[발췌: Hayek's] Profits, Interest and Investment & Other Essays

지은이: F.A. Hayek
출처: Profits, Interest and Investment, and Other Essays on the Theory of Industrial Fluctuations.
자료: http://mises.org/books/profits_interest_hayek.pdf

※ This is a reading note with excerpts taken, and with personal annotations and remarks added, in trying to understand the above text, so visit the source links to see the original.

* * *

OF WHICH:

Chapter 3_ The Maintenance of Capital

(Reprinted from Economica, Vol. II. (New Series) August 1935)

1. The nature and significance of the problem

It is not likely that in the whole field of economics there are many more concepts which are at the same time so generally used and so little analysed as that of a “constant amount of capital.” But while in the investigation of the effects of nearly any change this is almost without exception treated as a given datum, the question what this assumption exactly means is rarely asked. To most economists the answer to it has apparently seemed so simple and obvious that they have never attempted to state it. In consequence the difficulties involved have hardly ever been realised, still less have they been adequately investigated.

  The difficulties of the problem would undoubtedly have been realised if economists had been more generally conscious of its importance. But [:]

• although even in the analysis of a stationary equilibrium, the inclusion of the “quantity of capital” among the determinants of that equilibrium means that something which is the result of the equilibrating process, is treated as if it were a datum,[2] 
• this confusion was made relatively innocuous by the essential limitations of the static method, which while it describes the conditions of a state of equilibrium, does not explain how such a state is brought about.

[2] Cf. K. Wicksell, ^Lectures on Political Economy^, Vol. I, p. 202. London, 1934.

It was only with the more modern attempts to make the descriptions of the conditions of equilibrium the basis of the analysis of the dynamic processes, that the exact meaning of this assumption became of serious consequence. But we need not go far beyond the description of one state of equilibrium to see why it matters. [:] 

• Even the very simplest “problem of variation” as the effect of a shift in demand can only be answered on the basis of a definite assumption as regards the “supply of capital,” and the usual answer shows immediately how problematic is the assumption on which it rests.
• ※ What is such a usual answer like? 

  The idea implicit in the discussion of problems of this sort is that there is a quantitatively determined fund of capital, which can be distributed and redistributed in any way between the different lines of production without changing its aggregate value.[1] But [:]

• has this assumption of a perfectly mobile capital fund any definite meaning? 
• And if so, what determines whether this value remains constant, [?] 
• what are the conditions under which it will remain constant, [?]
• how is it measured, [?] 
• and by what is this constant magnitude represented? 

Certainly not by the concrete capital goods. It is the essential difference between them and factors like labour or land that they will not remain physically the same when prices change, but that the physical composition of the aggregate called capital will change in consequence of any change in the data. Any of the problems of variation with which we are concerned will involve what is commonly called a transformation of capital into other forms. It is not the list of the different pieces of individual capital goods which is assumed to remain unchanged, when we speak of a constant supply of capital, in the same way as we assume that the total supply of labour or land is composed of elements of the same kind, when we regard its supply as unchanged.

[1] This common assumption has recently been stated with particular emphasis by Professor Pigou in his latest article on the subject to be quoted later. He finds the basis of his deductions “in the concept of capital as an entity capable of maintaining its quantity while altering its form and by its nature always drawn to those forms on which, so to speak, the sun of profit is at the time shining." ^Economic Journal^, June, 1935, p. 239.

  Even if a definite meaning can be attached to the statement that the value of the capital goods has an existence independent of the capital goods themselves, it does not help us in any way to explain why, or under what conditions, this aggregate value should remain constant, when conditions change. There can be no doubt that the value of at least some of the existing capital goods will be very materially affected by almost any conceivable change. The question then is why should the capitalists, in spite of this change in the value of their concrete capital goods, be able and willing, by an appropriate adjustment of their investment activity, to maintain the total value of their possessions at exactly the same figure as before the change.

• Is there any justification for considering this in any sense to be the “normal” behaviour of the entrepreneurs, or is it even conceivable that they can, under all circumstances, behave in such a way? 
• Must we not rather assume that under some conditions it will be impossible for the capitalist to maintain the value of his capital constant, while under others it could hardly be considered in any sense “normal” if he did no more than this?

  (1) What, then, is that neutral state in which the owners of capital are supposed to take a merely passive attitude performing no new saving or dissaving? This question, already of considerable importance when we try to trace the consequences of any other change, becomes of special importance when we turn to the autonomous changes which can be said to originate on the side of capital. Then already (2) the initial problem, of what can be regarded as such an autonomous change on the side of capital, depends on our definition of that neutral state.

1. In the usual discussion of this sort of problem, it seems to be generally implied that there is a clear line between the normal process of maintaining and replacing the existing capital, and any net addition to it. 
2. It is assumed that it is always possible to decide in an unequivocal way whether capital remains constant, increases, or decreases, and that there are typical phenomena connected with each of these processes which, at least conceptually, can be clearly separated. 
3. Indeed, so long as all the other data remain constant, no difficulty arises in this connection. But as soon as one tries to apply these categories to a world where things are changing, all these alluringly simple concepts become dependent in more than one way on exactly what is meant by a given stock of capital. 
4. It is impossible to define income (or “earnings”) and therefore savings before one has separated from the gross produce those quantities which are required to “maintain” capital, and it is equally impossible to say what are additions to the stock of real capital before one has shown what capital goods are required to make up for current depreciation.

