Economics as a Coordination Problem: The Contributions of Friedrich A. Hayek
By the same act with which a bank increases the circulating medium of a country, it issues into the community a mass of fictitious capital, which serves not only as circulating medium but creates an additional quantity of capital to be employed in every mode in which capital can be employed (Lord Lauderdale, cited in Hayek, Profits, Interest, and Investment, p. 190).
Friedrich A. Hayek was invited to deliver the special university lectures at the London School of Economics during the 1930-31 session. In the space of four lectures, subsequently published under the title Prices and Production, he analyzed the microeconomics of the "typical nineteenth century business cycle."*1 What he actually did in less than 150 pages was potentially even more far reaching: he integrated monetary theory and price theory more fully than had been done before.*2 The goal of current work in the integration of macrotheory and microtheory is essentially the same as Hayek's, and I believe progress would have been much greater had the contributions of Hayek not been lost sight of in the aftermath of the Keynesian revolution.
The other major monetary theorists of the day were similarly engaged in a reexamination of business cycle theory. Notable among them were J. M. Keynes, D. H. Robertson, and R. G. Hawtrey in England; Irving Fisher and Henry Simons in the United States; Ludwig von Mises and Gunnar Myrdal on the Continent; and, indeed, virtually the entire membership of the Austrian and Swedish schools. The Continental and British monetary theorists—intellectually Hayek straddled these groups—were more intent than the Americans on finding microfoundations for their theories. These divergent thinkers had a common purpose: to achieve the integration of monetary and price theory.*3 All were profoundly dissatisfied with the state of monetary theory in its many variants.*4
Hayek's business cycle theory, as presented to his English readers, was a blend of monetary theory, capital theory, and price theory. The lectures were first published in 1931. (His 1928 work on the monetary questions raised by the study of cyclical fluctuations was not translated into English until 1933 and was therefore unavailable to his English audience.)*5 Hayek viewed the price mechanism as a system of signals and the "economic problem" as one of social coordination. However, these conceptions were undeveloped until his contributions to the Socialist-calculation debates and the tetrad on economic coordination, of which "Economics and Knowledge" was the first.*6 These articles for the most part were written after his initial work on cyclical fluctuations was translated into English.
Many of the substantive issues treated by Hayek can be examined without detailed development of Austrian capital theory. At times, Austrian capital theory can be so idiosyncratic that communication with those unfamiliar with it is impeded, as Hayek recognized in later writings.*7 Reference to its characteristic propositions will be made insofar as this aids in the subsequent analysis.
To Hayek there was one essential fact about the business cycle to be explained by any theory of economic fluctuations:
The task [of understanding the business cycle] is made rather easier by the fact that there does exist to-day, on at least one point, a far-reaching agreement among the different theories. They all regard the emergence of a disproportionality among the various productive groups, and in particular the excessive production of capital goods, as the first and main thing to be explained.*8
Gottfried Haberler, more than a decade later, had a similar view, though he no longer accepted the particulars of the Mises-Hayek account. He stated that there are "two features which can be observed in every cycle, probably without exception, although they are not implied by our definition of the cycle." They are (1) that cyclical fluctuations in production and employment are "accompanied by a parallel movement of the money value of production and transactions," and (2) "that cyclical fluctuations are more marked in connection with the production of producer's goods than in connection with the production of consumers' goods."*9
Hayek expanded on the theories of Wicksell and Mises to account for a particular phenomenon—the so-called volatility of investment, or the "disproportionality" in production.*10 A process of inflation, or more precisely "cyclical expansion," was seen to consist of a reallocation of resources from industries producing consumer goods (or "final output") to those producing intermediate products. If a cyclical expansion starts from conditions of actual general resource unemployment, then expansion could, for a time, occur in all industries simultaneously. Hayek did not see the expansion as being uniform in practice, and his theory attempted to account for this nonuniformity. He developed a theory of the transmission mechanism of monetary disturbances, in which a change in the growth rate of the money stock depresses the complex of market interest rates below an equilibrium level.*11 Investment is stimulated, and through the augmentation in money expenditures on capital goods factor incomes are bid up, and finally the prices of consumption output rise. The complex allocation questions involved in this process are analyzed in Prices and Production.
Hayek did not treat the "excessive production of capital goods" as the only feature of a business cycle, though he considered it the most important one. Any tendency for the prices of all goods to change was treated as a secondary phenomenon. He did not view changes in price levels as a necessary condition for a business cycle.*12
Hayek perceived the necessity for explaining the emergence of resource unemployment in any theory of cyclical fluctuations. However, his approach in Prices and Production created a good deal of controversy: he started with an assumption of full employment, that is, long-run equilibrium. He did not, however, assume that the economy would remain at full employment. His object was to focus on certain processes that he thought would occur during any actual cyclical upswing as full employment was approached. Indeed, if Hayek's analysis of the impact of changes in the growth rate of the money stock is correct, these processes are inevitable in such expansions.
