XXIII. The Effects of Unforeseen Changes and in Particular of Inventions

The grogg returns during the remainder of the life of thc capital good in question will now be likely to be greater or smaller than had been anticipated, either Factors to b. con-because these returns will now accrue at a sidered different rate, or because they will continue for a longer or shorter period than was expected before. In addition new alternative uses for the asset in question may have to be taken into consideration. To what extent this will be so will depend on the degree of specificity (')1' versatility of the a'>set. For some assets, e.g. most stocks of raw materials, the alternative uses will be very numerous 306 CR. XXlII Effects of Unforeseen Changes 307 and the usefulness of the good in these other uses will not be very much less than in that originally contem plated. Other assets such as specialised machinery may be highly specific; that is, there may be no other uses at all or only such as will yield a very much smaller return.

The availability of alternative uses to which the good may now be turned is a factor of great importance. But in considering it we must clearly realise from the beginning that there is no necessary connection Usefulness In alt.r between its possible usefulness for other natlve employments not necessarily con purposes and the function it was origin-nected with original ally destined to serve. This possibility of value using it for another purpose is, so to speak, an accident, and not a necessary consequence of, or identical with, its original function. What we can transfer to other purposes is in particular not the same "waiting power" which it represented while it was regarded as useful for the original purpose, but another and probably different contribution of which it happens to be capable in the new situation. In order to clarify some of the main points it will be useful to look first at the case of capital goods which are completely specific and for which therefore no alternative use need be considered. 1 Suppose that .. Windfall prollts"

the capitalist whose property is affected by (and iosses) made on • ., specillc assets the unforeseen change alms at keepmg hIS income stream in all circumstances constant, so that whatever happens to him, he wants no more than to keep his income permanently at the constant level which seems obtainable on the basis of his present knowledge. Then 1 As a rule, even in the case of the most specific type of plant or machinery its scrap value will, of course, have to be considered as an alternative return. We can, however, neglect this so long as we either ignore cases where the value of the capital good in question in its present use has not been lowered below its scrap value, or if we assume that the scrap value of the material is no higher than the cost of scrapping (as is often the case with houses).

308 Capitalistic Production under Competition PT. III he will have to treat the remaining gross returns of the asset as a terminable annuity of which he must consume no more than will enable him to build up a reserve from wh~ch, after the asset ceases to yield a return, he will be able to draw an income equal to that previously derived from that asset. Whether the particular unforeseen change is favourable or unfavourable to him, this will mean that as soon as he learns about its occurrence or imminence he will have to change his rate of consumption to the level at which it can now be permanently kept. If, for instance, his receipts increase in consequence of the change by £210, and the rate of return which he can obtain on re investment is 5 per cent, he must consume only £10, and must reinvest the remaining £200, which at 5 per cent will give him the same return in every future year. Such windfall profits are, therefore, not income in the sense that their consumption is compatible with "maintain ing capital intact". N or need consumption be reduced by the amount of corresponding windfall losses. In both cases only the current interest on the (positive or negative) capital gains ought to be counted as income. 1 After such a change, the capitalist who was previously in a state of individual equilibrium in which he merely Enects on time pre-wanted his income to continue constant, ference may very well change his attitude. The two most important data for his decision will have changed.

Not only will his income be different from what it was 1 In the case of "windfall losses" it would, of course, often be possible gradually to recuperate the value of the capital originally invested. In order to do this the owner of the capital would have to decide, after the unfavourable change had occurred, to make the same allowance for depreciation as before and to reduce current consumption by the full amount of the loss. But this could hardly be described as " maintaining capital intact". It would mean that the owner would have to reduce consumption for a period below the level at which it could be permanently kept, in order to raise it later above that level. It seems that this would more appropriately be regarded as new saving - saving, it is true, to make up for a loss, but for a loss which has already occurred. This loss was irrevocably incurred when the invest ment was made in ignorance of the impending change.