• ※ (4): 자본을 이전의 크기 그대로 유지하는 데 필요한 수량들을을 총 산출물에서 분리해내지 않으면 소득(혹은 ‘새로 얻은 무엇’)을 정의할 수 없고, 따라서 저축도 정의할 수 없다. 마찬가지로 당기의 자본소모(감가상각)을 보충하는 데 들어가는 자본재들이 어떠한 것들인지 밝히지 못하면 실물자본의 저량이 증가했다는 게 무엇인지 말할 수 없다. 
• ※ 이 대목만 보면, 총 산출물에서 자본소모를 공제하고 남은 것을 그냥(즉 본래의) ‘소득(income)’으로 부를 것인지, 아니면 ‘순소득(net income)’이라고 부를 것인지가 분명하지 않다고 볼 수 있다.

  If it should prove that serious difficulties arise in this connection, this would be a great importance for the theory of the trade cycle. (ㄱ) In the course of the last generation theorists in that field have tended more and more to agree that industrial fluctuations consist essentially in alternating periods of accumulation and decumulation of capital with all their typical consequences. More recently there has seemed to be considerable unanimity in seeking for causes of these fluctuations in the accumulation of capital in conditions which make the movement of investment to a certain extent independent of that saving. (ㄴ) But although the quantitative relationship between saving and investing, their correspondence or non-correspondence, has become the central point of attack, it seems that all the writers who have made use of that approach have failed to make clear exactly what they meant either by saving or investing. (ㄷ) This would have required a careful definition of that neutral position (a) in which neither positive nor negative saving were made, because all the income and no more than the income was consumed, and (b) in which capital goods were produced in exactly that quantity and composition that is required to keep the stock of capital intact.

• ※ (ㄱ)이지만 (ㄴ)이라서 (ㄷ)이 중요하다는 흐름: (ㄱ) 지난 한 세대 동안 경기순환을 자본의 축적과 축적의 후퇴가 서로 교대하는 현상으로 이해하는 경향이 강해졌고, 최근 이러한 자본축적의 변동 원인을 투자가 일정 정도 저축과 따로 놀게 되는 조건에서 찾는 견해가 보편화되고 있다. (ㄴ) 하지만 이 모든 분석이 저축과 투자를 정확히 정의하지 않았다.
• ※ (ㄷ) 자본이 변동하지 않는 중립 상태를 면밀히 정의해야 한다. 이 중립 상태란, (a) 첫째 저축이 양의 값도 음의 값도 아닌 상태다. 왜냐하면 [이 상태에서는] 소득은 전부 소비되고 딱 소득만큼만 소비되기 때문이다[그래야 저축=0이 된다는 얘기일 것]. {여기서도 하이에크는 소득 측면에서 본 (소비를 빼고 남은) ‘저축’과 생산 측면에서 본 ‘투자 수요’를 같은 선상에서 취급하는 인상을 줌} 그리고 (b) 둘째, 자본재가 자본 저량을 그대로 유지하는 데 필요한 크기와 구성대로 생산되는 상태다. 

  But while it cannot be denied that modern trade cycle theorists (including the present writer) have lamentably failed to provide a concept which is indeed indispensable if their deductions are to be have a clear meaning, it seems that they are not the only group of economists who have been deceived by the apparent simplicity of the problem. Even more than this group one should expect the writers on the income concepts to have provided a clear answer. And even if the general discussions of the income concept should prove disappointing, one would certainly feel entitled to expect a definite answer in the discussions of business profits, since profits of all things can evidently only be defined as the excess of the total business assets over the equivalent of the capital invested at the beginning of the period. But while we find in general discussions of the income concept, particularly in the writings of Professor Irving Fisher,[1] at least some references to our problem, the standard works on business profits, like Professor F.H. Knight's Risk, Uncertainty, and Profits, or Mr. C.O. Hardy's Risk and Risk Bearing, are almost bare of any reference to the problem.


2. Professor Pigou's treatment

In view of these circumstance the solitary attempt first made by Professor Pigou a few years ago in the third edition of his Economics of Welfare deserves far greater attention than it has actually received. Although Professor Pigou discusses the problem for a special purpose, the definition of a national dividend, he raises most of the problems that need investigation. And it is only an additional reason for gratitude to that distinguished author that he had apparently not felt satisfied with his first attempt towards an answer and that he has in the fourth edition of the same work given us an entirely new version of his solution.[1] In the third edition he had still considered that the problem could simply and only be answered by "employing money values in some way as our measuring rod," but that we cannot employed crude money values but must introduce "corrections."[2] Accordingly, he defined as constant quantities of capital collections of capital goods, whose aggregate money value, divided by an appropriate index number of general prices, remained constant. After making some further allowances for changes in this magnitude caused by changes in the rate of interest, which need not concern us here, he proceeded to apply this definition to cases where the value of capital goods "is destroyed through a failure of demand or through a new discovery which renders existing instruments obsolete,"[3] and suggests, consistently with the criterion adopted, that in such a case it is convenient, although arbitrary, to say that the stock of capital has decreased (or in the reverse case of an expansion of demand that it has increased).

