As we noted above, Say's definition of value and riches elicited objections from Ricardo early on. In August 1815 he wrote:
It appears to me ... incorrect to say as you do page 95 [of the Catéchisme] that that man is superlatively rich, although he has few valuables, who can procure easily or for nothing those things which he wishes to consume. He may only wish to consume bread and water and may be able to procure no more. He cannot be so rich as his neighbour who has abundance of valuables which he can exchange for all the luxuries of life, which it [is] his desire to consume. Riches are measured by the quantity of valuables which a man possesses, not by the moderation of his wants. (Ricardo, Works, vol. VI: 248)
Modern readers might be tempted to object that Ricardo's criticism of Say is mistaken, and that there is nothing wrong with saying that a man is 'rich' – that is, experiences high 'utility' – when his desires/wants/needs are very limited or modest. If low command over exchange value produces 'bliss' (for such a man), he can well be said to be 'rich' in the above sense although he possesses few valuables. But such an objection would miss Ricardo's point, which is Say's confounding of value and riches. Ricardo made this clear in the final passages of the chapter 'Value and riches' of his Principles, where he charged Say with having 'confounded two things which ought always to be kept separate, and which are called by Adam Smith, value in use and value in exchange'. He expounded:
If by an improved machine I can, with the same quantity of labour, make two pair of stockings instead of one, I in no way impair the utility of one pair of stockings, though I diminish their value. If then I had precisely the same quantity of coats, shoes, stockings, and all other things, as before, I should have precisely the same quantity of useful objects, and should therefore be equally rich, if utility were the measure of riches; but I should have a less amount of value, for my stockings would be only half their former value. Utility then is not the measure of exchangeable value. (Ricardo, Works, vol. I: 280 n.; emphasis in the original)
In his notes on Ricardo's Principles Say responded to Ricardo's criticisms in the chapter on 'Value and Riches' in several lengthy notes. To Ricardo's argument that he had made utility the measure of value, and value the measure of utility (ibid.), Say replied that he indeed considered value to be a measure of utility and that, if two commodities are of the same value, this proves that men derive the same degree of satisfaction from their consumption. He continued: 'I would be very wrong if one could infer from what I have said that when the price of a commodity falls its utility is diminished. The utility of a commodity which falls in price approaches the utility of the air, which costs us nothing, although it is extremely useful' (Say in Ricardo 1819, vol. II: 89).(4*) It is unclear how Say could consider these statements to be compatible with each other.
Say's response to Ricardo's argument that the confounding of 'riches' and 'value' must lead to contradictions reads as follows:
The costless action of the natural agents, when it replaces the costly action of men and of capitals, diminishes the value of the product. Since all value is relative, the value of the products cannot fall without increasing the value of the revenues (or of the productive funds from which these revenues are derived). The cheaper the products sell, the richer are the consumers. I have proved elsewhere that the fall of the products resulting from economising in the costs of production does not alter the revenues of the producers. (ibid.: 92)(5*)
It is not made clear how a fall in the value of commodities, due to the introduction of better methods of production, is compatible with a rise in the 'value of the revenues'. And when Ricardo pointed out that Say, having confounded riches and value, was inconsistent in maintaining that riches are augmented when the value of commodities is reduced because of better methods of production (see Ricardo, Works, vol. I: 288), Say replied: 'In the passage cited it is said that the revenues of society remained the same; whilst the riches are augmented by all that can be purchased additionally with the same revenue' (Say in Ricardo 1819, vol. II: 97).(6*) Say's 'response' thus consisted in reiterating the statement which had been shown to be inconsistent with his confounding of value and riches. He apparently had not understood Ricardo's argument.
