[ Avertissement : Les nombres entre crochets indiquent les pages de l'édition de référence. ]
Pour toute remarque concernant cette édition Internet de
Henry Sidgwick,The Wages Fund Theory,
Philippe Bazard , firstname.lastname@example.org
 There is no more burning question in the present state of Economic controversy than that of the Wages-fund, and none on which the opinions of teachers in full repute appear more widely divided. The doctrine that was held by J. S. Mill till 1869 (and then "presumed" by him to be found in every systematic treatise on Political Economy) is still apparently taught by Professor Fawcett; (1) it was restated with some additions and explanations, but without important qualification, by the late Professor Cairnes; (2) and it is at least continually implied in the advice tendered to the working classes from the capitalists' side, even by sympathetic and fair-minded writers, such as Mr. Brassey. On the other hand, the doctrine is altogether rejected by Mr. Jevons, Mr. Thornton, Mr. Cliffe Leslie, Mr. F. Walker, and other American and English economists; and, under the influence of these writers, I think there is a growing tendency in the organs of cultivated opinion to treat it as exploded. It seems, therefore, opportune to consider, first, what precisely is the theory of the determination of wages which is thus held by one set of economists and denied by others, and what is its real importance; secondly, to weigh the negative arguments urged against it by these latter; and thirdly, to inquire what positive doctrine is or may be proposed in place of that which is thus rejected. The discussion in the present article will fall under these three heads taken in the above order.
The first inquiry might seem to be superfluous, after the doctrine has been stated by such able expositors as J. S. Mill and Cairnes. But the former's statements have certainly been misunderstood both by friends and by foes; and I must confess that I have found some difficulty in ascertaining the exact limits of the latter's position. In the passage (B. II. c. xi. p. 1) in which Mill first speaks of the wages-fund he seems rather to describe the manner in which the whole sum paid in wages is distributed, than to state the law by which the total is determined. "What may be called the wages-fund of a country," he says, is made up of "that part of the circulating capital (of the country) which is expended in the direct purchase of labour," together with all other funds that are paid in exchange of labour. If we knew no more of the wages-fund than that it is a total thus heterogeneously made up, it might seem to be an  insignificant truism to affirm that wages depend upon the relative amount of the wages-fund and of population ; as it is merely an incontrovertible deduction from the principles of simple arithmetic that "the general rate of wages cannot rise but by an increase of the aggregate funds employed in hiring labourers or a diminution in the number of the competitors for hire."
Mill, however, means to assert something much more important than this elementary arithmetical proposition; (3) something which is more distinctly implied in the statement which he gives as roughly equivalent to the above, that "wages depend on the proportion between population and capital." He means that, since the great majority of the wage-earning class are labourers hired by employers for a profit, the amount of wealth devoted to the payment of wages is mainly determined by the law of increase of capital. (4) Hence, since capital is the result of saving, the wages-fund must be fixed independently of the discussion between individual employers and labourers as to the wages which the former are to give the latter. Mill was, of course, aware that the proportion of the whole capital of a country that is employed in wages is not strictly a constant one, but varies with the changes that invention introduces into the methods of production; but for his purpose this variation is not important, provided it is independent of the haggling of the labour market. His point was, that this proportion as well as the whole amount of capital must be taken as "predetermined" in considering the problem of Distribution.
And this is the point which Cairnes is still concerned to maintain, with a more careful statement of the relation of the wages-fund to the whole capital of the country. Cairnes explains that the character and condition of the national industries determine the proportion which labour will bear to the other part of capital, which is not wages-fund, but fixed capital, raw materials, &c. He allows that, since wages may be higher or lower while the amount of labour remains the same, this is not quite the same thing as determining the proportion of wages-fund to capital. But he argues that if the supply of labour, the total amount of capital, and the proportion of capital that is not wages-fund to labour, be all three given, the wages-fund and the average rate of wages must also be determined;  and that its determination gives us the rather remarkable result that not only the average rate of wages, but the whole money spent in wages is decreased by an increase in the supply of labour. (5)
Here, then, we have an abstract statement of the wages-fund doctrine in its latest and most precise form; but we shall gain a clearer view of it by considering briefly its practical bearings. From the manner in which Mill introduces the notion of a wages-fund in close connection with his discussion of remedies for low wages, he might perhaps be understood to imply that the fund is too rigidly limited to admit of being increased by any legislative action or by any moral pressure on employers or other rich persons. But, in fact, he is far from maintaining anything of the kind; and in his argument against the expediency of such remedies he does not refer to any supposed limits of the wages-fund, but to the Malthusian theory of population. So again he (as well as his disciples) plainly recognises, in the distinction drawn between "wages" and the "cost of labour," that the labourers' share of consumable commodities may be increased by an increase in their efficiency without any diminution of profits; though it is undoubtedly a defect in his exposition - unconsciously inherited from Ricardo and James Mill - that he keeps too much in the background the large possibilities of amelioration which this consideration opens up. In short, the only mode of enlargement of the wages-fund which he, or any other economist of repute, has believed to be rigidly excluded by its so-called determinateness, is enlargement through the successful bargaining of labourers with their employers.
