Almost every published refutation of the labour theory of value defeats one claim and then announces that it has defeated three. That is the single most useful thing to know before reading any critic or any defender, and it explains why an argument that has run for well over a century produces so little movement on either side. The theory is not one proposition. It does three separable jobs, and an objection that demolishes the first may leave the second and third untouched, while an objection that lands on the third may be irrelevant to a critic who only ever cared about the first.

Sorting the criticisms of the labour theory of value by what they actually target is therefore not a debating trick to protect Marx. It is the precondition for the argument being worth having at all. The six objections surveyed here range from one that is fatal to a claim Marx did not make, through two that land squarely and permanently, to one advanced by a philosopher who accepted the political conclusion the theory was built to support and argued that the theory was not needed to reach it. That last objection, from inside the tradition’s own sympathisers, is the most interesting thing anyone has said about the subject in decades, and it is almost entirely absent from the pages that answer this query.

What follows states each objection in the terms its own proponents would use, reconstructs the argument so its premises are visible, concedes what it gets right without defensiveness, gives the best reply rather than the most convenient one, and reaches a verdict of lands, partially lands, or misses. The strongest defence of the theory is also stated at full strength, together with the price that defence charges, because the honest position is that the theory can be saved on one reading and cannot be saved on another, and the reading that saves it gives up a good deal of what the theory was historically claimed to do.

The three jobs the theory is doing

The proposition under attack has to be stated before it can be attacked, and the statement has to be Marx’s rather than the textbook compression of it. The full treatment sits in the labour theory of value explained, which owns the definitional question; the compressed version needed here is that the theory carries three distinct workloads that ordinary summaries fuse into one.

The first job is the price-relation claim. The theory holds that the exchange ratios at which reproducible commodities trade are systematically governed by the socially necessary labour time required to reproduce them. Note the qualifiers, because every one of them does work: exchange ratios rather than individual prices, systematically governed rather than equal to, reproducible rather than all goods, and socially necessary rather than actually expended. The claim is about a regulating centre of gravity, not a formula that outputs a price tag.

The second job is the surplus claim. The theory holds that profit originates in the difference between the value of labour power, which is the cost of reproducing the worker, and the value that the exercise of labour power creates during the working day. This is the argument of Chapter Six of the first volume of Capital, where labour power is introduced as the peculiar commodity whose use value is the production of value, and it is the argument that makes exploitation a structural feature of a system of voluntary exchange rather than a matter of cheating. The full development belongs to the Marxist theory of exploitation.

The third job is the social-form claim. The theory holds that in a society where production is private and dispersed but the products are socially interdependent, the allocation of the society’s total labour among branches of production is accomplished behind the backs of the producers, through the movement of prices, and that value is the name for this form of social validation. This is the argument that runs from the first section of Chapter One into the fetishism section, and it is a claim about what kind of society this is rather than a claim about numbers.

These three are logically separable. A critic could accept the third and reject the first. A critic could accept the second and reject the third, which is roughly where the analytical Marxists ended up. A defender who wins on the third and thinks they have therefore won on the first has changed the subject without noticing.

What exactly does an objection to the labour theory of value have to hit?

An objection has to name which of the three jobs it targets. Objections to the price-relation claim say nothing about the social-form claim. Objections to the surplus claim are the only ones that touch the political conclusion. A refutation that does not specify its target has not been stated precisely enough to be assessed, let alone answered.

The target-check rule

The rule this article advances, and the one worth carrying into any argument on the subject, is the target-check rule: before assessing any objection to the labour theory of value, establish whether it targets the price-prediction claim, the source-of-surplus claim, or the social-form claim, because the theory has three jobs and almost every published refutation defeats one while assuming it has defeated all three.

Applied to the literature, the rule sorts the field immediately. Bohm-Bawerk’s celebrated critique targets the price-relation claim and the internal consistency of the derivation. The Sraffian redundancy charge also targets the price-relation claim, from a completely different direction, which is why treating the two as the same objection is an error. The land and antiques objection targets the price-relation claim under a misreading of its scope. The machines objection targets the surplus claim. Cohen’s argument targets the necessity of the whole apparatus for reaching the surplus claim’s political conclusion. And the value-form defence is a defence of the third claim that concedes ground on the first.

The rule has a second use, which is diagnostic rather than analytical. If a page presents a single objection and concludes that the labour theory of value has been refuted, without specifying which claim was refuted, the page has not done the work, and this applies with exactly equal force to a page that presents a single reply and concludes that the theory has been vindicated. The asymmetry that makes this subject unreadable online is not that one side argues badly; it is that both sides argue against a target the other side is not defending.

Objection one: the contradiction between the two volumes

The classical objection, and still the one most often invoked, is that the first volume of Capital and the third volume contradict each other. In the first volume, commodities exchange at their values, and the rate of surplus value is uniform across industries. In the third volume, commodities exchange at prices of production, which equalise the rate of profit across industries with different ratios of constant to variable capital, and which therefore systematically diverge from values. Since the whole apparatus of the first volume rests on exchange at values, and the third volume concedes that commodities do not exchange at values, the critic says the system has abandoned its own foundation while claiming to have completed it.

The critique was pressed most influentially by Eugen von Bohm-Bawerk in an essay of 1896 published after the third volume appeared, and translated into English as Karl Marx and the Close of His System. Bohm-Bawerk’s charge is not that Marx made an arithmetical slip. It is that the third volume solves the problem by abandoning the premise: if the deviations of price from value are systematic and are determined by the composition of capital, then the composition of capital and not labour time is doing the explanatory work, and the aggregate equalities Marx offers as compensation, total value equals total price and total surplus value equals total profit, are equalities of sums that no longer tell us anything about the individual exchange ratios the theory was introduced to explain.

The argument was sharpened technically by Ladislaus von Bortkiewicz in essays published in the first decade of the twentieth century, which showed that Marx’s own transformation procedure was incomplete because it transformed outputs into prices of production while leaving the inputs valued at their values, and that a consistent simultaneous solution generally cannot preserve both aggregate equalities at once. That is the formal core of what is now called the transformation problem, and it has its own dedicated treatment in the transformation problem explained, which owns that dispute in this series.

What the objection gets right, stated without defensiveness: the claim that individual commodities exchange at ratios proportional to their labour contents is not a claim the mature system makes, and any presentation of the theory that implies otherwise is presenting something Marx explicitly denied in the third volume. Textbook accounts do imply otherwise, constantly. A great deal of popular defence of the theory consists of defending a proposition its own author abandoned before the second volume of his own manuscript.

What the objection does not establish: that the system is incoherent. The best classical reply came from Rudolf Hilferding in an essay of 1904, and the shape of it survives all the technical developments since. Hilferding’s move is to deny that the value calculation was ever supposed to be a price calculation. The value analysis of the first volume establishes the origin and magnitude of the aggregate surplus; the price analysis of the third volume establishes how that aggregate surplus is redistributed among capitals through competition. Redistribution presupposes a magnitude to redistribute, and the magnitude has to be determined at the level of the total social capital before the question of its division among individual capitals can even be posed. On this reading, prices of production are not a correction of the value theory but its application at a lower level of abstraction.

The reply has force and it also has a cost. It concedes that the theory does not predict relative prices, and it stakes the theory’s claim to scientific standing on the aggregate proposition instead. Whether the aggregate proposition itself survives is precisely what the redundancy charge, discussed below, calls into question. The verdict on this objection is that it partially lands: it destroys the naive price-proportionality reading permanently, it does not by itself destroy the two-level reading, and it forces the defender onto ground where the next objection is waiting.

Did Bohm-Bawerk refute Marx?

Bohm-Bawerk refuted the claim that commodities exchange in ratios proportional to their labour contents, which the third volume of Capital had already denied. He did not refute the two-level reading on which value determines the aggregate surplus and competition redistributes it. Treating his essay as the end of the argument ignores both the reply and everything since.

Objection two: reducing skilled labour to simple labour

If the value of a commodity is determined by socially necessary labour time, and labour comes in kinds of wildly differing productivity and training, then the theory needs a conversion rate. An hour of a surgeon’s labour and an hour of an unskilled labourer’s are not the same quantum of value-creating substance, and the theory concedes as much: complicated labour counts as multiplied simple labour, so that a smaller quantity of the former equates to a larger quantity of the latter. The conversion is asserted in the second section of Chapter One and is never derived.

The objection is that the conversion rate can only be read off from the wage differential or the price differential, and that the moment it is, the theory has become circular. Values are supposed to explain prices; if the coefficient that converts skilled hours into simple hours is obtained from observed relative wages, then the price data is being used to construct the value magnitudes that are then presented as explaining the price data. Bohm-Bawerk made the point, and it has been restated in every generation since, most carefully in the technical literature on heterogeneous labour that developed in the nineteen seventies, where Samuel Bowles and Herbert Gintis argued that the reduction cannot be performed on independent grounds and that the theory’s aggregate labour magnitude is therefore not well defined.