In considering Hayek's approach, a distinction must be made between the methodological assumption of full employment and the empirical assumption of full employment. The methodological assumption of full employment has three distinct advantages. First, it is consistent with the general approach in other areas of economic theory. Starting from full employment (i.e., equilibrium) avoids attributing adjustments that would occur in any case to the disturbance under consideration. Next, it focuses attention on the problem to be analyzed—that of unemployment—and compels the theorist to deduce the emergence of unemployed resources rather than beg the question by assuming what needs to be explained.*13 Finally, it minimizes the possibility of constructing a theory—believed to be general—that is contingent on the existence of unemployed resources throughout the economy.*14 A number of critics, including a friendly expositor of Hayek's ideas, have failed to note the aforementioned distinction and Hayek's own strictures on the subject.*15
Hayek was not the first to be attacked for assuming full employment. After Keynes set the example of reconstructing theories out of whole cloth, the classical economists became subject to this charge. While economists of the nineteenth century undoubtedly wrote as though there were ordinarily no unemployed resources, I know of no major figure who denied the existence of unemployed resources. Furthermore many attempted to explain the phenomenon, even though it was not their chief policy concern.*16
Although older generations of economists had virtually no theoretical explanation for large-scale unemployment not due to wage controls or other direct market interventions, contemporary economic theory does little more. The worker-search hypothesis leaves much to be desired. As William H. Hutt noted, Keynesian macrotheory has not produced a theory of unemployment at all; at best it offers a theory of aggregate demand.*17 Seldom has the ancient admonition regarding the throwing of stones in glass houses been more applicable than to critics of classical and allegedly classical treatments of unemployment.
In Prices and Production Hayek developed a "goods in process" model, in which land and labor services pass through successive stages of production until consumer goods finally emerge. A "stage" consists of some productive activity (for example, stamping out parts with a mold). An entire process of production consists of many stages with the goods processed at each stage called intermediate products. A change in the number of stages or the real-location of factors among stages is described as a change in the structure of production.*18 A consumer good yields all its services in a single period; by this definition Hayek circumvented the durable-goods problem.*19 His purpose was to emphasize an aspect of capital to which he felt insufficient attention had been paid.*20 This is the distribution of capital goods in time. In Hayek's metaphorical analysis, intermediate products flow as from tributaries into successive stages of production, and the value of intermediate products at any point in the stream is a function of time, f(t). The total value of the intermediate products is thus the integral of this function over a period r, equal to the length of the adopted process of production. Beginning at time x, the total value of the intermediate products is:
According to Hayek, the output of consumer goods (that is, the rate at which consumer goods appear) is a function of this time interval, f(x+r). Thus the Hayek model is expressed purely in terms of flows. The reasons Hayek did not attempt to translate the model into one couched in terms of both stocks and flows are several.
First, there is the inherent difficulty of including durable goods in a model of this type: "The different instalments of future services which such goods are expected to render will...have to be imagined to belong to different 'stages' of production corresponding to the time interval which will elapse before these services mature."*21 Hayek also initially assumed that at each stage the intermediate products are exchanged for money; there is no vertical integration. If durable capital goods were included, the individual must be assumed to be renting the capital goods from himself.*22 Finally, and most important, Hayek felt that in ordinary analysis too much attention is accorded to the durability of particular capital goods in explaining the effects of changes in the rate of interest on its value.
It is not the individual durability of a particular good but the time that will elapse before the final services to which it contributes will mature that is regarded as the decisive factor. That is, it is not the attributes of the individual good but its position in the whole time structure of production that is regarded as relevant.*23
The model of Prices and Production emphasizes the "time structure of production." Hayek employed "Jevonian Investment Figures" as a pedagogic device to illustrate the production process:
In figure 4.1 the horizontal side of the triangle represents the output of consumption goods f(x+r); the hypotenuse, the production function, f(t); and area, the integral that appears above.*24 The various stages of production—the horizontal segments of the diagram—consist of circulating capital (measured in value terms).
Was Hayek attempting to measure the aggregate capital stock in Prices and Production? The answer is in doubt. Although he did not produce such a measure, he created the impression that there is some basis for doing so. Thus, he referred to the "proportion between the amount of intermediate products... and the amount of... output" increasing as the duration of the production process increases.*25 He also described production as becoming more "capitalistic" as "the average time interval between the application of the original means of production and the completion of the consumers' goods increases."*26 While his use of "capitalistic" is loose, it suggests a capital-output ratio. What he actually measured was the value (in consumption output) to which an input will grow (that is, the compounded value of an investment).
Hayek's lack of precision in Prices and Production on this matter is notable on several counts. Even in terms of his own endeavor, he had no need to make use of aggregate concepts; indeed, he disparaged such aggregative procedures. He generally objected to the treatment of capital goods as homogenous. In The Pure Theory of Capital he pointed out that in principle it is impossible to measure capital stock by reference to an average period of investment, even if the rate of interest is already given.*27
I believe Hayek adopted a most "un-Hayekian" procedure in Prices and Production. He treated consumer goods as homogeneous if they are available in the same time period. Analysis applicable to the production of a particular consumer good is carried over to the aggregate output of consumer goods. He in effect adopted a one output-good model, though he was quite explicit at other places about the difficulties that arise from the existence of heterogeneous capital goods. Yet having treated consumption output as homogeneous, he also treated capital goods as inchoate consumer goods even though he specifically rejected this view at other times. In every other context Hayek was a trenchant critic of attempts to treat "capital" as a fund or as an amount of waiting.*28
Hayek was quite open about the deficiencies of Prices and Production. He considered himself fortunate to have received an offer to deliver the University of London lectures when he did. When the invitation arrived, he was convinced the central ideas of Prices and Production were correct, and he presented them in their simplicity.