But there is a further qualification of fundamental importance which has been frequently forgotten in practical applications of the wages-fund theory. The most confidently "deductive" economists, so far as I know, have never supposed that the fund spent in wages in any particular trade was determined independently of the bargaining between employers and employed. Certainly Mill, in the days of his completest acceptance of the wages-fund doctrine, allowed most explicitly that workmen in certain trades might, by combining, keep their wages at a higher level than they would otherwise secure. He only argued that this could not be done without also causing the aggregate wages of the rest of the working class to be less than they would otherwise have been. But since the theory left perfectly indefinite the individuals and even the classes on which such loss would fall, this consideration could only be supposed to influence  the action of working men so far as they are governed not by self-interest or even by natural esprit de corps, but by a refined and abstract sympathy with the interests of certain unknown labourers. Now I am far from saying that it is undesirable to encourage this kind of sympathy, or that the working classes are incapable of being seriously influenced by it. But it must be allowed that this is a motive very difference in kind from those by which the "economic man" has been usually conceived to be governed. In most other departments of social organization, under the régime of free competition, we continually find individuals and classes getting rich in a manner which involves a corresponding diminution in the wealth of other men; and not merely of "other men" abstractly and indefinitely conceived, but of very definitely and familiarly known individuals - tradesmen in the same street, physicians in the same town, barristers on the same circuit. Of course it may be replied that the competition which brings about this result benefits society by keeping production generally at the highest pitch of efficiency. But the point that I am urging is that no economist has ever supposed the competing producers to be extensively influenced by this consideration, or has asked, with a serious expectation of being listened to, that they should refrain from all modes and expedients of competition except such as are socially useful. In all the other eager struggles for well-being which society everywhere presents, we are content to direct the force of social disapprobation against such diminution of other people's wealth as is due either to fraud or to mere recklessness. Hence, when we read the economic sermons on the text of the wages-fund which have so often been addressed to workmen in their struggles with employers, we must conclude that the preacher has either a confused notion of the doctrine that he is expounding or a remarkably high opinion of the moral standard of his audience. (6)
It seems clear, in short, that the wages-fund doctrine-or any other theory of the determination of general or average wage is of practical importance only so far as men's pursuit of self-interest well understood "admits of being restrained and modified by moral or philanthropic motives. So far as the labourers in any particular industry are ,economic men" of the ordinary type, the considerations which they have to take into account in regulating their  combined action for raising and sustaining wages are of a totally different kind. They ought to forecast carefully the effects of the competition of other actual or possible labourers in the same industry, and the decrease in the demand for their product that will result from a rise in its price. But they certainly need not trouble themselves much about the general wages-fund of the country, even if its "predeterminateness" were established on the most solid basis of scientific reasoning.