The replies split into three, and they are of unequal quality. The weakest is the appeal to training time, which says that a skilled hour embodies the labour time expended in producing the skill, so that the conversion coefficient is itself a labour-time magnitude and no circularity arises. This is not empty, and it captures something real about apprenticeship and education as production processes with inputs. It fails as a general solution because the coefficients it generates do not match observed differentials even approximately, because a great deal of skill differential reflects scarcity and social closure rather than training cost, and because the training-time account cannot handle natural differences in capacity at all.

The second reply concedes that the reduction is empirical and denies that this is circular. On this account, the reduction is a measurement problem rather than a definitional one: the theory asserts that a reduction exists and is stable, and the market provides the evidence of its magnitude, in the same way that a physical theory may assert that a constant exists and then determine its value experimentally. The response is respectable but it changes the status of the claim, because a constant determined from the very data the theory explains cannot then be cited as an independent explanation of that data unless the theory generates other testable implications, which returns the argument to the empirical literature discussed in the evidence on the labour theory of value.

The third reply is the strongest and it is the one that fits the target-check rule. It says the objection lands against the price-relation claim and is nearly irrelevant to the social-form claim. If the theory’s core assertion is that the total labour of a society is allocated among branches through a process the producers do not control, then heterogeneity of labour complicates the measurement of that allocation without touching the claim that the allocation happens in this form. The critic is entitled to answer that a claim which survives by becoming unmeasurable has bought its survival expensively.

Verdict: lands against the price-relation claim, and the empirical work discussed later in this series treats the reduction as an unsolved problem rather than a solved one. It does not touch the surplus claim, since exploitation can be stated in terms of the labour of a given category of worker without requiring a universal conversion.

Objection three: goods that embody little or no labour

The objection everyone reaches for first is the one about the object that took no labour and commands a high price. Land embodies no labour at all and sells for a great deal. A painting by a dead artist took a certain number of hours to produce and sells for a sum bearing no relation to them. Wine that has been aged in a cellar for thirty years is worth more than the identical wine was on the day it was barrelled, and nobody laboured on it in the interval. A signed first edition, a meteorite, a plot with a view.

Stated as a refutation, this objection misses, and it misses for a reason stated in the text it claims to refute. The theory is explicitly restricted to commodities that are freely reproducible by labour. The restriction is not a later patch; it is in the argument from the beginning, because the mechanism by which labour time regulates exchange ratios is competitive reproduction. If a good sells above the ratio that its labour content would command, capital enters that branch, output rises, and the ratio is pushed back. Where reproduction is impossible, the mechanism has nothing to work with, and the theory says so.

Marx’s own treatment of the non-reproducible cases is more developed than critics usually acknowledge and is scattered across the third volume rather than gathered in one place. Land is handled through the theory of ground rent in the sixth part of the third volume, where the price of land is treated as capitalised rent, a claim on a future revenue stream discounted at the prevailing rate of interest, and therefore not a value magnitude at all. Unique goods with fixed supply are handled through monopoly price, which the third volume distinguishes explicitly from price of production, and which is determined by what buyers can be made to pay rather than by conditions of production. Objects that have a price but no value, a category the third chapter of the first volume raises directly and without embarrassment, include claims on future revenue, and the honour and conscience examples Marx offers are not throwaways but the point: price is a form that attaches to things that were never produced.

What the objection gets right, and it is not nothing: the restriction to reproducible commodities is a substantial narrowing of scope, and it is narrower in a modern economy than in a nineteenth-century one. A large share of what is traded in a financialised economy consists of claims, licences, brands, spectrum rights, intellectual property, and locational advantage, none of which is freely reproducible, and much of which is priced by capitalised revenue rather than by conditions of production. A theory that covers the reproducible sector covers less of the economy than it did, and a defender who deploys the reproducibility restriction should acknowledge that they have just conceded a large and growing territory rather than won an argument.

Verdict: misses as a refutation, lands as a scope limitation. The person who raises the painting example has not refuted anything; the person who raises the share of contemporary asset value that consists of capitalised claims has raised a real problem about the theory’s reach.

Objection four: machines, automation, and where value comes from

If labour is the sole source of value, and production becomes more capital-intensive, then value creation should be squeezed as machines displace workers, which seems to say that the most technologically advanced industries should be the least profitable, which they are not. The intuitive version is simpler still: a factory with robots produces goods, the robots did the work, so capital produced value and the theory is false.

The theory’s answer to the simple version is the constant and variable capital distinction, and it is a genuine answer rather than a dodge. Means of production transfer their existing value to the product as they are consumed, and they transfer no more than they contain; a machine that cost the equivalent of a thousand hours transfers a thousand hours to the products it helps make over its lifetime, no matter how much output it enables. Labour power is the only input whose consumption creates value in excess of the value required to reproduce it, which is why it is called variable. The robot argument therefore does not touch the theory; it restates the theory’s own premise as if it were an objection.

The harder version does touch it. If individual capitals adopt labour-displacing technique because it lowers their unit costs and wins them market share, and if value creation depends on living labour, then the system generates a tension between the rationality of the individual capital and the reproduction of the aggregate. That tension is exactly what the falling rate of profit argument formalises, and this series treats it separately in the crisis cluster rather than here. Two things are worth registering at this point. The first is that the tendency in the third volume is stated together with counteracting influences in the same part, so any argument that treats it as a prediction of steady decline is arguing against a text that does not exist. The second is that the empirical status of the tendency is a measurement dispute rather than a settled question, and the honest thing to say is that the profit-rate series produced by different measurement choices do not agree.

There is a sharper technical form of this objection worth naming, because it is often confused with the transformation problem and is a different argument. Nobuo Okishio demonstrated in nineteen sixty-one that a cost-reducing technical change adopted by a profit-maximising capital at prevailing prices, with the real wage held constant, cannot lower the general rate of profit and will normally raise it. If sound, this severs the mechanical link between mechanisation and profit-rate decline. The temporal single-system reply is that the result depends on simultaneous valuation of inputs and outputs, and that once inputs are valued at the prices at which they were actually bought, the conclusion does not follow. Whether the temporal reading is a reconstruction of Marx or a different theory that reaches Marx’s conclusions is itself contested, and readers should treat anyone who reports this as settled in either direction with suspicion.

Verdict: misses in its popular form, since the constant capital reply is complete. Raises a genuine and unresolved problem in its technical form, which belongs to the crisis literature rather than to the value theory as such.

Objection five: the redundancy charge

This is the objection that did the real damage in the twentieth century, and it is routinely confused with the Bohm-Bawerk critique although it is a different argument reaching a more dangerous conclusion. Bohm-Bawerk said the value theory was inconsistent with the price theory. The redundancy charge says the value theory is consistent and unnecessary.

The argument descends from Piero Sraffa’s book of nineteen sixty, Production of Commodities by Means of Commodities, which showed that given the technical conditions of production and one distributional variable, either the wage or the rate of profit, the prices of production and the other distributional variable are simultaneously determined. Labour values do not appear anywhere in the determination. They can be computed from the same data, but computing them adds nothing, since the prices and the profit rate were already determined without them. Ian Steedman pressed the conclusion against Marxist economics directly in a book of nineteen seventy-seven, Marx after Sraffa, arguing that value magnitudes are at best a redundant detour and at worst positively misleading, because in cases involving joint production values can be negative, which is a meaningless result for a magnitude supposed to represent embodied labour.

The charge is dangerous because it does not need to show that Marx made an error. It concedes the arithmetic and asks what work the value concept is doing. If the answer is none, then the theory is not false but idle, and idle theories are abandoned rather than refuted.

The replies are three, and they correspond to the three modern defences described below. The first reply, from the temporal single-system interpretation, denies the simultaneity assumption on which the Sraffian determination rests: if inputs are valued at the prices ruling when they were purchased rather than at the prices that will rule when outputs are sold, the system is not simultaneous, the value magnitudes are not derivable from the price system, and the redundancy result does not go through. The second reply, from the New Interpretation associated with work by Gerard Dumenil and Duncan Foley in the early nineteen eighties, redefines the value of labour power as the labour commanded by the money wage rather than the labour embodied in a fixed consumption bundle, which preserves the aggregate identity between surplus value and profit without requiring the individual-commodity claim, and thereby makes the value category do a job the price system cannot do on its own. The third reply, from value-form theory, denies the premise that value magnitudes were supposed to be inputs to a price calculation at all, and holds that the category answers a different question about the form social labour takes.

What the objection gets right: if the theory’s job is to determine prices and the profit rate, the Sraffian system does that job with fewer assumptions and no detour through unobservable magnitudes, and no reply has overturned that result on its own terms. The Sraffian challenge to Marxist economics has its own treatment in this series, and readers who want the full technical exchange should go to Sraffa and Marxist economics rather than expecting it here.