In Prices and Production Hayek was almost offhand about the necessary conditions for equilibrium, or correspondence between the plans of consumers and producers: "The proportion of money spent for consumers' goods and money spent for intermediate products is equal to the proportion between the total demand for consumers' goods and the total demand for the intermediate product necessary for their continuous production."*29 In a lecture delivered less than three years after the University of London lectures, he was far more precise about the conditions for equilibrium. Describing his theory as "the 'Wicksellian' theory of crises" he remarked that it is "quite independent of any idea of absolute changes in the quantity of capital and therefore of the concepts of saving and investment in their traditional sense." Hayek continued:
The starting point for a fully developed [business cycle] theory...would be (a) the intentions of all the consumers with respect to the way in which they wish to distribute at all the relevant dates all their resources (not merely their 'income') between current consumption and provision for future consumption, and (b) the separate and independent decisions of the entrepreneurs with respect to the amounts of consumers' goods which they plan to provide at these various dates. Correspondence between these two groups of decisions would be characteristic of the kind of equilibrium which we now usually describe as a state where savings are equal to investments and with which the idea of an equilibrium rate of interest is connected.*30
For Hayek, the crucial question for business cycle theory was the mutual correspondence of the plans of savers and investors and those of consumers and producers. Before asking how an economy can be in disequilibrium, he sought to posit the theoretical conditions under which the system would be "in equilibrium."*31
An essential difference exists between a change in the structure of production resulting from increased saving and a change from an increase in the supply of money through bank loans to producers (at a lower rate of interest than previously was the case).*32 The alternative methods of financing the same increase in investment may be illustrated as a first approximation as follows:
If increased investment is the consequence of a change in the propensity to save, then savings (S) and investment (I) will be equalized at a market rate of interest, io, which will be equal to the natural, or equilibrium, rate, n. If the increased investment is financed, not by increased saving, but by an amount of bank credit being extended per period (equal to the difference between I1 and S1), then the loan rate, i will be below the natural rate, n. This natural rate is the rate of interest at which the flow of consumer and producer goods being produced and sold on the market is the same as the flow of expenditures on such goods. It is the rate at which the ex ante decisions of producers and consumers are mutually consistent.*33
This approach can, however, only be a beginning, for it is quintessentially macro in conception. Consequently, it ignores problems crucial to Hayek's analysis. The saving-investment analysis presented here involved the aggregation of the separate micro-investment functions. Hayek considered this procedure to be suspect in ignoring the shift effects on the separate investment demand curves of changes in the rate of interest (changes which merely move one down this constant aggregate investment function). Also according to Hayek:
Whether we are able to decide what savings and what investment are depends...on whether we can give the idea of maintaining capital intact a clear and realistic meaning... As soon, however, as one makes any serious attempt to answer this question, one finds not only that the concept of the maintenance of capital has no definite meaning, but also that there is no reason to assume that even the most rational and intelligent entrepreneur will ever in dynamic conditions be either willing or able to keep his capital constant in any quantitative sense, that is with respect to any of the measurable properties of capital itself.*34
To recast the saving-investment analysis in net terms, the concept of maintaining capital intact must be clarified; yet entrepreneurs do not necessarily keep the value of their capital intact. These difficulties with saving-investment analysis must be noted lest serious confusion arise as the result of what is said later.
An increase in the propensity to save, according to Hayek, will not initiate a cyclical expansion.*35 Rather, the phenomenon is useful in explaining the movement of relative prices during a credit expansion (that is, when voluntary saving has not changed). Whether an increase in saving is necessarily transmitted into increased investment is not considered. Like Wicksell, Hayek assumed that the financial sector does nothing to impede this process.*36
With an increase in saving, the impact of the concomitant decrease in consumption demand on producers who put the "final touches" on consumer goods is obvious—they find their receipts falling. However, along with the decreased receipts from current consumption there is an increase in funds available for future consumption output. In the final stages of production (that is, those nearest the production of consumer goods) the decrease in costs will not offset the decrease in receipts. Increased expenditures elsewhere maintain, even augment, factor costs. The price margin between output and input—from whence the return on capital emerges—narrows. Employment of funds in the final stage of production becomes less attractive than its employment at other stages. Funds will tend to be shifted to earlier stages of production at the same time new savings are being made available. The production stage preceding the final stage will also experience a narrowing of price margins, as will the stage prior to that, and so on. But more funds will be available in early stages, and at some point the tendency for a rise in the prices of the products of early stages (that is, those farthest in time from the production of consumption goods) will overcome the tendency for a fall in prices. Indeed, it will pay to begin new stages of production. An intertemporal reallocation of factor services occurs as a consequence of intertemporal changes in relative prices.*37
A narrowing of price margins leads to a shifting of capital among stages. The investment period increases.*38 A switch to the longer investment period occurs because it is not profitable to employ the older and quicker production methods. Investment projects with lower yields relative to prevailing rates of return acquire a positive present value. Economic losses are incurred on some old investments.
Hayek employed a discounted value of the marginal product theory of factor remuneration. Increased saving leads to a fall in the rate of interest. As a first approximation, the prices of land and labor services are bid up as the rate of interest falls. In this first approximation, land and particularly labor are treated as homogeneous.
To Hayek, a decrease in the interest rate may be translated into the proposition that both wage rates and land rents have risen relative to the prices of the products they help produce. As the interest rate falls, price margins between stages narrow, for real factor costs are increasing. The fall in prices of the products of the later stages of production is partly responsible for this narrowing of price margins.*39
Hayek's proposition must be distinguished from one particularly common misinterpretation. Were this a valid conceptual experiement, a fall in the rate of interest would, other things equal, increase the demand-price schedules for all assets. But formulating the experiment in ceteris paribus terms overlooks the mutatis mutandis conditions of the disturbance. The very process that affects the interest rate alters the size of the quasi-rents accruing to the heterogenous capital goods, which are combined according to particular expectations that will be falsified (by the assumed disturbance). This process is not to be relegated to the second order of magnitude, as is usually done.*40 In his criticism of Keynes's Treatise, Hayek observed:
Capitalization is not so directly an effect of the rate of interest; it would be truer to say that both are effects of one common cause, viz. the scarcity or abundance of means available for investment, relative to the demand for those means. Only by changing this relative scarcity will a change in the Bank Rate also change the demand price for the services of fixed capital.*41
Although fixed capital was excluded from the model of Prices and Production, the essence of the problem remained, that is, to demonstrate the effects of a change in intertemporal consumption demand—or an apparent change—and the consequent change in relative factor scarcities on the pattern of investment.