  In the fourth edition of the Economics of Welfare that entire chapter is rewritten. Although no explicit explanation is given why the former answer has been abandoned, it is fairly evident from the nature of the change made, and the general shift of emphasis, that the aim has been to make the decision more dependent on the reasons why it is thought desirable to maintain capital intact. Considerations relating to the constancy of the money value of capital which formerly occupied the chief place are now relegated to a subordinate place, and mentioned only to show that if, all other things remaining the same, in consequence of a contraction in the supply of money, money values all round are substantially reduced, and the money value of capital contracts along with the rest, nobody would consider this as a decrease of capital. The second case mentioned is again that of the effects of a change in the rate of interest on the value of capital, and it is decided (apparently on the assumption that such a change will not affect the return from the existing capital goods) that such changes are not to be considered as changes in the quantity of capital, so far as the estimation of the national dividend is concerned. But then the effects of changes in demand and of inventions are taken up, and here the decision is now the reverse from what it was before. In Professor Pigou's opinion, "we may say quite generally that all contractions in the money value of any part of the capital stock that remains physically unaltered are irrelevant to the national dividend; and that their occurrence is perfectly compatible with the maintenance of capital intact." [1] The same applies to actual destruction of the capital goods by "acts of God or the King's enemy," where the distinguishing criterion is that they are not incidential to the use of them, or as I shall suggest instead, because and in so far as they cannot be foreseen.[1] But all other physical deterioration in the capital stock, such as the ordinary wear and tear of machinery and plant, destruction by accidents like fire and storm, in so far as these are, as in the case of ships, incidential to their use, ought to be made good by adding "to the capital stock something whose value is equal to that which the machine, had it remained physically intact, would have now."[2] By this are explicitly excluded all the losses in value, which are not due to physical deterioration, but are due to causes like changes in demand or the new inventions mentioned before, to which Professor Pigou now adds foreign competition. This is expressly confirmed by the concluding sentence of the paragraph, in which, summarising the result, he says, "Maintenance of capital intact for our purposes means then, not replacement of all value losses (not due to acts of God and the King's enemies), but replacement of such value losses as are caused by physical losses other than the above."[3]

  We shall have to discuss these cases in greater detail in the systematic part of this article. Here only one difficulty arising out of Professor Pigou's answer may be mentioned, since it opens up the vista on a set of problems which he has left untouched. What seems most surprising in his classification is that obsolescence, even where it is foreseen at the time a capital good is produced or acquired, and where accordingly the investment is made in the expectation that the product will cease to have value long before it has physically decayed, should not have to be made good in order to maintain capital intact. This means that gradually all existing capital may be squandered, in the ordinary sense, by erecting durable structures for very transient purposes, and replacing them, when they become obsolete after a short time, only by capital goods of a value equal to that which they still possess after their temporary utility has passed. Surely there must be some cases where obsolescence, a decrease of usefulness of a capital good not connected with any change in its physical condition, has to be taken into account. And apparently, the distinction must somehow be based on whether that change can be foreseen or not. But if this is so, does it not provide a criterion of much more general use than the casuistic distinction drawn by Professor Pigou? This is the problem to which we have to turn.[1]


3. The Rationale of maintaining capital intact.

To "maintain capital intact" is not an aim in itself. It is only desired because of certain consequences which are known to follow from a failure to do so. And as we shall see, it is not even possible to attach a definite sense to this phrase if we try to apply it to a changing world, independently of why we want to do so. We are not interested in its magnitude because there is any inherent advantage in any particular absolute measurement of capital, but only because, ceteris paribus, a change in this magnitude would be a cause of a change in the income to be expected from it, and because in consequence of ^every^ change in its magnitude may be a symptom for such a change in the really relevant magnitude, income. Professor Pigou has abandoned the attempt to define in terms of a value dimension of capital the position, in which the stock of capital undergoes no changes that need to be added to or deducted from the output of consumers' goods for the computation of the national dividend. He has thereby not only acknowledged the inherent difficulties of any value measurement, which in the present case indeed are particularly serious,[*] but he has apparently also recognised the much more fundamental fact that such constancy of the value dimension has no necessary connection with the reasons why we wish to "maintain capital intact," and that in consequence there is no reason to assume that people will in fact normally act in such a way as to keep the value of the stock of capital constant.

  What, then, are the reasons why we wish capital to behave in a particular way? In the first instance, (a) this reason is evidently that the persons who draw an income from capital, want to avoid using up unintentionally parts of the source of this income, which must be preserved if income is to be kept at the present level. We want to avoid an unintentional temporary "splashing" or "stinting" which would have the effect of later reducing incomes below or raising it above the level at which we aim. Capital accounting in this sense is simply a technical device, an abbreviated method of solving the complicated problems arising out of this task of avoiding involuntary infringements upon future income. Whatever the time shape of the future income stream derived from the capital in his possession at which an individual aims, there still remains the problem of deciding what is the required action with regard to the individual parts of his possessions. And although we have certainly no right to assume that every person will normally aim at a permanent constant stream of income from his capital, there is probably some justification for regarding this case as one of special interest. In any case, even when the capitalist aims at some other shape of the income stream, the problem remains the same, and the case of the constant income stream might simply be regarded as the standard with which the other cases are compared.