Verdict: lands, and lands hardest. Of the six objections it is the one with the fewest good answers, and the three replies to it are not variants of one another but genuinely competing research programmes, which is itself a sign that the problem is unresolved.

Is the labour theory of value redundant for explaining prices?

For the narrow task of determining relative prices and the rate of profit from technical conditions and one distributional variable, yes: the Sraffian system does it without value magnitudes. Defenders reply that this narrow task was never the theory’s job and that the aggregate and social-form claims survive intact.

Objection six: the internal challenge

The most interesting objection to the labour theory of value was made by someone who wanted the exploitation conclusion to be true. G. A. Cohen, writing in nineteen seventy-nine, argued that the labour theory of value is not merely insufficient to establish the exploitation of the worker but is actually irrelevant to it, and that Marxists who tie the two together weaken a strong claim by attaching it to a weak one.

The argument runs as follows. The labour theory of value says that the value of a commodity is determined by the socially necessary labour time required to produce it. The exploitation claim says that the worker produces a product, does not receive the whole of it, and that the capitalist appropriates the remainder without having produced it. Cohen’s observation is that the second claim is about the product, a physical and social object, while the first is about value, an exchange magnitude, and that the second does not depend on the first. What grounds the charge of exploitation is that labour produces the things, and that some of the things are taken by people who did not produce them. Whether the exchange ratios among those things are governed by labour time is a separate question with no bearing on the transfer.

Cohen goes further and argues that the labour theory of value, taken strictly, is embarrassed by the exploitation claim rather than supporting it. Under the theory, the value of the product is determined by the labour time socially necessary at the moment of sale, not by the labour actually performed by the workers who made it. So a worker who takes longer than socially necessary produces less value than the labour they expended, and a worker in an unusually productive firm produces more. The magnitude the worker is said to have created and been deprived of therefore is not the magnitude of their own labour, which is what the moral argument wants it to be.

This is the objection that best illustrates the target-check rule, because it targets neither the price-relation claim nor the social-form claim. It targets the inference from the theory to the political conclusion, and it does so in order to make the political conclusion more secure by detaching it. Whether a Marxist should welcome the offer is a live question in the tradition. John Roemer built a whole research programme on the affirmative answer, formalising exploitation and class in terms of unequal ownership of productive assets rather than in terms of labour value, and demonstrating along the way that a formally identical exploitation result can be derived using any basic commodity as numeraire rather than labour, which is a further embarrassment for anyone who thinks the labour theory of value is doing the moral work.

Against this, defenders reply that detaching the exploitation claim from the value theory costs more than Cohen allows. The theory does not merely say that some people take things others made, which is true of every class society and needs no special apparatus. It says that in this particular society the taking occurs through exchanges that are, at every point, exchanges of equivalents, so that no cheating, no coercion at the point of sale, and no violation of the norms of the market is required. Removing the value analysis removes the explanation of how appropriation is compatible with equal exchange, which was the specific puzzle Chapter Six was written to solve. Cohen’s version can say that appropriation occurs; it has more difficulty saying why it does not show up as unequal exchange.

Verdict: partially lands, and it is the objection every serious defender should engage with first, because a critic who has read Cohen cannot be answered with the standard replies at all.

The six objections ledger

The table below is the article’s findable artifact, and it is designed to be usable in an argument. The verdict column reports where the objection leaves the theory, not whether the objector was a good economist.

Objection	Who pressed it	What it targets	Best reply	Verdict
Value and price contradict between the volumes	Bohm-Bawerk, sharpened technically by Bortkiewicz	The price-relation claim and the internal consistency of the derivation	Hilferding’s two-level reading: value determines the aggregate surplus, competition redistributes it	Partially lands. Destroys price proportionality permanently; leaves the aggregate claim standing
Skilled labour cannot be reduced to simple labour without circularity	Bohm-Bawerk, formalised later in the heterogeneous-labour literature	The price-relation claim and the measurability of the aggregate	The reduction is empirical rather than definitional, and the social-form claim does not need it	Lands against measurement. Does not touch the surplus claim
Land, antiques, and unique goods have prices without labour	Raised in every generation, rarely by economists	The scope of the price-relation claim	The theory is restricted to freely reproducible commodities; rent and monopoly price handle the rest	Misses as refutation. Lands as a scope limitation, and a growing one
Machines produce value, so labour is not the sole source	Popular form widespread; technical form from Okishio	The surplus claim	Constant capital transfers value and does not create it; the technical form belongs to the crisis debate	Misses in its popular form. Technical form remains open
Value magnitudes are redundant for determining prices and profit	Sraffa’s framework, pressed by Steedman	The price-relation claim and the theory’s necessity	Temporal valuation, the New Interpretation, or the value-form denial of the premise	Lands hardest. No reply commands consensus
Exploitation does not need the value theory at all	G. A. Cohen, extended by Roemer	The inference from value theory to the political conclusion	The value analysis explains appropriation through equal exchange, which the alternative cannot	Partially lands. The most serious challenge from a sympathetic source

The theorem that was supposed to settle it, and the theorem that unsettled it

Two formal results from the nineteen seventies and early nineteen eighties deserve more attention than they get, because between them they define the current state of the argument about whether the value theory is needed for the exploitation conclusion.

The first is the result Michio Morishima named the Fundamental Marxian Theorem in a book of nineteen seventy-three. Stated informally, it establishes that in a suitably specified linear production model, the rate of profit is positive if and only if the rate of exploitation is positive, which is to say that profit exists if and only if workers perform more labour than is required to produce the goods their wages command. This looked, for a decade, like the vindication of the theory’s central proposition in a form no marginalist could dispute, because it was derived within a formal framework the discipline recognised. Note carefully what it establishes and what it does not. It establishes an equivalence between two conditions. It does not establish that labour values determine prices, and Morishima himself was clear that the transformation of values into prices in Marx’s sense could not be sustained.

The second result cut the ground from under the first. John Roemer demonstrated that the same formal structure produces a parallel theorem for any basic commodity in the system. Take steel as the numeraire and define steel values, steel embodied in production; the rate of profit is positive if and only if the steel rate of exploitation is positive. The result holds for corn, for energy, for any input that enters directly or indirectly into everything. This is sometimes called the generalised commodity exploitation theorem, and its force is devastating in a very specific way: it shows that the formal equivalence Morishima proved is a property of the production structure rather than a discovery about labour. If steel is exploited in exactly the sense the theorem attributes to labour, then the theorem cannot be what makes labour special.

The defender’s reply is not weak and it is not formal. It is that labour is not one input among others because labour power is bought from people who must sell it, who then work under the direction of the purchaser for a period they do not control, and who are the only input that can resist, slow down, organise, or be disciplined. Steel does not have a working day. The asymmetry that grounds the exploitation claim is a social asymmetry between persons and things, and a formal theorem that treats labour symmetrically with steel has abstracted from precisely the thing at issue. This reply concedes a great deal, and what it concedes is that the moral and political content of the exploitation claim comes from the social relation rather than from the value arithmetic, which is Cohen’s conclusion arrived at from the other direction.

The two theorems together are the cleanest illustration in the whole literature of the target-check rule at work. Morishima’s result was taken as a defence of the value theory when it was in fact a result about the source of surplus. Roemer’s result was taken as a refutation of exploitation when it was in fact a demonstration that the formal apparatus does not pick out labour. Neither theorem touches the social-form claim at all.

Critics from inside the tradition

The most damaging criticism of the labour theory of value has consistently come from people who wanted Marx’s conclusions to hold. This is unusual in the history of ideas and it should change how the dispute is read.

Joan Robinson argued in an essay of nineteen forty-two that the value theory was a metaphysical apparatus that contributed nothing to Marx’s substantive achievements, and that his analysis of accumulation, crisis, technical change, and the reserve army stood without it. Her position was that the labour theory of value functions as a rhetorical rather than an analytical device, and that Marxists defending it were expending effort on the one part of the system a hostile economist could attack successfully while neglecting the parts that were strong. She was writing as an admirer, which is why the argument was hard to dismiss and why it has never been fully answered.

Isaak Illich Rubin’s essays of the nineteen twenties attacked the orthodoxy from the opposite side, arguing that the prevailing Soviet reading had reduced value to a technical quantity of embodied labour and thereby lost the whole point, which was the analysis of the social form that labour takes under commodity production. Rubin’s work is the source of the value-form tradition and it constitutes a criticism of most twentieth-century Marxist economics rather than of Marx.

The Uno school in Japan, developing from the nineteen fifties, restructured the entire presentation by separating the pure theory of a capitalist economy from stage theory and from the analysis of actual history, on the grounds that the standard treatment illegitimately mixed levels of abstraction. Whatever one thinks of the solution, the diagnosis identifies a real and recurring failure in Marxist argument: a claim established at one level of abstraction is deployed as though it applied at another.