CH. XXIlJ E./fect8 of Unfore8een Change8 before, but the returns to be expected from new saving will also be different from what they were before. In all proba bility, if the income from his old investment has increased, the return to be expected from the new investment will now also be greater. And this, as a rule, will encourage him to aim at a further increase of his income; that is, it will make him save. A decrease of the return on the old investment, on the other hand, will also frequently be accompanied by an increase of the returns to be expected from new investments, and therefore will also often lead to new saving. But the reverse effect, namely a reduction of income from the old investment, coupled with a decrease in returns to be expected from new investments, leading capitalists to aim at a decreasing income stream (that is to dissave), is by no means impossible. The nature and significance of the reactions of the capitalist entrepreneur to unexpected changes will become clearer if we consider in somewhat more detail the effects of a particular type of change. In many Err If' II ec S 0 mven ons respects the effects due to changes in to be discussed as h I . I kId " special Instance tec no oglCa nowe ge or InVentIOns are the most interesting, and they may therefore serve here as a concrete example of the general class of phenomena under consideration. The application of the argument to the other main types of change which are relevant here, i.e. shifts in demand between different types of consumers' goods, or changes in the supply of factors, will present no difficulty.

For the case of inventions, two kinds of effects will have to be considered: on the one hand the possibility of a loss of capital invested in plant that is made obsolete by an invention, and on the other hand, Two cases to be COD the possibility of a gain on plant and stock sldared which, at least during a transition period, may bring higher returns than was expected. Although there is some reason to suppose that any unexpected change is much more likely to lead to considerable capital losses 310 Capitalistic Production under Competition PT. III than to capital gains, it is not impossible that in particular instances the gains may exceed the losses. The case of capital gains can be treated more shortly and may therefore be taken up first. Capital gains con sequent upon an invention will mainly occur during such (s) capital gains transition periods as will elapse until it is possible to increase the supply of particular instruments which are now required in the newly invented process. 1 If, for instance, the new machinery required can be produced only in a particular plant, which before was expected to be used only little or discontinuously, and if it takes a long time to erect an additional plant of the same sort, the owner of the existing plant will clearly be able to make considerable and unexpected profits during the interval. Since these profits will be of a temporary character, he ought not to regard them as ordinary income. He ought to reinvest such part of them as will secure him an additional income in the future equal to that which he consumes during the transition period.

Losses due to technological progress require somewhat more careful consideration, since they are closely connected with questions which have been the subject of wide dis(b) It II cussion and some considerable confusion. It cap a osses: an example of "capltal-is again advantageous to begin by supposing saving" Inventions h . .. t at the old eqUIpment whIch, In consequence of an invention, will no longer be reproduced, is of a highly specific character. Since the example of the displace ment of the long-distance cables by wireless telegraphy has often been used as an instance of what was regarded as a "capital saving" invention, 2 we may keep this in mind 1 Weare here negl~cting increases in the capital values of wasting natural resources. 2 Cf. A. C. Pigou, The Economics of Welfare (4th ed., 1932), p. 675. The terms" labour saving" and" capital saving " inventions are here used in the older, more popular sense in which they are used by Pro fessor Pigou, who defines a capital saving invention (ibid. p. 674) as one" which reduces the ratio of capital to labour in the industry where it applies". Professor Hicks (The Theory of Wages, 1932, pp. 121 et CR. XXnI Effects of Unforeseen Changes 311 as a concrete instance, without troubling too much about the actual technical details. We shall feel free to make successively different hypothetical assumptions about the extent to ,which we suppose the introduction of wireless telegraphy to reduce the cost of transmitting telegrams, and about the relative share which prime and supple mentary costs constitute in either kind of telegraphy.

Consider first the position of the owners of the old type of equipment, that is, in our case of the cables. Their position will, of course; be affected only if the new. method makes it possible to provide the Effects on owners 01 same services at lower cost. But the fact old equipment that it has become possible to provide these services at a lower total cost does not necessarily mean that it will have become unprofitable to use the already existing equipment. If we assume the scrap value of the existing equipment to be practically zero (because, for instance, it may cost more to raise a submarine cable than can be realised for the material contained in it), then providing at least that mere operating costs are covered, it will clearly be more profitable to continue operating it than to shut down completely. The increased demand for the serviecs, at the lower cost at which they can now be 'provided, will make it profitable to install so many wireless stations in addition to the existing cables that the price of a telegram will fall relatively to the cost of transmission by the new service. But, at the beginning at any rate, and until they are worn out and need replacement, the existing cables will remain in use.

