Ask what caused a crisis and two words arrive together so reliably that they are often printed as if they were one word with a hyphen missing: overproduction and underconsumption. A factory system makes more than people can buy, the story runs, warehouses fill, workers are dismissed, and the machine stalls on its own abundance. The story has the great virtue of fitting the visible scene of a slump, the idle plant and the unsold stock and the queue outside the labour exchange, and it has the great defect of being two stories compressed into a single sentence. Overproduction names a surplus on the side of what has been made. Underconsumption names a shortfall on the side of what can be purchased. They point in opposite directions along the same circuit, they imply different mechanisms, they suggest different remedies, and in the writings of Karl Marx they receive strikingly different treatment, since he built the first into his account of crisis while attacking the second, in one of its most quoted passages, as an empty formula that explains nothing. A reader who keeps the two apart has hold of a distinction that reorganises the whole literature of capitalist crisis. A reader who merges them will misread Marx, misread his critics, and misread the claims made in his name by later schools that he would partly have disowned.

The organising claim of this article can be named, and the name is worth keeping because it converts a vocabulary lesson into an instrument. The claim is the relative-to-what test. Every statement about overproduction, and every statement about underconsumption, measures an excess or a shortfall against some standard, and the standard is doing the real work in the sentence. Production can exceed the demand that is backed by money, which is one measure. It can exceed the volume that can be sold at prices yielding the rate of profit capital requires, which is a second and quite different measure. Or it can exceed what human beings need, which is a third measure, and the one under which overproduction, in any abundant sense, almost never occurs. Marx’s usage belongs overwhelmingly to the second measure. When he wrote of overproduction he meant, in the decisive passages, overproduction of capital: an accumulation of productive capacity and of claims upon future profit beyond what could be valorised, set to work and made to yield an adequate return. Underconsumption, as an explanation of crisis, belongs to the first measure dressed in the language of the third, and Marx’s objection to it was precisely that it slid between denominators, borrowing the moral force of unmet need to explain a failure of monetary demand. State the denominator, and most of the confusion between the two terms dissolves. Refuse to state it, and the confusion is permanent, because the argument can always retreat to whichever measure the critic has not examined.

Three corrections follow from the test, and each cuts against the version of Marx that circulates most widely, including among his supporters. The first correction is that overproduction in his sense is not a physical surplus of goods relative to human need, and the spectacle of unsold goods beside unmet needs is for him a symptom to be explained rather than the explanation itself. The second correction is that Marx explicitly rejected the underconsumptionist theory of crisis in the form in which it is most often attributed to him, calling the claim that crises arise from a scarcity of consumption a pure tautology and observing, in the same passage, that crises are regularly prepared by periods of generally rising wages, a fact the crude theory cannot digest. The third correction is that the two words name rival explanations rather than two descriptions of one event, and a serious account of crisis must also make room for a third rival, disproportionality, in which nothing is excessive in aggregate and the failure is one of balance between branches of production. The wider map of these families, and of the falling rate of profit that stands as a further rival mechanism, is drawn in the marxist crisis theory guide, while the vocabulary across the whole field is fixed in the marxism glossary key terms. What follows here is the precision work: the denominators distinguished, the texts examined, the lineage traced, the strongest opposing case heard at full length, and a method left in the reader’s hands for judging any crisis claim that arrives using either word.

Overproduction is a relation, not a pile of goods

Begin with the grammar of the word itself, because the grammar conceals the trap. Overproduction sounds like a property of a quantity, as if a society could simply make too much, the way a baker can bake too many loaves for a single morning. But a quantity is only excessive against a standard, and capitalist production supplies several standards that diverge sharply. A mill can produce more cloth than its district can wear, more than merchants can sell at the old price, more than can be sold at any price covering costs and the expected profit, or more than the looms’ owners can keep profitably employed. These are four different excesses with four different causes, and only loose writing treats them as one. The word overproduction therefore behaves less like the name of a thing and more like the name of a comparison whose second term has been left unstated. Filling in the missing term is not pedantry. It is the difference between an explanation and an incantation, since each completion of the sentence commits the speaker to a mechanism, a body of evidence and a remedy, and the completions contradict one another.