In terms of Hayek's investment diagram, there will be a narrowing of the base and a lengthening of the vertical side. Increased investment will be distributed both as an increase in resources in early stages and an addition of new stages. Only by this process are rates of return on all investments equalized. The investment must take the form of capital "deepening" rather than capital "widening" in the absence of any unemployed land and labor.
Ralph Hawtrey introduced the terms capital widening and capital deepening:
The process of which the capital equipment of a community is increased may take two forms, a "widening" and a "deepening." The widening of the capital equipment means the extension of productive capacity by the flotation of new enterprises, or the expansion of existing enterprises, without any change in the amount of capital employed for each unit of labour. The deepening means an increase in the amount of capital employed for each unit of labour.*42
Hawtrey concluded that a reduction in the rate of interest "is a sign that the widening of capital is insufficient to use up the resources of the investment market."*43
The process of capital deepening will be accompanied by capital widening in particular industries. This fact is implied in the Hayek's schematic analysis in Prices and Production—more is invested in early stages and less in late stages. Capital deepening might be defined as capital narrowing in the later stages, concomitant with capital widening in earlier stages.*44
Pace Hawtrey, capital deepening may be analyzed without reference to aggregate capital measure. One need only say that capital will be invested in different ways and in different goods (and outline in what general direction changes will be made). Hayek particularly did not need to refer to an aggregate capital measure. But as late as 1941 he spoke, as did Hawtrey, of "providing an unchanged number of workmen with more (or more elaborate) equipment."*45 This situation results when the investment period tends to lengthen as a result of the same conditions that bring about a fall in interest rates. As interest rates fall, however, the relative prices of various capital goods change, and new types of capital goods are substituted. The units will be incommensurable, and we cannot be sure what will happen to the value of the capital stock.*46 Indeed, Hayek pointed out that changes in the structure of production may be analyzed without reference to aggregate capital stock or an average period of investment.*47 We need only speak of marginal adjustments in the pattern of investment.
Hayek was more circumspect in The Pure Theory of Capital than in Prices and Production. He pointed out that "most of the changes in productive technique are likely to involve changes in the investment periods of different units of input to a different degree and perhaps in different directions."*48 A "change in the length of the process" could be employed to describe a situation where changes are "predominantly in one direction."*49 Though Hayek slipped into less precise and less guarded modes of expression, his final opinion was as follows: "For our present purposes we do not need to know whether a whole process as such is longer or shorter than another."*50 Nor apparently do we need to know anything about the aggregate capital stock or capital per head.
A critical question for Hayek is whether the effects of an increase in saving in a money economy are the same as in a barter situation. Of the effects of an increase in saving in a money economy, he observed: "The effect thus realized...is one which fulfills the object of saving and investing, and is identical with the effect which would have been produced if the savings were made in kind instead of in money."*51 Of course, if the saving were made in kind, the process would not occur in the same way, unless we are to assume money does not facilitate exchange! "It is self-contradictory to discuss a proces which admittedly could not take place without money, and at the same time to assume that money is absent or has no effect."*52 The justifiable conclusion of Prices and Production is that changes from the increase in saving lead to a new stable equilibrium. This is not necessarily the equilibrium that would obtain in a barter situation, assuming attaining equilibrium would be possible.*53
Hayek always treated the distinction between real and monetary changes as fundamental. Real changes (for example, a change in tastes, or in time preference) lead to changes in money expenditures, which in turn bring about equilibrium reallocations. On the other hand, he treated monetary changes (for example, a change in the money supply) as "self-reversing" and the temporary equilibria created by such changes as "inherently unstable."*54 This is the principal thesis of Prices and Production, where Hayek contrasted the effects of an autonomous increase in the money stock with those of an increase in the rate of saving.*55
The immediate impact of an increase in the money stock (which by assumption is an increase in bank credit) is the same for transactors as a shift in the savings function. The reason is that both disturbances generate the same signals in the loan market—the interest rates fall.*56 In the case of a shift in the savings function, the resulting changes involve a transition to a new equilibrium of the kind assumed in barter theory. If fluctuations in savings are violent and frequent, cyclical fluctuations may be observed in the absence of additional monetary disturbances. But Hayek contended that "experience provides no ground for assuming that such violent fluctuations in the rate of savings will occur otherwise than in consequence of crises."*57
The impression from reading Prices and Production is that in the absence of credit creation by the banking system, the real economy is fundamentally stable. The actions of transactors in the aggregate are not "perverse." Increases in the propensity to save with resulting changes in the pattern of investment are not suddenly reversed. Transactors do not save in order to lure unsuspecting entrepreneurs into (ex post) foolish investments, and then withdraw their savings at the crucial moment, thus to wreak havoc on interdependent investment plans.
