XXVII. Long-Run Forces Affecting the Rate of Interest

CHAPTER XXVII LONG-RUN FORCES AFFECTING THE RATE OF INTEREST HAVING considered in the last chapter the impact effect of any change in investment demand, we must now turn to the further repercussions of the changes we have observed. We have seen that, in a money economy, one of the effects of a change in the profitability of investment will be a release of money from (or art absorption of money into) idle balances and a consequent change in the size of the money stream which meets the stream of goods. We have not yet considered the effects on returns of this change in the money stream, since they will make them selves felt only after the short period with which we were concerned in the last chapter. The returns curve or investment demand schedule which then we treated as a given magnitude, or as an independent variable, will clearly be affected by changes in the size of the money stream. Before we Iollueoces determlo can analyse these effects it will be necessary log the shape of the Investment demand to make a somewhat closer examination of curve the factors which determine the shape of this curve in general, and at the same time to distinguish between the different ways in which the amount of investment can change and the effects of such changes on the returns curve.

So far we have not explicitly discussed the relation between the returns curve as used in the last chapter, which refers to the returns from successive amounts of money invested, and our earlier discussion of the pro ductivity of investment in real terms. But so long as we treat prices as given, as we were able to do for the purpose of the analysis of the last chapter, the relationship is so 369 25 370 The Money Rate of Interest PT. IV obvious that it hardly needs further explanation. Just as successive amounts of investment expressed in terms of any other unit will bring decreasing returns, so the invest ment of successive doses of such quantities of input as can be obtained for a given amount of money will also bring decreasing returns. Somewhat more careful consideration is needed of what exactly we mean here when we speak of an increase in investment. Strictly speaking, if we start from an initial M I f h equilibrium position where the existence of ean ng 0 • anges In the" amount of unused resources 1 is excluded by defini Investment tJ tion, an increase or decrease of investment should always mean a transfer of input from the pro duction of consumers' goods for a nearer date to the production of consumers' goods for a more distant date, or vwe versa. But where we assume that this diversion of input from one kind of production to another is ac companied, and in part brought about, by changes in total money expenditure, we cannot at the same time assume that prices will remain unchanged. It is, how ever, neither necessary nor advisable to adhere for our present purposes to so rigid a type of equilibrium assump tion. At any rate, so far as concerns the impact effects of a rise in investment demand which we discussed in . the last chapter, there is no reason why we should not assume that the additional input which is being invested has previously been unemployed, so that the increase in investment means a corresponding increase in the employment of all sorts of resources without any increase of prices and without a decrease in the production of conS'Umers' goods. This assumption simply means that.

there are certain limited quantities of various resources available which have been offered but not bought at current prices, but which would be employed as soon as 1 This means unused resources which coul$1 be had at the ruling market price. There ].ViII of course always be further reserves which will be offered only if prices rise.

CII.XXvn Long-run 1 nfiuences 371 dema~d at existing prices rose. And since the amount of such resources will always be limited, the effect of making this assumption will be that we must distinguish between the effects which an increase of investments and income will have while there are unused resources of all kinds available and the effects which such an increase will have after the various resources become successively scarce and their prices begin to rise. The initial change from which we started our discussion in the last chapter, an invention which gives rise to a new demand for capital, means that with given prices the margin between the cost of production and E i I fleet of a r se n the price of the product produced with ,the Inveslment demand ·11 b h· h th th I· on Incomes new process WI e Ig er an e ru mg , rate of profit, i.e. that the marginal rate of profit on the former volume of production will have risen. The first result of this, as we have seen, will be that investment will increase, the marginal rate of profit will fall, and the cash balances will decrease till the desire for holding the marginal units of the decreased cash balances is again just balanced by the higher profits which may be obtained by investing them. This new rate of profit will be somewhere between the old rate and the higher rate which would exist if investment had not increased. But since this additional investment has been financed by a release of money out of idle balances, incomes will have increased, and as a consequence the demand for consumers' goods will also increase, although probably not to the full extent, as some of the additional income is likely to be saved.

If we assume that there are unused resources available not only in the form of factors of production but also in the form of consumers' goods in all stages of completion, and so long as this is the case, the increase E" t f I I nee 0 a r se n in the demand for consumers' goods will Incomes on Investr • 1 d.J . menl demand J.or some time ea merl:1J.y to an mcrease in sales without an increase of prices. Such an increase of the quantity of output which can be sold at given 372 The Money Rate of Interest PT. IV prices will have the effect of raising the investment demand further, or, ll;lore exactly, of shifting our returns curve to the right without changing its shape. The amount that it will appear profitable to borrow and invest at any given rate of interest will accordingly increase; and this in turn will mean that, though some more money will be released from idle balances, the rate of interest and the rate of profit will be raised further.

And since this process will have raised incomes still further, it will be repeated: that is, every further increase in the demand for consumers' goods will lead to some further increase of investment and some further increase of the rate of profit. But at every stage of this process some part of the additional income will be saved, and as rates of interest rise, any given increase in final demand will lead to proportionally less investment. (Or, what is really the same phenomenon, only seen from a different angle, successive increases of investment demand will lead to the release of decreasing amounts of money from idle balances.) So the process will gradually slow down and finally come to a stop. Where will the rate of interest be fixed in this final equilibrium? If we assume the quantity of money to have remained constant, it will evidently be above the Final position 01 rate rate which ruled before the initial change of return occurred and even above the somewhat higher impact rate which ruled immediately after the change occurred, since every revolution of the process we have been considering will have raised it a little further.