• ※ (a): cf. Keynes's mention in chapter 6 of the GT:

“It remains true, however, that net income, being based on an equivocal criterion which different authorities might interpret differently, is not perfectly clear-cut. Professor Hayek, for example, has suggested that an individual owner of capital good might aim at keeping the income he derives from his possessions constant, so that he would not feel himself free to spend his income on consumption until he had set aside sufficient to offset any tendency of his investment-income to decline for whatever reason.[6] I doubt if such an individual exists; but, obviously, no theoretical objection can be raised against this deduction as providing a possible psychological criterion of net income.” ([P14] of the part, I. income of the chapter 6, the GT)

  So long as we confine ourselves to the effects of the decisions of the capitalist on his own income stream, it may seem arbitrary to treat any one of the different sets of consistent decisions regarding his future income stream as in any way more "normal" than any other. It may even have a certain theoretical attraction simply to define whatever incomes he wants to have at different periods as equal incomes. We are, however, constrained by other reasons to abstain from such a pure subjectivism and to adopt a more objective standard. These other reasons are that we are interested in the maintenance of capital, not only because of the people who themselves deliberately distribute income over time and by doing so become capitalists, but also because of the effect of their activity on the incomes of other people. In the case of the workman, whose labour receives a greater remuneration because of the co-operation of capital, but who is not himself an owner of capital, we have no expression of his preference as regards the shape of his income stream, and we have to assume that he wants an income stream which does at objective sense might provisionally be defined as consisting at every successive moment of varying collections of commodities actually bought at an aggregate price, at which the collection of commodities actually bought at the beginning of the period might have been obtained.


4. The action of the capitalist with perfect foresight

The next task, then, is to find out how the individual owner of capital goods will behave, if he wants to keep the income he derives from his possessions constant, and if he has complete foresight of all relevant changes.[1] Complete foresight in this sense need not refer to all the relevant future. It is sufficient for our purpose to assume that at any one moment he foresees all changes that will affect the return of the investment he then make. His anticipation need only be correct for a period equal to that for which his investment runs, and if he makes investments in different fields, the extent of his foresight need only cover the relevant facts affecting the different investments during such different periods as these investments last. Beyond this, only some more general expectation as regards the rate of interest at which it will be possible to reinvest the capital recovered, will be necessary.

  Within these periods, his anticipations must in the first instance cover the relevant price changes. But such foreknowledge is hardly conceivable without some foresight of the real changes, which bring about the change in prices. The main types of changes, which he will have to foresee and the effects of which we shall have to discuss, will be changes in the demand for the products and the consequent changes, not only of the prices of the products, but also of the prices of the factors, changes in the quantities of the factors of production and changes in their prices caused in this way (including, in particular, in both these cases, changes in the rate of interest), and finally, changes of technical knowledge or inventions. With respect to this last case the idea of foresight evidently presents some difficulty, since an invention which has been foreseen in all details would not be an invention. All we can here assume is that people anticipate that the process used now will at some definite date be superseded by some new process not yet know in detail. But this degree of knowledge may be sufficient to limit investments in the kind of equipment, which is bound to be made obsolete by the expected invention, in such a way that the old equipment wears out as the new invention can be introduced.

  If we take first the case of an anticipated change in the demand, either away from or towards the product produced by our capitalist, his knowledge of this impending change, a full investment period in advance of its actual occurrence, will evidently put him in a position so to redistribute the earned amortisation quotas of his capital between the different industries that he will derive the greatest return from them possible under the new conditions. But it what sense will these amortisation quotas represent a constant magnitude, and under what conditions can we assume that the capitalist will invest them and no more and no less, if he merely aims at keeping his income from capital from now onwards constant?

There is no reason to assume that, if he just continues to reinvest, after the change has become know, the amounts he used to invest, in what appears the most profitable way in the light of the new knowledge, this will have that effect. The shift in demand between different products, if the co-efficients of production in the different lines of industry are not exactly the same, is bound to change the relative prices of the factors, and these changes will occur gradually as all the entrepreneurs redirect their resources in anticipation of the impending change. (...)

(... ... ...)

5. Obsolescence and anticipated risks

There remain two special cases to be considered before we abandon the assumption of correct foresight on the part of the capitalists. The first is obsolescence, as distinguished from wear and tear, as a cause of destruction of existing capital values. Although at first it may appear otherwise, this is a phenomenon which will occur even with perfect foresight. The second case on the other hand is somewhat intermediate between that of correct and incorrect foresight; it is the case where the probability of the occurrence of certain changes is correctly and equally foreseen and estimated by all member of the society. (...)

(... ... ...)

6. The reaction of the capitalist on unforeseen changes: (... ...)

7. The impossibility of an objective standard with different degrees of foresight

So far the criterion, for what is to be understood as maintaining capital intact on the part of the individual entrepreneur or capitalist, is purely subjective, because it depends on the extent to which the individual capitalist foresees the future. Entirely different action by two capitalists who hold different views on the future, but who are otherwise in exactly the same position, may satisfy our criterion. Both may, in the light of their different knowledge, do their best to obtain a constant income stream, and yet both will probably fail, earlier or later and in different degrees. Are we to say that neither has been maintaining his capital, and ought we to reserve this term to the case of action with perfect foresight? In a world where very imperfect foresight is the rule this would clearly lead to absurd results. We should not only have to say that nobody ever succeeds in maintaining his capital intactㅡwhich in a sense of course would be trueㅡbut we should also be prevented from using this concept of maintaining capital intact as a description of the actual behaviour of the entrepreneurs, who want neither to decrease nor to increase their income from their possessions. Taking into account the fact that human foresight is of necessity very imperfect, and that all economic activity must be based on anticipations, which will partly prove incorrect, it would still seem desirable to find a criterion which would enable us to distinguish between lossesㅡor rather missed opportunities of faster improvementㅡwhich are unavoidable in view of unpredictability of the change, and capital loss due to what appear to be avoidable mistakes.