Hayek focused on the market signals employed by savers and investors in making their decisions, which, though independently arrived at, are interdependent. He sought to demonstrate how a general inconsistency of plans could come about. As he phrased it, "We might have to distinguish between what we may call justified errors, caused by the price system, and sheer errors about the course of external events."*58 He had an hypothesis: "It seems...more likely that they[entrepreneurs] may all be equally misled by following guides or symptoms which as a rule prove reliable."*59
The signals followed by market decision makers are relative prices, precisely because this procedure has a demonstrated survival value. However, there are times when they prove unde pendable, that is, there are periods of economic disequilibria in which relative prices function so as to discoordinate the economic activity of savers and investors. This observation neither supports the contention that decentralized decision making is inherently "irrational" nor explains why the price system can misfunction. For these explanations one must turn to capital and monetary considerations. The problem was aptly phrased by an expositor of Hayek's theories:
Pricing...is...a continuous information-collecting and disseminating process, but it is the institutional framework that determines both the extent to which, and the degree of success with which, prices are enabled to perform this potential signalling or allocative function.*60
The assumption of Prices and Production is that the capital goods industries or earlier stages of production can expand (that is, switch to more "capitalistic" methods) only at the cost of present consumption. But in the case of a monetary expansion
this sacrifice is not voluntary, and is not made by those who will reap the benefit from the new investments. It is made by the consumers in general who, because of the increased competition from the entrepreneurs who have received the additional money, are forced to forego part of what they used to consume. It comes about not because they want to consume less, but because they get less goods for their money income. There can be no doubt that, if their money receipts should rise again, they would immediately attempt to expand consumption to the usual proportion.*61
Let us assume for the moment that consumers' command over resources again increases at some point. There is no reason to expect that, once consumers' incomes rise, there will be any change in the desired consumption-saving ratio out of that level of income. But the constancy of saving necessitates a return to the previous, "less capitalistic" method of production. It is this contraction process that Hayek explained most fully in a series of articles on the Ricardo effect, an effect that we will examine in the following chapter.
In terms of the investment diagram, the change to a narrower base and a heightened altitude will be reversed. But with the increased amount of money in circulation (assuming no endogenous deflation of the money stock as part of the cyclical decline in economic activity), the dollar value of goods sold in each stage will be greater. In other words, the standard long-run comparative static conclusions of the quantity theory are preserved. Indeed, once one admits that money is not neutral in the short run, and that there are relative price effects of sufficient duration that time-consuming production processes are altered, one is virtually compelled to accept the plausibility of the self-reversing features of a monetary disturbance; that is, one must so long as one believes that the absolute quantity of money is of no significance in the long run, and that the economy cannot be in neutral equilibrium for any conceivable set of relative prices.
Forced saving plays a crucial role in Hayek's malinvestment theory.*62 If saving, ex post, is greater than it would be at full employment, then so is investment. Hayek employed the theory of forced saving to demonstrate that under some conditions malinvestment would occur. The conclusion is hardly startling given the assumption that the rate of interest at which entrepreneurs can borrow is below the equilibrium (natural) rate of interest.
Yet a number of economists have had difficulty in perceiving how Hayek arrived at this conclusion. Hicks commented that, if forced saving takes place, "there has to be a lag of consumption behind wages." But he concluded that "obviously this lag is not acceptable."*63 Where did Hicks and Hayek part company?
Hicks continued to work on the problem. But their "parting of the ways was on the issue of the effects of money on decision making." Hicks argued subsequently that Hayek's analysis "was mixed up, in that exciting work, with monetary considerations that do not really belong."*64 No sense can be made of the arguments in Prices and Production if one adopts the view that the "monetary considerations...do not really belong." Indeed, the monetary considerations probably are the most original contribution of the analysis. One could disagree with the details of the production theory and still accept the monetary theory. Prices and Production simply does not "work" without the monetary analysis—at least not as a theory of economic fluctuations. This is what Hicks in fact demonstrated in "The Hayek Story" and why the Hayek story must be retold.
Without the doctrine of forced saving induced by bank credit expansion, there is no logic to the lag of consumption behind income. Hicks tried "sticky wages" (which are not in Prices and Production—at least not wages "sticky upwards"), but noted that "if one pursues this line of thought, one is led towards a theory which is more like that of Keynes, or perhaps of Robertson, than of Hayek."*65
Hicks finally tried to mend Prices and Production: "Suppose that one had not started with a 'credit expansion' but had begun with...a genuine increase in the propensity to save."*66 What follows is a tortured transformation of a theory of short-run disturbances into "a forerunner of the growth theory of more recent years."*67 It is no accident that only in this guise did Hicks contend "we can still make something of it [Prices and Production]."*68 For what Hicks outlined is a barter theory of growth-induced disturbances.*69
Forced saving occurs because entrepreneurs are given the means by the banking system to appropriate a larger portion of the economy's scarce resources. It must be emphasized that Hayek never dealt with a single "one shot" increase in the money supply.*70 In his response to Hicks's 1967 work, Hayek emphasized that the continual injection of money was in the form of business loans:
This process can evidently go on indefinitely, at least as long as we neglect changes in the manner in which expectations concerning future prices are formed. Whatever the lag between the impact effect of the new expenditures on a few prices affected immediately and the spreading of this effect to any other prices, the distortion of the "equilibrium" price structures corresponding to the "real" data must continue to exist. The extra demand which continually enters in the form of newly created money remains one of the constant data determining a price structure adjusted to this demand. However short the lag between one price change and the effect of the expenditure of the increased receipts on other prices, and as long as the process of change in the total money stream continues, the changed relationship between particular prices will also be preserved.*71
Two concepts come together in Hayek's theory: the Thornton-Wicksell analysis of the effects of a divergence between the loan and equilibrium rates of interest, and the doctrine of forced saving. Hicks tried to abstract from monetary considerations in his reconstruction of Hayek's argument:
The "reduction" of the market rate below the natural rate must therefore be interpreted as a disequilibrium phenomenon; a phenomenon that can only persist while the markets are out of equilibrium. As soon as equilibrium is restored, equality between market and natural rate must be restored. Thus there is no room for a prolonged discrepancy between market rate and natural rate if there is instantaneous adjustment of prices. Money prices will simply rise uniformly; and that is that.*72
It was to counteract this view that Hayek wrote the previously cited passage. Hicks's argument makes sense only in terms of a "one shot"increase in the money supply. There can be a discrepancy between the two rates of interest only if markets are out of equilibrium. Furthermore, equality "must be restored" once equilibrium is restored because that is how equilibrium is defined in the Wicksellian system! That "there is no room for a prolonged discrepancy between market rate and natural rate" is precisely the issue.*73 The inability of Hayek and Hicks to communicate on the issue stems from Hicks's contention that the monetary considerations "do not really belong." These considerations hold together the otherwise diverse strands of thought in the neo-Wicksellian analysis.*74 The process that Hicks summed up by saying "that is that" turns out to be the business cycle.