Although this is rather familiar ground 1 it is necessary to give a more exact statement of the conditions under seq.) has refined this original concept, but, as WI' shall seo, he hill' no1 altogether escaped the consequence of the lise of the "[lind" eOIlC't'pt of capita\. 1 Cf. A. C. Pigou, The Economics of Welfare (4th cd" 1!l:~2). p. 188; L. C. Robbins, An Essay on the Nature and 8igm'jic{lIIce (if k'cullolilic Science (1932), pp. 50 et seq.; HaYf'k, "Tho Trend of Eeol1olllie Thinking ", Economica (May 1933), and 1936c.

312 Capitalistic Production under Competition PT. III which it will be profitable to operate equipment of the old type and equipment of the new type side by side, and the conditions under which the new type of equipment General conditions will immediately supplant the old one. The under which Intro-relevant magnitudes for both types of equip ~uctlon 01 Invention will prove prolltable ment are as follows. First, there is interest and amortisation on the capital which would now be required to create the equipment in question; this we shall call" capital cost" and shall designate by C1 for the old equipment and C2 for the new equipment. Secondly, there are the operating costs of the old and the new equipment, which will be designated by 01 and 02 respectively.l The condition, then, for it to be at all profitable to introduce the newly invented process is that Cz +02<C 1 +01' So long, however, as Cz +02>01 it will still be profitable to use the old equipment. The return on this old equipment will have fallen below C l , the figure necessary to make its replacement profitable, but it will still be a positive figure, namely Cz +02 -01' Only when the total cost of the new process becomes as low as the mere operating cost of the old process, that is when C z +02 =01' will the old equipment cease to give any return and lose all value. And if C2 +02-<01 it will-be al together unprofitable to use the old equipment, since its mere operating cost will be higher than the price which is determined by the total cost of the new process.

The owners of the old equipment, finding the gross return from it reduced, will in the first instance have to decide what part of the return to reinvest and what part of it to consume. Let us suppose that they act in con1 "Capital cost" and "operating cost" as hero used, although closely related to, are not necessarily identical with, the concepts of "supplementary cost" and "prime cost" as genorally used. Tho reason why we prefer the terms used in the text is that for the purpose in hand the distinction oetweml the two kinds of costs will have to be made with respect to the use of the particular piece of equipment in question and may be different· [mill any such distinctions made from the point of view of tho firm as a whole ..

CH. xxm Effects of Unforeseen Changes 313 formity with a rational policy of maintaining capital intact, that is, that they aim at whatever constant income stream they can continue to obtain in the future. Then, as soon as they learn about the effects of the A rtf tl Ii mo sa on po cy new invention, they will reduce their con-of owners of old . I I h h equipment sumptIon to a eve suc t at what they reinvest will secure them the same net income in per petuity. They will, of course, reinvest not in the old type of equipment which it has become unprofitable to replace, but either in the new equipment or in an alto gether different line of business. Since, as the old type of equipment wears out, there will be opportunities for further investment in the new type of equipment and the owners of the old type will presumably have more scope for using their knowledge here than elsewhere, it is at least probable that they will attempt to invest as much as they can in this form. But how much will they have to invest 1 Is it possible that what they gradually recover now will be more than what can still be profitably invested in the new process, so that capital will be released for other purposes? If this excess which would now over flow into other lines of industry were greater than the amount which was initially brought into our industry from outside to start the additional new services, we should have the case which has usually been treated as a capital-saving invention. Is this a possible case? And, on the other hand, may it not also happen that so little capital is recovered as to be insufficient to replace the old equipment which is being worn out by new equipment which will render equivalent services, and that in conse quence, as the old equipment wears out, further and further doses of new capital will be attracted from outside?

The answer depends in general not on the absolute amounts of the two kinds of cost, but on the proportions in which they can be most profitably expended in order to produce a unit of the service in question. But it can 314 Capitalistic Production under Competition PT. III be shown that where there is complete specificity of the old equipment, the first case, that of the so-called capital saving invention, will occur only under very special and SlgnlHcance of pro-rather unlikely conditions. In such cases, porUons between hI' t I . . d £ th operaling and capital were ess capl a IS requIre or e cosis new process than was required for the old one, it is distinctly more likely that there will still be a need for additional capital from outside sources than that it will be possible to create all the new equip ment out of the amortisation quotas of the old. The reason for this is of course that very often as much, or more, of the value of the old capital will be destroyed by obsolescence than is being "saved" by the invention.