Underconsumption has the same grammatical shape with the sign reversed. It names a shortfall of consumption, but consumption relative to what? Relative to the output that producers have brought to market, relative to the consumption that would keep every plant running, relative to the needs of the population, or relative to some earlier level from which demand has fallen away? Each version has been held by somebody. The version that matters for crisis theory is the first: the claim that the consuming power of society, and specifically the restricted consumption of the working population, falls short of the expanding scale of production, so that goods pile up unsold and the circuit breaks at the point of sale. Held in that form, underconsumption is a genuine theory with a mechanism and predictions. Held in the looser form, in which any failure to sell is redescribed as somebody consuming too little, it explains every crisis by definition and therefore explains none of them, which is the burden of Marx’s complaint against it. Precision about the denominator is what separates the theory from the redescription.

Why does talk of too much production mislead?

It misleads because the phrase borrows a physical picture, a pile of goods passing some ceiling, for what is really a relation between output and a standard of comparison. Change the standard and the same output flips from excessive to inadequate: a harvest can exceed what merchants can sell profitably while falling short of what a hungry district needs.

The physical picture does further damage by suggesting that the remedy for crisis is simply less production, as if the system suffered from abundance the way a river suffers from flood. Every serious tradition in this debate denies that picture in its own way. For Marx, the paradox of capitalist crisis is that production is restricted by the relations of distribution and profitability amid palpable unmet need, so that the problem is never that society has provided too well for itself. For the underconsumptionist, the surplus is real but relative to purchasing power, and the remedy runs through redistribution rather than through any reduction of output. For the theorist of disproportionality, output in total may be exactly right while its composition is wrong, too much of one thing and too little of another, so that speaking of too much production in general is a category error. The shared rejection of the naive picture is the one point on which the rival schools agree, and it makes the survival of the naive picture in popular writing all the more striking. Newspapers reach for the warehouse image because it can be photographed. The mechanisms cannot be photographed, and so the mechanisms are what an article of this kind exists to supply.

A useful discipline, when either word appears, is to complete the sentence aloud. Overproduced relative to solvent demand: that is a claim about markets and incomes. Overproduced relative to profitable employment: that is a claim about capital and its required return. Overproduced relative to need: that is a claim about provisioning, and in a society where needs go chronically unmet it is very nearly a contradiction in terms. Underconsumed relative to output: a claim about the distribution of purchasing power. Underconsumed relative to some previous period: a claim about a change, which requires the change to be identified and dated. The completions are not interchangeable, they do not stand or fall together, and an argument that needs one of them gains nothing from evidence supporting another. Much of the published quarrel between Marxists and their critics consists of the two sides completing the sentence differently and then refuting the completion the other side never made. Making that manoeuvre visible at the moment it is performed is what the test is designed to do.

Nor is the ambiguity an accident of careless authors alone. It is built into the phenomena, because in an actual crisis all three denominators fail together in a tangle: goods go unsold, profits collapse, capacity stands idle, and needs go unmet in the same streets at the same time. The simultaneity tempts the observer to treat the words as interchangeable labels for the general wreckage. Theory begins where the temptation is resisted, in asking which failure is doing the causing and which are consequences travelling in its wake. A glut of unsold cotton goods might be the first domino or the last; a collapse of profit might originate in the sales failure or precede and provoke it; a disproportion between branches might ignite the whole sequence or merely transmit a shock begun elsewhere. The words overproduction and underconsumption are claims about the order of events disguised as descriptions of the scene, and the order of events is exactly what the rival theories dispute.

The relative-to-what test stated in full

The test itself is short enough to be carried in the memory and applied in the middle of an argument. When a claim of overproduction or underconsumption is made, ask three questions in sequence. First, what exactly is said to be in excess or in shortfall: finished goods, productive capacity, capital seeking investment, or consumption itself? Second, against what standard is the excess measured: the demand backed by money at ruling prices, the sale that yields an adequate rate of profit, or the needs of the population? Third, by what mechanism does the measured gap produce the crisis: a failure to sell, a failure to valorise, a breakdown of proportions between branches, or some chain linking several of these? A claim that survives the three questions with definite answers is a theory, and can be argued with on evidence. A claim that answers the second question by drifting between standards, or the third by pointing at the crisis itself, has failed the test, and its failure is diagnostic regardless of the politics of its author. The test is neutral between schools. It convicts careless Marxists as readily as careless critics of Marx, which is one mark of a genuine instrument rather than a party weapon.