It is now time to examine this reasoning more closely. As we have seen, the foundation of the whole theory consists in the supposed independence of two facts, (1) "saving," the process by which capital is increased, and (2) the discussion between employers and employed by which the wages of particular labourers are fixed. If (1) is altogether uninfluenced by (2), it will certainly follow that though the efforts of particular labourers to get higher wages may be successful, it can only be at the expense of other labourers; and similarly, though particular capitalists may beat down their workmen to a lower wage, the money they thus save, being destined for productive employment, must ultimately take the form of wages of other labourers. And this degree of rigidity in the wages-fund is certainly implied in Mill's language in his Political Economy (compare v. c. x. § 5 with II. cxi. §§ 1 and 3 - passages which, I observe, he did not think it worth while to alter even after his review of Thornton). Mill's disciples, however, seem to have admitted a greater degree of elasticity in the limitations of the fund long before the appearance of Thornton's criticism. Professor Fawcett (e.g.), Pol. Econ., II. c. ix. (ed. I), while arguing that combinations of workmen cannot permanently raise wages, affirms that they may do so temporarily if they demand an increase when trade is flourishing and profits high. In this passage he implies that this addition to the wages of one set of labourers will not be taken from the wages of another set ; and the same conclusion is reached by Cairnes (Some Leading Principles, Pt.II.c.3). But how is this result consistent with the chain of reasoning that we have just been considering ? If the total amount of capital is determined by saving, and therefore independently of the haggling of the market, and if the proportion of capital that becomes wages-fund is determined by the character of the national industries, &c., together with the supply of labour, how can any action of any set of labourers (so long as these determining conditions remain unchanged) increase the total wages-fund, as it must do if they raise their own wages without diminishing those of any other labourers ? The only possible answer to this question seems to be that which Mill gives in his review of Mr. Thornton's book "On Labour." Although the process of increasing capital is generally voluntary - what we ordinarily call "saving" - there is no economic law which prevents it from being compulsory; and, in,  fact, when an employer yields to the demands of a union and raises his workmen's wages, if he finds the money by cutting down his expenditure instead of taking it from a bank or some other investment, he does increase capital in this compulsory way. Whether we choose to call this saving or not is a mere question of words; it is at any rate a process not independent of, but determined by, the haggling of the labour-market.
But if this be so, what becomes of the wages-fund theory ? If this compulsory economy be possible at all, why should it not be on the whole successful ? In fact, neither Professor Fawcett nor Professor Cairnes really denies this possibility. When they say that combinations of labourers can only raise wages "temporarily," (7) they do not mean to assert that the temporary rise will inevitably be balanced by a consequent temporary fall in the wages either of the same or of other labourers; they clearly imply that this will not be the case, if only the combined action of the labourers be wisely directed, and their demands for advances only made when trade is exceptionally prosperous. It is true that Cairnes does verbally deny that a "permanent elevation of the average rate of wages" can be effected in this way; but the whole context distinctly shows that in this phrase he has momentarily fallen into a confusion between "average" and "usual." If the cases in which combinations can successfully raise wages are, as he assumes, "exceptional," it follows, of course, that they cannot raise the usual rate in any branch of industry; but if these exceptional rises are not inevitably compensated by consequent falls, it equally follows that they must tend to raise the average rate. Of course the force thus exerted by any single strike is very slight; but if we make the rather ideal supposition that the whole body of labourers in their several industries are wisely led, and thus never demand an advance unsuccessfully, it is clear that the level of average wages may be steadily elevated by a continual series of slight rises. And if we suppose the movement of wages to take place not in one trade only, but along the whole line of the labour-market, (8) what is there to prevent the compulsory enlargement of the wages-fund from being both rapid and extensive ?
 It may be replied that the fall of average profits involved in such a rise of wages will diminish the motives to save, and thus ultimately reduce the voluntary additions to the wages-fund so much as to balance the compulsory enlargement of the fund through the successful bargaining of the workmen. Here, I think, we have the last stronghold of the believers in a strictly determinate wages-fund; and, indeed, the only position which they seem seriously prepared to maintain. It is not the wages-fund at any given time, but the wages-fund in the long run, which they really hold to be independent of the haggling of the labour-market. In considering this position, we must bear in mind, as was before observed, that this fall in profits will not accompany the rise in wages, so far as the efficiency in the labourer is increased through the improvement in his physical health, due to better food, &c. The practical importance of this consideration seems to have been conclusively established by Mr. Brassey. (9) And however we may agree with Cairnes's strictures on the absurdity of formulating a "general law that the cost of labour is uniform" on the strength of Mr. Brassey's statistics, a considerable approximation to such uniformity, under certain conditions and over a limited range, has undoubtedly been made out; and this is quite sufficient to render nugatory all general statements as to the inevitable connection of a rise of wages with a fall of profits.