But under our present assumptions there is no reason why, even when this process comes to an end, the rate of interest need have risen to the full extent to which it would have risen in the beginning had the supply of in vestible funds been entirely inelastic. Thus, under the conditions we have considered, the release of money from idle balances (and the same would of course be true of an increase in the quantity of money) may keep the CR. XXVII Long-run Influences 373 rate of profit and interest lastingly below the figure to which it would have risen without any such monetary change. Let us be quite clear, however, about which of our assumptions this somewhat, surprising result is due to. We have assumed that not only the supply of pure input but also the supply of final and inter-I Nature of assumpt ons mediate products and of instruments of underlying this anaall kinds was infinitely elastic, so that lysis every increase in demand could be satisfied without any increase of price, or, in other words, that the increase of investment (or we should rather say output) was possible without society in the aggregate or even any single individual having to reduce consumption in order to provide an income for the additional people now em ployed. Or, in other words, we have been considering an economic system in which not only the permanent resources but also all kinds of nonpermanent resources, that is, all 'forms of capital, were not scarce. There is indeed no reason why the price of capital should rise if there are such unused reserves of capital available, there is even no reason why capital should have a price at all if it were abundant in all its forms. The existence of interest in such a world would indeed be due merely to the scarcity of money, although even money would not be scarce in any absolute sense; it would be scarce only relatively to given prices on which people were assumed to insist. By an appropriate adjustment of the quantity of money the rate of interest could, in such a system, be reduced to practically any level.

Now such a situation, in which abundant unused reserves of all kinds of resources, including all inter mediate products, exist, may occasionally prevail in the depths of a depression. But it is certainly Mr. Keynes' eco not a normal position on which a theory nomlcs of abundance claiming general a,Pplicability could be based. Yet it is some such "world as this which is treated in Mr. Keynes' 374 The Money Rate of Interest PT. IV General Theory of Employment, Interest and Money, which in recent years has created so much stir and confusion among economists and even the wider public. Although the technocrats, and other believers in the unbounded productive capacity of. our economic system, do not yet appear to have realised it, what he has given us is really that economics of abundance for which they have been clamouring so long. Or rather, he has given us a system of economics which is based on the assumption that no ~eal scarcity exists, and that the only scarcity with which we need concern ourselves is the artificial scarcity created by the determination of people not to sell their services and products below certain arbitrarily fixed prices. These prices are in no way explained, but are simply assumed to remain at their historically given level, except at rare intervals when "full employment" is approached and the different goods begin successively to become scarce and to rise in price.

N ow if there is a well-established fact which dominates economic life, it is the incessant, even hourly, variation in the prices of most of the important raw materials and of the wholesale prices of nearly all foodstuffs. But the reader of Mr. Keynes' theory is left with the impression that these fluctuations of prices are entirely unmotivated and irrelevant, except towards the end of a boom, when the fact of scarcity is readmitted into the analysis, as an apparent exception, under the designation of "bottle necks ".1 And not only are the factors which determine the relative prices of the various commodities systematic1 I should have thought that the abandonment of the sharp dis tinction between the " freely reproducible goods " and goods of absolute scarcity and the substitution for this distinction of the concept of vary ing degrees of scarcity (according to the increasing costs of reproduction) was one of the major advances of modem economics. But Mr. Keynes evidently wishes us to return to the older way of thinking. . This at any rate seems to be what his use of the concept of " bottlenecks" means; a concept which seems to me to belong essentially to a naive early stage of economic thinking and the introduction of which into economic theory can hardly be regarded as an improvement.

CR. XXVII Long-run I njluences 375 ally disregarded; 1 it is even explicitly argued that, apart from the purely monetary factors which are supposed to be the sole determinants of the rate of interest, the prices of the majority of goods would be indeterminate. Although this is expressly stated only for capital assets in the special narrow sense in which Mr. Keynes nses this term, that is, for durable goods and securities, the same reasoning would apply to all factors of production. In so far as " assets" in general are coricerned the whole argu ment of the General Theory rests on the assumption that their yield only is determined by real factors (i.e. that it is determined by the given prices of their products), and that their price can be determined only by capitalising this yield at a given rate of interest determined solely by monetary factors. 2 This argument, if it were correct, would clearly have to be extended to the prices of all factors of production the price of which is not arbitrarily fixed by monopolists, for their prices would have to be equal to the value of their contribution to the product less interest for the interval for which the factors remained invested. 3 That is, the difference between costs and prices would not be a source of the demand for capital but would be unilaterally determined by a rate of interest which was entirely dependent on monetary influences.

1 It is characteristic that when at last, towards the end of his book, Mr. Keynes comes to discuss prices, the " Theory of Price" is to him merely" the analysis of the relations between changes in the quantity of money and changes in the price level" (General Theory, p. 296). • Cf. General Theory, p. 137: "We must ascertain the rate of interest from some other source and only then can we value the asset by • capitalising' its prospective yield". 3 The reason why Mr. Keynes does not draw this conclusion, and the general explanation of his peculiar attitude towards the problem of the determination of relative prices, is presumably that under the influence of the .. real cost" doctrine which to the present day plays such a large r6le in the Cambridge tradition, he assumes that the prices of all goods except the more durable ones are even in the short run determined by costs. But whatever one may think about the useful ness of a cost explanation of relative prices in equilibrium analysis, it should be clear that it is altogether useless in any discussion of problems of the short period.

