We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
We use cookies to distinguish you from other users and to provide you with a better experience on our websites. Close this message to accept cookies or find out how to manage your cookie settings.
Terry Peach[1]
(Shanghai University of Finance and Economics and University of Manchester)
I have the impression that Roy Grieve is not convinced – “unfazed”, as he might say- by the response I gave (Peach 2020) to criticisms he made (Grieve 2019) of an interpretation of Adam Smith’s “labour theory of (real) value” that I advanced some years ago (Peach 2009). It is certainly true that I could not discover any merit in Grieve’s arguments, so it is perhaps understandable that he should wish to rejoin the fray, presumably in the hope of sharpening his critique. Yet, it is surprising that, for the most part, he should merely recycle the same criticisms which, as I shall demonstrate, have not improved with age. What is striking about this later effort is the greater prominence Grieve accords to his conception of “the labour theory of value” according to which, by Grieve’s definition, it would be impossible for Smith (or Ricardo, or anyone else) to have held or applied such a theory, in any variant form, in any context other than Smith’s “early and rude state”, distinguished by the absence of capitalists and landlords and their associated incomes of profit and rent. It is, indeed, this very conception of a “labour theory” that he invokes when confronted with evidence that he cannot dispose of (to his own satisfaction) in any other way.
To restate my original position, stripped of Grieve’s rhetorical gloss[2]:
My purpose ... [is] to question the widely held opinion that Adam Smith confined his use of a labour theory of exchangeable value to an ‘early and rude state of society’ in which independent labourers exchange the surplus products of their labour. As I propose to establish, Smith continued to apply labour theory reasoning to the later commercial society, both to explain static exchangeable values and their changes.
And as I had hoped to clarify in Peach (2020):
For the purposes of this article, and without claiming that this is the only possible construction, a labour theory of exchangeable value should be understood as a theory maintaining that (changes in) exchangeable value between members of some designated class of produced commodities are determined by (changes in) the quantities of labour expended on their production.
Regrettably, it seems that I had only obscured my position, at least on Grieve’s understanding:
the labour theory of value (or rather Peach’s version thereof) focuses not so much on a given (static) relationship between labour embodied and commodity prices, but attaches particular importance to changes in the quantity of labour embodied, as responsible for changes in commodity prices. (Grieve’s italics)
For the avoidance of doubt, the interpretation I advanced, and the one I defend here, is that “Smith continued to apply labour theory reasoning to the later commercial society, both to explain static exchangeable values and their changes”. Let me stress: it is the presence of both types of explanation that I take as evidence of Smith’s “labour theory reasoning”.
In what follows I consider the interpretation of specific passages that Grieve has extracted for critical comment, beginning with what is, it seems to me, a straightforward example of Smith explaining exchangeable value in a “commercial” society by quantities of labour expended in production:
The proportion between the value of gold and silver and that of goods of any other kind, depends in all cases … upon the richness or poverty of the mines, which happens at any particular time to supply the great market of the commercial world with those metals. It depends upon the proportion between the quantity of labour which is necessary in order to bring a certain quantity of gold and silver to market, and that which is necessary in order to bring thither a certain quantity of any other sort of goods.[3] (WN II.ii.105, italics added)
There is no reference here to “cost of labour” or “real” magnitudes that might, to an unwary reader, obscure Smith’s position. Indeed, there is no shadow of a doubt that he was using “quantity of labour” as the determinant of exchangeable value. Here is Grieve’s response:
while the richness or poverty of the mines will certainly be reflected in production costs and consequently the price of what is produced, recognition of the fact of higher labour costs causing higher commodity prices does not, of itself imply adherence to the labour-embodied theory of value. For that to be so, it would have to be supposed also that rent and profits did not contribute to costs of production. (Italics added)
Grieve gratuitously translates Smith’s proposition into the language of costs and prices, and then, perhaps with some inkling that a more conclusive argument might help his cause, he lands his “killer punch” that it would be impossible for Smith to have betrayed a continued adherence “to the labour-embodied theory” unless he “supposed … that rent and profits did not contribute to costs of production”; that is, unless the argument was confined to an “early and rude state”. Given that it was not so confined, ergo (for Grieve) the absence of “labour theory” reasoning. The possibility of an alternative inference seems not to occur to him, namely, that Smith (later to be followed by Ricardo, and others) continued to claim the applicability of “labour theory” reasoning even though costs of production include elements other than wages.