Notes for this chapter
Hayek, "Three Elucidations of the Ricardo Effect," Journal of Political Economy 77 (March/April, 1969): 282. Lord Robbins gave a brief account of the favorable impact of these lectures (Autobiography of an Economist [London: Macmillan & Co., 1971], p. 127). Hayek was almost immediately offered the long-vacant Tooke Chair at the London School of Economics.
The locus classicus for Hayek was Ludwig von Mises's Theorie des Geldes und der Umlaufsmittel (1912) (The Theory of Money and Credit, tr. H. E. Batson [Irvington-on-Hudson, N. Y.: Foundation for Economic Education, 1971]).
Hayek recognized the similarity of enterprise among these figures. In a letter to Milton Friedman he remarked: "We all had similar ideas in the 1920s. They had been most fully elaborated by R. G. Hawtrey who was all the time talking about the 'inherent instability of credit' but he was by no means the only one" (Friedman, Optimum Quantity of Money [Chicago: Aldine Publishing Co., 1969], p. 88n).
Irving Fisher does not fit this analysis as well as the others. His work on money is largely a restatement of the quantity theory, against which many of the other figures were reacting (Roy Harrod, Money [London: St. Martin's Press, 1969], p. 27).
Hayek, Monetary Theory and the Trade Cycle (New York: Augustus M. Kelly, 1966).
Hayek, Individualism and Economic Order (Chicago: University of Chicago Press, 1948), pp. 33-56. The other articles, all reprinted in this same volume, are "The Facts of the Social Sciences," "The Use of Knowledge in Society," and "The Meaning of Competition."
Hayek, Profits, Interest, and Investment (New York: Augustus M. Kelley, 1970), pp. 6-7.
Hayek, Monetary Theory, p. 54 (emphasis original).
Gottfried Haberler, Prosperity and Depression, 3d ed. (Lake Success, N. Y.: United Nations, 1946), pp. 277-78. Haberler was particularly careful to note that each business cycle is a unique historical event, and that there are dissimilarities among cycles. He only argued that there are certain characteristic features of cycles. This is an historical question, and is treated as such by virtually all concerned (ibid., pp. 274-76).
Obviously Spiethoff influenced all Continental economists studying the business cycles at this time. Although both Mises and Hayek emphasized the monetary factors operating causally in the cycle, both mentioned Spiethoff in their works,though often to disagree with him. Also both contributed articles to Spiethoff's Festschrift (Schumpeter, History of Economic Analysis [New York: Oxford University Press, 1954], pp. 815-17, 1126-28).
It would be more precise to speak of changes in the rate of growth of credit, as these were the analytically important changes in Hayek's Misesian (or neo-Wicksellian) theory. Thus, while he was chiefly concerned with changes in bank deposits or credit money, he did not treat the amount of credit as rigidly determined by the stock of money, even though he viewed economic fluctuations as being initiated by monetary disturbances.
Hayek, Monetary Theory, pp. 103-6. On the attitude of the modern Austrian school toward price levels, see Schumpeter, History, pp. 701n, 1089, 1095.
See Hayek's own justification of this procedure in Prices and Production, 2d ed. (London: Routledge & Kegan Paul, 1935), pp. 32-36.
Hayek felt that Keynes, for one, was guilty of this error in The General Theory. Hayek at first appraised that work as providing theorists with "the economics of abundance." Twenty-five years later, he argued that Keynes's analytical framework was one in which "the whole price system [was] redundant, undetermined and unintelligible" (Sudha R. Shenoy, ed.,A Tiger by the Tail [London: Institute of Economic Affairs, 1972], p. 103).
Haberler, Prosperity and Depression, p. 63n; and Fritz Machlup, "Friedrich von Hayek's Contributions to Economics," Swedish Journal of Economics 76 (1974): 506.
One example is to J. S. Mill's exposition of the effects of a sudden conversion of circulating into fixed capital (Principles, pp. 93-97). Mill's analysis is reflected in Hayek's monetary explanations of the following century.
W. H. Hutt, The Theory of Idle Resources (London: Jonathan Cope, 1939), p. 15.
Hayek, Prices and Production, pp. 36-40. Throughout Hayek was concerned with the allocation of consumption over time; he did not consider the effects of a change in tastes for consumer goods to be consumed in a given time period. Thus, at times, "homogeneous consumption services" could be substituted for "consumers' goods."
The durable-goods problem was of particular importance for the Austrian school. Wicksell was the first to point out that the durable-goods problem is identical with the Marshallian joint-supply problem (Wicksell, Lectures on Political Economy, ed. Lionel Robbins [London: Routledge & Kegan Paul, 1935], I:260). D. K. Benjamin and Roger Kormendi rediscovered this fact in "The Interrelationship between the Markets for New and Used Durables," Journal of Law and Economics 18 (October 1974): 381-401. See also Hayek, The Pure Theory of Capital (Chicago: University of Chicago Press, 1941), pp. 66-67.