Steedman’s redundancy argument, already discussed, belongs to this category too, and so does the analytical Marxist programme. The pattern is consistent enough to be worth stating as a finding. The external critics attack the theory’s consistency and its price predictions, which are the claims the tradition can most easily give up. The internal critics attack its necessity, its level of abstraction, and its relation to the political conclusion, which are the claims that matter. A reader who wants the strongest available case against the labour theory of value should read the sympathetic critics rather than the hostile ones, which is an unusual piece of research advice and a reliable one here.

Why the argument does not converge

Arguments in a discipline usually end in one of three ways: a decisive empirical result, a demonstration of inconsistency, or the abandonment of the question. None of these has happened here, and the reasons are structural rather than accidental.

The empirical route is blocked by a measurement problem rather than a shortage of data. Labour values are not directly observable and have to be computed from input-output tables under assumptions about the reduction of skilled labour, the treatment of unproductive sectors, the handling of fixed capital, and the choice of normalisation. Different defensible choices produce different magnitudes, which means an empirical test tests a conjunction of the theory and the measurement scheme, and a failure can always be attributed to the scheme. This is not a special vice of Marxian economics; it afflicts the measurement of capital in mainstream growth accounting too, and the Cambridge controversies were about exactly that. It does mean the empirical route is slower than either side hopes.

The consistency route is blocked because consistency turns out to be a function of the valuation convention. Under simultaneous valuation, Marx’s procedure is incomplete and the two aggregate identities cannot both be preserved in general. Under temporal valuation, they can. The dispute is therefore not about arithmetic but about which convention represents the object of study, and no arithmetic can settle that.

The abandonment route is blocked by the political stakes, which is the least respectable reason and the most operative one. If Cohen is right, the political conclusion does not depend on the theory, and the dispute could be conducted as a technical question in the history of economic analysis. Very little of it is conducted that way, on either side, because the theory has become a marker of allegiance. That is a fact about the sociology of the argument rather than about its content, and a reader who notices it will read both camps more accurately.

The four arguments that are not objections

A large share of what circulates as criticism of the labour theory of value consists of four arguments that fail before they reach the theory. They are worth disposing of properly rather than ignoring, because they occupy most of the space in public discussion and because a defender who cannot dispose of them quickly will spend an entire argument on ground where nothing is at stake.

The first is the mud pie. Take a quantity of mud, spend eight hours shaping it into a pie nobody wants, and observe that it has no value despite the labour expended. The conclusion drawn is that labour cannot be the source of value because labour was expended and value did not appear. What the example actually tests is whether the theory says that any expenditure of effort creates value, and the theory says the opposite in the same paragraph where it introduces the concept. Labour creates value only insofar as it is socially necessary, which requires both that the labour be performed under prevailing conditions of productivity and that it satisfy a social need, since a product that meets no need is not a commodity at all and its labour is not validated. The mud pie is a demonstration of the socially necessary clause, not a counter-example to it. It is worth adding that the example was not invented by hostile critics; the failure of privately expended labour to receive social validation is one of the things the theory is about, and the fetishism argument in the fourth section of Chapter One is largely concerned with the fact that producers learn whether their labour counted only after the fact, in the market.

The second is the assertion that value is subjective. This is offered as though it settled something, when it is a statement of the rival framework rather than an argument against this one. If it means that individuals differ in how much they want things, the labour theory does not deny it and has a category, use value, for exactly that heterogeneity, holding that use value is a precondition of exchange without being its measure. If it means that relative prices are determined by aggregated marginal valuations under constraint, that is a substantive theory with its own structure and its own difficulties, and asserting it is not the same as refuting the alternative. Since the two frameworks answer different questions, the comparison requires care rather than assertion, which is why it is handled separately in this cluster.

The third is the observation that firms set prices by adding a markup to cost, that sellers can charge what they like, or that a monopolist prices above the competitive level, from which the conclusion is drawn that prices are administered rather than determined by labour. The theory is not committed to denying any of this. It is committed to the claim that the costs a firm marks up are themselves the prices of inputs produced elsewhere, that the markup is constrained by competition and by the general rate of profit, and that a firm which sets prices wherever it wishes will discover the constraint through loss of market share. The proposition is about what happens to the firm that ignores the regulating centre, not about whether anyone consults it. A price that is administered is still a price that has to be validated by a sale.

The fourth is the charge that the theory is a tautology or is unfalsifiable, usually supported by the observation that any deviation of price from value can be attributed to some other factor. The charge lands against certain defences of the theory and not against the theory itself, and the distinction matters. A defender who explains every observed divergence by invoking a fresh countervailing factor has indeed made the position unfalsifiable, and this happens. But the theory as stated has implications that could fail: it implies that industries with higher ratios of constant to variable capital should show price-value deviations in a determinate direction, it implies that the aggregate identities hold under some consistent valuation scheme, and it implies that labour requirements should track relative prices more closely than arbitrary alternative input measures do. Each of those has been tested, with results that are disputed but not vacuous.

What would actually falsify the labour theory of value?

Three results would do it. A demonstration that the aggregate identities fail under every consistent valuation scheme. A demonstration that arbitrary non-labour input measures track relative prices as well as labour requirements do, once industry size is normalised out. And a coherent account of the origin of aggregate profit that does not appeal to unpaid labour in any form.

The point of specifying this is not to protect the theory. It is that a critic who has not specified what would count as disconfirmation is in the same epistemic position as the defender who explains away every anomaly, and both positions are exempt from evidence. The empirical programme that has attempted the second of those three tests is the most serious thing that has happened to this argument in a generation, and it is the subject of a dedicated article in this cluster.

How the six objections interact

The objections are usually presented as a list, which conceals the more important fact that they are not independent. Two of them reinforce each other, two of them pull in opposite directions, and one of them changes what the others mean.

The transformation charge and the redundancy charge reinforce each other in a specific way. The standard reply to the transformation charge is Hilferding’s two-level move: value determines the aggregate, competition redistributes it. That reply relocates the theory’s content to the aggregate identities. The redundancy charge then arrives at exactly that location and asks what the aggregate identities are for, given that the profit rate has already been determined without them. A defender who uses the first reply has walked into the second objection, which is why these two are the pair that does the real damage and why the modern defences all have to address both at once. The New Interpretation’s move is intelligible only as a response to this pincer: it secures one aggregate identity by construction, accepting that this makes the identity definitional, in order to keep the rate of exploitation measurable.

The skilled-labour charge and the empirical programme pull against each other. The empirical work has to compute labour requirements from input-output data, and it typically does so in employment or wage-bill terms, which means it either ignores the reduction problem or resolves it by using wages as the conversion, which is the circularity the second objection identified. A critic can therefore run the two objections together: the empirical results depend on a reduction the theory cannot justify. Defenders reply that results computed with different reduction assumptions do not differ dramatically, which is a claim about robustness that should be checked rather than assumed.

Cohen’s objection changes what all the others mean. If the exploitation claim does not depend on the value theory, then the political stakes of the transformation problem, the redundancy charge, and the skilled-labour reduction fall away almost entirely, and the argument becomes a technical dispute in the history and methodology of economics rather than a battle over whether capitalism is unjust. Very few participants on either side write as though this were true, which tells us something about why the dispute is conducted the way it is. A critic who has established the redundancy of value magnitudes for price determination has not established anything about exploitation, and a defender who has rescued the aggregate identities has not established anything about exploitation either.

The land and machines objections, finally, are the two that a serious critic should abandon, and the fact that they dominate popular discussion while the redundancy charge is almost unknown outside the specialist literature is the clearest single indication that public argument on this subject is not tracking the strongest available positions on either side.

What actually happened: displacement, not refutation

The popular history says that Bohm-Bawerk refuted Marx and the economics profession moved on. Two separate events have been compressed into one, and the compression matters because it makes the theory’s abandonment look like the outcome of a decisive argument rather than what it was.

The marginal framework arrived in the eighteen seventies, in three near-simultaneous publications by William Stanley Jevons and Carl Menger and, three years later, Leon Walras. The first volume of Capital had appeared in eighteen sixty-seven, before any of them, which is why Marx never replied to marginal utility theory and could not have. The reorientation of the discipline around marginal analysis was under way for two decades before Bohm-Bawerk’s essay appeared, and it was driven by the analytical power of the new framework for problems the classical surplus approach handled awkwardly, particularly the allocation of scarce resources among competing uses and the determination of demand.

The classical approach was not shown to be internally inconsistent and then discarded. It was displaced because the profession’s central questions changed. That distinction is not a rescue of the labour theory of value; it is a correction to the history, and it cuts both ways, since a framework that loses the profession’s interest for eighty years has lost something real regardless of whether a knockdown argument was ever produced. It also matters that the classical surplus approach did not die: it was revived in a non-Marxist form by Sraffa in nineteen sixty, which is why the modern alternative to marginalism in price theory is not Marx but the classical revival. The comparison between the two frameworks, and the misreadings that dominate it, are handled in Marx versus the marginalist theory of value.