To the sample of inconsistent statements from his Catéchisme, by which Ricardo had sought to show him his contradictions, Say responded with a long note in which he introduced the distinction between 'richesses naturelles' and 'richesses sociales'. In the former category he included all things from which utility is derived, including air, sunlight, or the affection of family members, while the latter, with which Political Economy is concerned, comprised only the products of human industry possessed of utility. He then repeated his statement that 'Exchangeablevalue and riches are synonymes' (ibid.: 99)(7*), and continued:
Now the main problem is to bring the principles of Social riches or Political Economy into a relationship with those of Natural riches. If the produce is augmented due to the better use that we make of our land, of our capitals, and of our industry, more utilities (or natural riches) are produced, and at the same time the production of social riches appears to be diminished, because the exchangeable value of the product falls. Social riches thus do not follow the same course as natural riches. (ibid.: 100)(8*)
According to Say's view, which appears to be rooted in physiocratic thought, the introduction of new methods of production, and in particular the introduction of machines, is associated with the substitution of 'services productifs', that is, services supplied by productive agents, which are costly, by 'services naturelles', that is, services supplied by nature, which are gratis. 'Richesse naturelles' and 'richesse sociales' may therefore move in different directions. Overlooking this divergence, Say maintained, had prompted Ricardo to advance his unjustified reproach of contradictions in his, Say's, writings. However, the importance of this divergence is far from clear, as Say continued his response with the following statement:
It remains to identify those who benefit from this augmentation, those who are more rich, not only in terms of natural riches, but also in terms of social riches, in terms of exchangeable values, by this increase in produced utility. (ibid.: 101-2)(9*)
After having first stressed the divergence of 'richesse naturelles' and 'richesses sociales', Say now suddenly asserted that both are augmented simultaneously. According to Say, two cases of an augmentation of riches must be distinguished. The first case is this:
If, from causes the discussion of which is extraneous to our present argument, the exchangeable value of each bushel of wheat remains constant, in spite of an augmentation in the quantity of wheat produced, the increase in the riches produced goes entirely to the profit of the producers; that is, the proprietors of the funds of capital, the funds of land, and the industrial funds, which generated fivehundred instead of fifty bushels. The revenue derived from these portions of the funds is tenfold. (ibid.: 102)(10*)
No explanation is given how the value of each unit of wheat may remain constant in spite of an augmentation of the quantity produced. Say then contrasted this case with a second one, in which the value of wheat is falling:
If ... the exchangeable value of each bushel of wheat has fallen because of the greater quantity produced of it, the profit obtained is still in the proportion of fivehundred to fifty; but this profit is made by the class of consumers, who are richer by what they pay less, that is, by as much as the producers would have had from selling more. Their revenue does not increase tenfold, because they do not spend it entirely on wheatbread; but the portion of the revenue which is ordinarily spent on wheatbread is raised tenfold, and all the portions of the revenue which are raised tenfold, summed up, give a sum that is equal to the value of the product which has increased tenfold, if the price is supposed not to have fallen. In both cases the society thus enjoys an augmentation of values as well as an augmentation of utility. (ibid.: 102)(11*)
Again, the contradiction pointed out by Ricardo has not been resolved, but merely restated. No wonder, then, that Ricardo did not feel inclined to continue this 'debate'.
Given the unsatisfactory nature of Say's response in his notes, Ricardo saw no reason to withdraw his criticism in the second edition of the Principles, despite the changes that Say had made in the third edition of the Treatise. And when he rewrote the final part of chapter 20 for the third edition, Ricardo made it clear that no matters of substance were involved: 'M. Say, notwithstanding the corrections he has made in the fourth and last edition of his work, "Traité d'Economie Politique," appears to me to have been singularly unfortunate in his definition of riches and value' (ibid.: 279). The rewriting of the final paragraphs of chapter 20 in the third edition was prompted by the changes made by Say in the fourth edition of his Treatise, which made it impossible for Ricardo to retain the extensive citations from the earlier editions. In revising this part, Ricardo now summarized Say's contradictions by confronting inconsistent statements from the fourth edition of the Treatise in a series of numbered paragraphs (see ibid.: 282-3).