376 The Money Rale of Interest PT. IV We need not follow this argument much further to see that it leads to contradictory conclusions. Even in the case we have considered before of an increase in the Basic Impol1ance 01 investment demand due to an invention, .. &relly the mechanism which restores the equality between profits and interest would be inconceivable without an independent determinant of the prices of the factors of production, namely their scarcity. For, if the prices of the factors were directly dependent on the given rate of interest, no increase in profits could appear, and no expansion of investment would take place,since prices would be automatically marked to make the rate of profit equal to the given rate of interest. Or, if the initial prices were regarded as unchangeable and unlimited supplies of factors were assumed to be avail able at these prices, nothing could reduce the increased rate of profit to the level of the unchanged rate of interest.

It is clear that, if we want to understand at all the mechanism which determines the relation between costs and prices, and therefore the rate of profit, it is to the . relative scarcity of the various types of capital goocts and of the other factors of production that we must direct our attention, for it is this scarcity which deter mines their prices. And although there may be, at most times, some goods an increase in demand for which may bring forth some increase in supply without an increase of their prices, it will on the whole be more useful and realistic to assume for the purposes of this investiga tion that most commodities are scarce, in the sense that any rise of demand will, ceteris paribus, lead to a rise in their prices. We must leave the consideration of the existence of unemployed resources of certain kinds to more specialised investigations of dynamic problems. This critical excursion was unfortunately made neces sary by the confusion which has reigned on this subject since the appearance of Mr. Keynes' General Theory. We may now return to our main subject, the effect of a rise CH. XXVII Long-run Influences 377 in incomes and final demand on the investment demand schedule and the rate of interest. The case which we shall now take up is the situation that will arise once the increased demand for consumers' goods can no EtJ. f 1 ec. 0 an ocrease longer be satisfied at constant costs because of 110al demand on at least some of the factors from which profit schedule additional consumers' goods would have to be produced become definitely scarce. It does not matter for our purpose whether this occurs immediately, as soon as incomes and the demand for consumers' goods increase, or not until later, for, as we have seen, the process by which increased investment increases final demand, and increased final demand increases investment further, will go on for some time. We are now concerned not with the transitory effects which occur while any unused capacity caused by a previous slump is being absorbed ~ the analysis of this is the proper subject of dynamic studies - but with the way in which the influence of scarcity win reassert itself once this slack in the system has been taken up.

Sooner or later, then, the increase in the demand for consumers' goods will lead to an increase of their prices 1 and of the profits made on the production of consumers' goods. But once prices begin to rise, the additional demand for funds will no longer be confined to the pur poses of new additional investment intended to satisfy the 1 We must not allow ourselves to be misled by the fact that for special reasons connected with the imperfectly competitive character of many retail markets, retail prices of consumers', goods are notoriously sluggish in their movements. The fact apparently is that for the individual retailer the price elasticity of the demand for his products is too low (and selling costs, so long as incomes of his customers are constant, too high) to make it worth his while to increase sales by lowering prices, although (if we exclude selling costs) he may be operating under decreasing costs. But this does not exclude the possibility that when demand increases he may be able to expand his sa1es at decreasing costs and increasing profits and that he will there fore be able to offer higher prices to the wholesalers. For this reason it is probably wholesale prices and not retail prices of consumers' goods which are relevant for the purposes of the present discussion.

378 The Money Rate of Interest PT. IV new demand. At first - and this is a point of importance which is often overlooked - only the prices of consumers' goods, and of such other goods as can rapidly be turned At lint th, ratil of into consumers' goods, will rise, and conse profit will rls. In th. quently profits also will increase only in lat. SUips of production only the late stages of production. In order that the rise of prices should become general and should exert a proportional effect on the prices of all the various factors of production, as appears commonly to be assumed to be the normal case, it would not only be necessary that producers in all stages should be put in a position at one and the same time to spend proportionately more; it would also be necessary that the increase in incomes which would be caused by this increased spending should not lead to any further increase in the demand for consumers' goods and a further increase of their prices. Otherwise the prices of consumers' goods would always keep a step ahead of the prices of factors. That is, so long as any part of the additional income thus created is spent on consumers' goods (i.e. unless all of it is ~aved), the prices of consumers' goods must rise permanently in relation to those of the various kinds of input. And this, as will by now be evident, cannot be lastingly without effect on the relative prices of the various kinds of input and on the methods of production that will appear profitable.

The general nature of the price mechanism that will be set in operation, and of the effects this kind of change will have on the volume of investment generally, is Th. rise of the rate of already familiar to us from discussion. profit cannot be wiped in an earlier Part of this book. It will, out by a proportional rlseo! all other prices however, be useful to re~state it now in monetary terms. The starting point for this analysis must be the fact that, whether the increase in investment 1 is 1 The reason why throughout the following argument we shall con· centrate on an increase in the demand for consumers' goods is that we want to bring out the significance of the scarcity of consumers' goods (or of" capital" - which amounts to the same thing) as clearly as possible. But the argument would of course, mutatis mutandis, CR. XXVII Long-run Influences 379 brought about by employing, in the production of invest ment goods, formerly unemployed input, or by employing apply equally to the case of a fall in the demand for consumers' goods.