The three modern defences and what each one answers

Anyone assessing the current state of the argument needs to know that there is no single Marxist position on value theory, and that the three main defences are not complementary emphases but rival programmes that disagree with one another about what Marx said and what the theory is for. Each answers a different objection well and none answers all of them.

The temporal single-system interpretation holds that inputs and outputs must be valued at the prices ruling at the moments of purchase and sale respectively, rather than simultaneously, and that value and price form a single system rather than two parallel accounting schemes. Its principal claim is that Marx’s own transformation procedure is internally consistent once read temporally, and that both aggregate equalities hold. It answers the internal-inconsistency charge and the Okishio result. Its cost is that it makes the theory’s results depend on a reading of the text that other Marxist economists reject, and its critics argue that it secures consistency by making the value magnitudes depend on prices in a way that concedes the substantive point.

The New Interpretation, developed in the early nineteen eighties, redefines two quantities: the value of money, taken as the ratio of total labour time to total money value added, and the value of labour power, taken as the labour commanded by the money wage. With these definitions the equality between aggregate surplus value and aggregate profit holds by construction, and the rate of exploitation becomes directly measurable from national accounts data, which is why the empirical literature has largely adopted it. It answers the transformation problem by narrowing what the theory claims. Its cost is that it abandons the individual-commodity claim entirely and makes one of the two aggregate equalities definitional rather than substantive.

Value-form theory, whose modern lineage runs from Isaak Illich Rubin’s essays of the nineteen twenties through the German and anglophone reconstructions that followed, denies that the theory is a quantitative price theory in the first place. Its subject is the form value takes, why the product of labour appears as an exchangeable thing, why social labour has to be validated through money, and what kind of social relation this is. It answers the redundancy charge by conceding the price-determination ground and relocating the theory’s content. Its cost is the largest of the three, because it gives up the empirical claims that made the theory testable, and its critics inside Marxism argue that it converts an economic theory into a philosophy of social form.

The honest summary is that the three defences cannot all be right, that each is a live research programme with serious people in it, and that the existence of three incompatible reconstructions of what the theory even claims is itself a fact a critic is entitled to make something of.

The strongest defence, stated at full strength

Because this article is a survey of objections, the steelman requirement reverses: the defence has to be stated at its strongest, in the terms its own proponents use, before any verdict.

The strongest defence is that the labour theory of value is not and never was an instrument for predicting individual prices, and that judging it as one is a category error produced by reading a nineteenth-century critique of political economy as though it were a twentieth-century price model. The theory’s subject is a society in which production is carried on privately, by independent producers who make no collective decision about what shall be produced or in what proportions, and yet whose products are in fact interdependent parts of a single social division of labour. Such a society faces a real problem: the total labour available has to be allocated among branches in something approximating the proportions that reproduction requires, and no one allocates it.

The theory’s claim is that this allocation is accomplished, imperfectly and after the fact, through the exchange of products, and that the magnitude regulating the process is the labour time socially necessary to reproduce each product. Value is the name for the social validation that private labour receives when its product exchanges. Money is not a convenience laid over barter but the necessary form in which that validation appears, which is the argument developed in Marx on money and the value form. Prices oscillate, deviate systematically under the pressure of competition and unequal capital compositions, and are pulled about by supply and demand, and none of this touches the claim, because the claim is about what the oscillation is oscillating around and, more fundamentally, about why a society organises its labour through the movement of prices at all.

On this reading, the objections lose most of their force. The value-price divergence is expected rather than embarrassing. The redundancy charge assumes the theory competes with the Sraffian system for the price-determination job and it does not. The land and antiques cases fall outside the theory’s stated scope. And no rival framework offers any account of the social form of labour allocation, because no rival framework treats that as a question: marginalism takes the existence of exchange, prices, and money as given and asks how a rational agent behaves within them.

That is the defence at full strength, and it is a serious position held by serious people. Its price should be stated with equal clarity. It gives up the claim that the theory explains relative prices, which is what the theory was taken to claim by Ricardo, by the socialist economists of the nineteenth century, by a great deal of twentieth-century Marxist economics, and by every introductory account of it. It converts a theory with testable implications into a theory about social form, which is more defensible and less useful. And it leaves the tradition with an awkward question: if the theory never claimed to explain prices, why did generations of Marxist economists spend their careers trying to show that it did?

What the exchange leaves standing

Taking the six objections together, the position that survives is narrower than either camp’s public account of it.

The price-relation claim in its strong form does not survive. Commodities do not exchange in ratios proportional to labour contents, the mature text says so, the transformation from values to prices of production cannot preserve both aggregate equalities under simultaneous valuation, and value magnitudes are not required to determine prices given technical conditions and a distributional variable. A defender who continues to assert strong price proportionality is defending a position their own tradition abandoned.

The price-relation claim in a weakened form is disputed rather than dead. The proposition that labour inputs track relative prices closely enough to be doing explanatory work has been tested empirically across many national input-output datasets, and the results are contested on statistical grounds that have nothing to do with the classical objections. That literature is worth reading before anyone announces a verdict, and it is treated in full in the evidence article in this cluster.

The surplus claim survives the six objections but is no longer uniquely supported by the value theory. Cohen’s argument and Roemer’s formalisation show that a coherent account of exploitation can be constructed without the labour theory of value, and even, in Roemer’s demonstration, with any basic commodity substituted for labour. The claim that this system generates unpaid surplus appropriated by owners of means of production does not stand or fall with the value theory, and both critics and defenders make an error when they assume that it does. That is the most consequential single result in this entire literature and it is almost never reported in popular treatments in either direction.

The social-form claim is untouched by five of the six objections and is the ground on which the most sophisticated modern defences stand. It is also the least falsifiable of the three, which is not by itself a criticism, but which does mean a reader should notice when a defence retreats to it.

What Marx inherited, and the difficulty he inherited with it

The labour theory of value did not begin with Marx and the objections did not begin with his critics. Adam Smith held that in a primitive state of society, before stock had accumulated and land had been appropriated, the quantity of labour required to acquire objects governed their exchange, and he then held that once profit and rent existed, price resolved into wages, profit, and rent added together. That is two theories, and Smith’s successors noticed. David Ricardo took the labour principle seriously as a general theory and spent the rest of his life on the difficulties it generated, revising the opening chapter of his Principles across three editions and leaving unfinished work on an invariable measure of value at his death.

Ricardo’s central difficulty is the ancestor of the transformation problem and is worth stating because it shows the problem is structural rather than a slip by Marx. If commodities exchange in proportion to labour embodied, and if capitals are advanced for different lengths of time or in different proportions between wages and equipment, then a rise in wages will change relative prices, because it affects capitals of different composition differently. Labour embodied has not changed, but the exchange ratio has. Ricardo could not eliminate this and he did not pretend to have eliminated it. He described the deviations as modifications of a rule rather than refutations of it, and he searched for a commodity of average composition whose price would be unaffected by distribution, precisely so that changes in relative prices could be attributed to their proper causes.

What Marx inherited, then, was a principle whose author already knew it could not deliver exact relative prices under capitalist conditions. What Marx added was an explanation of the deviation rather than an apology for it: the equalisation of the profit rate across capitals of unequal composition, which necessarily transfers surplus between branches and therefore necessarily produces prices that diverge from values in a determinate pattern. The third volume’s account is best understood as the completion of the classical project, an attempt to explain systematically what Ricardo had recorded as an anomaly.

Two conclusions follow for anyone assessing the criticisms. The first is that a critique which presents the value-price divergence as a discovery is presenting as news something the classical economists knew and worked on for half a century. The second is that the criticisms bite hardest against the strongest form of the classical project, which is the attempt to derive prices from a magnitude given prior to and independent of prices. That project has real difficulties, they were visible from the start, and no amount of textual exegesis makes them go away.

The nineteenth-century socialist writers who preceded Marx drew a different and simpler conclusion, and it is one Marx explicitly rejected. If labour creates value, they reasoned, the worker should receive the full value their labour creates, and the remedy is exchange at labour values through labour-time currency and cooperative exchange banks. Marx’s argument against this is one of the sharper things in his economic writing: the workers do receive the full value of what they sell, which is labour power rather than labour, and the difference between that value and the value labour creates is not a violation of exchange but a consequence of it. Anyone who has understood that argument can see why the theory cannot be turned directly into a fair-price scheme, and why the popular reading of it as a doctrine about deserved prices misses the point so completely.

A short history of the argument

The dispute has four phases, and knowing which phase a source belongs to prevents most confusion about what its author thought they were doing.

The first phase runs from the appearance of the third volume of Capital to the outbreak of the First World War. This is the classical exchange: Bohm-Bawerk’s essay of eighteen ninety-six, Hilferding’s reply of nineteen hundred and four, and Bortkiewicz’s technical papers in the years immediately following, which reformulated the transformation as a system of simultaneous equations and demonstrated the incompleteness of Marx’s own procedure. What is striking about this phase is that the technical result came from a sympathiser of neither camp and was largely ignored by both for forty years.