  At first one might feel inclined to base the definition of what is to be regarded as adequate maintenance of capital on such a degree of foresight as the intelligent capitalist can reasonably be expected to possess. But closer examination of the problem soon reveals that any attempt to find an objective test of what can be regarded as maintenance of capital, short of the case of absolute foresight, must necessarily fail. It seems common sense to say that if an entrepreneur expects a change in taste, e.g., because he hopes to interest the public in a novelty, but is disappointed in his expectations, and loses the capital invested in the venture, this is a loss of capital which must be made up out of new savings, if capital is to be kept intact. If this did not happen, and similar failures were frequently repeated, the capital available for the production of things which people want would be considerably reduced by conversion into equipment for making things which nobody wants. On the other hand, the loss of capital due to an unforeseen change in taste seems merely the incidental and unavoidable concomitant of a process leading to what is now a preferred income. Yet if entrepreneurs had correctly anticipated the changeㅡand some entrepreneurs may have done soㅡthe wants of the public would have been supplied even better. Is, therefore, the loss of the entrepreneurs, who did not foresee correctly, to be counted as a capital loss to be deducted from gross output? and does it already mean that they might have foreseen the change if a single happy speculator chanced to do so, who, according to all reasonable expectations should have proved to be a waster of capital?

  It is not possible to base the distinction here on the concept of a change, and to say that to invest in anticipation of a change which does not occur is wasting capital, while to invest in the mistaken assumption that things will remain as they are is only a cause of unavoidable loss. In the first instance, it is by no means evident what is to be regarded as a change. If a temporary change is mistakenly considered as permanent, or if the expectation that the seasonal fluctuations of the past will be repeated is disappointed, are these to be regarded as mistaken expectations of a change, or as a mistaken expectation that things will remain constant? Clearly in economic life, and outside of a fictitious stationary state, the concept of a change itself has frequently no meaning except in the sense of a change relative to expectations. In the second place, and even more important, an approximately correct anticipation of the majority of "changes" in the usual sense is an indispensable condition of that degree of progress which is observed in actual life. One need only consider for a moment what would happen if entrepreneurs always acted as if things would remain forever as they are at present, and changed their plans only after a change in demand (or some other change) had actually occurred, in order to see what would necessarily be the effect on general productivity. Every change would mean an enormous loss, or rather, the adaptation of production to the change would become so expensive (not because the loss on existing investment would have to be counted as cost, but because the "free capital" required for the new production would be so scarce) as to make it in may cases impossible. How rich, on the other hand, should we now be if all past changes had been correctly foreseen from the beginning of things!

  All this means simply that the mobility of capital, the degree to which it can be maintained in a changing world, will depend on the foresight of the entrepreneurs and capitalists. If this is a commonplace, it is at least a commonplace to which far little regard is paid in usual reasoning.

• It means nothing less than that the amount of capital available at any moment in a dynamic society depends much more on the degree of foresight of the entrepreneurs than on current saving or on “time preference.”
• This is simply a corollary to the equally obvious and neglected fact that “capital” is not a factor, the quantity of which is given independently of human action even in the comparatively short run. 
• How great a contribution to the possibility of satisfying human wants a given stock of capital goods will still represent some time later, will depend largely on how correctly the entrepreneurs foresee the situation at this moment. 
• Their anticipations in this respect are quite important a “datum” for the explanation of the dynamic process as the “stock of capital,” and the latter concept has in fact little meaning without the former. 
• As an enumeration of individual capital goods existing at the beginning, the “stock of capital” is, of course, an important datum, but the form in which this capital will still exist some time afterwards, and how much of it will still exist, depends mainly on the foresight of the entrepreneurs and capitalists. 
• It would probably be no exaggeration to say that to maintain his capital so as to receive the greatest lasting return, is the main function of the capitalist-entrepreneur.[1] 

But not only is in this sense the size of the productive equipment of society dependent on the success of the entrepreneur, it is also dependent in a world of uncertainty on his capitalising capital gains ("windfall profits"). It should be recognised that much of the new formation of capital equipment (which may, but need not, represent net additions to capital in the traditional terminology) does not arise out of savings proper, but out of those gains of individual capitalists which are part of the process of capital maintenance. This process will, as shown above, always involve unforeseen profits on the part of some and unforeseen profits on the part of other entrepreneurs, changes on capital account which are part of an ever-proceeding process of redistribution of wealth, not to be confused with distribution of income. The entrepreneur who finds that a risky undertaking succeeds, and who for a time makes extraordinary profits because he has restricted the amount of investment so as to give him in case of success a margin of profit over cost which is proportionate to the risk, will not be justified if he regards the whole profit as income. If he aims at a constant income stream from his investment, he will have to reinvest such part of his profits as will be sufficient to give him an income equal to the part he has consumed, when the rate of profit in what has now proved to be a successful line of business falls to normal.[2] It is in such a way that, in case of changes in demand or technical progress, etc., capital is newly formed without new saving in place of that which is lost elsewhere. There is, of course, no reason to assume that the capital lost and that which is newly formed will correspond in any quantitative sense; and it is exactly for this reason that the usual concept of a net change in capital, which is supposed to correspond in some way to saving, is of little value. There has in this case been no abstention from consumption which could have been maintained at that level. If anybody can be said to have refrained from consumption which would be compatible with enjoying the same income permanently, it is not the entrepreneurs, but rather the consumers who for a time had to pay a price in excess of the cost which the production of the commodity entails after it has proved an assured success. But this "saving" is, of course, neither voluntary nor does it represent an abstention from consumption, which could have been regarded as permanently possible so long as the outcome of the venture was uncertain. It can hardly be questioned that in the actual world a great deal of the equipment which is made necessary by some change is financed out of these temporary differences between cost and price. But it may appear somewhat paradoxal that where it can be provided in this way this ought not to be classed as saving but as a capital gain a kind of transfer of capital which means that not only new capital is formed in place of that lost elsewhere, but that it is formed exactly where it is most needed, and in the hands of those most qualified to use it; this follows, however, necessarily from the consistent use of the definition of maintaining capital and saving, which we have adopted. It will be shown in the next section that this use of the terms proves convenient in other directions also.