Grieve deploys similar “reasoning” to my interpretation of a passage in which Smith is explaining the consequence of differences in transportation arrangements in Europe compared with China and “Indostan”:
Through the greater part of Europe … the expence of land-carriage increases very much both the real and nominal price of most commodities. It costs more labour, and therefore more money, to bring first the materials, and afterwards the compleat manufacture to market. In China and Indostan the extent and variety of inland navigations save the greater part of this labour, and consequently of this money, and thereby reduce still lower both the real and the nominal price of the greater part of their manufactures.” (WN I.xi.g.28, italics added)
Leaving aside temporarily the reference to “real price”, the way I interpret this passage is that the difference in nominal prices (and exchangeable value) of manufactures between the two regions is explained by Smith in terms of differences in transportation, as defined by differences in the quantity of labour required to market the commodities. As for the reference to the cost of labour, both here and in other contexts this is taken by Smith as a proxy for quantity of labour. As he states earlier in the same chapter, requiring “nearly equal quantities of labour [in production] … comes to the same thing [as] the price [or cost] of nearly equal quantities” (WN I.xi.e.28, italics added). That is, on the tacit assumption that money wage rates are given, equal costs of labour (the total wage bill) incurred in production are taken as an indication of equal quantities of labour expended, just as differences ###/i
Our response is again that, while it may be indisputable that these differences in transport costs, reflect differences in labour expended, this nevertheless cannot be cited as a “case of labour-embodied thinking” - unless it were understood that no elements of profit or rent entered, along with wages, into the costs of the commodities transported.
It “cannot be cited as a ‘case of labour-embodied thinking’” in Grieve’s conceptual universe, but that is all.
Returning to the “real price” allusion in WN I.xi.g.28 (quoted above), Smith is claiming that produce requiring a greater or lesser quantity and (therefore) cost of labour in its marketing will have a greater or lesser “real price”, the latter obtained by dividing the nominal price of produce by a nominal wage rate. There are many similar passages in The Wealth of Nations[5], where differences and changes in “real price” (or “real value”, or sometimes “real cost”) are associated with differences and changes in quantities of labour expended and, as a linguistic variation on that theme, with differences and changes in the cost of quantities of labour, as with WN I.xi.g.28. Taking the case of fish:
it will generally be impossible to supply the great and extended market without employing a quantity of labour greater than in proportion to what had been requisite for supplying the narrow and confined one. A market which, from requiring only one thousand, comes to require annually ten thousand ton of fish, can seldom be employed without employing more than ten times the quantity of labour which had before been sufficient to supply it. … The real price of this commodity, therefore, naturally rises in the progress of improvement. (WN I.xi.m.15, italics added)
Evidently, an intertemporal rise in the “real price” of fish is the consequence of a rise in the quantity of labour expended in the supply of fish. Turning to manufacturing:
It is the natural effect of improvement ... to diminish gradually the real price of almost all manufactures. That of the manufacturing workmanship diminishes, perhaps, in all of them without exception. In consequence of better machinery, of greater dexterity, and of a more proper division and distribution of work… a much smaller quantity of labour becomes requisite for executing any particular piece of work (WN I.xi.o.1, italics added).
Here, the lower “real price” of manufactures is taken as the effect of a falling quantity of labour expended on their production.
As to “cost of labour” formulations, in one case Smith poses the question of why the real price of cloth should have been higher in ancient times. He answers:
It cost a greater quantity of labour to bring the goods to market. When they were brought thither, therefore, they must have purchased or exchanged for the price of a greater quantity. (WN I.xi.o.13, italics added)
Similarly, in the case of “different sorts of rude produce”, the rise over time in their “nominal or money-price” is the reflection “of a rise in their real price”, itself the effect of a change in their conditions of production:
As it costs a greater quantity of labour and subsistence to bring them to market, so when they are brought thither, they represent or are equivalent to a greater quantity.” (I.xi.l.13, italics added)
References to “costs” notwithstanding, in these examples Smith is positing a causal link between changes and differences in quantities of labour expended in production and resulting changes and differences in “real prices/values/costs”, the latter calculated by his “real measure of exchangeable value”. However, while it may be unobjectionable to claim a causal link between changes/differences in quantities of labour and “real” magnitudes ceteris paribus, it does not follow that changes and differences in “real prices/values” are determined exclusively by changes/differences in labour quantities. It was Smith’s own recognition of this fact, and his (failed) attempt to resolve it -his forlorn effort to cement, as it were, the association between “real” magnitudes and quantities of labour expended- that I took, and take, as further evidence of his attachment to “labour theory reasoning”.