The second phase runs through the middle of the twentieth century and is dominated by the socialist calculation debate and by the professionalisation of economics around the marginalist framework, during which the value controversy was mostly conducted outside academic economics altogether. Paul Sweezy’s rediscovery and popularisation of Bortkiewicz’s solution in the nineteen forties brought the technical problem back into anglophone Marxist discussion, and the terms of the argument that most readers encounter today were set in this period.

The third phase begins with Sraffa’s book of nineteen sixty and runs through the nineteen seventies. This is when the argument changes character. The Cambridge capital controversies weakened the marginalist account of capital and distribution, which might have been expected to help the Marxian side, and instead the classical revival that emerged supplied the framework from which the redundancy charge was launched. Steedman’s book of nineteen seventy-seven is the pivot: it comes from the left, it is sympathetic to the political project, and it argues that the value apparatus should be abandoned as an obstacle to that project rather than a support for it. Marxist economics spent the following decade responding.

The fourth phase runs from the early nineteen eighties and is the one that has not closed. The New Interpretation appeared as a way of preserving measurable content while conceding the individual-commodity claim. The analytical Marxists, with Cohen’s argument and Roemer’s formal models, reconstructed exploitation without value. The temporal single-system writers argued that the entire ninety-year technical literature had misread Marx by imposing simultaneous valuation. Value-form theory relocated the whole subject. And the empirical programme using input-output data began producing results that neither the classical objections nor the Sraffian ones had anticipated, which generated a statistical dispute of its own.

How did the argument reach its present state?

Through four phases: a classical exchange around the turn of the twentieth century, a long dormancy while economics professionalised elsewhere, a Sraffian challenge from within the left in the nineteen sixties and seventies, and a still-open fourth phase in which three incompatible reconstructions and an empirical research programme compete without resolution.

Reading the critics without being managed

Anyone researching this topic will encounter four kinds of source and each requires a different discount.

The advocacy pages, on both sides, are identifiable by a single feature: they present one argument and reach a general verdict. An Austrian-school page that presents Bohm-Bawerk and concludes that the theory is refuted, and a socialist page that presents the two-level reply and concludes that the objections are answered, are making the same mistake in opposite directions, and neither will mention the redundancy charge, because it is inconvenient for both. It embarrasses the defender by conceding consistency and attacking necessity, and it embarrasses the Austrian critic because it comes from a framework that is not marginalist either.

The textbook treatments are usually accurate and always incomplete. They present the theory in its first-volume form, note that Marx modified it in the third volume, and move on. What they omit is that the modification is the whole argument, and a student who learns the theory from a textbook and then meets a serious critic will discover that they have been taught the version the critic is attacking.

The technical literature is where the argument actually lives, and its difficulty is not mathematical but terminological: the same words carry different definitions in different programmes. The value of labour power means one thing in a traditional treatment and another in the New Interpretation. Simultaneous and temporal valuation produce different magnitudes from identical data. A reader who does not establish which definitional scheme a paper is using will find results that appear to contradict each other and in fact do not address each other.

The fourth category is the polemical history, which recounts the argument as a story with a winner. These accounts are unreliable in a specific way: they compress the marginal revolution and the Bohm-Bawerk critique into a single event, they attribute to the profession a verdict no survey establishes, and they typically do not mention that the classical surplus approach was revived in the nineteen sixties by a Cambridge economist with no interest in defending Marx. The correction is not that the theory won; it is that the story has more moving parts than the telling allows.

The practical rule for a reader is to establish four things about any source: which of the three claims it addresses, which definitional scheme it uses, whether it states the opposing position in a form the opponent would accept, and whether it names what would change its verdict. A source that fails all four is an advertisement. A source that satisfies all four is worth the time regardless of which side it comes down on.

Three contemporary cases where the objections are pressed

The objections are not confined to nineteenth-century examples, and the versions that arrive from contemporary economic life are worth treating separately because they are the ones a reader is most likely to meet and because two of them are stronger than the classical objections they resemble.

The first is software and digital goods. A program is written once at considerable labour cost and thereafter copied at a cost approaching zero, so the labour required to reproduce a unit falls toward nothing while the price does not. The objection is that the theory predicts a price collapse that does not occur. The reply available to the theory is that the price is sustained by legally enforced exclusion rather than by conditions of production, which makes it a case of monopoly price rather than a counter-example to a claim about competitive reproduction, and that where exclusion fails, the price does collapse toward the reproduction cost, which is what happens to unprotected digital goods. That reply works, and it costs something: it concedes that an expanding share of output is priced by legal exclusion rather than by the mechanism the theory describes. A defender who deploys the reproducibility restriction here should notice they are deploying it against a sector that keeps growing.

There is a further wrinkle the popular versions miss. The theory’s own category for this case is not exclusively monopoly price. Software production involves large fixed labour outlays recovered across a run of output, which is a problem the theory shares with any account of increasing returns, and the classical apparatus handles increasing returns poorly. That is a real limitation and it is not answered by invoking intellectual property.

The second case is financial assets. A bond, an equity, a derivative, or a securitised claim has a price and embodies no labour in the relevant sense. The theory’s answer is that these are titles to future revenue streams, priced by capitalisation at the prevailing rate of interest, that the third volume treats such titles under the heading of fictitious capital, and that the category is designed precisely for claims whose price movements have no direct relation to conditions of production. The answer is textually solid. What it establishes is that the theory has a place to put these objects, not that it explains their prices, and in an economy where the market value of financial claims dwarfs the value of the output they are claims on, having a category is a thinner achievement than it once was. The honest statement is that the theory’s account of the reproducible sector is unaffected by financialisation and that the reproducible sector is a smaller share of what gets priced.

The third case is unpaid work, whether the household labour that reproduces labour power or the user activity that generates data and attention for platforms. Here the objection runs in the opposite direction: not that the theory counts labour that creates no value, but that value appears to be generated by activity the theory does not count as labour at all, because it is not performed under a wage relation or organised for exchange. This is a genuine problem, and it has produced a large and serious literature inside the tradition rather than outside it, which is treated in this series in the social reproduction cluster rather than here. Two things can be said at this level. The theory’s own resources allow a distinction between activity that produces value and activity that is a condition of value production without itself producing it, and that distinction does real work for household labour. It works much less well for platform activity, where the boundary between using a service and producing the thing that is sold has genuinely blurred, and where a defender should concede that the categories were not built for the case.

What all three cases share is that they test the theory’s scope rather than its internal logic, and the pattern is consistent: the theory holds where competitive reproduction by labour is the mechanism setting the regulating price, and it has categories rather than explanations everywhere else. Whether a framework whose domain is shrinking is a framework in trouble depends on how large the reproducible commodity sector remains, which is an empirical question and not a philosophical one, and it is the question a serious critic should be asking instead of the one about paintings.

Do digital goods break the labour theory of value?

Not internally, but they narrow its reach. Digital reproduction costs approach zero, so prices are sustained by legal exclusion rather than by conditions of production, which places the case under monopoly price rather than under value. The theory has a category for it; what it does not have is an explanation of the price level.

The objection nobody makes, and should

There is a difficulty in the theory that the standard critical canon almost entirely ignores, and it is more troubling than three of the six objections surveyed above. It concerns the second sense of socially necessary labour time.

The first sense is the familiar one: labour performed under the prevailing conditions of productivity and average skill, so that the sluggard’s excess hours do not count. The second sense appears in the third volume’s discussion of market value, where the question is not how efficiently a given quantity was produced but how much of the product the society needed. Labour expended producing a quantity in excess of social requirements is not socially necessary either, however efficiently it was performed, and the value of the whole output is affected accordingly. Marx states the point clearly and does not develop it.

The difficulty is that this second sense smuggles a demand condition into the determination of value, and the theory has no account of where the social requirement comes from or how it is measured. Say the amount of a good the society needs is itself a function of prices, incomes, and the distribution of both, all of which are supposed to be explained downstream of value rather than fed into it. If the quantity socially required depends on price, and value depends on the quantity socially required, then value depends on price in a way that the theory’s whole architecture is designed to avoid. The circularity is structurally similar to the one identified in the skilled-labour reduction and it has attracted a fraction of the attention.

The reply available is that the second sense operates at the level of branches rather than individual commodities, and that it explains why capital migrates between branches rather than entering into the determination of a single commodity’s value. That is a reasonable reading of the text and it does contain the damage, but it leaves the theory silent about proportions, which was the problem the social-form claim identified as central. A theory whose distinctive achievement is explaining how the total labour of a society gets distributed among branches cannot be entirely indifferent to what determines the branch proportions.

