Smith's Labor Theory Thought Experiment

There is a persistent conventional wisdom that in his (1776) An Inquiry into the Nature and Causes of the Wealth of Nations, (hereafter Wealth of Nations or WN), Adam Smith holds a labor theory of value (LTV) (see here). In what follows, I deny this claim and explain the kernel of truth in it. All my references are to the Glasgow edition by paragraph and page-number, but I have also linked passages to Liberty Fund’s Cannan edition.

The LTV is a class of views that hold a family resemblance to each other. I like the Wikipedia definition that according to the LTV the value (or price) of a good, commodity, or service is determined by the total amount of labor required to produce it. I call variants on this version of the view ‘the strict LTV.’ As the late Steven Horwitz notes, “sometimes the LTV is generalized a bit more to include other inputs, turning it into a “cost of production theory of value.” In this wider account the LTV claims that “value of outputs as being determined by the value of the inputs that went into producing them.” This I call ‘the broad LTV.’

In what follows, I am agnostic about the truth of the LTV. However, sometimes I show that Smith clearly does not hold the strict LTV by seeming to attribute the wider LTV to Wealth of Nations. In such cases I ask you to suspend your judgment and read to the end of this essay.

Here are some passages that, taken in isolation, seem to be clear evidence that Smith did hold the LTV. I list the most important ones and they are labeled to facilitate discussion:

I grant that IV especially looks like an endorsement of something very close to the strict LTV. And that the use of 'natural' here also looks like (to use a by now familiar charge) a naturalization of social relationships.

My general strategy with such passages is broadly the same. If you look at the wider context, Smith is (a) either describing/stipulating a particular measure or (b) using a thought experiment to introduce a much more complex claim than the LTV. And (c) Smith has a number of measures of value. This last point (c) suggests that (d) strictly speaking Smith does not have a single theory of value at all in the sense that is at stake when people attribute a LTV to him.

For, at the end of WN 1.4, Smith distinguishes between two notions of value: utility and purchasing power (or exchange). He is explicit (“three following chapters”) that the material in chapters WN 1.5-7 from which we have been quoting ([I-IV]) are the principles that govern exchange value only. In fact, he “very earnestly” entreats “both the patience and attention of the reader!” (WN 1.4.18, 46)

In so far as Smith holds a LTV at all, it is divorced from the use value of a commodity. I could end this essay here. But I also deny that Smith holds, all things considered, a LTV for exchange value.

In order to explain the principles of exchange value, Smith distinguishes among three measures of exchange value: (i) natural or ordinary prices, (ii) real prices, and (iii) nominal or market prices. He introduces natural prices at the end of WN 1.4 and discusses them at length in WN 1.7. Natural prices are (to use anachronism) the counterfactual prices that would obtain if the factors of production (rent, capital, wages, etc.?) were left free. I defend this claim here (and in chapter 12 of my book on Adam Smith).

The natural price is crucial to Smith’s scientific and political programs, and the natural price is at odds with a strict LTV. For Smith would deny that rent and capital reduce to labor embodied in some sense. This suggests support only for a broad LTV. But even a broad LTV is awkward, because according to Smith, the natural price is almost never instantiated in the real world (WN 1.7.24-32, 78-80). So, let’s leave it aside here.

Second, the nominal price of things can only be measured at a given time and place: “[A]t the same time and place, therefore, money is the exact measure of the real exchangeable value of all commodities. It is so, however, at the same time and place only” (WN 1.5.19, 55). Taken by themselves nominal prices are orthogonal to a commitment to a LTV. All this suggests is that in a monetized economy, the money price is the measure of exchange value at a given time and place.

It's worth noting that passage [IV], when quoted in full, explicitly involves a qualification that shows a circumstance prior to the invention of money and the development of property/rent: "IN that early and rude state of society which precedes both the accumulation of stock and the appropriation of land." (WN 1.6.1, 65) So, [IV] involves a situation where nominal prices are absent, and where capital and rent are absent! It's akin to an early Lockean state of nature. In Marxist terms, if Smith holds a strict LTV as a measure of exchange value, then it is only well before the stage of primitive accumulation not in a capitalist economy. This illustrates my claim that the LTV gets introduced only in the context of thought experiments that go on to make more complex claims.

Of course, if the measure of exchange value would track a LTV that is embodied in real prices (as suggested by [I-III]), then I would have to concede I am wrong. So, let’s take a closer look at the nature of a real price and its purpose in Smith’s hands.