The problems surface in the “Digression concerning the Variations in the Value of Silver during the Course of the Four last Centuries”, in which Smith was at pains to refute the opinion “of the greater part of those who have written upon the prices of commodities in antient times, that … the value of silver was continually diminishing” (WN I.xi.e.15). As the argument proceeds, it becomes clear that Smith’s focus is, more precisely, on the historical course of the real value or price of silver. But how are these real values (or prices) ascertained? Ideally, as Smith reminds us, they should be calculated in terms of labour commanded (WN I.xi.e.26). However, that would require knowledge of the historical course of nominal (silver) wage rates, and as Smith had forewarned earlier in the book, “the current prices of labour [nominal wages] at distant times and places can scarce ever be known with any degree of exactness” (WN I.v.22). That consideration had prompted Smith to propose “corn commanded” as the “nearest approximation” to “labour commanded” for the purpose of long-period comparisons, on the grounds that historical corn prices “are in general better known”, and that corn is the principal subsistence commodity. It therefore comes as no surprise that Smith should declare his use of the “corn commanded” proxy in the “Digression”, where long-period calculations of real price/value (of silver) are at issue:
Corn ... is, in all the different states of wealth and improvement, a more accurate measure of value than any other commodity or sett of commodities. In all those different stages, therefore, we can judge better of the real value of silver, by comparing it with corn, than by comparing it with any other commodity, or sett of commodities. (WN I.xi.e.28).
What is surprising is that he should introduce a completely new consideration in favor of corn’s suitability as a proxy “real measure”:
In every different stage of improvement, besides, the raising of equal quantities of corn in the same soil and climate, will, at an average, require nearly equal quantities of labour; or what comes to the same thing, the price of nearly equal quantities (WN I.xi.e.28, italics added).
Consideration of the quantity of labour required to raise corn is irrelevant to corn’s suitability as a “proxy”, which should depend solely on corn’s significance as the principal wage good. So, why introduce it as a relevant consideration in the “Digression”?
There is no point in looking to Grieve for an answer: once again, he has nothing to say on the matter, understandably so. For, if the purpose of the “real measure” was confined to ascertaining “labour commanded” either directly (using wage rates) or indirectly (using corn prices), as Grieve doggedly maintains, no purpose is served by drawing attention to the “nearly equal quantities of labour” expended on its cultivation in “every different stage of improvement”. But if, as I have suggested, Smith was attempting to restrict intertemporal changes in “real value” (of silver) to those caused by changes in the quantity of labour expended on its production, the newly introduced consideration makes perfect sense.
Unfortunately for Smith, the linkage between quantity of labour and real value/price holds only if the “nearly equal quantities of labour” required in its cultivation are uniquely defined at any point in time[6], and if it is possible to abstract from “transitory and occasional” events such as those that may result from sudden changes in supply (WN I.xi.g.17), yielding changes in exchangeable value unrelated to a commodity’s conditions of production.
Violation of the uniquely defined “nearly equal quantities of labour” condition occurs in the following case:
In great towns corn is always dearer than in remote parts of the country. This, however, is the effect, not of the real cheapness of silver, but of the real dearness of corn. It does not cost less labour to bring silver to the great town than to the remote parts of the country; but it costs a great deal more to bring corn. (WN I.xi.e.37)
For Smith, “real cheapness/dearness” should signal the lesser or greater quantity of labour required in production and marketing, as reflected in the cost of labour. But, if corn is the real measure, it must follow that silver is “really cheaper” in towns, while the “real value” of corn is, necessarily, the same (one unit of corn commands one unit in all locations). It was precisely such results that Smith may have hoped to avoid by invoking the “nearly equal quantities of labour” consideration. But now, because of the non-uniqueness of the labour required to produce corn[7], he was forced to identify “real dearness/cheapness” directly with labour expended. The corn measure had failed him.