Consider how the test sorts three assertions that sound identical in a headline. A report says that an industry suffers from overcapacity. Completed against solvent demand, the claim means that buyers at current prices will not take the output, and the evidence would be inventories, order books and price cutting. Completed against profitability, the claim means that the installed plant cannot earn the return its owners require, and the evidence would be margins, write-downs and the flight of new investment toward other sectors. Completed against need, the claim would mean that the population’s requirements are saturated, and for most industries in most places the evidence would be embarrassingly absent. The three completions describe one industry at one moment, yet they are different diagnoses with different prognoses: the first may clear with a price fall, the second persists until capital is destroyed or devalued, and the third, were it ever true, would mark a condition capitalism has never reached. A single word, overcapacity, carried all three because the denominator went unspoken. Speaking it aloud is the whole technique.

How can goods stand unsold while needs go unmet?

They stand unsold because a market registers not need but demand, and demand means need armed with money at the point of sale. A coat unsold in a shop window and a child cold in the street outside are compatible facts, joined not by any scarcity of coats but by the absence of purchasing power where the need resides.

Marx treated that compatibility not as a paradox to be marvelled at but as a defining feature of production for exchange, in which the social connection between producer and consumer runs entirely through sale. Goods are produced as commodities, as values seeking realisation in money, and their usefulness to anyone is a precondition of their being made at all but no guarantee of their reaching the person who could use them. When sale fails, the goods are, in the only sense the system recognises, superfluous, even while the need they could meet persists untouched. The scandal of the scene therefore proves less than it appears to prove. It shows that distribution through the market separates production from need, which nobody in this debate denies. It does not by itself show why sale failed in a particular crisis, whether the originating failure sat in consumption, in profitability, or in the proportions among branches. Writers who present the unsold coat and the cold child as a complete theory of crisis have described the stage set and mistaken it for the plot.

The point matters because the moral force of the image is often used to smuggle in the underconsumptionist mechanism without defending it. If the scene itself were the explanation, then every crisis would have the same cause, the poverty of the mass of consumers, and the differences among crises, their timing, their sectoral origins, their financial or industrial character, would be decoration. The historical record refuses that flattening. Crises have begun in railway shares and in cotton, in property booms and in harvest failures, amid rising wages and amid stagnant ones. A theory that explains all of them by one unchanging shortfall of consumption explains the constant and misses the variable, and it is the variable, the timing and the trigger, that a theory of crisis exists to catch. Marx’s objection to the crude doctrine, as the following sections show, was aimed at exactly this flattening, and his own alternative located the trigger in the movement of capital itself.

Gluts of particular goods: the case everybody grants

Start where agreement is widest, with the glut of a particular commodity. A season’s fashion changes and a warehouse of last year’s cut stands unsold; a harvest technology doubles the yield of one crop and its price collapses; a railway mania lays track toward towns that never grow. In each instance too much of one thing exists relative to the demand for it, capital embodied in that thing is devalued, and its producers suffer losses or ruin. Classical political economy, whatever its quarrels elsewhere, never denied gluts of this kind. They are the ordinary friction of production carried on by independent producers who must guess at a market they cannot survey, and their existence proves nothing about the system as a whole, because the losses in the glutted branch are mirrored, in the textbook version, by unmet demand and rising prices somewhere else. Capital and labour are supposed to migrate from the swollen branch to the starved one until proportions restore themselves.

Even at this modest level the relative-to-what test earns its keep, because a particular glut is always measured against demand for that commodity at a price, and the price matters. The same physical stock is a glut at the old price and a bargain at a lower one, which is why the partial glut is normally cleared by devaluation: the goods are sold at a loss, the capital their owners advanced is written down, and the branch contracts. Marx’s analysis of crisis takes this clearing mechanism seriously and extends it, since for him the destruction and devaluation of capital is not an accident that befalls the system from outside but one of the ways the system restores the conditions of profitable production. The partial glut thus introduces two themes that return at every larger scale: excess is relative to price and profit rather than to physical quantities alone, and the cure the system applies is the destruction of value, not the redistribution of goods to those who need them.