The reason this objection is absent from the popular canon is instructive. It cannot be used to show that Marx was a bad economist, since he raised it himself, and it cannot be used to show that the theory is vindicated, since he did not resolve it. It is a genuine unfinished problem in an unfinished work, which is the least rhetorically useful kind of finding and the most scholarly one. The third volume is an editorial construction assembled by Engels from manuscripts Marx left unprepared, and the passage in question is among the material where the argument is visibly in progress rather than complete, a point worth remembering whenever anyone reports what Marx concluded about market value.

Anyone looking for a research question in this area rather than another restatement of the transformation problem should start here. The relation between the two senses of social necessity is underdeveloped in the secondary literature, it connects directly to the demand-side gap that critics attack from outside, and unlike the transformation dispute it has not been argued to exhaustion.

What a fair-minded critic and a fair-minded defender should each concede

The most useful way to close a survey of this kind is to state what each side ought to give up, because the concessions are asymmetric and specifying them shows where the real disagreement sits.

A critic should concede that the land and artwork examples are not arguments, that the machine objection in its popular form restates the theory’s own premise as an objection, that the marginal framework and the labour theory answer different questions so that the profession’s reorientation was a displacement rather than a refutation, that no survey establishes any professional consensus about the theory’s status, and that the exploitation claim can be constructed without the value theory, so that defeating the value theory does not defeat the political conclusion. A critic should also concede that a theory can be false in its quantitative claims and still have identified a real object, which is what the social-form defence asserts, and that dismissing that defence requires an argument rather than a sneer about unfalsifiability.

A defender should concede that commodities do not exchange in ratios proportional to labour contents and that Marx says so, that Marx’s own transformation procedure is incomplete under simultaneous valuation, that the redundancy result stands within its own framework and has not been overturned on its own terms, that the reduction of skilled to simple labour has no non-circular derivation, that three incompatible reconstructions of what the theory claims coexist inside the tradition, and that the strongest defence available converts a quantitative economic theory into an account of social form, which is a substantial retreat from what the theory was taken to claim for most of its history.

The residue after both sets of concessions is the actual dispute, and it is narrower and more interesting than the public argument. It concerns whether the aggregate identities have substantive content under any consistent valuation scheme, whether the empirical correlation between labour requirements and prices survives normalisation, and whether an account of the social form of labour allocation is a genuine explanatory achievement or an unfalsifiable residue. Those three questions are open, they are tractable, and none of them is what the argument is usually about.

For students writing on this in an examination

Examiners on this topic reward evaluation that names its target. An answer that says the labour theory of value has been criticised, mentions Bohm-Bawerk, and asserts that Marxists have responded, sits in the middle band no matter how fluently it is written. An answer that says the transformation problem targets the internal consistency of the value-to-price derivation while the Sraffian redundancy charge concedes consistency and attacks necessity, and that the two therefore require different replies, is doing the work the mark scheme is looking for.

The single distinction that earns marks here is the one between refuting a theory and rendering it redundant. Candidates almost never make it, and it takes one sentence. The second most valuable move is naming the reproducibility restriction when the land or artwork example appears in a question, because that example is frequently offered as though it were decisive and a candidate who disposes of it in two sentences has shown they know the theory’s actual scope.

The standard trap is asserting that Marx claimed price equals labour time. Whoever wrote that in an answer has told the marker they have not read past the first chapter, and no amount of subsequent evaluation recovers the position. The second trap is treating the objections as a list to be recited rather than as arguments with different targets and different verdicts; length spent listing is length not spent evaluating. Answer structures, marks logic, and worked skeletons for questions on Marx’s value theory belong to the exam and essay guide for Capital Volume One, which owns that guidance in this cluster, and this article should be used as the evaluation material feeding into it.

For teachers running this as a seminar

The misconception that dominates a class on this topic is that the labour theory of value is a claim about how much things ought to cost, usually arriving as a moral intuition about fairness. Students who hold it will read every objection as an attack on the idea that workers deserve more, and will defend or attack accordingly, which produces a seminar in which nobody discusses the theory.

The question that surfaces it reliably is to ask what the theory predicts about the price of a good produced by a firm using obsolete, slow equipment. Students holding the moral reading say the good should be expensive, because more labour went into it. The theory says the opposite: the producer’s excess hours are not socially necessary and do not count, so the firm loses. That single example separates the theory from the fairness intuition faster than any amount of exposition, and it opens directly onto the socially necessary clause.

The format that works for the objections themselves is a structured disagreement with sides assigned rather than chosen. Assign the redundancy charge to one group and the two-level reply to another, with the explicit instruction that each group must state the other’s position to the other group’s satisfaction before arguing against it. The productive discovery, which usually arrives without prompting once the rule is enforced, is that the two groups are arguing about what the theory is for rather than about whether it is true, which is the actual state of the literature and is a more valuable thing for a class to find out than any verdict a lecturer could hand them.

For journalists, researchers, and anyone citing this dispute

Three claims circulate confidently in public writing on this subject and none of them is safe. The first is that the labour theory of value was refuted, usually credited to Bohm-Bawerk or to the marginal revolution. What can be said accurately is that the strong price-proportionality version does not hold, that the profession displaced the classical approach rather than disproving it, and that the redundancy charge is the strongest single argument against the theory’s necessity. The second unsafe claim is the mirror image: that the objections have all been answered. Three incompatible reconstructions exist precisely because they have not been.

The third and most common error is attributing to Marx the claim that prices equal labour values. Before citing anything on this topic, verify which volume and which chapter the claim is being drawn from, because the first volume and the third volume operate at different levels of abstraction and quoting the first as though it were the whole system is the standard mistake in both directions. Note also that the second and third volumes are editorial constructions assembled by Engels from manuscripts Marx did not prepare for publication, which is relevant when the question at issue is what Marx would have said about a difficulty he was still working on.

For attribution, the safe form is to name the objection, its author, and its target rather than to summarise the state of the argument. Writing that Sraffa’s framework showed prices and the profit rate can be determined without value magnitudes, and that Steedman pressed this against Marxist economics, is verifiable. Writing that economists consider the theory discredited is not, since no survey establishes it, and it invites a correction from anyone who reads the heterodox literature. The same discipline applies to dates: the classical critique, the classical reply, the Sraffian framework, the redundancy book, and the analytical challenge each belong to a specific decade, and getting the sequence wrong makes a reply look like an anticipation or an anticipation look like a reply, which reverses the meaning of the exchange. Readers building a research file on this dispute can save these arguments as citation-linked notes on VaultBook and keep the objection, the target, the reply, and the source attached to one another, which is exactly the discipline this topic requires.

What is settled and what remains open

Settled: commodities do not exchange in ratios proportional to labour contents, and Marx said so. Settled: the reproducibility restriction disposes of the land and artwork examples, and those examples are not arguments. Settled: value magnitudes are not necessary for determining relative prices and the profit rate from technical conditions and one distributional variable. Settled: the exploitation claim can be constructed without the labour theory of value, which means the political conclusion does not depend on winning this argument.

Open: whether the aggregate identities can be preserved under a temporal reading, and whether that reading is Marx’s. Open: whether the empirical correlation between labour inputs and prices, once industry size is normalised out, constitutes evidence for anything. Open: whether the social-form claim is a defensible core or a retreat to unfalsifiability. Open: whether a theory that survives by giving up its quantitative content is worth having.

The distribution of these results is the reason the argument does not end, and it is also why the honest verdict has to be reported claim by claim rather than as a headline. Both camps have real victories, and each camp’s victories are on claims the other camp does not principally care about. Anyone entering the dispute should establish which claim is at stake before anything else, which is the target-check rule and the only piece of equipment needed to read the entire literature without being misled by either side. Applied consistently, it turns an argument that appears interminable into three separate questions, two of which have answers and one of which is still worth working on.

Frequently Asked Questions

Q: Has the labour theory of value been debunked?

Partly, depending on which version is meant. The claim that commodities exchange in ratios proportional to their labour contents does not hold, and the third volume of Capital says so, so anyone defending that version is defending something Marx abandoned. The claim that value magnitudes are necessary to determine prices and the profit rate was undermined by the Sraffian redundancy argument and no reply commands consensus. The claim that labour is the source of the aggregate surplus, and the claim that value is the social form private labour takes in a market society, are disputed rather than debunked. A blanket verdict either way misstates the position, because the theory does three separable jobs and the results differ for each.

Q: What about goods that take no labour, like land?

Land is outside the theory’s stated scope, which covers commodities that are freely reproducible by labour. The mechanism by which labour time regulates exchange ratios is competitive reproduction: if a price rises above the regulating level, capital enters, output expands, and the price is pushed back. Where reproduction is impossible, that mechanism has nothing to work on. Marx treats land prices through the theory of ground rent in the third volume, where the price of land is capitalised rent, a discounted claim on future revenue rather than a value magnitude. The objection therefore misses as a refutation, though it does highlight that a growing share of modern asset value consists of capitalised claims the theory does not cover.

Q: Why do rare paintings sell for so much under the labour theory of value?

