Individualism and Economic Order

XI. The Ricardo Effect1

Machinery and labour are in constant competition, and the former can frequently not be employed until labour rises.

—David Ricardo.

1

WHEN in a recent essay on industrial fluctuations the author introduced “the familiar Ricardian proposition that a rise in wages will encourage capitalists to substitute machinery for labour,”2 this was done under the illusion that thus an argument he had long employed could be stated in a more familiar and readily acceptable form. That illusion has been dispelled by the various comments on that essay;3 and a re-examination of the earlier literature on the subject has revealed a rather peculiar situation: while the proposition has been supported and used by numerous writers ever since it was first enounced by Ricardo,4 it seems never to have been adequately expounded. In particular, although it is fundamental to the discussions of interest in the works of Böhm-Bawerk, Wicksell, and Mises, none of these authors develops it at any length. The frequent brief references to it in other general theoretical works in modern times,5 which seemed to confirm the impression that it was widely accepted, prove on examination not only to be inadequate but often to be based on faulty reasoning. Although it used to be treated as a commonplace in realistic studies of the influence of high wages on the use of machinery, there, too, we search in vain for a reasoned argument.6 The relatively fullest discussion in recent times is to be found in some German publications.7 But when in England a few years ago Professor Hicks made use of the proposition in a chapter of his Theory of Wages, Mr. Shove in his review of that work produced what has become the standard reply—that, as long as the rate of interest is unchanged, a general change in wages will affect the cost of production of the different methods of production in the same proportion (which is undeniable) and that, therefore, it cannot alter their relative advantages (which does not follow);8 and Professor Hicks’s later withdrawal of the whole chapter in which occurred the passage criticized seemed to imply that he abandoned the contention.9 Still more recently Mr. Kaldor, in an article to which we shall have to refer later, while admitting the principle, seemed to restrict its significance to rather special conditions.10

The proposition in question is of importance far beyond the special context in which it has been used in recent discussions. It is not surprising that those who completely reject it seem at the same time to be unable to attach any meaning to the conception of a given and limited supply of real capital11 because it is through this effect that the scarcity of real capital will make itself ultimately felt, however much the rate of interest may be affected by purely monetary factors, and that the volume of investment must ultimately be adjusted to a level compatible with the demand for consumers’ goods. The proposition thus is an essential part of the elementary theory of production. If this is true, the lack of agreement on it would go far to explain the sharp and apparently irreconcilable conflict of economists on the more complex problems of industrial fluctuations, and an attempt at a fuller statement of the argument on which the proposition is based seems to be urgently needed.

Such a statement will be attempted here in terms which should make it as far as possible independent of disputed points of the theory of capital, and without direct application to the problems of industrial fluctuations beyond (1) a general stress on short-run rather than long-run effects, and (2) concentration on the effect of a fall rather than a rise of wages relative to product prices, because it is in this form that the principle seems to be particularly relevant for the exploration of industrial crises. In order to separate the various parts of the argument, the problem will be approached in stages. The next section will be devoted to an explanation of the concepts used and an exposition of the general principle on assumptions which will enable us to disregard the money rate of interest. After the general principle has thus been established, the concrete ways in which it is likely to affect investment demand will be discussed in Section 3. The interplay between the Ricardo effect and the rate of interest on money loans will be taken up in Section 4 and will be discussed first on assumptions concerning the supply of credit which are approximately true of the real world. In Sections 5 and 6 the same problem will be considered on the assumption of a “perfectly elastic supply of credit,” which, though highly unrealistic, raises theoretical problems of considerable interest. In a final section will be added a few considerations which have to be taken into account in any attempt at statistical verification of the theorem.

2

The proposition here described as the Ricardo effect asserts that a general change in wages relative to the prices of the products will alter the relative profitability of different industries or methods of production which employ labor and capital (“indirect labor”) in different proportions. In its original form it asserts that a general rise in wages relative to the prices of the products will not reduce the profitability of the industries or methods employing relatively more capital to the same extent as those employing relatively less capital. We are here more particularly concerned with the inverse of this, namely, with the proposition that a general fall in wages relative to product prices will have the opposite effect.12

A general change in the relationships between wages and product prices may be caused by a general change in product prices, by a general change in wages, or by a change in technical knowledge or the physical quantities of other factors available which change the productivity of labor. While any of these changes may serve for our purpose as the independent variable, we must, of course, not treat in the same way changes in the productivity of labor which are the consequence of variations in the proportion between capital and labor employed, since this is the dependent variable of our problem.

The particular kind of change on which we shall here illustrate the proposition will be a general rise in the prices of final products (or consumers’ goods, henceforth briefly referred to as “commodities”) while money wages are assumed to remain constant (and thus to fall relatively to commodity prices). We shall assume this rise of prices to be due to an increase of demand, caused by a growth of incomes earned from producing investment goods, and exceeding the amount beyond which the output of commodities can be readily increased. We shall further assume that entrepreneurs expect commodity prices to remain at least for some considerable time at the new higher level. No assumption will be made concerning any change in the price of capital goods, this being part of our problem.

The assumption of a general rise in the prices of commodities while wages remain unchanged means, of course, that all wages fall relative to commodity prices. It is important to emphasize this because the theorem has often been misunderstood to refer to a situation where only the wages of labor co-operating with machinery change relative to prices, while the wages used in the production of machinery remain unaffected.13 It should be at once admitted that, with such a general change in the wage level relative to final prices, the costs of producing final goods by different methods must, if we assume a uniform rate of interest, be changed in the same proportion. Our contention is that nevertheless the attractiveness of investing in different industries or methods of production will be affected differently.

In order to exclude, for the purpose of the present section, any influence exercised by the money rate of interest, we shall assume for the time being that there is no lending of money of any kind during the period with which we are concerned: entrepreneurs either owning all the capital they employ and being effectively prevented from lending any of it, or being limited by a strict rationing of credit. We shall assume, however, that before the rise in commodity prices occurred, the rates of return on capital had been the same in all the firms. By excluding any consideration of the rate of interest in discussing the effects of the changes in commodity prices, we are, for the moment, deliberately avoiding what will later become our main problem. This temporary shelving of the central issue will, however, help us to isolate the more elementary parts of the argument which seem still to need explicit statement.

Our present problem thus is how, with unchanged wages, the rise in commodity prices will affect the current distribution of the funds at the command of the entrepreneurs between expenditure on wages (or investment in “circulating capital”) and expenditure on machinery (investment in “fixed capital”). To avoid complications arising from changes in the prices of raw materials, etc., which I have discussed elsewhere,14 we can assume that the firms with which we are concerned are all of the type represented by a brickyard on marginal land in which the labor employed produces not only all the raw material but also the fuel.

It remains to introduce an unambiguous and, so far as possible, uncontroversial measure of the proportions in which capital and labor are combined in the various firms and possible methods of production. For the purpose in hand, the most convenient measure which has also the advantage of being familiar to businessmen, is the concept of the “rate of turnover,” applied either to the whole or to any part of the capital of a firm. That some firms can expect to “turn over” their capital (i.e., to reinvest out of current receipts an amount equal to their capital) once in two months, while others can expect to do so only once in five or even ten years, and that this rate of turnover depends, at least in part,15 on the nature of the business and the character of the methods adopted is a familiar fact. Similarly, it will also be true that within any given firm some parts of its assets will be “turned over,” or wholly turned into cash and reinvested, twelve times a year, while others may be thus completely amortized and replaced only once in twenty years. The “rate of turnover” expresses (as an integer or fraction) the number of times the capital is turned over in the course of one year. As it will be convenient to have an adjective describing firms or methods with a relatively high or relatively low rate of turnover, we shall, for reasons which will be obvious, occasionally employ the technical term “more capitalistic” for firms or methods with a relatively low rate of turnover, and “less capitalistic” for firms or methods with a relatively high method of turnover.