What makes a glut partial rather than general?

A glut stays partial when failure to sell is confined to branches while demand elsewhere holds, the surplus in one line answering to shortage in another, so the disturbance is one of composition. It turns general when unsold stocks, falling prices and idle capacity spread across branches together and no expanding sector remains to absorb what the others release.

The distinction sounds tidy, and nineteenth century economists wielded it as a weapon, conceding the partial case in order to deny the general one. Their argument, examined in the next section, held that production itself creates the demand for production, so that a universal glut is a logical impossibility and every apparent instance must be a disproportion in disguise. The distinction nonetheless survives its polemical origins, because it identifies a real fork in explanation. A disproportion can in principle be corrected by movement between branches, painful as that movement is for the capitals trapped in the wrong one. A general glut, if such a thing exists, cannot be corrected that way, since there is nowhere for capital to flee to; it requires either a destruction of capital values on a broad front or an expansion of demand from some source outside the circuit of production itself. Which of the two a given crisis is, and which mechanism drove it, are empirical questions, and the relative-to-what test is the instrument that keeps them separate. Naming a disturbance a glut settles nothing until its scope and its denominator are both declared.

There is a further subtlety in the partial case that textbooks glide past. A branch can suffer a glut because its customers are other branches, and their demand is derived from their own expectations of sale. The market for machine tools, for steel, for industrial chemicals, is made of producers buying in order to produce, so a hesitation in final demand echoes backward through the chain with amplification. What began as a modest surplus in one consumer good can present, a few links upstream, as a spectacular glut of productive capacity, and observers stationed at different links will describe the same event with different words and different implied causes. The observer among the unsold consumer goods sees underconsumption. The observer among the idle machine shops sees overproduction of capital goods. Both are reporting local facts. The theory of the crisis is the account of how the local facts connect, and no local fact, however vivid, supplies that account by itself.

The general glut and the law of markets

The classical denial of the general glut rests on the proposition associated with Jean-Baptiste Say, that products are paid for with products, so that aggregate production generates an equivalent aggregate demand and a universal excess of supply cannot occur. Money, on this view, is a veil and a convenience; behind it, producers exchange their goods against one another, and since every seller is necessarily a buyer, the totals must balance. Ricardo accepted the proposition, and it hardened into orthodoxy: crises were real, but they were the offspring of disproportion, of sudden shifts in demand between branches, of war and harvest and government blunder, never of production as such outrunning consumption as such. Against this orthodoxy stood the two great heretics of the early nineteenth century, Sismondi in Geneva and, from a very different direction, Malthus in England, both of whom insisted that the totals need not balance in the way the doctrine required, and both of whom built the insistence into theories of crisis grounded in the distribution of income.

Marx’s route into this old quarrel ran through a distinction the classicals had underweighted: the distinction between barter and the circulation of commodities through money. Where exchange is mediated by money, sale and purchase fall apart in time and space; a producer can sell and hold the money, interrupting the chain, and the identity of aggregate supply and aggregate demand dissolves from a necessity into a possibility that may or may not be realised. In the first volume of Capital this separation is presented as the abstract possibility of crisis, the simplest form in which a general failure to sell can be conceived at all. Possibility is not yet mechanism, and Marx never rested the theory on the mere separability of sale and purchase. But the point dismantles the logical veto on general gluts. Once money intervenes, there is no contradiction in commodities as a whole failing to find buyers at prices their producers require, and the question of crisis moves from logic to the concrete study of what makes the failures synchronise.

Say’s proposition also concealed a choice of denominator that the relative-to-what test exposes. Products pay for products, in the doctrine, at whatever prices clear the markets, including prices that ruin the producers. Against the standard of sale at some price, a general glut is indeed nearly impossible, since prices can fall far enough to move almost any stock. Against the standard of sale at prices reproducing capital with its accustomed profit, a general glut is a plain recurring fact, since whole branches periodically discover that their output cannot be sold on those terms. The classical argument secures its conclusion by quietly adopting the first standard, while the crises it purports to explain are defined by the second. Ricardo’s system, for all its rigour, slides between the two at exactly the point where the argument needs to be watched. Marx watched it, and the habit of watching for the slide, generalised, is the test this article is named for.