Hayek, Prices and Production, pp. x-xii; see also Hayek, Pure Theory of Capital, pp. 46-49.
Hayek, Prices and Production, p. 40n.
Ibid., pp. 40-41n.
Hayek, Pure Theory of Capital, p. 48.
Hayek, Prices and Production, pp. 38-42. Hayek said that Jacob Marshak suggested the term "Jevonian Investment Figure." Jevons as well as Wicksell and Ackerman used similar figures (ibid., p. 38n). The similarity to Wicksell's approach is particularly striking (Wicksell, Lectures, 1: 151-54). See also W. Stanley Jevons, Theory of Political Economy, ed. R. D. C. Collison Black (Baltimore: Penguin Books, 1970), p. 231.
Hayek, Prices and Production, pp. 41-42.
Ibid., p. 42.
Hayek, Pure Theory of Capital, pp. 199-200.
Ibid., pp. 3-13, 93-94. Much of this reflects the still very strong influence of Böhm-Bawerk's thinking on Hayek.
Hayek, Prices and Production, p. 46.
Hayek, "Price Expectations, Monetary Disturbances, and Malinvestment," Profits, Interest, and Investment, pp. 153-54 (hereafter, "Price Expectations"). In pointing out that his theory was "quite independent of any idea of absolute changes in the quantity of capital," Hayek noted that his theory did not depend on being able to measure the capital stock or (to deal with the question that concerned him at this point) on giving any determinate meaning to the maintenance of capital.
Hayek was following his own dictum in "Economics and Knowledge" that "before we can explain why people commit mistakes, we must first explain why they should ever be right" (Individualism, p. 34).
The two methods involve hypothetical experiments: the increase in the propensity to save assumes a constant money supply, and the increase in the money supply assumes a given propensity to save. Hayek took into account the complexities introduced by an elastic supply of trade credit (Prices and Production, pp. 115-18).
Hayek, "Price Expectations," pp. 152-154.
Ibid., p. 152.
Hayek placed no emphasis on the interest elasticity of saving. "The factors which affect an individual's willingness to save are the regularity and certainty of his income, the security of the investment opportunities available to him, and the possibility of investing in his own business... It seems that in the short run the willingness to save varies very little and that it is particularly not much affected in the aggregate by changes in the rate of interest" (Profits, Interest, and Investment, p. 169). The inclusion of this essay in Encyclopedia of the Social Sciences in 1935 suggests that it must have represented the consensus of the profession, for new theories are not usually introduced in such articles. The dependence of saving on income is referred to as "self-evident commonplace" (ibid., p. 53n).
To Wicksell, the adaptation of entrepreneurs to a changed propensity to consume is as a rule "of secondary importance in comparison with the main phenomenon," changes in the struture of production. He did assume, however, an "adaptability and a degree of foresight in the reorganization of production which is far from existing in reality" (Lectures, 2: 193).
This analysis is from Hayek, Prices and Production, pp. 75-77. An application of resources to the early stages would allocate circulating capital to more productive operations. Production for consumption would take longer (measured from the first application of labor and land services). Fewer consumption goods would be available immediately and more would be available ultimately. This is precisely what consumers desire when they increase their propensity to save. If net value productivity is involved in extending the number of stages, the output of consumer goods will eventually increase (see note 39 below).
An investment period is "the interval between the application of a unit of input and the maturing of the quantity of output due to that input" (Hayek, The Pure Theory of Capital, p. 69). The concept is most applicable to what Frisch called a "point input-point output" model. For a continuous input-point output, or point input-continuous output model, Hayek used joint-demand analysis (for factors in the first case) and joint-supply analysis (for the services in the second case) (ibid., p. 67). Hayek did not give the Frisch citation there.
Hayek, Prices and Production, pp. 79-83. According to Hayek, the discounted value of the marginal product of nonspecific factors will increase for a second reason: the superiority of "roundabout," or "capitalistic," methods of production, which insures that total output of consumer goods will increase once the new process has been completed. This controversial and typically Austrian proposition is not essential for what follows, though one wonders why investment would ever become "more capitalistic" if this were not true.
"A change in the bank-rate is not calculated to have any effect (except, perhaps, remotely and of the second order of magnitude) on the prospective real yield of fixed capital" (J. M. Keynes, A Treatise on Money, 2 vols. [New York: Harcourt, Brace & Co., 1930], 1:202).
Hayek, "Reflections on the Pure Theory of Money of Mr. J. M. Keynes," part 2, Economica 11 (February, 1932): 25; and Gerald P. O'Driscoll, Jr., "Hayek and Keynes: A Retrospective Assessment," Iowa State University, Staff Paper no. 20, 1975), esp. pp. 24-26.
Hawtrey, Capital and Employment, 2d ed. (London: Longmans, Green & Co., 1952), p. 31; see also Hayek, Pure Theory of Capital, p. 286.
Hawtrey, Capital and Employment, p. 36. Wicksell spoke of the "breadth" and "height" of capital as capital accumulation occurs; he was apparently borrowing Ákerman's terminology (Lectures, 1:266).
Hayek, Pure Theory of Capital, pp. 286-87.
Ibid., p. 286.
It is questionable whether we want to measure market value when we try to measure capital. A number of writers (for example, Robert Dorfman and Abba Lerner) questioned this procedure, noting that we do not follow it with other factors, and that, if we did, some results would be paradoxical. For instance, were the demand for labor to be inelastic, an increase in supply would diminish the value of labor employed in production (Kirzner, Essay on Capital [New York: Augustus M. Kelley, 1966], p. 135).
Hayek, The Pure Theory of Capital, pp. 69-70, 76-78.