Grieve sees matters differently: “we reckon that Smith is (appropriately) employing labour-commanded [rather than corn-commanded]”; moreover, “in terms of Smith’s labour-commanded standard, far from being lower, as supposed by Peach, the real value of silver (labour commanded) is actually not much altered.” But suppose that Smith had been using “labour-commanded”, even though this would be in the same chapter section in which he had flagged his use of “corn commanded” and had given no explicit indication that he was switching back to “labour commanded”. In view of corn’s status as the principal subsistence commodity, its higher money price in towns might be expected to result in a correspondingly higher money wage-rate, in which case its “real price” is not greater than its “real price” in the country. The outcome would be otherwise if real wages in the town are, for some unknown reason, lower than they are elsewhere, so that the higher corn price is not reflected in a proportionally higher money wage. In that case, however, the comparative “real dearness” of corn in towns would have nothing to do with it costing “a great deal more” labour to transport corn, which is what Smith thinks it should reflect, just as an approximate equality between the real price of silver in town and country would not be an indication of the cost/quantity of the labour required to circulate silver, which was Smith’s preferred explanation; rather, the results follow from a wholly arbitrary assumption about differences in the real wage between town and country, for which there is not a shred of evidence.
Similar issues arise, for Smith and for Grieve, at a later point in the “Digression”:
It does not cost less labour to bring silver to Amsterdam than to Dantzick; but it costs a great deal more to bring corn. The real cost of silver must be nearly the same in both places, but that of corn must be very different. (WN I.xi.e.38)
Once again, it was evidently Smith’s wish that “real cost” should reflect the “cost of labour” (as a function of labour quantity) incurred in bringing a commodity to market. However, because the cost/quantity of labour is different for corn in the two places, the real cost of silver, measured by corn-commanded, would also be different according to Smith’s reasoning (lower in Amsterdam, higher in Dantzick), while the real value of corn would be the same. For similar reasons as before -interspatial differences in the labour required to cultivate and market corn- the corn-commanded standard would yield the “wrong” results, leaving Smith with no option but to assert the desired association between “real cost” and the cost/quantity of labour expended in “production”.
Grieve will not have it: “What Peach does not appreciate, is that, in all probability, Smith was all the time (and appropriately) employing labour-commanded.” The qualification “in all probability” is significant, alerting us to the fact that explicit evidence for Smith’s use of “labour commanded” is wanting. However, suppose that Smith was using “labour commanded”. In that case, “the real cost of silver must be the same in both places” only if the money wage is the same, while the “real cost” of corn “must be very different” only if the price of corn is “very different”, which again implies “very different” magnitudes for the real wage in the two places (it must be lower in Amsterdam). As with the town/country comparison, however, there is no evidence whatever that Smith was basing his position on some completely arbitrary and unstated assumption about differences in real wages. But he is clear in his desire to associate “real cost” with the cost/quantity of “labour embodied”.
Finally, consider the following, also from the “Digression”:
The high price of corn during these ten or twelve years past, indeed, has occasioned the suspicion that the real value of silver still continues to fall in the European market. This high price of corn, however, seems evidently to have been the effect of the extraordinary unfavourableness of the seasons, and ought therefore to be regarded, not as a permanent, but as a transitory and occasional event. (WN I.xi.g.17)
Of course, the “real value of silver” does fall if the silver price of corn rises and the “real measure” is corn, because a given quantity of silver commands, or purchases, less corn. The fall is not so much a “suspicion” as a fact. But the fall is unrelated to a change in the conditions of producing silver; hence, I suggest, Smith’s rather feeble efforts to detract from its significance (“a transitory and occasional event”). His problem, yet again, was that corn had failed in its role as a measure of “labour embodied”, this time because of an “extraordinary” reduction in corn’s supply.