The main source of the erroneous conceptions which rule in this field is a false analogy to the effect of changes in the expected return and the rate of interest on the price of assets (mainly securities) which are capable of giving only one particular kind of return. The price of a fixed interest-bearing bond, e.g., will (disregarding for our purpose the effects of various degrees of risk) always be equal to the value of the expected yield, discounted at the current rate of interest; and its price will therefore change in inverse proportion to the rate of interest and in direct proportion to any change in returns if such should occur. The situation is, however, altogether different with regard to factors of production which can be used in, various ways so as to give different returns at different dates. And this holds even for com pletely specific factors of production which can be used only for one particular purpose, provided they co-operate with other factors which fall in the former category. (The only kinds of real assets which in effect would be similar to securities in this respect would be durable consumers' goods which neither need the co.operation of any other factors to yield their services nor can be used more or less intensively so as to last a shorter or longer time.) In order to obtain a valid analogy to the determination of the prices of capital goods in the field of securities, we should have to conceive of securities which not only entitled the owner to different options of various sorts (corresponding to different uses to which productivr resources can be put), but some at least of which would bring a return only if owned in certain combinations with other securities. If this were the case it would clearly be possible for, e.g., a given rise of the rate of interest to reduce the price of one group of securities which carried a title only to one fixed series of returns by much more than it reduced the price of another group which conferred an option on a shorter series of larger returns as an alternative to the same series of fixed returns. And if some securities could be used only in com bination with others so as jointly to entitle the owner to a certain return, it might well be the case that a rise in the rate of interest would lower the price of the first kind of securities a great deal and at the same time raise the value of the second kind of securities. This would happen if the value of the first kind largely depended on a long series of small returns which could be obtained as an alternative to the use in com bination with the second kind of security, this latter use providing an outlet for only a very small part of the total amount of the first kind.

In this case the value of the first kind of security would depend almost exclusively on this independent use and would be reduced a great deal by a rise in the rate of interest. If, on the other hand, the return from the joint use of both kinds of security were one large sum in the near future, a rise in the rate of interest would affect the present value of 380 The Money Rate of Interest p'r. IV input formerly used in the production of consumers' goods, the remuneration of input in general in terms of con sumers' goods must fall unless the owners of the input voluntarily reduce their consumption. In the first case a given output of consumers' goods will have to be divided among a greater number of income-receivers, while in the second case a reduced output of consumers' goods will have to be divided among the same number of income-receivers. In such a situation no monetary change can alter the fact that relatively to every unit of input employed there is less output available, and that, therefore, unless people spontaneously decide to save correspondingly more, the price of input in terms of final output must fall. l But if it is impossible in such a situation for the prices of all kinds of input to rise in proportion to the rise in the price of output, and as the value of nearly all input this joint return very little. But as the part of this joint return that would have to go to the first kind of security (determined by its value in other uses) would be reduced much more, the value of the second kind of security would increase.

Now this sort of thing, which in the realm of securities would be a freak case and very unlikely to be of any importance, may well occur with productive resources. And even if here too the rule should prove to be that a rise in the rate of interest will reduce the value of income· bearing assets, this will be true to so varying an extent and subject to so many exceptions, that the analogy to the normal case of securities will be very misleading. The point to keep in mind is that, with real productive resources, the yield can as a rule be varied and will be deliberatcly varied in response to changes in their prices, and that it will not be the greatest absolute yield but the highest time rate of yield which will guide their use. We shall later see that on this last point, which of course distinguishes the theory of capital from timeless productivity analysis, it is analogies to the latter which have provided the second important source of error.

1 Even if present money prices were instantaneously" marked up " in full proportion to the rise in the price of the product (or even the expected rise in the price of the output), this could lead only to a continuous and progressive rise in the price of output which would always exceed entrepreneurs' expectations till they realised that, how. ever great the increase in the prices paid for the input, they could not prevent these prices from falling relatively to the price of output, and that therefore it would be better to adapt their methods of production to the new price relations.

CR. XXVII Long-run Influences 381 in terms of output must fall to some extent, it is also im possible for the rise in the price of output to leave the relative prices of the different kinds of input unaffected. If this were so, i.e. if the pre-existing prices of the different kinds of input continued to prevail, the given rise in the price of out put would mean very different changes in the rates of profit earned on different kinds of input. For, although a given rise in the The Increase In the dillerence between the price of output and the prices of input generally must iead te changes in the reialive prices of dif ferent kinds o/Input price of all output would, of course, increase the difference between the price of any unit of input producing a given marginal product and the price of that marginal product by the same amount, it would clearly change the time rate of profit earned on different units of input to very different degrees according to the periods for which the different units of input were invested. 1 If in the previous 1 The argument can be illustrated by a simple diagram. If along the abscissa Ot we represent investment periods, and along the ordinate ---~ '" <> r U' T' 5' __ -:;:::---, ., , ~p T U ~ p~~~==~~ ____ -+5 ________ ~ __ _ '" ~ o 2 3 year years years InlJestment period FIG. 28 t Or values of units of input and of their marginal products (using for this axis a logarithmic scale), we can represent the value of the 382 The Money Rate of Interest PT. IV equilibrium position the margin between the price of a unit of input and the price of its marginal product corre sponded everywhere to a uniform time rate of 6 per cent, the difference between the price of a unit of input invested two years before the completion of the product and the price of its marginal product would be 12 per cent,l while the difference bet'ween the price of a unit of input invested only one month before the completion of the product and the price of its marginal product would be only one-half of 1 per cent. A rise in the price of the product by 2 per cent, which would increase these margins to 14 and 2! per cent respectively,.· would increase the per annum rate of profit on the former to only 7 per cent but would increase the per annum rate of profit on the latter to 30 per cent. 2 marginal products of equal units of input invested for various periods by an upward sloping line P U. The difference between the price of a unit of input (OP) and the value of its marginal product corresponds in all cases to a uniform rate of interest represented by the slope of the line PU.

Assume now that the price of all output is raised by a given pro portion while the price of the input remains unchanged. The increase in the price of the unchanged quantities of output due to the various inputs can be shown in the diagram by raising the line PU, without changing its slope, to some position such as P' U'. The time rates of profit that will now be earned on input invested for various periods are shown by the slopes of the lines PR', PS', PT', etc., and it will be seen that the rate of profit now earned on one year's investment will be greater than that earned on two years' investment, the latter greater than that on three years' investment, and so on, all these rates being higher than the previous uniform rate of profit represented by the slopes of the lines PU and P'U'. And as the rate of profit on invest ment for different periods will have changed to a different extent, so the demand for input for investment for different periods will have changed.