8. "Saving" and "Investing"

The upshot of the discussions of the last sections is [:]

• that if changes in the data occur (such as new inventions, shifts in consumers' demand or changes in the supply of factors), the amount of capital (conceived as a multiple of the income of a given period, orㅡwhat amounts to the same thingㅡthe result of a certain "average" waiting period, or in any other conceivable quantitative sense), which is available and required to maintain income from then onwards at a constant level, will change also, and 
• that in consequence there is no reason to expect that any of the conceivable dimensions of capital will remain constant. 

It remains true, of course, that ceteris paribus it is necessary to maintain or replenish a reservoir of goods of a constant size, in order to maintain a given output. But when conditions change so as to make a smaller or larger reservoir necessary for the same purpose, its content will tend to change automatically in such a way as to preserve income constant from the moment when the change becomes known. The fact that an impending change is likely to become known to different people at different times will lead to capital gains and capital loss of individuals, with the effect that persons who have shown the greatest foresight will command the greatest amount of resources. But in a world of imperfect foresight not only the size of the capital stock, but also the income derived from it will inevitably be subject to unintended and unpredictable changes which depend on the extent and distribution of foresight, and there will be no possibility of distinguishing any particular movement of these magnitudes as normal.

  These conclusions have rather far-reaching consequences with respect to the much used, or much abused, concepts of saving and investing.

• (a) If (a1) the stock of capital required to keep income from any moment onwards constant cannot in any sense be defined as a constant magnitude, (a2) it becomes also impossible to state that any sacrifices of present income in order to increase future income (or the reverse) must lead to any net changes in the amount of capital.

※ (a): (a1) 미래소득을 일정하게 유지하는 데 필요한 자본 저량을 불변의 크기로 정의할 수 없다면, (a2) 미래소득의 증가나 감소를 위해 희생해야 할 현재소득과 순자본 저량의 변화, 이 두 가지의 관계에 대해서도 말할 수 없다.

• (b) Saving and investment in the ordinary meaning of the terms are, of course, one of the causes, but by no means the only cause, which affect the magnitude of capital (in any conceivable quantitative sense), and the changes in the size of the capital stock cannot therefore be regarded as indications of what sacrifices of present income have been or are being made in the interest of future income. 

※ (b): 물론, 저축과 투자가 자본량에 영향을 미치는 유일한 원인은 아니며, 자본량의 변화가 미래소득을 목적으로 희생해야 할 현재소득이 얼마인지를 가리키는 지표가 될 수는 없다. 

• (c) This idea, [?however,] appropriately enough for the analysis of the effects of a change under otherwise stationary conditions, must be completely abandoned in the analysis of a dynamic process.

※ (a)는 저자가 말하려고 하는 주된 논지다. (b)는 이 (a)에 대한 단서나 제한이다[즉, (a)가 옳지만 (b)와 같은 제한이 있다는 식]. (c)는 다시 주된 논지 (a)로 돌아오기 위한  요소다. 즉, 논증의 흐름은 (a)를 피력한 다음, (b)와 같은 제한은 있지만 ‘무엇 때문에(혹은 b와 같은 제한이 있다고 하더라도)’ (c)라는 문제가 있다는 식으로 이어진다. 하지만 (b)라는 단서에서 탈출해 (c)로 넘어가기 위한 논증의 고리 하나가 누락된 것 같다. 

(d) If (d1) we want to retain the connection between the ideas of saving and investment, and that of a sacrifice of potential present income in the interest of future income (and it will be shown that it is this concept which is of important in the connection in which those terms are commonly used), (d2) we cannot determine the size of either saving or investment by any references to changes in the quantity of capital. And (e) with the abandonment of this basis of the distinction, there will, of course, have to fall the habitual practice of economists of separating out such part of general investment activity as happens to leave the capital stock in some sense constant, as something different from activities which add to it, a distinction which has no relationship to anything in the real world.[1]

※ (d): (d1) ‘저축과 투자라는 개념’과 ‘미래소득을 겨냥한 현재소득의 희생’이라는 개념의 관계를 유지하고 싶어도  (d2) 저축과 투자의 크기를 자본량의 변화와 결부지어 파악할 수 없다. (e) 포기해야 하는 기본적인 구분이란 ‘저축과 투자의 크기’와 ‘자본량의 변화’의 관계를 지칭하는 듯.

※ (d1)에서 ‘저축’과 '미래소득의 겨냥한 현재소득의 희생’을 서로 다른 개념으로 언급:..

  To deny that the usual distinctions between new and merely renewed investment[net investment and investment for depreciation], and between new savings out of net income[?net savings, ?current savings] and merely maintained savings[previously accumulated savings], as distinctions based on the idea of quantitative increases or decreases of capital, have any definite meaning, is not to deny that they aim at a distinction of real importance.