The concept of the rate of turnover of capital provides a specially useful starting-point for our discussion, because changes in the wage-price relationship will evidently in the first instance affect the gain made each time the product of a given expenditure can be sold. As long as the prices of commodities remain high relative to costs, the difference will be a source of a given proportional profit on the capital every time the capital is turned over, and any given rise in product prices relative to costs will enable entrepreneurs to make higher profits per unit of time from their given capital according as they are able to turn over their capital more frequently.

In the situation of long-period equilibrium which we assume to have existed before prices rose, a situation in which the rate of return on capital will be the same for all firms, the relation between the rate of turnover and the proportional gain on each turnover is very simple. In order to avoid the ambiguous term “profit,” we shall henceforth employ the following terms: (1) The per annum net percentage return on the whole capital of a firm (or on any part of it for which we find it necessary to compute it separately), net of “wages of management” and of risk premium, we shall designate as the “internal rate of return.”16 In the position of long-term equilibrium to which we have just referred, these internal rates of return will be the same for all firms and each part of the capital of any firm. (2) The proportional gain on each sale, and therefore on the capital at each turnover, expressed in percentage, we shall designate as the “profit margin.” When it is remembered that the rate of turnover expresses the number of times total sales (or rather the costs of the products sold in a year) exceed the value of the capita) of the firm, it is clear that, if the internal rate of return is to be uniform for all firms, profit margins will have to vary inversely with the rates of turnover. Thus, if we call the internal rate of return I, the rate of turnover T, and the profit margin M, the relationship will be represented by

I=TM or M=I/T.

If, for example, the internal rate is 6 per cent, the profit margin of a firm turning over its capital six times a year will have to be 1 per cent, while a firm turning over its capital only once in two years will have to earn 12 per cent on all sales, and a firm turning over its capital only once in every ten years will have to earn a profit margin of 60 per cent.17

How will these internal rates of return of the different firms be affected by a general rise of prices of, say, 5 per cent? As such a rise means a proportional increase of the receipts from the sale of any quantity of commodities, the cost of production of which is unchanged, it will mean a clear addition to the profit margins earned on each turnover equal to the amount of the rise. For the three firms which we have just considered by way of illustration, the first (with an annual rate of turnover T = 6) will find its profit margin increased from 1 to 6 per cent; the second (with T = 1) from 6 to 11 per cent; and the third (with T = 1/10) from 60 to 65 per cent. Multiplying these profit margins by the corresponding rates of turnover, we obtain the new internal rates of return of 6 x 6 = 36 per cent for the first, 1 x 11 = 11 per cent for the second, and 1/10 x 65 = 6.5 per cent for the third firm.18

In the circumstances assumed these differences of the internal rates of return of the different firms cannot, in the short run, bring about any change in the capital at their disposal (beyond any reinvestment of profits)—although the effect these differences would have in the real world on the distribution of capital between the firms will be readily seen. Let us therefore turn from the differences between the effects on the return of the different firms to the differences between the effects of the same change on the rate of return on the different parts of the capital of any one firm. The concept of separate and ascertainable rates of turnover of, and rates of return on, different parts of the capital of any firm (certainly known, although probably never precisely determined in practice) depends on the possibility of ascertaining the marginal contribution to the product of the different parts of the capital; and this, in turn, depends in the familiar manner on the possibility of varying the proportions in which the different forms of capital are combined. We shall in the next section explain why we think that in the relevant sense this variability is fairly high, even in the short run. For the present we shall proceed on the assumption that this is so and that we are consequently in a position to determine the rate of turnover as well as the marginal product of, and therefore the profit margin earned on, any part of the capital of the firms.19

We can use for the purpose of this analysis the same numerical illustration which we have just used in connection with different firms; that is, we can assume that, for the major component parts of the capital of the particular firm which we now consider, the rates of turnover are 6 for the sums invested in current wages, 1 for the operating parts of the machine tools, etc., and 1/10 for the heavier machinery, buildings, etc. We shall again assume that after a uniform internal rate of return of 6 per cent had been established, product prices rise by 5 per cent and that in consequence the internal rates of return earned on the different kinds of capital rise as before to 36 per cent, 11 per cent, and 6.5 per cent, respectively. This can clearly be only a temporary position if the proportions between the forms of capital with different rates of turnover can be varied. It will now pay to redistribute current outlay so as to increase investment in capital with a high rate of turnover and to reduce investment in capital with a low rate of turnover. This change will be continued until the expected rates of return are once more the same on all forms of investment, and current investment will continue in this new form so long as the same conditions prevail, until ultimately all the capital of the firm has been adapted to the new conditions. As a result, a new and once more uniform internal rate of return for the firm will be established somewhere between the extremes of 6.5 and 36 per cent, and at this new rate of return the total yield that can be earned from the limited resources of the firm (augmented only by any reinvested extra profits) will have reached its maximum.

Although once more uniform for any one firm, the internal rates of return will, however, remain different for different industries and (to a lesser extent) different firms in the same industry. Where the internal rate will be fixed for any one firm will depend on the original composition of the capital of the firm and on the degree to which costs will be raised by any transition to less capitalistic methods. But, generally speaking, the rates of return will remain higher in the industries which, because of the nature of their product, need relatively less capital, and lower in the industries which need relatively more capital, although both kinds of industries will tend to change so far as possible to less capitalistic methods of production.20

3

Before going further, it will be advisable to consider briefly the probable quantitative significance in the short run of the phenomenon considered. The belief has been expressed, and appears to be widely held, that although the argument may be correct, the practical importance of the effect in question could only be small. Although it would be convenient to postpone these considerations of a more concrete kind until the theoretical argument is complete, it is probably as well to forestall a feeling of impatience on the part of the reader who may feel that all this lengthy argument is wasted on a point the practical significance of which, even if proved, would be negligible.

This widely held belief seems, however, to be based on a misconception. Of course, the proportion in which fixed capital and circulating capital (or more and less durable or laborsaving machinery) are used in production can be changed only gradually and slowly over a long period of time. But this is not the point. We are interested not in the proportions between the existing stocks of fixed capital and circulating capital but in the relative rates at which firms will spend their current outlay on renewing (or adding to) the two kinds of capital assets. Here both common experience and general considerations suggest that this proportion is highly variable in the short run.

The mistaken impression is probably caused by the kind of illustrations of the transition from less to more capitalistic methods commonly used in textbooks and describing alternative positions of long-term equilibrium. The familiar instances of “changes in the method of production” through replacement of all machines by those of another kind, of less durable by more durable, of less laborsaving by more laborsaving machinery, or of processes which are altogether shorter by processes which are altogether longer, stress an aspect which would indeed seem to be relatively unimportant in the short run. To realize how the same tendency operates no less strongly in the short run, we must overcome our prepossession with the “comparative statics” of the textbook and try to think more realistically of the concrete decision which entrepreneurs will continuously have to make.