The additional amount of capital embodied in the now antiquated machinery will be lost at the same time as it becomes" superfluous". This has, however, yet to be shown to apply to each of the various conceivable cases. Let us begin with the case most favourable to " saving" of capital, that is, the case where the absolute amount of operating costs, as well as their relative share in the total Case 1: operating cost of a unit of the product, is greater in the cosls of tbe new pro-case of the new process than in the case of cess greater tban in old process the old process. This is perfectly compatible with the total costs of the new process being lower than those of the old process, the saving on capital cost being greater than the extra expense on operating cost. The conditions of our case can then be stated thus: C1 +° 1 > C2 +0 2, 01 < 02 01 0z ----_._-- < -_._-_._, C1 +01 cz+o z from which it follows that and Ci-C 2> 02-01 _C_l_>~.

C1 +01 C2 +0 2 OR. XXIII Effects of Unforeseen Changes 315 The gross returns of the old equipment will be reduced by the introduction of the new process to C 2 +0 2 - 01 < Cl , but this may still be greater than C2 • That is to say, the amortisation quotas recovered from the old equipment may be larger than what is required to replace it by a piece of new equipment which (with a greater expenditure on operating cost) will provide the same services for the same period.! In this case, where the value of the old capital is preserved to an amount greater than the amount of the new capital required, in addition to the new equip ment, a further amount of input will be required to produce the same final services. The owner of the old equipment will, it is true, find himself in command only of such a reduced total of resources as is now required to produce the same amount of final services. But as he will now find it profitable to use a greater quantity of input than before, he will substitute such input for capital and will actually be able to release capital for other purposes.

It is essential, however, to be quite clear about the way in which "capital" in this sense can be released from one use and transferred to another. What actually happens is that, as the old equipment wears The .. release" 01 out and is replaced by equipment requiring capllal for olher pur the investment of a smaller amount of poses input, but co-operating with a greater amount of current input, the aggregate of the investment periods of this input is shortened. If this were not compensated by the lengthening of the investment period of some other input, the effect would be that during some period the amount of output currently maturing would be due to a greater 1 In this case, where the mere operating cost of the new process is larger than the operating cost of the old process, it is of course imposs ible for the total cost of the new process to be smaller than the operating cost of the old process, and therefore for the invention of the new process to lead to the instantaneous abandonment of the old one and the complete destruction of the 'value of the old equipment. The new kind of equipment will in this case be installed only to satisfy the additional demand, called forth by the lower price of the product, but it will replace the old equipment only as the latter wears out.

316 Capitalistic Production under Compet'ition PT. III quantity of input than is currently applied. In fact, if a total amount of input equal in value to the amortisation quotas currently earned from the old equipment plus input (the operating cost) used in conjunction with it were currently used to provide for the supply of the final services by the new process, the effect would necessarily be that the supply of such services would temporarily be increased beyond the level at which it could be per manently maintained. For, since in the aggregate (or on the average) this input would mature sooner than it did in the past, the flow of the final services obtained from it would not dovetail but would partly overlap with those of the old equipment. To make up for this fact that the old equipment wears out more slowly than the new, it will be necessary to reinvest in the new equipment at a slower rate but over a longer period. 1 This means that for a time resources can, as it were, be lent to other industries, which will be enabled to start tlie production of the products which are to be re~dy at a particular date earlier than would other\vise have been the case.

The case where the newly invented process makes it profitable to expend a greater absolute amount of operat ing cost to produce a given amount of services is, however, Case 2: operating probably not of very frequent occurrence, costs In new process and is certainly not the only case which is smaller absolutely; but larger in propor-usually regarded as capital-saving. But it tion to capital cost is easy to show that as soon as the amount of operating cost required to produce a unit of the final services is no more (or is less even) than in the old process, even if it is higher relatively to the capital cost with which it is combined, no capital will be " saved". That is, no 1 I am here neglecting the case where, in consequence of a fall in price, the demand for the product increases fully in proportion or more (that is where the elasticity of demand for the final product is greater than unity). In t!his case reinvestment may proceed at the same rate as or faster than amortisation, and the additional input required will have to be attracted away frOlll other uses.

