understanding and expositing the equilibrium conditions of a production economy, even though it did not illuminate the dynamics of the market process. Hayek used the simple economy device—exposition in terms of the optimizing principles to be observed by an omniscient and benevolent planner of a non-exchange economy—through much of The Pure Theory of Capital of 1941.

A puzzle remains. If Hayek’s position in 1926 was that a GE approach to value imputation was appropriate—or even obligatory—for subjective value theorists developing a model of a production economy, why did he think that his position clashed with that of marginal productivity theorists who stressed a need for competitive markets to price inputs appropriately in the real world? The 1926 Hayek was talking about a purely logical construct with production techniques assumed by the modeler, a Wieserian one-mind omniscient-planner simple economy.¹³ The non-Wieserian marginal productivity theorists such as Mises, and the 1945 Hayek, were talking about an economy of many entrepreneurial minds containing various bits of knowledge and conjectures about efficient production techniques. One could hold both positions, as indeed Hayek did in 1945 when he said that GE analysis was a useful preliminary for theoretical understanding although it disregarded everything that is important and significant in the real world for the economy’s guidance. When Hayek in 1926 declared it a mistake—tantamount to abandoning subjective value theory as a satisfactory explanation for economic processes—to claim that the determinants for the prices of the factors of production exist only in an exchange economy and that for this reason an imputation of value is not possible, he was mistakenly viewing Wieser’s logical construct as the only factor-pricing theory consistent with subjective value theory. Mises, in rejecting the usefulness of Wieser’s one-mind given-techniques approach for pricing factors in the real world of competitive bidding for factors, did not reject equilibrium constructs per se and certainly did not reject subjective value theory.

Could Hayek have had some other critic of Wieser in mind besides Mises? Joseph Schumpeter is a candidate. R. C. McCrea argued that Schumpeter in his Theorie der wirtschaftlichen Enlwicklung [Theory of Economic Development] held the position that, in McCrea’s words, the problem of determining factor incomes cannot be solved by an explanation of value phenomena, for income phenomena are price phenomena.¹⁴ That is, values in Wieser’s sense inhabit economic theories, whereas factor prices inhabit markets. But Hayek discussed Schumpeter by name elsewhere in the article, so it would have been odd not to identify him in these passages if Schumpeter’s was the position under criticism. For Hayek not to wish to criticize his new job supervisor Mises by name in print, on the other hand, would have been understandable. So it seems likely that Mises and not Schumpeter was the object of Hayek’s criticism.

Although Hayek’s position in the 1926 article can be partly reconciled with his position in the more famous 1945 article, it clashes with another of his well-known later articles—namely, Economics and Knowledge of 1937.¹⁵ In the 1937 article, Hayek insisted that the observation of how people gather and communicate information through markets is a crucial empirical element to understanding market processes. Hayek later characterized the article as a criticism of Misesian apriorism, a demonstration that there are sharp limits to the range of phenomena explainable by the pure logic of choice alone. But Mises’ position on socialism—that without actual market exchange by private owners an economy cannot appropriately price factors of production—is in an important sense more empirical and less aprioristic than Hayek’s 1926 position that factor pricing and optimal allocation can be understood by Wieser’s simple economy construct of a single decision-maker. Wieser’s construct is after all based on nothing but the pure logic of choice. Hayek in 1937 was thus also implicitly criticizing the Wieserian apriorism that he himself had embraced just over a decade earlier.

In what respects Hayek’s well-known later critique of socialist calculation (found in his book chapters and journal articles of the 1930s and 1940s, collected in Individualism and Economic Order) was complementary to Mises’ critique, and in what respects they were at odds, is a thorny question beyond the scope of our discussion here.¹⁶ However, it should be clear that Hayek’s 1926 article on imputation does not show that Hayek’s later writings on socialism are Wieserian rather than Misesian. It was written in 1923 before Hayek became more fully exposed to Mises’ thinking. And as we have just seen, Hayek in his later writings had clearly moved away from his 1926 position in crucial respects.¹⁷

If Wieser, and Hayek in 1926, wanted to impute values to factors of production in a model without market exchange, and therefore without market prices, in what units are the values to be measured? Samuel Bostaph has raised the units question with regard to Wieser’s equations, commenting: Assuming that they are subjective value units begs the question of how such subjective values have been objectified.¹⁸ Hayek in 1926, following Wieser, clearly regarded them as subjective value units, what other economists have called utils, objectified by the analyzing economist and assumed to be measurable and interpersonally addible.