Within limited space we can illustrate this only by selected examples. But they will, I hope, show the width of the range of variations possible in the very short run.

We have to think of entrepreneurs equipped at any given moment with a given stock of durable machinery of which only a small part needs replacement during any short period of time. If conditions had remained what they had been, they would have continued period after period to invest their earned depreciation allowances in machinery of the same kind. But they will do so merely because this would be the most profitable method of using their funds, and we must not assume that they will continue to do so after conditions have changed. Particularly when demand increases there will be any number of possible ways of increasing output other than by multiplying machinery of the kind they have been using before. If they cannot borrow so as to bring their internal rate down to the former level, some of these will appear more profitable than those used before.

There will be mainly two kinds of changes which will now appear advantageous: an entrepreneur may use his existing machinery more intensively (i.e., with more labor)—employing for this purpose part of the funds which would have otherwise been invested to replace the machinery by new machines of the same kind—or he may replace those machines that wear out by a larger number of cheaper ones. Both of these methods will probably be resorted to, although the first is probably the more important one.

The obvious methods of quickly increasing output as commodity prices rise is to work overtime, to introduce double or treble shifts, to provide extra assistance to relieve the workers on the existing machines from ancillary operations, etc. This will normally raise labor costs per unit of output, and this fact will have prevented these devices from being used before prices had risen. But if the adoption of any of these methods increases marginal labor cost per unit of output by, say, 4 per cent,21 this would, with a 5 per cent rise in prices, still leave an extra profit of 1 per cent, which, with a rate of turnover of 6, would make the internal rate of return on this form of investment still 12 per cent compared with the 6.5 per cent on the machine with a rate of turnover of 1/10. This more expensive method of production will therefore now become the one through which, with the limited resources at the disposal of the entrepreneur, the largest profits can be realized; and the labor co-operating with the machinery will be increased until the fall of the return on funds invested in more labor and the rise of returns on funds invested in machines make the two rates of return once more equal at an intermediate figure.

The kinds of changes in the machinery used which have to be taken into account even in the short run will be equally numerous and will also all have the effect of raising marginal costs. There will, in the first instance, be the possibility of less perfect maintenance and attention, makeshift instead of thorough repairs, shorter or fewer periods of laying off for inspection and overhaul, which will reduce the efficiency and shorten the life of the existing machinery but may well be worth while if current output can thus be increased. There will be, second, the possibility of outright nonrenewal, not, of course, of essential parts of the equipment, but of the many auxiliary laborsaving devices such as automatic feeders and other gadgets performing operations which can also be done by hand. Third, there will be the possibility of using obsolete or secondhand machines instead of new ones. Many older factories have a certain amount of such old machinery for temporary use to meet peak-time demands or in an emergency, for which it would not pay to keep a new machine in reserve. There exists in many branches a supply of secondhand machinery which can be used in the same way. Fourth, and last, there will be the possibility of replacing those machines that wear out by new but cheaper and less efficient ones. So long as the internal rate of return of any firm remains above what it has been before, it may well be profitable to buy two less efficient machines at the price of one more efficient one, if the two less efficient machines enable the firm, though with the co-operation of much more labor, to increase output more than with the more efficient one.

If we consider the effect of all these possible changes, not on the proportion in which the stock of capital of any firm is composed of different parts, but on the rates at which its current outlay is spent on different kinds of resources, or on the proportion in which total outlay is distributed between fixed and circulating capital, it seems to be clear that, in consequence of a general change in commodity prices, large changes in the latter magnitudes may be brought about in a comparatively short time. In extreme cases it may even be profitable for entrepreneurs temporarily to discontinue all demand for machinery and yet for a considerable period greatly to increase output. But while this extreme result may not be probable, it does not seem unlikely that the demand for certain kinds of new equipment will be absolutely reduced. This would seem to be likely particularly where, as is true in the case of buildings and most heavy machinery, the equipment has to be made to order and large sums will have to be locked up in it by the buyer during the period of production without bringing any current yield; the same would seem to be true wherever a gradual transition to some new (e.g., more laborsaving) but more expensive kind of machinery has been under way, which will now be stopped; and generally wherever the change in the methods of production adopted will involve a change from equipment made by one group of people to that made by another group of people.22 In so far as any labor is specific to the production of the kind of equipment, the demand for which now decreases or ceases entirely, the consequence of the rise in final demand will thus be unemployment in the capital goods industries.

4

We have now to introduce the possibility of borrowing money at rates of interest determined by the market and not necessarily changed in response to an increase in the demand for funds. In the present section we shall consider how far this modifies the conclusions thus far reached if we make assumptions which in the most important respects approximately correspond to conditions in the real world and which will therefore enable us to judge what the practical significance of our conclusions is likely to be. The theoretically very interesting but practically irrelevant case of a “perfectly elastic supply of credit” will be deferred until the next section.

The sharp distinction between the two cases and the order of treatment is indicated by the frequent but misleading application to this problem of the category of “perfect competition”; this concept is quite inappropriate to it, simply because successive (additional) loans to the same borrower will never represent the “same commodity” in the sense in which the term is used in the theory of competition. While in a commodity market “perfect competition” means that any single buyer can buy at the given market price any quantity he likes, it would, of course, be absurd to assume that even in the most perfectly competitive money market every borrower (or, for that matter, any borrower) can at the given rate of interest borrow any amount he likes. This is precluded by the fact that in given circumstances the security a borrower has to offer is not so good for a large amount as for a small one. In consequence, every prospective borrower will have to face an upward sloping supply curve of credit—or, rather, not a continuous supply curve, but an upward stepped “curve,” showing that the rate of interest, while constant within certain limits, will go up by distinct steps whenever one of the limits is reached up to which he can borrow at a given rate.

The most important, though not the sole, factor limiting the borrowing capacity of a firm at any given rate of interest will be the size of the capital owned by it. Bankers, as a rule, will not be willing to lend to any one firm more than a given proportion of its own capital and take very good care that no firm borrows at the same time from more than one bank; and beyond this limit the firm will be able to obtain funds only at a higher rate of interest, or, what comes to the same thing, on more onerous conditions of some other kind. This limitation of the amount of funds any firm can raise to increase its output will be further strengthened where banks provide loans only for investment in circulating capital and effectively refuse to provide funds for the investment in fixed capital. The general fact we have to remember in this connection is that, in the existing institutional framework, lending (in the strict sense of the word), and particularly short-term lending, will secure mobility of capital only to a limited extent and that, in a world where risk is ever present, it will by itself not be sufficient to bring about an equalization of the rates of return on capital invested in different firms or completely to adjust these rates to a given market rate of interest. For this, in addition to lending, transfers of capital by way of full participation in the risk of the business, i.e., changes in the share capital or what we may quite generally describe as the “own capital” (as distinguished from the borrowed capital) of the firms will be necessary. But this latter process is necessarily much slower than the provision of additional bank loans, and it will therefore frequently be true that in the short run most firms will not be able to raise as much capital as they could profitably use or that they will be able to do so only at rates much higher than the “market rate.”