For Grieve, almost needless to say, “labour-commanded” was Smith’s “natural (and perfectly adequate) choice”, never mind that the case for using corn as the standard is re-stated in the paragraph immediately preceding the one in question, and that there is no mention of money wages in the relevant context. In this case, however, the use of “labour commanded” is at least logically consistent with Smith’s statement, to the extent that the same “suspicion” will arise if nominal wages rise with the price of corn. However, “labour commanded” would raise the possibility of additional “suspicions”, with the “real value” of silver subject to reductions from a rise in the real price of labour (real wages), and from “permanent” increases in the price of wage-goods (plural), quite independently of changes in its (silver’s) “real dearness”. I think we may be confident that Smith would not seek willingly to adopt a standard that could multiply the very “suspicions” he was trying to dispel.
In conclusion, I find no reason to withdraw or amend my contention that Smith did not restrict the relevance of a “labour theory of value” to an “early and rude state” of society. He continued to suggest, in various contexts, that exchangeable value, and changes in exchangeable value, could and should be related to quantities of labour expended in production (including marketing) and to changes in those quantities. And, significantly, he made a concerted (if unsuccessful) effort to link “real” magnitudes -value, price, dearness, cheapness- with quantities of expended labour, even to the point of asserting the connection in defiance of the results yielded by his own “measure”. Granted, one might quibble with his choice of wording, especially with his use of “cost/price of labour” as the determining element. But his articulation of unambiguous “quantity of labour” formulations, both with regard to exchangeable values and their changes, and his use of “linking” formulations, such as “the cost of a quantity of labour”, should remove any doubt that “cost/price of labour” was taken as anything other than a monetary indication, or manifestation, of labour quantity, at least in the passages on which my interpretation is based.
Finally, I am not so naïve as to imagine, or even hope, that anything I have said could make the slightest impression on Grieve. If, as he insists, the labour theory of value is only applicable to an “early and rude” state of society, that is the end of the matter, although I must register my astonishment at his assertion that his understanding corresponds to “what is generally understood by ‘the labour-embodied theory of value’”. But never mind. If others have objections to the name of the “theory”, let them substitute their own. The question will remain: did Smith, or did he not, restrict the determination of (changes in) exchangeable value by (changes in) quantities of expended labour to an “early and rude” state? My answer will also remain the same: he did not.
References
Grieve, Roy H. 2019. “On Terry Peach’s Unconvincing ‘Reconsideration’ of Adam Smith’s Theory of Value.” History of Political Economy 51 (4): 753-77.
Peach, Terry 2020. “Adam Smith’s Labor Theory of (Real) Value: The Case of a Misfiring Critique.” History of Political Economy 52 (1): 171-90.
Peach, Terry, 2009. “Adam Smith and the Labor Theory of (Real) Value: A Reconsideration.” History of Political Economy 41 (2): 383-406.
Smith, Adam [1776] 1996. An Inquiry into the Nature and Causes of the Wealth of Nations. Oxford: Oxford University Press.
[1] Correspondence may be addressed to [email protected] or [email protected].
[2] According to Grieve, my thesis is that Smith “was a convinced and consistent proponent of the labour theory of value”. The idea that there was, or is, any such thing as the labour theory of value is a fiction of Grieve’s own authorship; as for “consistency”, Grieve seems to forget my argument, revisited below, that Smith was not consistent, particularly when it came to dealing with “unwanted” results that came from using his own choice of “real measure”.
[3] As Smith had stated earlier: “A mine of any kind may be said to be either fertile or barren, according as the quantity of the mineral which can be brought from it by a certain quantity of labour, is greater or less than what can be brought by an equal quantity [of labour] from the greater part of other mines of the same kind.” (I.xi.c.11, italics added)
[4] Complications will arise with allowance for heterogeneous labour, paid at different wage-rates, but this does not seem to have troubled Smith unduly: see Peach (2009) 386-87.
[5] As with I.xi.o.13, I.xi.l.13, I.xi.m.15 and I.xi.o.1, all quoted below, passim. See also I.xi.m.15, and I.v.7 which, although not mentioning “real price/value” explicitly, is making the same point.
[6] This “problem” was to be resolved by Ricardo, who identified the greatest quantity of labour required at a point in time as the relevant quantity for determining the exchangeable value of the entire class of otherwise homogeneous produce.
[7] And, it might be added, his apparent inability to deal with this in the manner of Ricardo (see n.4, above).