Since Lionel Robbins’ 1932 book The Nature and Significance of Economic Science—heavily influenced by Mises’ methodological subjectivism—economists have commonly recognized that choice-theoretic utility does not require measurability or interpersonal comparability in order for the theorist to derive demand curves and explain market prices, and so those properties can be excluded by the principle of Occam’s razor. But Wieser clearly conceived of utility as something measurable and interpersonally comparable when he spoke of maximizing total social utility and when he endorsed progressive taxation on the grounds that a poor man values the marginal dollar of his income more than a rich man.¹⁹ Hayek acknowledged that Wieser considers the calculability of utility as basic for dealing with the whole problem, as can be hardly avoided.²⁰ If we follow Robbins in excluding calculability (measurability or interpersonal comparability) from choice-theoretic utility, then the need to assume calculability is a fatal flaw in Wieser’s concept of imputation without market prices. It is not surprising then that Hayek in his later writings would have moved away from Wieser’s concept.

The Problem of Interest Theory

Hayek’s 1927 essay On the Problem of Interest Theory is notable for the contrast between its critique of time-preference-based interest theory and Hayek’s own later position, represented in this volume by his 1936 essay Utility Analysis and Interest. In an appendix to The Pure Theory of Capital, Hayek essentially repudiated his 1927 article, commenting that in it he had used a particular approach without being aware of the illegitimate assumptions which it involves.²¹

Hayek in 1927 criticized what he called the "ad hoc assumption that future needs are invariably valued lower compared to the same needs at the present time."²² In 1936 Hayek took note of the progress which has been achieved by Professor Irving Fisher’s application of the modern apparatus of utility analysis.²³ In The Pure Theory of Capital, he would declare that the interest theory exposited by Fisher in The Theory of Interest (1930) was formally unimpugnable.²⁴ The consumer in Fisher’s theory has intertemporal preferences that, in an ordinary equilibrium, imply the subjective discounting of marginal future consumption relative to marginal current consumption. That an individual ordinarily (not invariably) values future satisfaction of needs below the present satisfaction of the same needs then becomes an implication of equilibrium and not an "ad hoc assumption. Subjective or psychical (as Hayek called it in 1936) discounting by consumers is the same concept that Hayek in 1927 characterized as systematic under-valuation of future needs," but without the connotation (found also in Böhm-Bawerk’s exposition) that the consumers are committing an error.²⁵

Hayek in 1927 lamented the way in which Böhm-Bawerk’s approach had excluded productivity explanations of interest. Fisher’s theory, alongside subjective discounting, incorporates productivity in the form of an intertemporal production possibilities frontier. Hayek in 1927 also criticized what he characterized as Böhm-Bawerk’s approach, in which the structure of production generally did not constitute an object of analysis but was taken as given.²⁶ This was an uncharitable view of Böhm-Bawerk’s approach. Böhm-Bawerk’s model establishes an equilibrium in which the degree of roundaboutness in production—a scalar measure of the time period that will elapse between the first application of input and the arrival of output under the chosen production plan—is endogenously determined along with the equilibrium interest rate. The equilibrium interest rate equates the marginal subjective discount rate to the marginal physical rate of return from extending the degree of roundaboutness (for example, the marginal percentage increase in lumber yield from delaying the harvest of a growing tree). Thus the length of the structure of production was an object of analysis. Böhm-Bawerk offered a pair of diagrams that depicted longer and shorter structures of production as larger and smaller sets of concentric rings. Hayek used triangular diagrams for the same illustrative purpose. Hayek’s diagrams were more readily adapted to other uses, particularly to illustrate the progress of a business cycle in which credit expansion fuels an investment boom.