This is not to say that the maximum a firm will be able to borrow at a given rate of interest will be rigidly fixed in proportion to its own capital. The director of a firm who can convince his bank manager that he has an exceptional opportunity of making large profits on additional capital and thus can provide a large margin of safety in case his optimism should prove not to be quite justified will be able to borrow proportionally more than another. In general, when prospects are good all firms may be able to borrow more in proportion to their own capital than when prospects are poor.23 The stepped supply “curves” of credit which all firms face will be shifted to the right as general prospects get better (and to the left when prospects get worse), and such sideways shifts of the supply curves will frequently act, and often be deliberately used, in exactly the same way as an outright change in the rate of interest (i.e., as a raising or lowering of the whole curve).

But, although any general increase in expected profits is likely to increase the amounts firms can borrow, it will in many instances increase the amounts they would like to borrow at the current rates of interest still more and thus bring firms up to the limit beyond which they can raise capital only at higher costs. Though there will be, at the ruling rate of interest, an unsatisfied demand, this demand will not be “effective” demand, because it will not fall within the categories to which the ruling rates apply, and these rates will therefore remain unchanged. The situation is similar to that caused by credit rationing, although it will arise, without the intervention of authority or a monopolist, merely as a result of the views the banks hold about the “credit-worthiness” of the borrowers.

There is no need to explain at any length that, whenever the amounts people would like to borrow at the current rate of interest are larger than the amounts they can obtain at that rate, it will be these amounts and not the ruling market rate which will determine the internal rates of return of the different firms. As in the situation discussed in the last section these internal rates will be different for different firms according to the circumstances then enumerated (to which we must now add the limitations on the borrowing facilities of any particular firm), and the investment of each firm will be governed by its own internal rate, which may be very much higher than the market rate, which may not have changed at all. The rise in internal rates would lead to a general change to less capitalistic methods of production, different in extent according as the internal rate has changed in the different firms.

There remains, however, the question whether, to the extent that the firms are able to procure additional credits, this will reduce the degree to which their internal rates will rise, and therefore the degree to which they will change to less capitalistic methods of production, compared with the case where no additional credit at all was available. The problem which arises here is the same as that which we intend to consider in its more general form in the next section, since, if our proposition holds true even when the supply of credit is completely elastic, it must apply still more in the present case. We can, therefore, immediately proceed to this “stronger” case.

5

The assumption that the supply of credit at a given rate of interest is perfectly elastic is not only unrealistic but, when we contemplate its implications, perfectly fantastic; and it makes the analysis rather complicated. But, as it brings us face to face with a fundamental theoretical problem, it is well worth undertaking. It raises in its purest form the question of the relationship between the monetary and the real factors affecting the relative profitability of different methods of production.

The contention that if the supply of credit is perfectly elastic it must be the money rate of interest which will determine which forms of investment are most profitable may be based on either of two assertions which ought to be clearly distinguished: it may be asserted that in this case the cost-price relations (or the relation between wages and commodity prices) must necessarily be so adjusted by either a change in wages or a change in commodity prices as to make the difference correspond to the money rate of interest; or it may be asserted that even when this does not happen and wages remain, for example, too low relative to commodity prices, it will still be the money rate of interest and not the cost-price relationship that will govern the form of investment.

With regard to both these arguments, but particularly with regard to the first, it is important to remember that the situation which we consider is eminently not one of equilibrium but one in which the causes of continuous and cumulative change are inherent. It is indeed the classical instance of a cumulative process with which we are dealing; the perfectly elastic supply of credit at a rate of interest lower than the internal rate of all or most of the firms will be the cause of continuous changes of prices and money incomes where each change makes further changes necessary. There is no point in saying with respect to such a situation that “in equilibrium there must” exist such and such a relationship, because it necessarily follows from the assumptions that the relationship between at least some prices must be out of equilibrium. This is important particularly with reference to the two propositions: first, that prices must be equal to marginal costs, and, second, that the prices of the factors must be equal to the expected price of their marginal product discounted at the rate of interest at which credit can be freely obtained. All we need to say with respect to the first proposition is that, except in a very special and for our purpose irrelevant sense,24 it is just not true in the very short run, although a dogmatic belief that prices must always be equal to marginal costs in the relevant sense is probably responsible for a great many confusions in this field. The second proposition is the one with which we are here more directly concerned.

The belief that, if the supply of money at a given rate of interest is perfectly elastic while investment demand is inelastic, the former will uniquely determine the rate of return at which supply and demand will be equal, is derived by analogy from the general rule that, if either the quantity demanded or the quantity supplied of anything is completely elastic at a given price, it follows necessarily that this will be the price. But, while this statement is true enough when we discuss demand and supply in “real” terms, it neglects an essential difference of the present case where the “price” in question is the relationship between the prices of two groups of goods (labor and commodities), while the supply which is infinitely elastic is not that of one of the two goods but merely of the money that is in the first instance to be spent on one of the two goods; it neglects the fact that any increase of money expenditure on the one kind of good is bound to cause an increase of money expenditure on the other kind of good.

When it was said before that we are dealing with a position of disequilibrium, this meant precisely that we had to deal with two sets of forces tending to fix the same price (or rather the same relationship between two groups of prices) at different figures. On the one hand, we have a given output of consumers’ goods (only slowly variable) and a given propensity of the people to spend a certain proportion of their income on consumers’ goods, which together would for each volume of employment (and therefore of total income) determine a definite ratio between the prices of commodities and the prices of all factors; and, on the other hand, we have an infinitely elastic supply of money which tends to determine the prices of factors in a certain fixed relation to the prices of the products which is different from that determined by the first set of factors.

It is not to be denied, of course, that through changes in the money stream the relation between the prices of goods as determined by the real factors can be very considerably modified. The problem is merely whether there is no limit to the extent to which, and the period of time for which, the price structure as determined by the “real” factors can be thus distorted, or whether the fact that the extra money which has first raised one group of prices will soon work round to affect the other group of prices in the same direction does not set a limit to the possible degree of distortion. The question is rather similar to that whether, by pouring a liquid fast enough into one side of a vessel, we can raise the level at that side above that of the rest to any extent we desire. How far we shall be able to raise the level of one part above that of the rest will clearly depend on how fluid or viscid the liquid is; we shall be able to raise it more if the liquid is syrup or glue than if it is water. But in no case shall we be at liberty to raise the surface in one part of the vessel above the rest to any extent we like.

Just as the viscosity of the liquid determines the extent to which any part of its surface can be raised above the rest, so the speed at which an increase of incomes leads to an increase in the demand for consumers’ goods limits the extent to which, by spending more money on the factors of production, we can raise their prices relative to those of the products.25 The problem arises most sharply when we assume that the money rate is arbitrarily lowered to a very low figure in a new country with little capital and a very high “marginal efficiency of capital.” If the proposition we are considering were true at all, it would have to hold in this case also, i.e., the availability of an unlimited quantity of money at the low rate of interest would bring it about that wages would be driven up to the discounted value—not merely of the present marginal product of labor but of the marginal product which labor could be expected to produce after the machinery had been installed which it would be profitable to install at the low rate of interest. The aggregate value of the services of labor at that real wage might be very considerably larger than the total current output of consumers’ goods and certainly very much larger than the whole current output of labor. The effect of this must be that the demand for consumers’ goods and their prices would rise accordingly. If this rise of prices stimulated entrepreneurs to borrow and invest still more, this would only make prices rise still further, and the faster entrepreneurs expected prices to rise, the more they would necessarily speed up this price rise beyond their expectations. Although they might succeed at times in driving wages up to the discounted value of the expected price of the marginal product of labor, they could not possibly, whatever their effort, actually raise real wages to the figure corresponding to the low rate of interest, because the stuff to provide this real income just would not be there.