In May 1821 Ricardo wrote to Say, informing him that in his new edition he had 'pointed out the particular difference which exists between us, respecting the meaning which should be attached to the word "value". You use it in the same sense as "riches" and as "utility" and it is this part of your valuable work which I am very anxious should have the benefit of your further consideration' (Ricardo, Works, vol. VIII: 379). In one of the rewritten passages in chapter 20 Ricardo had asked rhetorically: 'When I give 2000 times more cloth for a pound of gold than I give for a pound of iron, does it prove that I attach 2000 times more utility to gold than I do to iron?' (Ricardo, Works, vol. I: 283). To this Say replied in a letter of 19 July 1821:
No, but if we suppose for a moment that a pound of gold and a pound of iron render a service to man that is perfectly equal in spite of the inequality in their value, then I say that in a pound of iron there are:
1999 units of natural utility which are a part of riches for which nature does not make us pay, and which do not concern political economy;
______ exchangeable riches, the sole object of our investigation.
2000 units of utility altogether, contained in a pound of iron. Whereas in a pound of gold there are:
2000 units of utility, all of which are exchangeable [value] because all of them are the fruits of our industry, our capitals and our lands; utility which is a part not only of our natural riches but also of our social riches, the only ones with which political economy can be concerned. (Ricardo, Works, vol. IX: 33-4)(12*)
Ricardo's response was communicated to Say in a letter of 5 March 1822.18 It is apposite to quote him in full:
I am happy to observe that the difference between us is much less than I had hitherto considered it. You speak of two different utilities which commodities possess, one, which they derive from nature, without any of the labour of man, the other, which they derive exclusively from his labour. You say that for the first of these, which you call natural utility, nothing valuable can be obtained in exchange, and it is only for that portion of utility which is given to a commodity by labour or industry, for which anything valuable can be obtained. ... You explain on these principles the case I had put to you of a pound of iron and a pound of gold, which I had supposed had exactly the same utility, though the gold was 2000 times more valuable. If we give 2000 times more for the gold than for the iron, you say, it is because that particular utility of which only Political Economy treats, namely that given by labour, is 2000 times greater than that given to iron, and you add that the iron has 1999 portions of natural utility for which nothing is given; of which the gold has none.
Although I cannot quite approve of the terms used to explain this truth, yet I do now, and always have substantially agreed in the reasoning which proves it, for I have always contended that commodities are valuable in proportion to the quantity of labour bestowed upon them, and when you say that they are valuable in proportion as they are useful, and they are useful in proportion to the quantity of labour or industry bestowed upon them, you are in fact expressing the same opinion in other words.
In your last edition of The 'Catechisme' you say page 2 that the riches of a person are in proportion to the value of the commodities which he possesses, and not in proportion to their quantity: so far you repeat the same opinion, but when your pupil calls upon you to explain what is the measure of the value of things, your [sic] answer that it is the quantity of all other things that the proprietor is enabled to command by their means, if he consents to exchange them. Now in this I think there is a contradiction, for we are told that riches are in proportion to value, and value in proportion to the quantity of things, therefore riches are in proportion to the quantity of things; and yet you say that riches are in proportion to value, and not in proportion to the quantity of things. (ibid.: 169-70)
Ricardo's letter continues:
Let us suppose that the same cause, namely, an economical process, which lowers the value of gold one half, lowers at the same time, in the same degree, and by the same means, hats, shoes, cloth and linen. In this state of things a pound of gold will command just as many hats, shoes, cloth, and linen as before any of the economical processes were discovered. I ask is the man equally rich as before who has a pound of gold? You first answer no, because he has not a commodity of equal value, and you secondly answer yes, because he can command an equal quantity of various other commodities. (ibid.: 170)
Touché! This demonstration was Ricardo's final attempt to explain to Say that his concepts of value and riches ended in complete muddle.19
According to Say, the utility of a thing 'is the primary source of value' (T: 284). He opted for taking the needs and wants of people 'as existing data, and reason upon them accordingly' (ibid.: 285). Interestingly, Say stressed that only those objects are possessed of a value (in contradistinction to free goods provided by nature, such as air or solar light) which represent utilities necessarily procured by human agency, 'because the very act of production implies an act of mutual exchange, in which the producer has given his personal agency for the product obtained by its exertion' (ibid.: 286). Hence, as in Ricardo, the attention focused on produced and reproducible commodities. The producer is said to put up with the productive effort only on the assumption that he receives 'an equivalent' for it.