If the price of input had risen in proportion to the rise in the price of output, i.e. to Opl, the rate of profit earned on the different invest Inent periods would still be uniform and the same as that which was earned before the rise in the price of the output. And the relative demand for input for these various forms of investment would have been unchanged. But although this assumption is implicitly contained in the usual analysis of these phenomena, it is, as we have seen, an illegitimate assumption to make. 1 Disregarding compound interest. S The example is worked out lnore fully in Hayek, 1939, p. 8.

OR. XXVII Long-run Influences 383 If we assume that at first the rate of interest at which money can be borrowed remains unchanged or rises only very little, it is clear that we have here a state of affairs which cannot last. And we have already Elrect 01 dllYerenoe 01 seen that the situation cannot be remedied by simply raising the prices of all input in proportion to the rise in the price of the output, for there is not enough output various magnitudes between the value of Input and the dis counted value of It. marginal produot to go round. The given output which has now to be distributed among a larger number of claimants (or the decreased output which has now to be distributed among an unchanged number 6f claimants) will still have to be distributed according to the discounted marginal pro ductivity of the various kinds of input. Entrepreneurs will still tend to bid up the prices of the various kinds of input to the discounted value. of their respective marginal products, and, if the rate at which they can borrow money remains unchanged, the only way in which this equality between the price of the input and the dis counted value of its marginal product can be restored, is evidently by reducing that marginal product.

This conclusion may at first appear paradoxical because it means, firstly, that input will have to be switched from uses where its marginal product is higher to uses where its marginal product is lower, and, secondly, that the marginal pro ductivity of all kinds, or of nearly all kinds, of input will have to be lowered at the same time. But if we cling to the Changes In pro ductive combinations (methods 01 produc tion) In order to adjust marginal produetlv ltIes two basic considerations: (a) that it is impossible under the conditions considered for prices of all input to rise in proportion to the prices of output, and (b) that until equality between the price of input and the discounted price of its marginal products is restored, it will be the rate of profit earned on the various uses of the input which will guide entrepreneurs in making their decision, the answer is not difficult to find. Perhaps we may 384 The Money Rate of Interest PT. IV begin by pointing out that, although this case is not usually considered in elementary marginal productivity analysis, the marginal productivity of all kinds of concrete input can of course be lowered at the same time by chang ing over to less capitalistic, and therefore less productive, methods of production. And to this consideration we need only add that, as we have already seen, if the difference between the price of a unit of input and the price of a unit of output increases, a smaller marginal product maturing at a nearer date may well represent a higher time rate of profit, and therefore appear more attractive, than the larger marginal product in the more distant future which, before the rise in the price of the product, promised the higher rate of return.

With the help of these general considerations we can now show in more detail what will happen to the prices of the different kinds of input. The r'ate of profit to be I II earned at the pre-existing prices will have lnlluence on re a ve prices 01 dIfferent increased most, and demand will therefore kinds 01 Input. t £ th k' d f' t InCreaSe mos , lor ose In s 0 lnpu which can be rapidly turned into consumers' goods. Whether and to what extent their prices will be raised in consequence, will depend on how easily the quantity of these kinds of input available for the rapid production of consumers' goods can be increased by transfers from other uses where demand is less urgent, from unused reserves, etc. Those kinds of input of which the supply for these purposes is very elastic will be used in very much greater quantities in proportion to others, so that their marginal productivity will be much reduced, and the gap between their price and the discounted price of their marginal product will be closed mainly by a decrease of that marginal product. For others, of which the quantity used for the production of consumers' goods in the near future cannot be easily increased, the marginal pro ductivity may be decreased only a little, and the gap will be mainly closed by a rise in their price, although this rise eH.XXVII Long-run Influences 385 will be smaller than that of the final product. And for still others, the quantity of which cannot be increased at all in the short run, the marginal productivity may actually be raised by the increased use of other co operating factors, and in order to adjust the margin between their price and the increased price of their marginal product their price may have to rise a great deal.

Experience shows that this happens during booms with respect to some raw materials, the price of which rises proportionately more than the price of the final product in the production of which they are used. Generally speaking, we may say that resources of which the greater part has already been used befol'e in what we have called late stages of production will rise the more in price the nearer they are to the final output, and will ha ve to be economised to a correspondingly greater extent. And in so far as such resources can be reproduced, their production will become relatively more or less profi ta ble according as they are nearer to, or further from, final output. Resources which can be directly transferred nearer to the consumption stage will generally be used in a much greater proportion for invest ments for shorter periods than for investments for longer periods. In the end we shall find that, for nearly all factors, the productivity has beer reduced by changing to productive combinations where they bring a snlaller marginal product at a nearer date. And, in general, the demand will have increased for those factors which can be made to yield in the nearer future a return not very much smaller than that they yielded before, while the demand will actually have fallen for those which in the near future can bring no return, or only one which is very small compared with that which they can bring in the distant future. And while all these changes have been brought about by the increase in the margin between the price of a given unit of input and the price of a given unit of output (or the rate of profit), in eonsequence of these 386 The Money Rate of Interest PT. IV adjustments the marginal rate of profit (or the marginal price margins) will have been reduced again so as to correspond to the given rate of interest.