In the situation to which we have to apply these considerations, such as will exist in a modern society in the late stages of a boom, the position will be different only in degree. It still remains true that entrepreneurs, by offering higher money wages, cannot effectively raise real wages to the level which would correspond to the low money rate of interest, because the more they raise money wages, the more the prices of commodities will rise.26 In this case, too, the limiting factor is simply that the consumers’ goods are not there and that, so long as all investment takes highly capitalistic forms, every increase in employment adds only a fraction of its value to the output of consumers’ goods. This brings us to the second hard fact which dominates the situation we are considering: that in this situation there will not be enough labor available to increase at the same time the current output of consumers’ goods and to push investment to the limit indicated by the rate of interest. So long as unused reserves of labor are available, there is indeed, as we shall presently see, no reason why the entrepreneurs should not use the unlimited funds to do both: to increase the output of consumption goods for the near future by the expensive but quick methods and to provide for cheaper production by investing on a large scale. This is the reason why in the early stage of a boom the money rate of interest will control the situation. But, although this would mean that in these circumstances the low rate of interest was effective so far as the volume of investment was concerned, it would still not mean that, once consumers’ goods prices began to rise, real wages could be maintained by proportional adjustments in the money wages.

6

That in the circumstances considered it will sooner or later become inevitable that real wages should fall, and investment expenditure be reduced, will be evident if we consider for a moment the paradoxical results that would follow if things worked as appears to be assumed by the contrary view. The increase in the prices of commodities, with unlimited amounts of money available at a fixed rate of interest, would lead to an increase of investment expenditure and of real investment, which, since no reserves of labor are available, could take place only at the expense of the output of consumers’ goods in the near future. The consequent increase of money incomes and of final demand, coupled with the decrease in the output of consumers’ goods, would cause a further rise in their prices relative to wages. This further rise in the prices of consumers’ goods would, according to that view, bring about a further increase of investment at the expense of the output of consumers’ goods, and so on, presumably until there are no people left producing consumers’ goods and everybody is engaged in providing machinery destined to produce consumers’ goods in some distant future when the men will all have died of starvation in the interval. While some tendency in this direction probably exists during the earlier part of a boom, it hardly needs any superstitious belief in the self-righting forces of the economic system to suspect that, some time before that extreme result is produced, counterforces will operate to check such a development. This brings us to the second version of the argument according to which it must be the money rate of interest which rules the roost.

This version, which admits that wages may remain relatively too low compared with the rate of interest but insists that in spite of this, if the supply of money is perfectly elastic, it will be the rate of interest and not the level of wages which will govern the form of investment, is represented mainly by Kaldor and Wilson.27 These two writers, however, as we shall try to show, so much simplify their task that they do not prove what they mean to demonstrate. All they do prove, in a quite unnecessarily elaborate manner, is that, so long as an unlimited amount of money can be obtained at the given rate of interest, it will depend solely on the rate of interest which method will bring the highest current profit above current costs after the equipment appropriate to that method has been procured. This is no more than another version of the truism which we have emphasized from the beginning that, so long as the rate of interest remains constant, a change in real wages cannot alter the relative costs of the different methods of production. What Kaldor and Wilson completely disregard is that, in comparing the profits obtained from producing with different methods, they are comparing methods employing different amounts of capital without counting in any way the cost of creating the extra real capital required for one of the two methods. They do this by omitting to give any attention to what will happen during the period of transition before the new equipment is available. Whether this equipment will ever be available will depend, however, precisely on what happens during this interval. The problem is not answered by the statement that, if we adopt a certain course, the excess of current receipts over current outlay will be largest from a certain future date onward, if we are not also told what happens between now and that future date. In choosing between the two alternative methods, we cannot decide merely on the basis of what the position would be after some new long-term equilibrium has been established, but we must also consider what will happen between now and then, because which long-term equilibrium will be established will depend on this. What the procedure of Kaldor and Wilson amounts to is to leave out from their data the real factors which determine the supply of capital and to assume that the quantity of capital will in the long run necessarily adjust itself so as to bring its “marginal efficiency” to the level of a rate of interest determined solely by monetary factors.

To speak more concretely, Kaldor and Wilson assume that, if only the funds were available, it will, in the circumstances assumed, necessarily be most profitable to meet an increased demand for the product by increasing equipment in proportion to the increase in the amount of the product that can be sold at a given price, although in this way it will as a rule be possible to catch up with increased demand only after some considerable interval. Only if (and to the extent that) we can assume the extra equipment needed to be waiting in the shops ready to be bought and instantaneously to be installed would no such interval occur. This assumption (which amounts to presupposing that all the real capital required for an expansion is already in existence) is evidently one which might be true for any one firm, but which will not be true when all firms are simultaneously in the same position. In the situation with which we are concerned the additional equipment and still more the output produced by it will be available only after considerable delay. In the interval until this output is available profits which might have been made by quicker methods will be lost and ought to be counted as part of the cost of the production for the more distant future.

To this it will no doubt be answered that there is no reason why the entrepreneurs should not do both: provide for the output in the near future by the quick but expensive methods and provide for the more distant future by ordering more machinery. But this brings us up to the fundamental issue whether the amount of real resources, and particularly of labor, will be sufficient to make both possible at the same time. Or, in other words, the question arises, which it is now fashionable to disregard in discussion of these problems, whether an unlimited supply of funds secures an equally unlimited supply of real resources. We have already seen that, in the kind of situation with which we are concerned, this is not likely to be the case.

It is instructive, however, to examine a little more closely how some economists manage to gloss over this difficulty and thereby apparently succeed in eliminating the given supply of capital from the relevant data of the problem. Kaldor’s treatment of the question in the article referred to is in this respect most illuminating. He explicitly claims to deal with all the cases where the output of individual firms is limited by “falling demand curves for the products and/or rising supply curves for its factors”28 as the only possible alternative to a limitation by an inelastic supply of credit. But, when he comes to discuss the case, he assumes, and, finally, even introduces in a footnote, the explicit assumption that “the elasticity of supply of factors, to the individual firm, is infinite.”29 But, although he has thus in fact confined his proof to only part of the group of cases for which he has originally undertaken to provide it, he proceeds as if he had proved it for all and continues to treat his original alternative (inelastic supply or credit or falling demand curves for the product and/or rising supply curves for the factors) as corresponding to the distinction between situations where either only the rate of wages or only the rate of interest determines the methods of production that will be profitable.

Kaldor’s omission to face the effects of limitation in the supply of labor is so significant because it is through the rise in the supply price of labor that the shortage of real resources available for investment (caused by the competing demand of the producers of consumers’ goods) makes itself felt. His conclusion follows solely from the assumption that, and is true only if, the elasticity of the supply of labor (and other factors) is infinite. When this is true, there is indeed no reason why entrepreneurs should not succeed in using the unlimited money funds to increase output quickly by costly methods and at the same time making arrangements for a more economical production of a larger output at a later date. So long as unused reserves of labor are available at an unchanged price, unlimited funds mean unlimited control over resources. But this is not the condition which is relevant for the position of full employment that will prevail near the top of a boom.