According to Smith, the real price is constituted by the “necessaries and conveniencies of life which are given [in return] for” this exchange value, that is the quantity of money (WN 1.5.9, 51). So, the real price tracks something like the purchasing power of a basket of goods.1

In fact, it is notable that 'luxuries' are absent here. In the Wealth of Nations -- and this is common in 18th century political economy – Smith divides goods among necessaries, conveniences, and luxuries. So, Smith's treatment of real price suggests that it is a measure of the purchasing power of the working poor and not the rich (something Malthus and the Webbs also discerned). In fact, the way Smith uses real prices suggests that he thinks it is the measure of purchasing power of the working poor (who are the vast majority in his day and age) over time in a given society with same stable and given technology. I have defended this interpretation in chapter 8 of my book, and originally in chapter 10 of The Street Porter and the Philosopher: Conversations on Analytical Egalitarianism, edited by Sandra Peart and David Levy. (This volume also includes the first publication of some of the correspondence between John Rawls and James Buchanan, so well worth your money!)

That the real price is a measure of the purchasing power of the working poor over time in a given society is best illustrated by the very long Digression on Silver, and this has been documented also in a lovely new paper by Maria Pia Paganelli here.

It is true that one may well argue that for Smith the real price also is a measure of the labor “commanded” or “purchased” (I.v.19, 55). (Recall [I-III].) Why labor never varies in this sense is a complex question. Presumably it is connected to toil and what we may call anachronistically, ‘disutility.’ But you might claim that [III], which really sums up [I-II], strongly suggests a LTV.

But notice two things about [III]. First, the LTV is introduced for comparison purposes. It’s not a precise value or an exact measure, but only approximate (“estimated”);“the current prices of labour at distant times and places can scarce ever be known with any degrees of exactness” (WN 1.5.22, 56) Obviously, this doesn’t settle anything about the status of the LTV. But if it is so inexact why use it and hang (purportedly) so much on it?

Second, and more important, labor is only such a standard when it comes to commodities, it is not obviously pertinent to services and capital or rent. So, [I-III] cannot be taken to defend the strict LTV at all. So, it is far more natural to attribute to Smith a broad LTV:

Now, a proponent of the strict LTV might argue that profit and rent are just ways of rewarding different kinds of labor. But Smith explicitly rejects this: “The profits of stock, it may perhaps be thought, are only a different name for the wages of a particular sort of labour, the labour of inspection and direction. They are, however, altogether different, are regulated by quite different principles, and bear no proportion to the quantity, the hardship, or the ingenuity of this supposed labour of inspection and direction. They are regulated altogether by the value of the stock employed, and are greater or smaller in proportion to the extent of this stock.” (WN 1.6.6, 66; He goes on to say something similar about the price of rent in subsequent paragraphs.) This turns out to be an important passage for my overall argument.

It shows that Smith is committed to the claim that there are different kinds of labor. This makes a strict LTV untenable. When it comes to a broad LTV, the cost of the component parts of a real price, is itself a feature of the supply (and demand) of these parts. And Smith appeals to this fact, for example, to explain different land rents for same kind of land at different distances from cities. So, often what looks like a broad LTV is really a short-hand for a much more complex process of price formation.

That Smith thinks there are different kinds of labor where the real price of labor is wholly irrelevant to exchange prices is best seen in many “liberal and honourable professions” (services), where this standard functions very badly. (WN 1.10.b.22, 123) These function more like unfair lotteries.

This variation in the nature of labour is a feature not a bug in Smith’s approach. Smith only relies on something like a LTV at all when he is trying to figure out the purchasing power of the working or laboring poor over time, and he assumes these people barely consume services or luxury goods.

In fact, according to Smith the value we put on the same job necessarily varies from one society to another or even within a society depending on one’s social rank (e.g., WN I.x.b.2–3, 117-118). And, in fact, jobs that were once considerable hardship for most people – like hunting -- become a status enhancing luxury pass-time in later stages of development.

When he discusses real prices, Smith is primarily interested in comparing the purchasing power of the working poor over time. This is also clear from his main proxy measure for the LTV, namely the price of corn, which in every civilized country, the principal part of the subsistence of the labourer” (WN 1.xi.e.29, 206; see also p. 194. In Asia, Smith would use rice for the same purpose.)

So, yes, Smith uses something like a LTV for particular practical purposes. In such cases the LTV is imperfect and, in principle, dispensable if there are improved measures available. But Smith does not hold a LTV as such, and certainly not for the kind of economy like ours.