At this stage of the exposition it is scarcely necessary to explain at length why such an increase in the demand for investments for short periods, combined with a decrease in the demand for investments for long Elfect on the proportional amount 01 In-periods, will decrease the total amount of vestment investment that will be made to provide a given output. But as one is easily misled by con siderations which apply only to stationary conditions where of course the current input that is required to maintain a given output is smaller with a large stock of capital than with a small one - it may be useful briefly to restate the reasons why, d'uring the transition from more to less capitalistic methods of production, the amount of input that will be demanded for investment purposes will fall. That this must be so is easy enough to see in simple cases. If a given amount of machine service, which in the past has been provided by machines lasting ten years, is from a certain date onwards main tained by replacing every machine that wears out by a cheaper one that lasts only five years, this will for a time reduce the amount of input that has to be invested in machines in order to maintain the stream of machine service at an unchanged level. The same applies to every other kind of investment, no matter whether we have to deal with the substitution of less for more durable goods, of less labour-saving machinery for more labour-saving machinery, or of shorter for longer processes of production in the literal sense of the term. In all these cases the amount of investment for short periods that is made more profitable by this transition will for some time be smaller than the amount of investment for long periods that is made less profitable.

We can describe this effect in terms of the investment demand schedule by saying that the curve describing the OK. XXVII Long-run Influences 387 demand for real input for investment purposes as a function of the rate of interest is tilted so that its upper end is raised and its lower end is lowered. It is diffi cult to express this exactly, since we are Th" II .. f h • til ng 0 I • not dealing with one homogeneous kind Investment demand f . d' th 1 t' 1 f schedule o Input an sInce e re a lve va ues 0 the different kinds of input will necessarily change in the course of the process. But in terms of any given system of prices (or if we i assume that there is only one homogeneous kind of input) we can say that as a conse quence of the rise in price margins (and therefore of the rate of profit on the given volume and method of production) the curve describing the amount of input (measured 0 along Op), which will be demanded at any p FIG. 29 rate of interest (measured along Oi), will shift from a posi tion like the one represented by curve a in the diagram, to a position like that represented by curve b or c. The reason for this tilting of the curve is that for every quantity of input that is more intensely demanded at a given rate of interest, a larger quantity of input will be less intensely demanded. The change in the relative profitability of the different kinds of investment will mean that the various investment opportunities will change their relative position on the investment demand schedule, and since for every quantity of input for which the demand increases there will be a larger quantity for which the demand decreases, the shape of the whole curve will be altered in the way indicated in the diagram.

388 The Money Rate of Interest PT. IV The conclusion which we must draw from these con siderations may at first appear somewhat paradoxical. It is that at any given rate of interest (except a very high The amount of in-one) the proportional amount of invest vestment per unlt of ment that will be called forth by any given output changes inversely with rate of final demand will be smaller with a high profit " rate of profit " (that is, large price margins or a low value of input in terms of output) and larger with a low rate of profit. Or, in other words, the amount of investment that, with a given final demand, will be required to bring the marginal rate of profit down to a figure equal to any given rate of interest, will be smaller when the "rate of profit" (price differences between given quantities of input and output) is high, and larger when the" rate of profit" is low. But, however paradoxical this conclusion may appear to those who have been brought up in the popular under-consumptionist views, it is no more paradoxical than the undeniable fact that certain kinds of investment which were profitable at a high rate of interest will cease to be profitable at a low rate of interest. It is evident and has usually been taken for granted that methods of production which were made profitable by a fall of the rate of interest from 7 to 5 per cent may be made unprofitable by a further fall from 5 per cent to 3 per cent, because the former method will no longer be able to compete with what has now become the cheaper method. It is true, however, that it is scarcely possible adequately to explain this, if one thinks only of the direct effect of a change in the money rate of interest on cost of production, and does not pro ceed to consider the changes in relative prices which ultimately govern the profitability of the various methods of production. It is onJy via these price changes that we can explain why a method of production which was profitable when the rate of inkrest was 5 per cent should become unprofitable when it falls to 3 per cent. Similarly, it is only in terms of price changes that we can adequately CII. XXVII Long-run I nfiuences 389 explain why a change in the rate of interest will make methods of production profitable which were previously unprofitable.

The most important conclusion, then, which emerges hom this discussion is that the method of production that will be adopted, or the proportional amount of capital that it will be profitable to use, will depend not on the rate of interest at which money can be borrowed but on the relations between different prices and· the shape of the profit schedule (or investment demand schedule) as determined by these price differences. And these relative prices will in turn depend on the relative scarcity of the various kinds of resources compared with the direction of demand. The rate of interest will, in the main, deter mine only to what point on the schedule investment will. be carried, that is, it will determine only the marginal rate of profit, and, through the latter, it will exercise a minor influence on the volume of ontput that it is profitable to produce with a given demand. The volume of invest ment, however, will depend as much if not more on how much investment it will be profitable to undertake in order to obtain a certain output. And with a high "rate of profit" any given marginal rate of profit will be reached with relatively little investment per unit of output, because with a high rate of profit the invest mend demand schedule will be steep, while with a low "rate of profit" the same marginal rate of profit will only be reached with much more investment per unit of output, because the investment demand schedule will be flat. l 1 For a discussion of the significance of these effects for the ex planation of industrial fluctuations and particularly their relation to the so-called " acceleration principle of derived demand", see Hayek, 1939. In particular it is shown there that the " rate of profit" deter mines the "multiplier" with which the "acceleration principle"

operates (or Mr. Harrod's "Relation "), and that changes in this mul.tiplier are likely to have a greater effect on the volume of invest ment than the second of the two factors which determine the accelera tion effect (the" multiplicand "), namely final demand.