We shall see the problem involved more clearly if for a moment we assume that each firm represents a completely integrated process of production, that is, not only that the production of the final commodity and of all the various raw materials, etc., used but also that of all machinery required for the production of that output is produced within the firm. In the circumstances we are considering, each of these integrated firms would be able to attract additional labor only by offering higher wages; and, although the relatively less capitalistic industries might find it profitable to increase their labor in this way at the expense of the more capitalistic industries, this will, if there are no unemployed available, not be possible for all firms or for industries or firms of average “capital intensity.”

For each of these firms, which for our purpose may serve as representatives of a general trend, the problem will therefore be how to distribute its given labor force between the production of commodities and the production of machinery. The way to maximize the excess of current receipts over current outlay for all periods after the change had been completed would be temporarily to transfer labor from the production of commodities to the production of machinery. This would involve a reduction of the current output of commodities, and therefore of current profits, not only below what they would be if the past volume of output had been maintained, but still more below the level which could be achieved if current output were increased, by adopting quick and more costly methods until marginal cost just equaled price. These profits, which will have to be foregone if the additional machinery is to be provided, would have to be regarded as costs of, and would therefore have to be offset against, the larger profits which in consequence could be continuously earned from some future date onward. It is this item which represents the costs of the extra waiting which the more capitalistic methods involve and which nowhere enters into the calculations of Kaldor or Wilson. Since these profits which will be earned during this interval are, as we have seen, likely to be very considerable, it is more than likely that they will turn the scales against the more capitalistic process. In other words, profits will be higher on the method with the higher rate of turnover, not because they would accrue at a higher rate after the new equilibrium envisaged by Kaldor had been established (which they would not do), but because the profits on the less capitalistic method will begin to accrue earlier than those on the more capitalistic method. It is the profits from now onward, not merely profits after the additional equipment has been created, which must be considered in deciding whether that additional equipment is to be created at all. It is for this reason that our integrated firms, if their internal rates of return only rise high enough, will certainly not transfer labor from the production of commodities to the production of machinery, but, on the contrary, will transfer labor from the production of machinery to the production of commodities. This change will not be merely temporary but will evidently have to be maintained so long as the conditions continue which made it appear profitable in the first instance, that is, so long as the prices of consumers’ goods remain high relative to wages.

Before leaving the integrated firms, it is worth while to consider a little more closely exactly what will happen in their machine shops. These machine shops will have to give up some of their labor that can also be used indirectly to produce consumers’ goods, and they will have to turn to the production of less elaborate, less costly machinery. Both of these changes will have the effect of making superfluous other kinds of labor which are specific to the production of the more elaborate kind of machinery or to jobs (such as the extraction of certain raw materials used in the production of machinery) which are wanted in a rigid quantitative proportion to the total output of machinery. In other words, the result of the shortage of the more generally employable labor will be unemployment of certain particular kinds of labor—that which is highly specific to the production of some kinds of machinery.

While it seems fairly evident that the results must be the same if we abandon the assumption of complete integration of the different industries, I must admit that I find it difficult to visualize precisely how it will be brought about. The physical conditions of the problem are, of course, the same: it will still be true that there will not be enough labor available at the same time to increase the output of consumers’ goods quickly and to provide more machinery to produce a still larger output by more efficient methods at a later date. And it will also still be true that, if entrepreneurs decide for the costlier but speedier methods, this will bring them the larger profits. The problem is what will enable them to foresee this result; because, so long as they believe that at the ruling price they will be able both to get the labor to increase the output immediately and to get the manufacturers of machines to produce machines for them, it will appear profitable to try to do both; the individual entrepreneur will no longer be directly faced with the problem of using the same labor either to produce more commodities or to produce more machinery; and it will be only when he and all the other entrepreneurs who are in the same position try to do so that they will find out that it cannot be done.

The answer, I think, is to be sought, first, in the fact that the provision for the near future will necessarily have the first attention of the entrepreneur, because, if the profits which might be made in the near future are not obtained, they (and perhaps a certain amount of permanent business) will be lost for good to a competitor, while delay in obtaining the more efficient machinery will affect the volume of output less and merely postpone the date when its costs will be lower. Closely connected with this will be the effect of the increasing uncertainty concerning the more distant future. Although the entrepreneur may expect the higher prices to continue indefinitely, he will be less certain that this will be so in the more distant future than in the near future. On the principle of “making hay while the sun shines,” provision for the profits to be made in the near future will take the precedence.

Second, there is the fact that, since in the short run the more capitalistic methods will require more labor for any given increase of output than the less capitalistic methods, the rising supply price of labor will make itself felt more with the former than with the latter, i.e., the attempt to procure the machinery necessary for a given increase of output will meet with a rise in the price of the machinery comparatively greater than the rise in wages which would be caused by employing the number of men required to produce the same amount by less capitalistic methods.

Third, there is the point that, in so far as the producers of commodities increase their output in the first instance, not without any additional machinery but by the use of a cheaper kind of machinery, the need for the more elaborate machinery will arise only after the machinery provisionally installed wears out, and that therefore the demand for the more elaborate kind of machinery may for a time cease completely.

Finally, and perhaps most important, there will be the fact that so long as the producers of commodities do not succeed in actually increasing output quickly to the extent necessary to bring marginal returns down to a level they can expect to prevail in the long run, they will be uncertain which of the various elements in the picture will change so as to create a new equilibrium position. In other words, so long as profits on the quick methods do not actually fall and further endeavor appears to be needed to gather all the high profits that can be made immediately, the more elaborate preparations for future profits at a lower rate (though higher in the aggregate) and involving greater risk will not appear very attractive. But so long as people try to do both, to increase output quickly and to order more machinery, incomes and final demand will continue to run ahead of the expectations of the producers of consumers’ goods. It will be only after investment has been considerably reduced that the cost of the expensive methods will catch up with prices and that thus the more capitalistic methods will once more appear attractive.

I am fully aware that all this is not very satisfactory and that a clearer picture of the precise process by which competition brings about this result would be very desirable. But I am not sure whether this is possible. We are dealing with a position of disequilibrium in which developments depend on the precise order in which the various changes follow one another in time and where the situation at any moment is likely to be, as we learned to say during the war, “confused.” We cannot say precisely when entrepreneurs will abandon the self-defeating attempts at the same time to build up elaborate equipment and to increase production quickly. All we can say is that, the longer the effect with which we are concerned is delayed, the stronger must become the forces tending to bring it about (i.e., the longer increases in final demand are allowed to bring about proportionally larger increases in investment, the greater must become the rise of prices of final goods relative to costs), and that therefore they are bound sooner or later to become the dominating element in the picture.

7

Any attempt to discover from the available statistical information whether the Ricardo effect does in fact operate as these considerations suggest encounters considerable difficulties. We can do no more here than show what are these difficulties of an attempt at verification and why the evidence so far available does not seem to allow any definite conclusions.