None of this makes the classical position worthless. The insistence that supply and demand are not independent magnitudes, that the income with which goods are bought is generated in producing them, remains a permanent corrective to naive pictures of demand as a rain that falls or fails to fall from outside the economy. The underconsumptionist tradition had to absorb that corrective or perish, and its sophisticated versions did absorb it, as later sections show. What the critique of Say removes is only the a priori veto, the claim that a general failure of realisation is unthinkable. With the veto gone, the field is open for competing mechanisms, and the competition among them, between underconsumption, disproportionality and the overproduction of capital, is the real subject of crisis theory from Marx onward.

Overproduction of capital: the usage that carries the argument

When Marx’s own usage is tracked across the mature work, a pattern emerges that surprises readers raised on the warehouse image. His characteristic phrase is not overproduction of goods but overproduction of capital, and the shift of noun changes everything. Capital is not a stock of things but a sum of value committed to a circuit whose end is more value; it exists in the forms of money, of productive plant and materials, and of finished commodities awaiting sale. To say that capital has been overproduced is to say that the accumulated mass seeking valorisation exceeds the opportunities to valorise it at the going rate of profit, so that a portion of it must stand idle, accept a diminished return, or be destroyed in value. The excess is measured against profitability, full stop. It is compatible with warehouses being half empty and with populations being badly fed, clothed and housed, because the standard is neither physical stock nor human requirement but the capacity of the system to keep its accumulated value expanding at the required pace.

The third volume of Capital gives this usage its most developed form in the discussion of the absolute overproduction of capital, where Marx considers a condition in which additional capital produces the same total profit, or even less, than the capital already employed. At that point, he argues, capital and the population it sets in motion have grown out of proportion to the valorisation they can sustain; a portion of capital lies fallow, competition sharpens into a struggle to shift the loss onto others, and the crisis performs its grim office of devaluing the excess away. Whatever a reader concludes about the falling rate of profit, whose treatment belongs to the guide on the tendency rate of profit to fall, the denominator of this argument is unmistakable. Nothing in it requires consumption to have fallen, or wages to have been squeezed, or goods to exceed needs. The excess is of capital relative to its own expansionary requirements, and the crisis is the mechanism by which that particular excess is corrected.

What did Marx mean by overproduction of capital?

He meant an accumulation of value seeking profitable employment beyond the openings to employ it at the rate of profit capital requires, so part of the mass stands idle, earns a diminished return, or loses value outright. The measure is valorisation, the self-expansion of value; neither the volume of goods nor the scale of human need enters the comparison.

Two clarifications keep this usage from being misunderstood. The first is that overproduction of capital is not the same as a physical surplus of plant, though it often appears as one. Capacity that could earn its keep at a lower required return, or under different ownership, counts as excess capital purely because of the return its owners must extract; the excess is a social relation wearing the disguise of an engineering fact. The second clarification is that the phrase names a tendency within accumulation rather than a permanent condition. Capital is overproduced at the peak of a boom, characteristically after a period in which credit has let investment run ahead of the profits available to validate it, and the crisis is the correction. Between crises the same physical capacity may be fully employed and unproblematic, which is why photographs of idle plant, taken at the trough, identify the phenomenon so imperfectly. The overproduction was consummated at the height of activity, when the commitments were made, and the idleness is its aftermath.

This usage also explains a feature of Marx’s crisis writing that puzzles readers expecting an underconsumptionist: his relative lack of interest in the absolute level of working class consumption as a crisis trigger. Consumption by workers figures in his system as a magnitude governed by the value of labour power and by the state of the labour market, rising when employment is plentiful and falling when the reserve of unemployed grows, and it enters the reproduction analysis as one flow among others. The engine of crisis, on his account, sits in the contradictions of accumulation itself: production expands under the compulsion of competition without regard to the limits of the market, each capital racing to enlarge its share, until the accumulated claims on profit outrun the profit available. The market’s narrowness is a condition of the drama, its restricted basis in a population whose consumption is held within bounds, but the trigger is pulled by the movement of capital, and triggers, not conditions, are what theories of timing must explain.