※ 통상적으로 순투자와 감가상각 투자를 구분한다든가 당기 저축과 누적된 저축을 구분하지만 이러한 구분을 자본의 양적인 증감에 기초하여 파악할 도리가 없다는 것이 내 주장이지만 그렇다고 이러한 구분이 중요하지 않다는 것은 아니다.

• There can be no doubt that (1) the decision of the consumers as to the distribution of consumption over time are something separate from (2) the decisions of the entrepreneur-capitalist as regards what quantities of consumers' goods to provide for different moments of time, and that the two sets of decisions may or may not coincide. 

※ (1)은 곧 소비자가 시간의 흐름에 따라 순소득에서 얼마를 소비하고 그 나머지로 얼마로 저축할 것이냐, (2)는 기업가가 시간의 흐름에 따라 소비재를 얼마나 공급(따라서 총산출에서 그 나머지인 자본재를 얼마나 공급)할 것이냐. 당연히 이 두 가지 결정은 서로 조응할 수도 있고 아닐 수도 있다. 

• What I do want to deny is only that the correspondence or non-correspondence between these two sets of conditions can be adequately expressed in terms of a quantitative correspondence between (net) saving and (net) investment. 

※ 이 두 가지 결정이 서로 조응하느냐 조응하지 않느냐 하는 문제를 (순)저축과 (순)투자의 양적인 조응 관계로 적절히 표현할 수 없다는 게 내 생각이다.

※ cf: Keynes's mention in chapter 6: “But when Professor Hayek infers that the concepts of saving and investment suffer from a corresponding vagueness, he is only right if he means net saving and net investment. The saving and the investment, which are relevant to the theory of employment, are clear of this defect, and are capable of objective definition, as we have shown above.” (chapter 6 of the GT)

But if this distinction is not to be formulated in this particular way, what are we to put in its place? In general terms the answer is not difficult. If we must no longer speak in terms of absolute increases and decreases of capital[,] we must attempt a more direct comparison of the time distribution of income. Capital accounting, as has been mentioned before is itself only an abbreviated method of effecting this comparison in an indirect way, and when this indirect method fails it is only natural to go back to its rationale, and to carry out the comparison explicitly. [Consequently:]

• (a) Instead of comparing them each with the supposed standard case of capital remaining “constant,” and so arriving at the concepts of net saving (net income minus consumption) and net investment, and then to juxtapose these derived concepts, 
• (b) we shall have to compare directly the intentions of the consumers and the intentions of the producers with regard to the income stream they want to consume and produce respectively.

※ (a)는 포기해야 할 접근. 여기서 net saving＝net income－소비라고 명시; 즉,＝(gross income－depreciation)－소비. net investment＝gross investment－depreciation일 것임. 그리고 (b)는 새로 채택하겠다는 접근.

  The question, then, is essentially whether the demand for consumers' goods tends to keep ahead of, to coincide with, or to fall behind the output of consumers' goods irrespective of whether either of the two magnitudes is increasing, remaining constant or decreasing in an absolute sense. But in order to give this question a clear meaning we have yet to settle in terms of what are the demand for and the supply of consumers' goods to be measured, in order to establish whether they coincide or whether the one exceeds the other. In a sense, of course, demand and supply are always equal, or made equal by the pricing process, and to speak of their relative magnitudes presupposes some unit in terms of which their magnitude is measured independently of the prices formed on the market.

  Consider first the decision of the "savers," or the body of consumers as a whole. What we can assume of them is, not that they will under all conditions aim at an income stream of a particular shape, but that if they are offered a present income of a given magnitude, plus the sources of a future income of a certain magnitude, they will attach certain relative values to these incomes. For every such combination of a given present income and the sources of a certain future income we must assume these relative valuations to be determined.

  If the relative value consumers attach to the sources of future income, compared with present values, should be higher (or lower) than the cost (in terms of present income) of reproducing new sources of future income of the same magnitude, more (or less) such sources will be produced. And assuming that the relative valuations of the consumers do not change abruptlyㅡwhich is least to be expected when in each successive period the available income is equal to that for which they have plannedㅡthen production will tend to provide in each successive period such amounts of present income and sources of future income, that their relative cost (in terms of each other) will approximately correspond to the relative values attached to them by the consumers. But if for some reasonㅡsay because additional money has become available for investment purposesㅡthe price of the sources of future income have been raised out of correspondence with the valuation of consumers, more sources of future income and less current income will be provided for the next period than consumers will be then willing to take at prices corresponding to their relative cost. Consumers will find that they get less current real income, and in consequence will attach a greater value to it compared with the sources of future incomeㅡinvestment will have exceeded saving in the usual terminology.

  Under the assumption of otherwise constant conditions (i.e., unchanged knowledge, taste, etc.) this process can be described in the familiar way in terms of changes of the investment period, to which correspond changes in the quantity of capital (in terms of income of a period). We would say that, by increasing the waiting period and thereby accumulating more capital, producers have caused a temporary gap in the income stream which leads to a relative rise in the prices of consumers' goods. But as soon as we drop this ceteris paribus assumption this is no longer true. The correspondence between the value attached to the source of future income and their cost is then no longer dependent on the cost of reproducing the same amount of capital, which under changed conditions will make it possible to produce the same future income.