But, secondly, even if we assume the efficiency of labour to be unchanged, it does not seem that we have any means of predicting à priori the extent to which a fall of profits will operate in decreasing the additions to capital. In examining this point we have in the first place to remove a confusion between "profit " and "interest" which pervades the treatment of this subject by Mill and his school, notwithstanding the express distinction between the two notions in Mill's analysis of profit (10). The employer's profit, as Mill explains, consists (besides indemnity for risk) of "wages of superintendence" as well as interest on capital; so that primâ facie any increase in the remuneration of his workmen may operate to reduce the former element rather than the latter; whereas it is the latter and not the former that constitutes the general inducement to save, which would be comparatively unaffected by a reduction in the wages of superintendence. There is nothing, I believe, in the extremest wages-fund doctrine to lead us to conclude that the average remuneration of employers is incapable of decrease. Suppose that average employers' profit, exclusive of risk, is now eight per cent., of which three per cent. represents absolutely safe interest on capital, and five per cent. the "wages of superintendence :" if an increase in workmen's wages should reduce average profits to seven per cent., have we any grounds for concluding that it will not in the  long run be possible to get the employers we want for four per cent.? What, then, are the persons who now become employers to do under the circumstances supposed ? Will they, to any considerable extent, live idly on three per cent. instead of working for seven ? Perhaps it may be said that they will enter the professions; and no doubt there is a certain competition between the learned professions and "business" for the services of young men of ability with some inherited property, so that this tendency would operate to some extent; but then this only leads to the wider question, Why should there not be a general fall in the average remuneration of the whole first grade of services, that includes all kinds of professional work as well as all kinds of employment of industry ? I know of no economic law that renders such a fall impossible.
At any rate it is clear, I think, that the blow caused by increase of workmen's wages will not fall immediately on the interest of capital, but will only reach it through a medium that will absorb at least a good deal of its force. But if we grant that the rate of interest will be reduced through the refusal of capitalists to become employers at the lower remuneration, we have really no definite knowledge of the extent to which the fall would check accumulation in England. All the saving that takes place to provide for old age, bad times, children, &c., is nearly unaffected by the rate of interest. Most of the saving accomplished by the poorer classes is of this character, and we may observe that a rise in wages would have a direct tendency to increase this quota. Again, many persons have a nearly fixed standard of living, and so long as their income is more than sufficient to provide for this they will save the surplus, whatever that may be : in proportion as this is the case their saving is only affected by the rate of interest so far as their income is affected by it. But, farther, so far as men in business and the professions save with a view of ultimately retiring on a certain income, it is obvious that a decrease in the rate of interest may tend to make them save more rather than less; as they will require a larger amount of accumulated capital to obtain the same amount of annual income. I do not wish to exaggerate the force of these considerations; I quite allow that a decrease in the rate of interest would on the whole tend to check accumulation of capital within the country; but I submit that we have no such means of measuring this tendency as would enable us to affirm that its operation must necessarily keep down average wages (cæteris paribus) to their present level. The same may be said of the tendency of capital to seek investment abroad in consequence of any fall in average profits confined to one particular country. No doubt it is of the utmost importance, in considering any concrete case, to take this consequence into account. But it seems clear that this tendency cannot be estimated à priori with any  such exactness as would be necessary if it were used as a basis for the wages-fund theory; so long as we are contemplating a state of things in which the mobility of capital is as imperfect as it is actually shown to be by the very different rates of interest and profit permanently maintained in different countries.
In the preceding discussion I have assumed that the wages of labour are to be regarded as a certain share of the produce of previous labour; the remainder being consumed by the capitalists or landlords themselves, or unproductive labourers whom they pay, or the Government. This is the view commonly taken by English economists; and I have adopted it because it seemed to me that the supposed rigid limitations of the wages-fund could be shown to be illusory on this view as well as on any other. I must now point out, with Professor Walker, that there is no absolute necessity that workmen's wages should be paid entirely out of the saved results of past industry. In fact, in newly colonised countries, where capital and labour are at once scarce and highly productive, the most natural and convenient plan is to pay the labourer out of the product of his industry, whatever sum he requires for subsistence while labouring being merely advanced (11). No doubt in old countries like England wages are for the most part completely paid out of the results of previous industry; still it is worth observing that there are large exceptions to this rule. For example, a large part of the labourers employed in transportation (railway porters, bus conductors, &c.) and in distribution (shop-boys, &c.) are paid by their employers after the services for which they are paid have been remunerated by the consumers; and a considerable number of small artisans (tailors, shoemakers, &c.), working each on his own hand and owning the small capital that he requires, do not receive the return for their labour till some time after they have sent home the finished product. But even if we put out of sight these cases, and consider all wages as actually paid out of the produce of previous labour, saved and devoted to production, it still is not true that the wages-fund at any moment is rigidly incapable of increase without trenching on the consumption of other members of the community. For the stock of finished products available at any given time is always somewhat than would be required to supply the consumption of society at the existing rate, during the varying periods of time that must elapse before it can be replaced by fresh production; and though the margin given may not be large, it is sufficient to invalidate the conclusion that an increase in the supply of labour without an increase in saving cannot possibly enlarge the wages-fund. For the new  labourers can be employed productively; and the prospect of this increased production will tend to increase immediately the sale of goods already finished, and therefore the whole fund of commodities in consumers' hands, and therefore probably the wages-fund (12).