The tautology passage: Marx against the underconsumptionists

The twentieth chapter of the second volume of Capital, in the long discussion of simple reproduction, contains the passage on which this whole question turns, and it deserves prominence because summaries of Marx so often invert it. Marx there considers the claim that crises arise because consumption is too restricted, because the mass of producers cannot buy back what they produce, and he rejects the claim in terms of unusual sharpness. To announce that crises are caused by a scarcity of effective consumption, he writes, is “a pure tautology”, and he adds that dressing the formula in deeper dress, by dwelling on the small share of the product that falls to the workers, gives it an appearance of profundity without adding an explanation. The argument embedded in the insult is precise. If the scarcity of consumption is inferred from the unsold goods themselves, then the crisis has been renamed rather than explained; the unsold goods were the phenomenon awaiting a cause, and pointing at them again in different vocabulary supplies none. A theory must identify something that changes, some mechanism that tightens or loosens, and the bare assertion that consumption is restricted identifies a permanent feature of the system and calls it the cause of a periodic event.

Did Marx reject underconsumption theory?

Yes, explicitly. In the second volume of Capital he calls it a pure tautology to say crises are caused by insufficient consumption, and points out that crises are typically preceded by periods in which wages generally rise, which the underconsumptionist account cannot accommodate.

The observation about wages is the empirical edge of the passage and the part of it that competing coverage almost never reports. Marx notes that crises are regularly prepared by a period in which wages generally rise and the working class receives a greater share of the annual product assigned to consumption. On the crude underconsumptionist account, crisis should approach as workers’ purchasing power is squeezed and recede as it recovers; the sequence Marx points to runs the other way. The boom, with its tight labour markets and improving pay, is the antechamber of the crash, and the crash arrives at the moment when the supposed cause, restricted consumption, is at its weakest relative to the cycle. A doctrine whose trigger is found loosening precisely when the event fires has failed as a timing mechanism, whatever truths it may contain about the system’s general structure. That is the form of Marx’s objection at its strongest: not that working class consumption is irrelevant to the economy, which would be absurd in a system where it forms the bulk of final demand, but that its restriction is too constant, and its movements too perversely timed, to explain why crises break when they do.

Care is needed about what the passage does and does not rule out, because both camps in the later dispute have overread it in their own favour. It rules out, decisively, the attribution to Marx of the simple doctrine that crises happen because workers are underpaid relative to output. Any presentation of Marxism that teaches that doctrine as Marx’s own is teaching against his text, and the passage is the standing refutation. What the passage does not do, by itself, is refute every theory in which the realisation of surplus value through sale is the weak point of the system, since the sophisticated versions of underconsumptionism are built to survive the wages observation, as the objection section below concedes in full. Nor does the passage deny that the restricted consumption of the mass of the population is a permanent limiting condition on capitalist production; Marx says almost the opposite elsewhere, treating the narrow basis of consumption as one pole of the contradiction that drives production to overshoot its market. The distinction he is policing is between a background condition and an efficient cause, between what makes the system crisis-prone across its whole history and what makes a crisis erupt in one season rather than another. Readers who grasp that distinction hold the key to the passage. Readers who miss it will keep recruiting Marx to a theory he attacked, or wielding his attack against theories it does not reach.

Why did rising wages before a crash embarrass the crude theory?

Because the crude theory makes restricted consumption the trigger of crisis, and a trigger must tighten as the event approaches, whereas the wage share of the product improves during the boom that precedes the crash. If consumption were the binding constraint, its relaxation should postpone the breakdown; the record Marx cites shows the breakdown arriving on the relaxation’s heels.