  Additional investment, in the sense that total output is reduced for a time in order to increase it at a later date, may take place, although at the same time the quantity of capital is reduced (and the "period of production" shortened). [1] Breaks in the even flow of consumers' goods, which make corresponding changes in the attitude of the consumers necessary if disturbances are to be avoided will[,] only occur when that quantity of capital is not maintained which under conditions prevailing at any moment is required to provide such a constant flow.

  The correspondence between the supply of current consumers' goods and the demand for them, which is what has been aimed at by the saving-investment equations, can only be stated adequately if we measure both supply and demand in terms of the alternatives open to consumers and producers under the circumstances existing at the moment.[1] To do this it seems necessary to abandon entirely the concepts of saving and investment as referring to something beyond and outside the normal process of maintaining capital quantitatively intact, and to substitute an analysis on the lines suggested, which does not try to separate "old" and "new" investment, and "new" and "maintained" saving as distinguishable phenomena. Or, if we want to refrain the familiar terms and to use them without reference to changes in the quantity of capital, we might say that "savings" correspond to "investment" when the value of the existing capital goods (in terms of consumers' goods) is such that it becomes profitable to replace them by the capital goods that are required to produce the income in the expectation of which people have decided currently to consumer as much as they actually do. I believe that some theories, which used to be stated in terms of net saving and net investment, can be restated in terms of these concepts, and I have tried to sketch such a formulation of my own views in another place.[1] But I am rather doubtful whether the same is possible with some other theories which seem more dependent on the concepts of saving and investment as absolute magnitudes.

[1] Cf. "Price Expectations, Monetary Disturbances and Malinvestments" now reprinted as the fourth essay in this volume, see below et seq.

All this is, of course, no reason for not using the concepts of absolute increases or decreases of capital under any circumstances. In the discussion of comparatively long-term changes it may sometimes be quite innocuous. And in the discussion of short-run changes it would be equally legitimate to speak of causes, which ceteris paribus would lead to increases or decreases of capital. But we must be very careful not to assume that they actually do and not to base any distinctions on supposed net changes in the quantities of capital which do not actually take place. Particularly the phenomenon of the trade cycle is probably largely concerned with changes within that region of indeterminateness between clear increases and clear decreases of capital, inside which the concept of an absolute change has no meaning. But it remains probably true that net accumulations and decumulations of capital in the usual sense will present similar phenomena as booms and depressionsㅡat least when real accumulation proceeds faster and real decumulation proceeds slower than saving and dissaving. [2]


9. Capital accounting and monetary policy

This discussion of the problems connected with the concept of the maintenance of capital has by no means been exhaustive. We have touched on many points which have not been cleared up, and there are many others equally important which have not even been mentioned. This is, however, unavoidable in an essay which treats what is in many respects one of the central problems of economic dynamics. But there is one further problem of great importance on which some remarks must be added in this concluding section

  Up to this point we have been largely concerned with an attempt to derive the appropriate action from the ^rationale^ of "maintaining capital intact," but we have said little about the effects of the actual practice of the entrepreneur, that "abbreviated method" which consist in regarding capital as a money fund of definite magnitude, and on which actual capital accounting is based. This practice, of course, largely due, and partly justified, by the fact that the capital of the individual enterprise is to a great extent furnished in the form of money loans, and that in consequence the entrepreneur has in the first instance to provide for a repayment of the money loans.

  It was also this practice of treating capital as a money fund which has given rise to the theoretical concept of capital as a quantitatively determined fund. But while the actual use of this concept in real life does not mean that we have to accept it as the basis of theoretical analysis, and does not relieve us from the duty of going back to the ^rationale^ of its use, the results of this theoretical investigation are little more than a starting point for a study of the effect of the actual practice.

  A more exhaustive investigation would, therefore, have to proceed after this preliminary clearing up of the fundamental concepts to the main task of explaining what the effects are of the actual accounting methods used by capitalists, to what extent, and under what conditions, they fulfil their purpose, and when they fail. That there are cases where the rigid application of the money fund concept fails is, of course, generally recognised and to some extent taken account of it the distinction made between changes on income account and changes on capital account. It is also obvious that the result achieved by this method will be largely dependent on monetary policy. Of course, no monetary policy can make the money value of capital behave in such a way that a constant value will always correspond to a quantity of capital which will give the same real income, and that all attempts to increase the future output of consumers' goods at the expense of the present and vice versa will lead to corresponding changes in the money value of capital; and in consequence a policy of the capitalist-entrepreneur which aims at nothing but this, will always err to a considerable extent, i.e., will lead to positions where their distribution of resources between current consumption and the provision of future consumption is not in accord with consumers' preferences. But the degree to which capital accounting in terms of money will prove will prove deceptive will depend on the particular monetary policy followed.

  To realise these effects in different situations would required a fairly detailed analysis of a number of representative instances. It is only possible to discuss here in the most general way one case, probably the most important and the one which has received the greatest attention in recent discussion; that of a continued increase in output due to "technical progress." Now, in spite of all the complications discussed before, it still remains true that ^in any given situation^ the value of capital required to provide an income stream at a certain constant or increasing rate will have to stand in a definite proportion to the value of current income. This proportion will change with any change in the relevant data, but where we have to deal with a development from moment to moment it is still approximately true to say that in order that the replacement of capital be sufficient to maintain the income stream at least at a constant rate its value should maintain a constant proportion to that income. (...) p. 139.

(... ... ...)