In a footnote, Hayek offered an interesting rationale for the thesis that more roundabout production processes yield a greater physical output from given inputs: "The longer the time period in question, the greater the number of methods available for roundabout production, so that now production methods with a higher yield may become feasible, which previously have not been available."²⁷ Hayek’s rationale is best understood as an alternative way of stating the general proposition that the greater productivity of more roundabout production processes is a logically implied characteristic of the frontier of non-dominated choices. Any positive discount rate implies that a process taking longer to bear fruit must produce more fruit to be potentially preferred to a shorter process.

In his 1936 article Utility Analysis and Interest, Hayek used Fisherian diagrams to analyze an economy without psychical discount, but instead exhibiting Schumpeter’s assumption that, when anticipated future income equals present income, marginal future consumption is not discounted relative to marginal current consumption. Combining Schumpeter’s assumption with the assumptions that capital accumulation is the only source of higher future incomes (technical progress being assumed absent) and that the returns to capital diminish toward zero with a growing capital stock, Hayek not surprisingly found that the economy’s net saving, economic growth, and interest rate must all converge to zero. Hayek’s analysis of the convergence to equilibrium persuaded him that the marginal rate of return to capital accumulation (intertemporal transformation or productivity) was the dominant determinant of the equilibrium interest rate, with time-preference relegated to the minor role of determining the rate of saving along the path to stationary equilibrium.

Hayek drew heavily on his 1936 analysis in appendix 1 to 1941’s The Pure Theory of Capital, titled Time Preference and Productivity. But he had second thoughts about the dominance of intertemporal transformation over time-preference in determining the interest rate in Time Preference and Productivity: A Reconsideration (1945), reprinted as appendix 5 to the latest edition of The Pure Theory of Capital.²⁸

Capital Consumption and Saving

In his essays on Capital Consumption (1932) and Saving (1934), Hayek developed some long-run implications of his version of Austrian capital theory, the short-run (cyclical) implications of which he had sketched in Prices and Production.

In Capital Consumption, Hayek expressed concerns about the evident shrinkage of the capital stock in Austria. Such concerns were shared by other Austrian School economists, especially Ludwig von Mises, Fritz Machlup, and Oskar Morgenstern. Hayek cited Morgenstern’s statistical study of the problem for the Austrian Institut für Konjunkturforschung, published in 1931. Although he did not cite them, Hayek was probably aware of some of the other efforts, in Vienna and elsewhere, to diagnose and quantify the extent of Austria’s capital consumption. Ludwig von Mises collaborated with Engelbert Dollfuss and Edmund Palla on an official report analyzing the sources of weakness of the Austrian economy, anonymously published in December 1930, and Mises devoted his university seminar in winter 1931 to the topic of capital formation, maintenance, and consumption.²⁹ Nicholas Kaldor, who at the time was a Hayekian (or Robbinsian—he had graduated from the London School of Economics the year before Hayek arrived there), discussed Austria’s capital consumption in an article published in the Harvard Business Review in October 1932. Machlup would publish his research on the topic in the Review of Economic Statistics in 1935.³⁰

Hayek emphasized the trade-off between consumption today and consumption tomorrow (via saving and investment today): The physical quantity of consumer goods per capita can only be increased by consistently devoting a larger part of productive resources to capitalistic investment rather than to immediate consumption.³¹ One can illustrate the trade-off by drawing a production possibilities frontier between consumption and investment, as Roger Garrison has done in his important work developing Hayekian macroeconomics, particularly in the book Time and Money.³² Although the trade-off derives directly from the assumption of scarcity, and is today taken for granted by economists in the context of growth theory, Hayek noted that it is implicitly denied by all those economists who assume that the demand for capital goods changes in proportion to the demand for consumer goods.³³ The most prominent such economists in 1932 were the underconsumption theorists of economic depressions, including John Maynard Keynes in his Treatise on Money of 1930, which Hayek cited in this connection. Keynes amplified the underconsumption theme in his General Theory of 1936.³⁴ The assumption that consumption and investment move in the same direction over the business cycle (by contrast to the trade-off acknowledged in the analysis of long-run growth) has been fundamental to Keynesian macroeconomics up to the present day.