So far I entirely agree with Professor Walker that wages may be philosophically regarded as paid out of current produce. But this does not help us to determine what share of the product will go to the labourer; and when Professor Walker goes on to say that "the product furnishes at once the motive to employment and the measure of wages," I cannot help thinking that he has confounded the notions of "measure" and "limit." He seems, indeed, to hold that on the assumption of perfect competition and perfect mobility of labour, the determination of average wages is quite easy and straightforward. "In a state of active competition, each labourer will sell his labour at the highest price which any employer can afford to give, since the employers are in competition among themselves for labour. Each employer will get his labour at the lowest price at which any labourer can afford to sell it, since the labourers are in competition among themselves for employment." (13) Phrases like these frequently occur in economic discussion, and they certainly seem to give a delightfully clear and simple account of the matter. I think, however, that the intellectual satisfaction that they afford depends upon an essentially vague notion of the effects of competition. Competition tends to equalise the terms on which several similar exchanges of commodities are made (including under the term "commodity" services of all kinds). Thus it tends to prevent there being two prices in the same market; two rates of interest to capital, risk being equal; two rates of profit or wages in different branches of industry, except so far as effective work in one branch involves qualifications of a rarer kind or requiring more expensive training than effective work in another. So again competition between different commodities limits the price of any commodity, even when monopolised, unless it be absolutely indispensable. But it is not clear what effect competition can have in determining the rate of exchange of any two commodities owned by two parties (external influences being excluded) when each commodity is incapable of being a substitute for the other, while yet it is the urgent interest of both to exchange. Now, when we are inquiring into the determination of "general" wages, we conceive the whole class of (employing)
 capitalists on the one hand, and the whole class of labourers on the other, as two parties to a general bargain; so that it becomes as difficult to settle the terms of the bargain by competition as if there were but one capitalist and one labourer. In short, it seems to me that while Professor Walker's argument gives a coup de grâce to the old wages-fund theory, it supplies no substitute for it; it leaves us with no theoretical determination whatever of the average proportions in which produce is divided between labour and capital.
Nor does it seem that any of the English economists who agree in rejecting the "Wages-fund Theory" has been successful in filling the gap thus created in economic doctrine, though more than one of them made an elaborate attempt to fill it. The most ingenious of such attempts is that presented by Professor Jevons, in his Theory of Political Economy, c. vii. Mr. Jevons considers that the industrial bargain to be contemplated from the side of labour rather than (as is ordinarily the case) on the side of capital; in fact that, philosophically speaking, it is labour that hires capital, and not capital labour. If we take this point of view, we find that the fundamental function which capital fulfils is, that by enabling the labourer to wait longer the produce of his labour, it enables him to apply it in a manner ultimately more productive. Of course the degree of this utility depend on the development of the mechanical and other arts of invention; but supposing these given, supposing that inventive ingenuity has devised instruments and processes calculated to render more profitable while delaying its remuneration, it is capital, the saved product of past labour, that enables labour to realise this productiveness. Since, however, capital - in countries such as ours -is absolutely indispensable to labour, it may seem that this does not yet enable us to determine the conditions on which it will be hired. But Mr. Jevons points out, that though some capital is thus indispensable, all that is actually employed is not : the difference between the whole amount actually used and a somewhat smaller amount - the "last increment of capital," as Mr. Jevons calls it - is something that labour could do without, though it will be rendered more productive by using it. Hence, in the increment of produce which this last increment of capital enables labour to obtain we may find a theoretical measure of the normal remuneration of capital, and therefore of the normal remuneration of labour.