In the first instance it must be pointed out that, although the phrase “real wages” is sometimes used in this connection, the relation between wages and product prices with which we are concerned has no close connection with “real wages” in the sense in which this term is commonly used. While in most contexts when real wages are discussed what is meant is the relation between wages as received by the worker and the prices of the commodities on which he spends these wages, we are concerned with the cost of labor to the entrepreneur and their relation to the prices of the products he produces. We shall merely mention that even the wages paid to the workers and the cost of labor to the entrepreneur may sometimes move differently.30 The more important difference is, however, that between the prices of the goods on which the workers spend their wages and the prices of the goods in the production of which the labor is used. The following are the main sources of this difference:

1. While the “cost of living” is affected largely by the prices of agricultural products, it is for our purpose mainly the prices of manufactured articles that are of importance. More generally, the importance for our purpose of the change in price of any particular product varies with the relative amount of capital used in its production (which is relatively low in agriculture and relatively high in manufacture). The significance of this will be seen when it is remembered that for our purposes a mere shift in demand from articles requiring comparatively little capital in their production to articles requiring a great deal would have the same effect as an increase in total demand. Any statistical investigation would probably do well to confine itself in the first instance to the effect of changes in the relation between the prices of the product and wages in any given industry on investment in that industry. In so far as a more general investigation is attempted, prices ought probably to be weighted according to the proportional amount of capital used in the production of the different goods. Where we have to deal with an “open system,” as will regularly be the case in statistical investigations, we shall have further to distinguish between prices of home-produced and of imported commodities.

2. While from the“cost of living” point of view it will be retail prices that will be relevant, for our purpose it will be prices received by the manufacturers that will count; and it is in general true that, for reasons into which we need not go here, the latter (or at least wholesale prices) fluctuate more than retail prices.

3. While from the “cost of living” point of view it is the relation between wages and the price of a fixed quantity of commodities which is relevant, we are here concerned with the relations between the costs of labor and the marginal product of that labor. This marginal product, however, is itself not a constant but a variable, and either may vary as a consequence of the effect with which we are concerned or may by its change become the cause of this effect. In other words, changes in the marginal product may appear either as dependent variables, when they are the consequence of a change in the proportional combination of capital and labor, or as independent variables, when they are brought about by changes in the “data,” particularly by changes in technological knowledge. Technological change, at least when it is rapid and general, may here cause serious difficulties.

So long as technological knowledge remains constant, the relations between the cost of labor and the price of its product which are relevant for our purpose will generally be the same as the relations between the costs of a fixed amount of labor and the price of a fixed amount of the product—although, when we have to deal with an “open” system, changes in the price of an important raw material may disturb even this simple relationship. But, once changes in technical knowledge have to be taken into account, the problem becomes very much more complex. It is evident, to take an extreme case, that if an advance in knowledge enabled us to produce with exactly the same machinery and other outlay 20 per cent more than before, the immediate effect would be very similar to that of a rise in the price of the product. So long as such a change occurs in isolation, there is no special difficulty about it. But, when it is combined with price changes, a problem arises for which it is difficult to see a practicable solution. To judge the significance of any price change occurring together with technological change, we should have to know which price relationship now “corresponds” to the price relationship which existed before, that is, which relation between the cost of labor and the price of the product will now make investment no more and no less attractive than the price relationship which existed before the technological change. At the moment I have no solution of this difficulty to offer.

It was the fatal mistake of Böhm-Bawerk (and to much less an extent of Wicksell) that, although he was quite aware that the existing stock of capital goods was capable of producing more than one single output stream, he attempted to simplify his exposition by identifying the stock of capital goods with a definite quantity of consumers’ goods and to represent this in his illustrations by a fixed amount of available money capital. The analysis of the famous final chapter on “The Market for Capital in Its Full Development” of the Positive Theory makes perfect sense if we remember this simplification but must seem to have no relevance to anything in the real world to anyone who takes literally the representation of the supply of capital by a sum of money.

It remains true, of course, that if we compute the rate of turnover (or the “average period of investment”) and the rate of return for the whole of the capital of a firm, the changes in either will be small. But the point is precisely that at any moment the decision has not to be made for the whole of the capital and that the alternative gains to be made on the sums currently to be reinvested will differ very considerably, absolutely, as well as expressed as percentages of these amounts currently to be reinvested.