The business cycle theory of Hayek’s Prices and Production, as Garrison has emphasized, is a theory of an unsustainable investment boom. The boom is unsustainable because an unwarranted credit expansion finances the investment of inputs into processes too roundabout to be completed by the available voluntary savings (resources made available by voluntary abstention from consumption by income-earners). Hayek labeled the misallocation forced saving. As Garrison has pointed out, such a label unfortunately suggests that investors are compelling an immediate reduction in consumption by bidding resources away from consumers, or in other words that the economy remains on the same consumption-investment frontier even during the expansion phase of the business cycle. But an immediate reduction in consumption is difficult to reconcile with the fact that the interest rate, which represents the reward for saving rather than consuming, falls at the outset of the credit expansion. Garrison’s own development of the Mises-Hayek cycle theory, building on Mises’ identification of overconsumption during the early boom, provides greater coherence. It recognizes explicitly that credit expansion can temporarily push the economy beyond its sustainable frontier by financing over-full employment of labor and machines. The employment rates of laborers and machines (capacity utilization) rise above what economists today call their natural rates. During such a credit boom total investment can for a time increase somewhat (an increase that will later be regretted) without any reduction of current consumption.³⁵

In 1932 Hayek had published a Note on the Development of ‘Forced Saving’ that traced the history of the concept among earlier economists. In his 1934 encyclopedia article on Saving, included in this volume, he expanded his categorization of types of saving. Interestingly, the encyclopedia article criticized the use of the label forced saving for the case that Hayek himself had used it, calling such usage an instance of the misleading practice of treating the term [saving] as equivalent to ‘capital formation.’³⁶

The Debate with Frank Knight over Interest, Investment, and Capital

In a valuable 1970 restatement of neoclassical capital theory, Jack Hirshleifer formalized the three leading approaches to the relationship between capital goods and output.³⁷

(1) The simplest approach is Frank Knight’s Crusonia-plant model, in which capital consists of a single good that is physically homogenous with (and thus always trades 1:1 against) the consumption good. Imagine a huge fungus that can be eaten or left to grow. The single labor-using production process is the removal of pieces of fungus for immediate consumption. Production is instantaneous and simultaneous with consumption. The unconsumed remainder of the fungus grows naturally by itself (no labor required) at a known rate. The representative agent faces a single trade-off: higher consumption today versus a permanently higher flow of future consumption. Every future period’s sustainable consumption, consistent with neither consuming nor accumulating capital, is exactly equal to one period’s natural expansion of the remaining fungus.

(2) In the more complex second approach, the capital good is distinct from the consumption good. The price of the capital good relative to the consumption good can now vary. Production is not instantaneous but proceeds in two steps. In the current period, some share of inherited capital and labor are combined to produce the current consumption good, while the remaining share is used for investment—that is, to produce the end-of-period capital stock to be carried forward to the next period. In the next period the choice repeats with the carried-forward capital stock becoming the new inherited capital stock.

(3) In the most complex (Austrian) approach, the capital stock is not one good but a time-phased chain of intermediate goods—that is, goods-in-process at different distances from the dates at which they will mature into the consumable output good. The relative price of any particular capital good can vary against that of the consumption good, and also against prices of the earlier and further-along capital goods. There are arbitrarily many production stages, in the limit a continuous temporal stream of inputs, leading to consumption.

Hirshleifer treated the alternative approaches as employing different but equally legitimate sets of abstractions, making them useful for tackling different but equally legitimate analytical questions. Knight did not see it that way. The Crusonia-plant theory of capital he favored was correct and sound, whereas the Austrian capital theory, particularly as represented by Hayek’s 1931 book Prices and Production, was incorrect and useless, lacking the essentials of a sound theory of capital. In freeing economics from the incubus of Austrian capital theory, "the doctrines to be eliminated include all notions of any definite relation between quantity of capital and the ‘length of the production process,’