On this theory it is to be observed, in the first place, that it does not attempt to settle the distribution of produce as between employers and employed, except so far as the employer's share consists of interest. That is, it does not help us to determine what Mill calls the "wages of superintendence." Now it is just this latter that in our practical discussions usually appears as the most prominent element of the problem : what English workmen grumble at is not the rate of interest, but the undue extra profit which they  suppose the employers to be making. But there is a more fundamental defect in Mr. Jevons's reasoning. He has apparently overlooked the fact that the entrepreneurs will require to be remunerated for the trouble and anxiety of employing the last increment of capital : so that the increment of produce which this enables labour to obtain cannot be taken as the measure of interest, any more than the additional value produced by the last increment of labour can be taken as the measure of wages. "Superintendence" must take its toll of both; and therefore, until we have settled the proportion that is to go to superintendence, we have not really advanced a step towards solving the abstract problem of distribution.
In short, the very node of this problem lies in determining the entrepreneur's normal remuneration; a complicated question, the difficulties of which English economists do not seem to have fully recognised. This is strikingly shown in the assumption, which they commonly make as a matter of course, that industrial competition tends to equalise the rate of profit (as well as interest) on capitals of different amount. There is no à priori ground for supposing this; since, on the one hand, the labour of managing capital certainly does not increase in proportion to the amount managed, while, on the other hand, the owner of capital has a qualified monopoly of the opportunity of employing it. Therefore, even if the assumption above mentioned is borne out by experience, this fact ought itself to be taken as a result requiring careful analysis and explanation.
What, then, the reader will ask, so long as this central point is undetermined, can abstract economic theory do for us ? What answer can it give to questions as to the amount of the aggregate wages fund and the average rate of wages ?
It can furnish us with no doctrine so simple and definite as that which we have been considering. Still, it can give us positive results of two different kinds, as follows:-
I. It can show limits on either side within which it is the common interest of employers and employed that the variations in wages should be confined. For (1) an inferior limit is given by the point at which any further decrease in wages would diminish the labourer's efficiency, so as to decrease the value of his work by an amount greater than the decrease in his wages (14). I do not say that wages never do fall below this limit, owing to the short-sighted greed of employers; but it is obviously not the employers' real interest that they should fall below it, and therefore the action of free but enlightened competition tends to prevent such a fall. (2) Similarly a superior limit is given by the point at which a fall in interest  consequent on any further rise in wages would check accumulation so powerfully that the portion of the prevented saving which would have gone to wages exceeds the total amount of the rise (15). I have already said that I see no ground for believing that we have practically reached this point in England. But it hardly belongs to abstract economic theory to investigate the point at which the limit will be attained, as this will obviously depend upon conditions varying from country to country, and in the same country at different times. It is conceivable that in a society where there prevailed generally a keen susceptibility to the pleasures of exertion, and a preference for simplicity in the satisfaction of physical needs and appetites, the limit might not be reached until the "reward of abstinence" had been reduced to zero; so that the price paid for borrowed capital would represent nothing but compensation for risk. It is conceivable that at the same time, owing to the general spread of education, the average rate of that part of each employer's profits which represents his wages of management might be reduced so low that it would only exceed the wages of ordinary labour by an amount just sufficient to compensate for the extra anxiety of his work and the extra expensiveness of his training. In this way our individualistic society might find itself peacefully conveyed to the goal at which socialism aims without any disturbance in the existing organization of industry. I do not, however, wish to imply that we are at present anywhere near this consummation.
II. Within the limits just stated economic theory shows us forces of an equilibratory or compensatory nature, which tend to reduce the effect of any upward or downward movement in wages. It is true that a fall in wages, accompanied as it will be, cæteris paribus, until the lower limit is reached, by a rise in profits, will give a certain stimulus to accumulation, and so tend ultimately to raise wages again. It is similarly true that a rise in wages tends to give a certain check to accumulation, probably increasing in strength as the rise approaches the superior limit above mentioned. What is not true, as I have tried to show, is that we have any calculus for measuring à priori the amount of these forces; or any grounds for affirming that they will ultimately bring back wages, after any movement, to the point from which the movement began. There is no question here of a discrepancy between "Deduction" and "Induction," between abstract theory and actual fact. It merely requires a careful consideration of the assumptions on which ordinary economic reasoning proceeds, to convince us that the definiteness of its conclusions on this point has been gravely overrated.