390 The Money Rate of Interest PT. IV We must now return to the problem of the effect of all these changes on the money rate of interest (and the marginal rate of profit) which, in the discussion of these The determination of changes, we have so far treated as given. the money rate of The effect will clearly depend on what Interest and the marginal rate of proDt happens (a) to the shape of the investment demand schedule in monetary terms, and (b) to the supply of money. So far as the latter is concerned, we shall here continue to assume that the supply of basic money is fixed, so that all we need to concern ourselves with is the increase in the supply of investible funds at increasing rates of interest due to the release of money from idle balances. (This can, of course, be interpreted to include any increase in the credit superstructure erected on the given cash basis.) The supply side may therefore be represented by a curve like the one used before in Fig. 27 (Chapter XXVI, p. 363 above), and our main problem will be what will happen to the monetary investment demand schedule. This we can deduce from what happens to this demand schedule described in real terms. All we need to do is to show the effect of the price changes we have already discussed on the demand for investible funds, or to redraw this demand schedule, in terms of the new prices, instead of using the pre-existing price (which is what expressing it in " real terms" essentially means in this connection).

We have seen that, when expressed in real terms, the demand schedule will be tilted by a rise in the price of the final product, that is, that it will be raised on the Ch 1 left and lowered on the right. But as the anges n the monetary Investment de-various kinds of input that will now be in mand schedule d d ·11 I . . . . greater eman WI a so rlse In prlce In different degrees, less real investment will be associated with the investment of any given amount of money, i.e. the monetary investment demand curve, besides being tilted, will also be shifted to the right. Any given amount of real investment, i.e. any amount of investment correCR. XXVII Long~run Influences 391 sponding to a given marginal rate of profit and leading to a given amount of product, will require more money and will therefore cause the rate of interest to rise to an extent which will depend on the shape of our liquidity preference schedule. If the supply of investible funds were completely inelastic, and the amount invested could not be increased at all, this would clearly mean a con siderable rise in the rate of interest and a decrease in the amount of real investment. The tilting of the demand schedule would in this case have the effect of making it profitable to employ less input of the kind which rises ilL price and more input of the kind which falls in price.

(This, incidentally, also illustrates how misleading it is to concentrate on the existence of unemployed resources or " full employment" as the case may be, and to argue in terms of changes in the general price level, the movements of which are supposed to depend on whether full employ ment exists or not. What is relevant is not whether full employment exists, but whether the particular kinds of resources needed exist in the proportions corresponding to the state of demand.) If, however, the supply of investible funds is not altogether inelastic, and the increase of demand for in vestible funds brings forth an increased supply, the rate of interest will not rise to the full extent of the ElIect on Intere.t rates rise in the demand curve, and real invest-when supply of money ·11 b ·1 d h Is eIaslic ment WI not e curtal e so muc or may not be curtailed at all, and may even rise further. But as this means a further increase in money incomes, it will lead to a new increase in the monetary demand for con sumers' goods, a further rise in their prices, and con sequently a further tilting of the real investment demand schedule and a tilting and shifting of the monetary invest ment demand schedule. The elasticity of the supply of money, which in the short run tends to keep the rate of interest low, has thus the effect - at least for some time of simultaneously raising the rate of return on the invest392 The Money Rate of Interest PT. JV ment of any given amount of money and lowering the amount of real investment that will correspond to it.

And since every further increase in money incomes will strengthen this tendency, this process must go on till the combined effect of the tilting of the investment demand curve and of the rise in the rate of interest finally out balances the effect of the further rise in demand, and thus prevents a further increase in the amount of money invested. Either of these two factors alone may bring about this effect. Elsewhere 1 we have tried to show that, even if the supply of money were perfectly elastic and the rate of interest therefore kept constant, the "tilting" effect by itself would in the end bring further expansion to a stop. And we have seen before that, if the supply of money were perfectly inelastic, the rise in the rate of interest would prevent the expansion before the tilting effect could occur. In real life, however, the two factors, the tilting of the investment demand curve and the rise in the rate of interest, will as a rule work conjointly. In such circumstances the process of expansion will come to an end only when the rate of interest and the marginal rate of profit have been kept low by monetary expansion for a long enough time to allow the repercussions, through changes in relative prices and price margins, to have so changed the slope of the investment demand curve as both greatly to reduce the amount of real investment which is profitable at a given rate of interest, and greatly to increase the amount of money which is required for that amount of investment, thus raising the rate of interest corresponding to any given supply of money.

We cannot at this stage attempt any more exhaustive treatment of this complex mechanism. For it would comprise a discussion of the whole subject of industrial fluctuations and we should require a separate book to deal with it adequately. Some further considerations which are 1 See Hayek, 1939, pp. 24 et seq.

CH.XXVII Long-run Influence8 393 relevant in this connection will be added in the final chapter of this book. We will conclude the present treat ment by once more stressing the fact that, though in the short run monetary influences may delay The basic Importance the tendencies inherent in the real factors of the real factors from working themselves out, and temporarily may even reverse these tendencies, it will in the end be the scarcity of real resources relative to demand which will decide what kind of investment, and how much, is profitable. The fundamental fact which guides production, and in which the scarcity of capital expresses itself, is the price of input in terms of output, and this in turn depends on the proportion of income spent on consumers' goods compared with the proportion of income earned from the current production of consumers' goods. These pro portions cannot be altered at will by adjustments in the money stream, since they depend on the one hand on the real quantities of the various types of goods in existence, and on the other hand on the way in which people will distribute their income between expenditure on con sumers' goods and saving. Neither of these factors can be deliberately altered by monetary policy. As we have seen, any delay by monetary means of the adjustments made necessary by real changes can only have the effect of further accentuating these real changes, and any . purely monetary change which in the first instance deflects interest rates in one direction is bound to set up forces which will ultimately change them in the opposite direction.