They sell under monopoly price rather than under value, and the third volume of Capital distinguishes the two explicitly. Monopoly price is determined by what buyers can be brought to pay, given fixed and unreproducible supply, rather than by conditions of production. A painting by a dead artist cannot be reproduced by expending the same hours again, so no competitive mechanism pushes its price toward a labour-time centre. The example is the most frequently offered refutation of the theory and it is the weakest, because it applies a claim about reproducible commodities to a non-reproducible one. What the example does establish, fairly, is that the theory covers a narrower slice of economic life than its popular presentations suggest.

Q: Is socially necessary labour time a circular definition?

The charge has real force where skilled labour is concerned. The theory requires a conversion of complicated labour into multiples of simple labour, and it asserts that such a conversion exists without deriving it. If the coefficient is read off from observed wage differentials, then price data is being used to build the value magnitudes that supposedly explain price data. Defenders reply that the coefficient is an empirical constant rather than a definitional stipulation, which is respectable but changes the claim’s status. The socially necessary clause itself, applied to identical labour under different conditions, is not circular: it says the labour of an inefficient producer does not count, which is a substantive claim with a clear implication.

Q: Do Marxists still defend the labour theory of value?

Yes, and in at least three incompatible ways. The temporal single-system interpretation defends Marx’s own transformation procedure by denying that inputs and outputs should be valued simultaneously. The New Interpretation defends a narrowed version by redefining the value of labour power as labour commanded by the money wage, which makes the rate of exploitation measurable from national accounts. Value-form theory defends the theory by denying it was ever a price theory, treating it instead as an account of why social labour must be validated through money. These are rival research programmes rather than variations on a theme, and their mutual disagreement about what the theory claims is itself something critics reasonably point to.

Q: Is the labour theory of value redundant for explaining prices?

For the narrow task of determining relative prices and the rate of profit, the Sraffian result says yes: given technical conditions of production and one distributional variable, both are determined without any reference to labour values. Ian Steedman pressed this conclusion against Marxist economics, adding that in joint-production cases computed values can come out negative, which is meaningless for a magnitude representing embodied labour. This is the strongest objection in the whole literature because it concedes consistency and attacks necessity. Defenders reply either by rejecting the simultaneity assumption, by relocating the theory’s claim to aggregates, or by denying that price determination was ever the theory’s job.

Q: Do you need the labour theory of value to prove exploitation?

No, and the argument establishing this came from a philosopher sympathetic to the political conclusion. G. A. Cohen argued that the exploitation claim concerns the product, a physical and social object that workers make and owners partly appropriate, while the value theory concerns exchange magnitudes, and that the first does not depend on the second. John Roemer then formalised exploitation and class in terms of unequal ownership of productive assets, showing that a formally parallel result can be derived using any basic commodity in place of labour. Defenders reply that the value analysis explains how appropriation coexists with exchange of equivalents, which the alternative accounts handle less well.

Q: What is the strongest defence of the labour theory of value?

That the theory explains how a society whose production is private and dispersed nonetheless allocates its total labour among branches, without anyone deciding to, and that value names the social validation private labour receives when its product exchanges. On this reading, price deviations are expected rather than embarrassing, the redundancy charge misses because the theory is not competing for the price-determination job, and no rival framework offers any account of the social form of labour allocation because none treats it as a question. The defence is serious and it is expensive: it concedes that the theory does not explain relative prices, which is what most of its historical proponents believed it did.

Q: Does supply and demand refute the labour theory of value?

No, and the objection misunderstands what the theory claims. Marx’s account treats supply and demand as determining the oscillation of market prices around a regulating centre, and as explaining deviations rather than the centre itself. The theory’s proposition concerns what prices gravitate toward when supply and demand are in balance, which is precisely the situation in which supply and demand explain nothing. Defending the theory does not require denying that supply and demand affect prices, and any defence that does deny it has misread the text. The genuine objections come from elsewhere: the transformation of values into prices of production, and the demonstration that prices can be determined without value magnitudes at all.

Q: Why do most economists reject the labour theory of value?

Careful writers should avoid asserting what proportions of any profession believe, since no survey establishes it. What can be said is that the discipline reoriented around marginal analysis following publications in the eighteen seventies, that this reorientation preceded the best-known critiques rather than following them, and that the classical surplus approach was displaced because the questions the profession asked changed, not because a decisive refutation was produced. The strongest analytical reason for rejection available today is the redundancy result: if prices and the profit rate can be determined from technical conditions and a distributional variable, value magnitudes add nothing to the calculation.

Q: What is the difference between the transformation problem and the redundancy charge?

They are different arguments with different conclusions and are constantly conflated. The transformation problem is an internal-consistency charge: it says Marx’s own procedure for converting values into prices of production is incomplete, because inputs remain valued at values while outputs are transformed, and that a consistent simultaneous solution generally cannot preserve both aggregate equalities. The redundancy charge concedes consistency entirely and asks a different question: even if the value calculation works, what does it add, given that prices and the profit rate were already determined without it? The first says the derivation is flawed; the second says the derivation is unnecessary. A reply that answers one leaves the other untouched.

Q: Did Marx claim that prices are equal to labour values?

No, and the third volume of Capital denies it explicitly through the theory of prices of production, which shows that competition equalises profit rates across industries with different capital compositions and therefore causes systematic divergence between prices and values. The first volume proceeds on the simplifying assumption of exchange at values because it is analysing the origin of surplus at a high level of abstraction, where the redistribution of surplus among capitals has not yet been introduced. Reading the first volume’s assumption as the system’s final claim is the single most common error in this literature, and it is made by critics and defenders in roughly equal numbers.

Q: How should a teacher handle the fairness intuition students bring to this topic?

Surface it before teaching anything else, because students who read the theory as a claim about what things ought to cost will misread every objection as an attack on workers. The quickest diagnostic is to ask what the theory predicts about a good made by a firm using slow, obsolete equipment. The fairness reading says it should be expensive because more labour went into it; the theory says the excess hours are not socially necessary, do not count, and the firm loses. Once that gap is visible, the socially necessary clause teaches itself, and the class can discuss the objections as arguments rather than as loyalty tests.

Q: Which criticism of the labour theory of value is the hardest to answer?

The redundancy charge derived from Sraffa’s framework and pressed by Steedman. Its power comes from conceding everything the older critiques disputed: it grants the arithmetic, grants the internal consistency on a suitable reading, and then asks what the value category contributes to a determination that has already been completed without it. Idle theories are abandoned rather than refuted, which makes this the most dangerous form of attack. The three defences that exist against it are mutually incompatible, which is itself evidence that the problem is unresolved. Anyone claiming this objection has been answered should be asked which of the three answers they mean.

Q: Does rejecting the labour theory of value mean rejecting Marx?

Not necessarily, and several of the most productive lines in twentieth-century Marxist scholarship proceeded on exactly that basis. The analytical Marxists rebuilt exploitation and class analysis on property relations rather than value magnitudes. Historical materialism, the theory of ideology, the account of the state, and the analysis of accumulation and crisis all have arguments that do not depend on the value theory holding in its strong form. What rejecting the theory does cost is the specific explanation of how surplus appropriation coexists with exchange of equivalents, which was the problem Chapter Six of the first volume was written to solve, and any replacement account has to say something about that puzzle.

Q: What should a researcher verify before citing this debate?

Three things. Which volume of Capital a claim is drawn from, since the first and third operate at different levels of abstraction and citing one as the whole system is the standard error. The manuscript status of the source, since the second and third volumes were assembled editorially by Engels from material Marx did not prepare for publication, which matters whenever the question is what Marx would have concluded. And which of the three modern defences a cited author belongs to, since temporal single-system, New Interpretation, and value-form writers disagree about what the theory claims and citing one as representing Marxist economics generally will misrepresent the field.

Q: Is there any empirical test of the labour theory of value?

There is a substantial empirical literature, mostly using national input-output tables to compute labour requirements by industry and compare them with observed prices, and the reported correlations have generally been high across many countries. The literature is more contested than the headline results suggest, because both the price vector and the labour-value vector are scaled by industry size, so a large share of the correlation may reflect variation in sector size rather than any relation between price and value. That statistical objection, and the replies to it, are treated at length in the dedicated evidence article in this cluster, which is where anyone assessing the empirical claim should start.

Q: Did Ricardo hold the same labour theory of value that Marx did?

No, and the difference matters for assessing the criticisms. Ricardo treated labour embodied as governing exchange ratios and knew that capitals of different composition or duration break the rule, so he classed the deviations as modifications and searched for an invariable measure that would let changes in relative prices be attributed to their proper causes. Marx accepted the deviations and explained them, through the equalisation of the profit rate across capitals of unequal composition, which necessarily transfers surplus between branches. Marx also introduced the distinction between labour and labour power, which Ricardo lacked and which carries the surplus argument. A critique aimed at Ricardo’s version does not automatically reach Marx’s, and the reverse holds too.