Ibid., pp. 69-70.
Ibid., p. 70.
Ibid., p. 76.
Hayek, Prices and Production, p. 53.
Hayek, Pure Theory of Capital, p. 31.
Two points are implicit here. First, in a barter world, there is no medium of exchange by definition. Thus one commodity will be demanded in equilibrium in a money economy, that will not be demanded in a barter economy. Second, no alternative mechanism for attaining equilibrium is specified in barter constructions (Pure Theory of Capital, p. 31). See Ludwig von Mises, Human Action, 3d ed. (Chicago: Henry Regnery Co., 1963), pp. 249, 398-99, 416-19.
"Monetary changes are...in a peculiar sense self-reversing and the position created by them is inherently unstable.For sooner or later any deviation from the equilibrium position—as determined by the real quantities—will cause a swing of the pendulum in the opposite direction" (Hayek, Pure Theory of Capital, p. 34).
He made the simplifying assumption that the entire increase in the money stock took the form of "credits granted to producers"; this was "the case most frequently to be encountered in practice" (Hayek, Prices and Production, p. 54). In fact Hayek's assumption is not unrealistic, even today. Most loans granted by commercial banks are for productive purposes. Consumer and personal loans are of growing importance in commercial bank portfolios (from approximately 18% in 1947 to approximately 25% in 1970). But even for this category of loans, shifts in consumer spending are induced by interest rate changes. For the portfolio statistics, see Colin D. Campbell and Rosemary G. Campbell, An Introduction to Money and Banking (New York: Holt, Rinehart & Winston, 1972), pp. 84, 89.
Wicksell demonstrated that the following analysis is no less valid if the changes in the money stock are induced by changes in the natural rate of interest, the money rate being constant (Lectures, 2:202-18). Hayek at one point chided Mises for emphasizing the autonomous nature of changes in the money stock (Monetary Theory, pp. 148-52). Machlup was too generous in crediting Hayek with an amendment to Wicksell in this issue. Machlup would have Hayek correcting Wicksell by pointing out that a cumulative process can be initiated by a rise in the natural rate of interest (Machlup, "Friedrich von Hayek's Contributions to Economics," p. 501). Yet Wicksell treated this case as typical (Lectures, 2:205).
Hayek, "Price Expectations," p. 143.
Ibid., p. 141. For a similar argument, see Ludwig von Mises, "'Elastic Expectations' and the Austrian Theory of the Trade Cycle," Economica, n.s. 10 (August 1943): 252.
Hayek, "Price Expectations," p. 141.
Shenoy, A Tiger by the Tail, p. 8.
Hayek, Prices and Production, p. 57.
Haberler called the class of theories of which Hayek's is an instance "Monetary Over-Investment Theories" (Prosperity and Depression, p. 33). Malinvestment is both illuminating and descriptively more accurate. The cyclical process in Hayek's work is generated when appropriate investments (given the equilibrium rate of interest) are made. Whether in some sense more capital is purchased with the increased investment expenditures is of secondary importance.
Hicks, "The Hayek Story," in Critical Essays in Monetary Theory (New York: Oxford University Press, Clarendon Press, 1967), p. 208.
Hicks, "A Neo-Austrian Growth Theory," Economic Journal 80 (June 1970): 277. Sir John is speaking here about Prices and Production. It must be pointed out that he amended his views in successive reassessments of Hayek and the Austrian school. But I believe that Hayek and Hicks still disagree about the role of Hayek's monetary considerations.
Hicks, "The Hayek Story," p. 210. Hicks characterized this approach as downright "un-Hayekian."
Ibid., p. 211. Streissler, in reinterpreting Hayek, follows a path similar to that taken by Hicks (Roads to Freedom [New York: Augustus M. Kelley, 1969], pp. 245-85).
Hicks, Critical Essays, p. 211.
Ibid., pp. 211-15. Hicks subsequently became intrigued with Ricardian analysis of the effects of excess investment in fixed capital, and his approach to capital theory must be seen as part of a general retrogression toward Ricardian macroanalysis. It is particularly unfortunate that Hicks subtitled his book "A Neo-Austrian Theory." It could be more aptly described as "neo-Ricardian" (Capital and Time, pp. 97-99). On the radical dissimilarities between the neo-Ricardian macro approach and the Austrian micro approach to capital theory, see Ludwig M. Lachmann, Macro-economic Thinking and the Market Economy.
"Of course, if the expenditure of the additional money in investment were a single non-recurrent event, confined to a single month, the effects would be of transient character" (Hayek, "Three Elucidations," p. 279).
Ibid., p. 280. Even if a given rate of increase in the money supply, and hence prices, came to be correctly anticipated, relative prices would not be at their equilibrium values. This situation would then not be one of equilibrium. More will be said on this in the next chapter.
Hicks, "The Hayek Story," p. 206.
Hicks emphasized that prices in Prices and Production were "perfectly flexible, adjusting instantaneously, or as nearly as matters" ("The Hayek Story," p. 206). This flexibility is not a necessary condition for Hayek's theory, and I do not think he assumed any such thing. Even if prices were perfectly flexible, his conclusions would not change. But Hicks did not appreciate the analogy Hayek employed ("Three Elucidations," pp. 281-82).
Hayek and Hicks apparently parted company over their interpretations of Wicksell. Hicks interpreted Wicksell's system as being in neutral equilibrium ("The Hayek Story," pp. 205-7). I believe that most students of Wicksell would have to disagree with Hicks's interpretation. In any case, the difference in interpretations may be reduced to disagreement over the importance of monetary analysis.
Chapter 5. The Ricardo Effect
End of Notes
Return to top