Ultimately, therefore, it is the rate of saving which sets the limits to the amount of investment that can be successfully carried through. But the effects of the rate of saving do not operate directly on the TheslgnlftcaneeoUhe rate of interest or on the supply of in-rate of saving vestible funds, which will always be influenced largely by monetary factors. Its main influence is on the demand for investible funds, and here it operates in a direction 394 The Money Rate of Interest PT. IV opposite to that which is assumed by all the under consumptionist theories. It will be via investment demand that a change in the rate of saving will affect the volume of investment. Similarly, it will be via investment de mand that, if monetary influences should have caused investment to get out of step with saving, the balance will be restored. If throughout this discussion we have had little occasion to make explicit mention of the rate of saving, this is due to the fact that the effects considered will take place whatever the rate of saving, so long as this is a given magnitude and does not spontaneously change so as to restore the disrupted equilibrium. All that is required to make our analysis applicable is that, when incomes are increased by investment, the share of the additional income spent on consumers' goods during any period of time should be larger than the proportion by which the new investment adds to the output of consumers' goods during the same period of time. And there is of course no reason to expect that more than a fraction of the new income, and certainly not as much as has been newly invested, will be saved, because this would mean that practically all the income earned from the new in vestment would have to be saved.!

1 The rate at which a given amount of new investment will con tribute during any given interval of time to the output of consumers' goods stands of course in a very simple relation to the proportion between any new demand and the amount of investment to which it gives rise: the latter is simply the reciprocal value of the former. For a fuller discussion of this relationship between this "quotient" and the " multiplier" with which the" acceleration principle of derived demand " operates I must again refer to Hayek, 1939, pp. 48-52. It cannot be objected to this argument that, since investment automatically creates an identical amount of saving, the situation contemplated here cannot arise. The irrelevant tautology, that during any interval oj time the amount of income which has not been received from the sale of consumers' goods, and which therefore has been saved (namely, by those who spent that income), must have been spent on something other than consumers' goods (and therefore ex definitione must have been invested), is of little significance for this or for any other economic problem. What is relevant here is not the relation between one classification of money expenditure and another, but the OR. XXVII Long-run Influences 395 The relative prices of the various types of goods and services, and therefore the rate of profit to be earned in their production, will always be determined by the impact of the monetary demand for the various The supply or capital k· d f d d th l' f th and the rate or profit In s 0 goo s an e supp les 0 ese and Interest In dlsgoods. And unless we study the factors equilibrium limiting the supplies of these various types of goods, and particularly if we assume, as Mr. Keynes does, that they are all freely reproducible in practically unlimited quantities relation of two streams of money expenditure to. the streams of goods which they meet. We are interested in the amount of investment because it determines in what proportions (in terms of their relative costs) different kinds of goods will come into existence. And we are interested to know how these proportions between quantities of different kinds of goods are related to the proportions in which money expenditure will be distributed between the two kinds of goods, because it depends on the relation between these two proportions whether the production of either kind of good will become more or less profitable. It does not matter whether we put this question in the form of asking whether the distribution of income between expenditqre on consumers' goods and saving corresponds to the proportion between the relative (replacement) com of the total supply of consumers' goods and new investment goods, or whether the available resources are now distributed in the same proportion between the production of consumers' goods and the pro.

duction of investment goods as those in which income earned. from this production will be distributed between the two kinds of goods. Which· ever of the two aspect3 of the question we prefer to stress, the essential thing, if we want to ask a meaningful question, is that we must always compare the result of investment embodied in concrete goods with the money expenditure on these goods. It is never the investment which is going on at the same time as the saving, but the result of paet invest· ment, that determines the supply of capital goods to which the monetary demand mayor may not correspond. Playing about with the relation· ships between various classifications of total money expenditure during any given period will lead only to meaningless questions, and never to any result of the slightest relevance to any real problem. I do not wish to suggest that the recent discussions of the various meanings of these concepts have been useless. They have helped us to make clear the conditions under which it is meaningful to talk about relations between saving and investment. But now that the obscurities and confusions connected with these concepts have been cleared up, the meaningless tautological use of these concepts ought clearly to disappear from scientific discussion. On the whole question, and the recent discussions about it, compare now the excellent exposition in the new chapter eight of the second edition of Professor Haberler's Prosperity and Depression (Geneva, 1939).

396 The Money Rate of Interest Fl.'. IV and without any appreciable lapse of time, we must remain in complete ignorance of the factors guiding production. In long-run equilibrium, the rate of profit and interest will depend on how much of their resources people want to use to satisfy their current needs, and how much they are willing to save and invest. But in the comparatively short run the quantities and kinds of consumers' goods and capital goods in existence must be regarded as fixed, and the rate of profit will depend not so much on the absolute quantity of real capital (however measured) in existence, or on the absolute height of the rate of saving, as on the relation between the proportion of the incomes spent on consumers' goods and the proportion of the resources available in the form of consumers' goods. For this reason it is quite possible that, after a period of great accumulation of capital and a high rate of saving, the rate of profit and the rate of interest may be higher than they were before - if the rate of saving is insufficient compared with the amount of capital which entrepreneurs have attempted to form, or if the demand for consumers' goods is too high compared with the supply. And for the same reason the rate of interest and profit may be higher in a rich community with much capital and a high rate of saving than in an otherwise similar community with little capital and a low rate of saving.l 1 For some further discussion of this point see Hayek, 1937.