Value, its component parts, and utility
Value (or price), Say contended, was governed by 'demand' and 'supply'. He gave these concepts the following analytical meanings. As regards the former, the reference is to 'aggregate demand' for a commodity and it is argued that it depends negatively on the commodity's price. This is due to two effects: first, as the price of the commodity is increased those less well off will gradually have to withdraw from the market: 'the number of its consumers is reduced'; secondly, those who can still afford to buy the product will buy less of the product: 'the consumption of each consumer is reduced also' (ibid., pp. 288-9). It seems that Say thought exclusively in terms of the income effect of a change in the price of a commodity; there is at any rate no explicit mentioning of the (relative) price effect. Therefore the passages under consideration must not be interpreted as anticipating the principle of substitution in consumption as developed by marginalist theory.
By 'supply' Say meant the total or aggregate 'supply or amount in circulation', that is, 'the quantity attainable at a given time, and ready for the satisfaction of those who are in want of the specific article' (ibid.: 288). He maintained that 'the supply will be more abundant, when the current price is high, and more scanty when that price has declined' (ibid.: 290). However, no compelling reason is given why this should be so. At any rate there is no element in Say's discussion resembling the principle of substitution in production as conceptualized by marginalist theory.
Say's reasoning culminated in the proposition that 'the relative intensity of supply and demand ... is the ground-work of price on every act of exchange'. He stressed: 'Demand and supply are the opposite extremes of the beam, whence depend the scales of dearness and cheapness: the price is the point of equilibrium, where the momentum of the one ceases, and that of the other begins' (ibid.: 290). There is a modern ring to this formulation, but there is not yet a sufficiently developed theory behind it. What can at most be said is that Say contributed to shunting the car of economics away from the classical on to a different track which was eventually to lead to marginalist theory.
The link between demand and supply, on the one hand, and cost of production, on the other, Say specified as follows: 'The price paid for every product, at the time of its original attainment or creation, is, the charge of the productive agency exerted, or the cost of its production' (ibid.: 298; similarly p. 315). Hence the price of each commodity is composed of, and thus can be resolved into, rents, profits and wages, or rather, as Say preferred to call them, into the 'profit of land', the 'profit of capital', and the 'profit of labour' (ibid.: 316-17). With technical progress, cost of production will fall and if, in the short run, the market price remains the same, this will be reflected in a 'larger profit than ordinary in this particular channel'. This above-normal profitability 'is said to naturally attract a larger proportion of productive agency, the exertion of which, by increasing the supply, reduces again the current price to a level with the bare cost of production' (ibid.: 298) – and, we may add, brings profitability back to a normal level. Say's argument concerns the 'gravitation' of market prices to their natural levels as contemplated by Smith and the classical English economists (see also Say in Ricardo 1819, vol. I: 79 n.; Béraud 1992: 378).20 The question is whether the new level of profitability equals the old one or is different from it. Say is not clear about this, as his following discussion shows.
The adding-up concept of value is a treacherous ground for the inattentive scholar because he may easily slip into the view that value rises or falls with a rise or fall of each single component. As is well known, Smith in places fell into the trap and so did Say. Thus, like the Scotsman Say was of the opinion that a low (high) wage rate meant low (high) prices. In one place we read: 'In proportion as the value he [the labourer] consumes is small, his ordinary wages may be low, and the product [!] of his labour cheap. If his condition be improved, and his wages raised, either his product becomes dearer to the consumer, or the share of his fellow producers is diminished' (T: 336; similarly p. 339). As Ricardo was to clarify, a fall (or rise) in the remuneration of any productive factor of necessity involves a rise (or fall) in the remuneration of some other factor, given the technical conditions of production, whereas the variation in the price of a commodity in terms of some standard of value cannot be ascertained a priori, but depends on the technical conditions of production (direct and indirect) of the commodity relative to those of the standard. There is no reason to presume that a lower (higher) real wage generally involves lower (higher) prices of commodities. And it is not true that a higher real wage rate involves 'either' a higher price 'or' a smaller share of some other factor: it always involves the latter.