CH. XXIII Effects of Unforeseen Changes 317 capital will be made available to produce additional quantities either of the same, or of some other product. The conditions of these cases whioh we now have to consider can be stated shortly, in the notation used before, as follows : C1 +01 > C2+0 2, C1 > c2, 01 ~ 02' _~L>~ .. C1 +01 == C2 +0 2 In this case the gross return on the old equipment, i.e. the difference between the prices of the final services as determined by the total cost of the new process, and the operating cost of the old equ~pment (C 2 +0 2 -0 1), can at most (if 01 = 02) be equal to the capital cost of the new equipment, and may (if 01> 02) be smaller than the capital cost of producing with the new equipment. This means that, whatever the original value of the capital equipment which was required for the old process, no more. and perhaps less will be recovered than is required for the construction of such new equipment as will provide the same services. If the operating costs of the old and the new processes are the same, no input of any sort will be released, nor will any additional input be required to produce the same volume of output as was produced previously. For the additional output which it will now be profitable to produce at the lower cost, capital as well as current input will have to be attracted from elsewhere.

But no general statements can be made about the effect of this on the proportions between capital and labour in the rest of the system. For this will depend, not on the relation between the proportions in which the two sorts of resources are now required in the particular industry concerned and the proportions in which they were formerly required, but on the relation between the proportions in which they are now required in that industry and those 318 Capitalistic Prod'Uction 'Under Competition PT. III existing in the rest of the general economic system. If the operating cost, that is the amount of input required to produce a given amount of final output, is lower in the new process than in the old, less capital will Case 3: operating be recovered from the old equipment than costs absolutely and is required to produce the new, and it will proportionately smaller In tlie new be necessary to attract new capital from process outside the industry. But, on the other hand, some quantity of input will be released. What will happen, therefore, is that in the industry in question capital will be substituted for" labour" (i.e. pure input), and that consequently in the rest of the economic system capital will become relatively more scarce compared with pure input. This, as will now be clear, will very often be the effect of an invention, even though the invention causes a much smaller absolute amount of capital to be used in the industry directly affected. .

It might be concluded from the above that inventions which are actually capital-saving have to be regarded as a rather exceptional case, confined to inventions which, Ell ts Ii d bl while decreasing the absolute amount of ee were ura e Instruments are not capital required to produce a given output, completely spccillc 11 . ( 1 h h f actua y mcrease a t oug ,0 course, to a lesser extent) the amount of input required to co-operate with that capital. But we must remember that we have as yet only considered the case where the old equipment, which became obsolete because of the invention, was of a highly specific type. This means that we have still to consider the cases where the concrete nonpermanent resources, which became less useful in their original use because of the invention, can easily be turned to other uses. So long as we are thinking mainly of machinery these cases are not very likely to occur, although they are by no means impossible. If an industry which in the past had used a great number of electromotors could, because of an invention, suddenly dispense with the greater part of them, these could probably without great CK. xxm Effect8 of Unfore8een Change8 319 loss of value be absorbed in other industries. 1 But of much greater importance in this connection is what is commonly known as working or circulating capital such as stocks of materials, fuels, etc.

If, to give only one example, the paper industry, owing . to some invention, no longer requires timber as raw material, the stocks of standing trees raised in the expectation that they would be needed as raw material for paper-making, will not simply become superfluous. The timber may, for instance, be used in the production of artificial silk. This does not, of course, mean that it will represent the same value, i.e. produce the same income. Even in a highly favourable case like the one mentioned (i.e. where the identical material has already been used before in another extensive industry) the increase in supply relative to the now more restricted use will cause a considerable reduction in value. What can be transferred to other uses is, of course, not the "abstract quantity of value" or the "command over resources" which that stock represented before the invention occurred, but only the utility which it possesses in these other uses. This will probably mean (unless the input used in the reproduction of these raw materials is highly specific} that it will not be profitable to reproduce the full quantity of these nonpermanent resources when they are used up_ In that case the other industry will enjoy the advantages of a cheap additional supply of one of its raw materials only temporarily. But even this does not prove that the capital value which this stock of raw material temporarily represented for the industry to which it became unexpectedly available will not be preserved in some form. The situation is here exactly analogous to the case of wasting natural resources, which, of course, can never be reproduced in identical form.