• 1[*] The twelfth Finlay Lecture, delivered at University College, Dublin, on December 17, 1945. Published by Hodges, Figgis & Co., Ltd., Dublin, and B. H. Blackwell, Ltd., Oxford, 1946.
• 2Both the term “individualism” and the term “socialism” are originally the creation of the Saint-Simonians, the founders of modern socialism. They first coined the term “individualism” to describe the competitive society to which they were opposed and then invented the word “socialism” to describe the centrally planned society in which all activity was directed on the same principle that applied within a single factory. See on the origin of these terms the present author’s article on “The Counter-Revolution of Science,” Economica, VIII (new ser., 1941), 146.
• 3R. Bisset, Life of Edmund Burke (2d ed., 1800), II, 429. Cf. also W. C. Dunn, “Adam Smith and Edmund Burke: Complimentary Contemporaries,” Southern Economic Journal (University of North Carolina), Vol. VII, No. 3 (January, 1941).
• 4Carl Menger, who was among the first in modern times consciously to revive the methodical individualism of Adam Smith and his school, was probably also the first to point out the connection between the design theory of social institutions and socialism. See his Untersuchungen über die Methode der Sozialwissenschaften (1883), esp. Book IV, chap. 2, toward the end of which (p. 208) he speaks of “a pragmatism which, against the intention of its representatives, leads inevitably to socialism.”
• 5Edmund Burke, Reflections on the Revolution in France (1790), in Works (World’s Classics ed.), IV, 105: “Thus the commonwealth itself would, in a few generations, be disconnected into the dust and powder of individuality, and at length dispersed to all winds of heaven.” That Burke (as A. M. Osborn points out in her book on Rousseau and Burke [Oxford, 1940], p. 23), after he had first attacked Rousseau for his extreme “individualism,” later attacked him for his extreme collectivism was far from inconsistent but merely the result of the fact that in the case of Rousseau, as in that of all others, the rationalistic individualism which they preached inevitably led to collectivism.
• 6Alexis de Tocqueville, Democracy in America, trans. Henry Reeve (London, 1864), Vol. II, Book II, chap. 2, where De Tocqueville defines individualism as “a mature and calm feeling, which disposes each member of the community to sever himself from the mass of his fellows, and to draw apart with his family and friends; so that, after he has thus formed a little circle of his own, he willingly leaves society at large to itself.” The translator in a note to this passage apologizes for introducing the French term “individualism” into English and explains that he knows “no English word exactly equivalent to the expression.” As Albert Schatz pointed out in the book mentioned below, De Tocqueville’s use of the well-established French term in this peculiar sense is entirely arbitrary and leads to serious confusion with the established meaning.
• 7In his excellent survey of the history of individualist theories the late Albert Schatz rightly concludes that “nous voyons tout d’abord avec évidence ce que l’individualisme n’est pas. C’est précisément ce qu’on croit communément qu’il est: un système d’isolèment dans l’existence et une apologie de l’égoisme” (L’Individualisme économique et social [Paris, 1907], p. 558). This book, to which I am much indebted, deserves to be much more widely known as a contribution not only to the subject indicated by its title but to the history of economic theory in general.
• 8In this respect, as Karl Pribram has made clear, individualism is a necessary result of philosophical nominalism, while the collectivist theories have their roots in the “realist” or (as K. R. Popper now more appropriately calls it) “essentialist” tradition (Pribram, Die Entstehung der individualistischen Sozialphilosophie [Leipzig, 1912]). But this “nominalist” approach is characteristic only of true individualism, while the false individualism of Rousseau and the physiocrats, in accordance with the Cartesian origin, is strongly “realist” or “essentialist.”
• 9Adam Ferguson, An Essay on the History of Civil Society (1st ed., 1767), p. 187. Cf. also ibid.: “The forms of society are derived from an obscure and distant origin; they arise, long before the date of philosophy, from the instincts, not from the speculations of man. . . . We ascribe to a previous design, what came to be known only by experience, what no human wisdom could foresee, and what, without the concurring humour and disposition of his age, no authority could enable an individual to execute” (pp. 187 and 188).
• 10Cf. Schatz, op. cit., pp. 41–42, 81, 378, 568–69, esp. the passage quoted by him (p. 41, n. 1) from an article by Albert Sorel (“Comment j’ai lu la ‘Réforme sociale,’” in Réforme sociale, November 1, 1906, p. 614): “Quel que fut mon respect, assez commandé et indirect encore pour le Discours de la méthode, je savais déja que de ce fameux discours il était sorti autant de déraison sociale et d’aberrations métaphysiques, d’abstractions et d’utopies, que de données positives, que s’il menait à Comte il avait aussie mené à Rousseau.” On the influence of Descartes on Rousseau see further P. Janet, Histoire de la science politique (3d ed., 1887), p. 423; F. Bouillier, Histoire de la philosophie cartésienne (3d ed., 1868), p. 643; and H. Michel, L’Idée de l’état (3d ed., 1898), p. 68.
• 11The decisive importance of Mandeville in the history of economics, long overlooked or appreciated only by a few authors (particularly Edwin Cannan and Albert Schatz), is now beginning to be recognized, thanks mainly to the magnificent edition of the Fable of the Bees which we owe to the late F. B. Kaye. Although the fundamental ideas of Mandeville’s work are already implied in the original poem of 1705, the decisive elaboration and especially his full account of the origin of the division of labor, of money, and of language occur only in Part II of the Fable which was published in 1728 (see Bernard Mandeville, The Fable of the Bees, ed. F. B. Kaye [Oxford, 1924], II, 142, 287–88, 349–50). There is space here to quote only the crucial passage from his account of the development of the division of labor where he observes that “we often ascribe to the excellency of man’s genius, and the depth of his penetration, what is in reality owing to the length of time, and the experience of many generations, all of them very little differing from one another in natural parts and sagacity” (ibid., p. 142).
• 12Réné Descartes, A Discourse on Method (Everyman’s ed.), pp. 10–11.
• 13On the characteristic approach of the engineer type of mind to economic phenomena compare the present author’s study on “Scientism and the Study of Society,” Economica, Vols. IX-XI (new ser., 1942–44), esp. XI, 34 fï.
• 14Since this lecture was first published I have become acquainted with an instructive article by Jerome Rosenthal on “Attitudes of Some Modern Rationalists to History” (Journal of the History of Ideas, IV, No. 4 [October, 1943], 429–56), which shows in considerable detail the antihistorical attitude of Descartes and particularly his disciple Malebranche and gives interesting examples of the contempt expressed by Descartes in his Recherche de la vérité par la lumière naturelle for the study of history, languages, geography, and especially the classics.
• 15James Bonar, Philosophy and Political Economy (1893), p. 85.
• 16A. W. Benn, in his History of English Rationalism in the Nineteenth Century (1906), says rightly: “With Quesnay, following nature meant ascertaining by a study of the world about us and of its laws what conduct is most conducive to health and happiness; and the natural rights meant liberty to pursue the course so ascertained. Such liberty only belongs to the wise and good, and can only be granted to those whom the tutelary authority in the state is pleased to regard as such. With Adam Smith and his disciples, on the other hand, nature means the totality of impulses and instincts by which the individual members of society are animated; and their contention is that the best arrangements result from giving free play to those forces in the confidence that partial failure will be more than compensated by successes elsewhere, and that the pursuit of his own interest by each will work out in the greatest happiness of all” (I, 289).
• 17Edmund Burke, Thoughts and Details on Scarcity (1795), in Works (World’s Classics ed.), VI, 15.
• 18This phrase is used over and over again by De Tocqueville to describe the effects of socialism, but see particularly Oeuvres completes, IX (1886), 541, where he says: “Si, en definitive, j’avais à trouver une formule générale pour exprimer ce que m’apparait le socialisme dans son ensemble, je dirais que c’est une nouvelle formule de la servitude.” Perhaps I may be allowed to add that it was this phrase of De Tocqueville’s which suggested to me the title of a recent book of mine.
• 19John Locke, Two Treatises of Government (1690), Book II, chap. 4, §22: “Freedom of men under government is to have a standing rule to live by, common to every one of that society and made by the legislative power erected in it.”
• 20Lerner, op. cit., p. 5.
• 21Lord Acton, “Nationality” (1862), reprinted in The History of Freedom and Other Essays (1907), p. 288.
• 22The actions a government can expediently take to reduce really avoidable uncertainty for the individuals are a subject which has given rise to so many confusions that I am afraid to let the brief allusion to it in the text stand without some further explanation. The point is that, while it is easy to protect a particular person or group against the loss which might be caused by an unforseen change, by preventing people from taking notice of the change after it has occurred, this merely shifts the loss onto other shoulders but does not prevent it. If, e.g., capital invested in very expensive plant is protected against obsolescence by new inventions by prohibiting the introduction of such new inventions, this increases the security of the owners of the existing plant but deprives the public of the benefit of the new inventions. Or, in other words, it does not really reduce uncertainty for society as a whole if we make the behavior of the people more predictable by preventing them from adapting themselves to an unforeseen change in their knowledge of the world. The only genuine reduction of uncertainty consists in increasing its knowledge, but never in preventing people from making use of new knowledge.
• 23The difference between the rationalistic and the true individualistic approach is well shown in the different views expressed by French observers on the apparent irrationality of English social institutions. While Henri de Saint-Simon, e.g., complains that “cent volumes in folio, du caractère plus fin, ne suffiraient pas pour rendre compte de toutes les inconséquences organiques qui existent en Angleterre” (Oeuvres de Saint-Simon et d’Enfantin [Paris, 1865–78], XXXVIII, 179), De Tocqueville retorts “que ces bizarreries des Anglais pussent avoir quelques rapports avec leurs libertés, c’est ce qui ne lui tombe point dans l’esprit” (L’Ancien régime et la révolution [7th ed.; Paris, 1866], p. 103).
• 24Is it necessary to quote Edmund Burke once more to remind the reader how essential a condition for the possibility of a free society was to him the strength of moral rules? “Men are qualified for civil liberty,” he wrote, “in exact proportion to their disposition to put moral chains upon their own appetites; in proportion as their love of justice is above their rapacity; in proportion as their own soundness and sobriety of understanding is above their vanity and presumption; in proportion as they are more disposed to listen to the councils of the wise and good, in preference to the flattery of knaves” (A Letter to a Member of the National Assembly [1791], in Works [World’s Classics ed.], IV, 319).
• 25W. Dibelius, England (1923), pp. 464–68 of 1934 English translation.
• 26E. Vermeil, Germany’s Three Reichs (London, 1944), p. 224.
• 27Lord Acton, “Nationality” (1862), reprinted in The History of Freedom, pp. 270–300.
• 28Lord Acton, “Sir Erskine May’s Democracy in Europe” (1878), reprinted in The History of Freedom, p. 78.
• 29Lord Acton, Lectures on Modern History (1906), p. 10.
• 30Lord Acton, “Sir Erskine May’s Democracy in Europe,” reprinted in The History of Freedom, pp. 93–94.