Measure of value
This leads directly to the problem of Ricardo's search for an 'invariable measure of value'. Say, like other commentators, had difficulty in understanding the analytical significance of this concept. This is not surprising, because without an intimate knowledge of the different steps reflecting Ricardo's consecutive attempts to simplify the problem of distribution and relative prices his concern with an invariable standard was bound to look like chasing a will-o'-the-wisp. According to Sraffa (1951: xxxi-xxxiii), Ricardo proceeded in four steps. The first consisted of eliminating the problem of the rent of land in terms of the theory of differential rent in the Essay on Profits, published in 1815 (see Ricardo, Works, vol. IV). The second step consisted of trying to get rid of the problem of value in determining the general rate of profits by assuming the 'corn model'. The third step consisted of adopting the labour theory of value. This allowed him to ascertain the values of commodities independently of income distribution and thus helped him to dispel the adding-up doctrine, suggested by Smith and adopted by Say, of price as a sum of wages and profits (and rents). It thereby put into sharp relief the constraint binding changes in the distributive variables. However, as Ricardo became increasingly aware, the labour theory of value could not generally be sustained as a theory of relative prices. This is because with unequal proportions of labour to means of production in different industries and unequal degrees of durability of fixed capital, relative prices would not only depend on the quantities of labour 'embodied' in the various commodities, but also on the level of the rate of profits, and would change with that level: with compound interest the weight of the profit component in prices depends on the rate of profits. Ricardo's search for a measure of value that is 'invariable' with respect to changes in distribution was the final step in his efforts to simplify the theory of distribution.21
Say can be expected to have known little about these consecutive attempts and especially about the fact that the concept of a measure that was meant to be invariable with respect to changes in income distribution was essentially a device designed by Ricardo in order to see through the complexities of the dependence of relative prices on the rate of profits. Say objected to the very idea of Ricardo's search for such a measure: 'An invariable measure of value is a pure chimera, because one cannot measure values other than in terms of values, that is, by a quantity essentially variable. This does not mean that value itself is chimerical; it is no more so than the temperature of bodies, which cannot be fixed beforehand' (Say in Ricardo 1819, vol. I: 12-13; see also ibid., vol. II: 69-70).(13*) This criticism anticipated an objection levelled by Samuel Bailey at Ricardo's concept. As we know today, while Ricardo's idea of a measure that is invariable with regard to changes in the distribution of income can be given a precise expression in well-specified conditions, his overall search was indeed futile (for details, see Kurz and Salvadori 1993).22
Gross and net revenue
Smith was accused of having put forward statements in his discussion of the component parts of value which can be read as confounding gross and net produce. This kind of criticism can certainly be levelled at Say, who stressed vis-à-vis the doctrine of the Physiocrats:
Whence it appears, that the term net product applies only to the individual revenue of each separate producer or [entrepreneur]23 in industry; but that the aggregate of individual revenue, the total revenue of the community, is equal to the gross produce of its land, capital, and industry. Which entirely subverts the system of the economists of the last century, who considered nothing but the net produce of the land as forming revenue, and therefore concluded that this net produce was all that the community had to consume, instead of admitting the obvious inference, that the whole of what has been created, may also be consumed by mankind. (ibid.: 318)
While Say is justified in criticizing the concept of 'produit net' in the Physiocrats as too narrow, his alternative conceptualization suffers from a confusion with regard to the definition of the net produce. For if the gross produce is defined as the 'total revenue of the community' (or total 'value added'), consisting of wages, profits and rents, then these income components cannot in turn be taken to form the costs of production which must be deducted from the gross produce in order to obtain the net revenue. Ricardo pointed this out in the third edition of the Principles: 'Of net produce and gross produce, M. Say speaks as follows: "The whole value produced is the gross produce; this value, after deducting from it the cost of production, is the net produce." ... There can then be no net produce, because the cost of production, according to M. Say, consists of rent, wages, and profits. ... Take a whole from a whole, and nothing remains' (Ricardo, Works, vol. I: 421 n.).