1 That the mobility of many individual capital goocIB is considerably greater than is commonlj)! supposed, is shown in the interesting article by L. H. Seltzer, 1932.

320 Capitalistic Production under Competition PT. III Yet a policy which aims at keeping the future income stream at a given level will have to see that such re£ources are replaced by some produced means of production which will provide services that are completely equivalent to the former ones. And, as has already been pointed out several times, the existence of such wasting resources always makes it possible to provide for such replacement. It is difficult to estimate the extent to which it is likely that inventions directly affecting a particular industry will in this way increase the supply of capital available to the rest of industry. That in some parThe probability 01 capital-saving effects ticular cases it may be not inconsideroble 01 Inventions h dl b d . d B h hi' can ar y e enle. ut on t e woe It seems that, even if we add the cases now< being discussed to the more exceptional cases discussed before, capital saving inventions are distinctly less likely to occur than is usually supposed. It may be added here that what has been said about inventions as a typical example of unforeseen changes in the conditions of production, applies equally well to unforeseen changes in the supply of pure input, or to unforeseen changes in tastes. Both might be treated on exactly the same lines and classi fied according to their capital-saving or labour-saving tendencies.

If we conclude from the preceding discussion that capital-saving inventions are on a priori grounds unlikely to be more than comparatively rare exceptions, must we Effect. of Inventions also draw the same pessimistic conclusions on wages as have been drawn by others concerning the probable effect of technological progress on the income of labour? It seems that here too the concept of capital as a fund of fixed magnitude has led to erroneous conclu sions. It has been argued that if on the whole inventions are likely to make capital relatively more scarce in com parison with labour, and therefore to increase the return on a unit of capital relatively to the return on a unit of labour, the absolute and the relative share in the national CR. xxm Effect8 Of Unforeseen Change8 321 income of the capitalists as a class will as a rule be in creased by an invention, while the relative and sometimes even the absolute share of labour will be decreased. But· this argument seems unconsciously to assume that in the course of these changes the aggregate value of capital always remains the same, and that in consequence a higher percentage return on capital must also mean an increase in the aggregate income of capital.

We may here take it for granted, without explicit proof,1 that it will only be profitable to introduce an invention if aggregate output is thereby increased. But if at the same time the supply of capital and the supply of labour remained the same, but the remuneration of a unit of capital increased relatively to that of a unit of labour, it would appear that in any case the capitalists as a class would draw an increased product, and that what was left to labour would be a smaller relative share and perhaps even a smaller absolute amount. But, as we know, there is no reason to assume that in the course of such changes the aggregate value of capital will remain the same. In fact, to say the least, this is rather unlikely to be the case. Inventions Unlikelihood ihai In'11 . d d 1 t d t . th veotloos will decrease WI In ee as a ru e en 0 Increase e Ihe relative share 01 return on the capital available for invest-labour ment, but they will also decrease or destroy the return on some of the existing capital equipment. The increase in the percentage return on new capital available for invest ment does not therefore in any way prove that the aggregate income of capitalists will increase. It is not at all unlikely that the decrease in the absolute returns on the old capital goods will exceed the additional return on the capital which can be invested in new forms. In this case, although the invention is labour-saving in the ordinary sense, it may very well decrease not merely the 1 cr. Wicksell, Lectures, vol. i, pp. 133.143; J. R. Hicks, 1933, p. 121 ; and Kaldor, " A Case against Technical Progress? " Economica, No. 36 (May 1932).

2.2 322 Capitalistic Production under Competition PT. m relative share but even the absolute share of the capitalists. And there is in any event little reason to suppose that the return on the old capital goods will increase as the rate of interest increases; consequently the share of the capitalists will hardly ever increase in proportion to the increase in the rate of interest.