Later writers in the underconsumptionist tradition have answered the wages point in two ways, and fairness requires both to be stated. The first answer distinguishes levels from movements: consumption may rise during a boom and still lag behind the productive capacity the boom installs, so that the gap between what can be produced and what can be consumed widens even while consumption itself grows. The second answer relocates the shortfall from workers’ consumption to the system as a whole, arguing that neither workers’ spending nor capitalists’ spending nor investment can be relied upon to absorb the full surplus, so that realisation depends on outlets whose availability is precarious. Both answers abandon the crude doctrine in order to save the tradition, which is itself a concession to the passage’s force. The tradition that survived Marx’s attack is a different and stronger theory than the one he attacked, and the sections that follow treat it in that stronger form, as the neutrality of this subject requires.

The underconsumptionist lineage, taken seriously

Underconsumptionist explanation did not begin as a socialist doctrine, and remembering its pedigree is the quickest way to see that it answers to evidence and argument rather than to allegiance. Sismondi, writing in the aftermath of the Napoleonic wars and watching machine production displace the workers who were also its customers, argued that production which outruns the revenue available to purchase it must end in glut, and that the separation of the producer from any claim on the product made such outrunning systematic. His remedy looked backward, toward protected small property, and the Marxist tradition never adopted either his nostalgia or his policy. What it inherited from him, through the sheer priority of his insight, was the question itself: who buys the expanding product, and what guarantees that their means to buy expands with it?

Why is Malthus an awkward ancestor for a radical tradition?

Because the famous statement of the effective demand problem came from Thomas Robert Malthus, a cleric who defended landlords’ rents and held that unproductive consumption by a wealthy class was necessary to keep demand adequate to supply. A theory of crisis grounded in deficient demand thus descends partly from a thinker whose purpose was justifying an idle consuming class.

Malthus lost his famous controversy with Ricardo in their lifetimes, and the law of markets ruled the textbooks, but the question he pressed would not die. It resurfaced in the later nineteenth century in writers outside the Marxist camp, most systematically in J. A. Hobson, whose analysis of underconsumption traced gluts to oversaving by the rich and whose work on imperialism proposed overseas outlets for surplus capital and surplus goods. It resurfaced inside the camp in a transformed and formidable shape in the accumulation theory published in the early 1910s by Rosa Luxemburg, who pressed the realisation question harder than anyone before or since: if workers buy back only the equivalent of their wages, and capitalists consume only part of the surplus, who buys the remainder, the portion of the product embodying the surplus value destined for accumulation? Her answer, that capitalism requires non-capitalist markets outside itself and expands into them by imperialism, made the realisation problem the hinge of a whole theory of the system’s geographical and historical limits. The answer was disputed from the moment it appeared, by Marxists wielding the reproduction schemas no less than by outsiders, but the question she sharpened has never been satisfactorily buried.

The tradition’s third great form arrived with the monopoly capital school of the 1960s, above all in the work of Paul Baran and Paul Sweezy, who recast underconsumption as a problem of surplus absorption under conditions of monopoly. On their account the mature system generates a rising surplus that its normal outlets, capitalist consumption and investment, cannot absorb, so that waste in its many forms, military spending, the sales effort, and periodic state intervention become structural necessities rather than policy choices. That recasting belongs to its own treatment, the guide to monopoly capital stagnation theory, and is noticed here to complete the lineage: from Sismondi’s revenues and Malthus’s gluts, through Hobson’s oversaving and Luxemburg’s external markets, to the surplus absorption problem, the underconsumptionist idea has shown a capacity for self-revision that entitles it to be answered in its strongest forms. Marx’s tautology passage, aimed at the doctrine of his own century, is the beginning of that conversation and not its end.

The realisation problem stated at full strength

Beneath the quarrel of slogans sits a genuine technical difficulty, and a fair article must state it as sharply as its proponents do. Production under capitalism is production of value, and the surplus value extracted in production exists initially as a heap of commodities. For the circuit to continue, those commodities must be sold: the surplus must be realised in money, converted back into the means of further production and into revenue. The difficulty is arithmetical on its face. The wages paid out in production equal only part of the value produced, so the product cannot be bought back by the workers who made it. Capitalists’ personal consumption absorbs another part, but under any plausible assumption it does not absorb the whole of the remainder, since the drive of the system is to accumulate rather than to consume the surplus away. The residue, embodying the surplus value earmarked for expansion, must find buyers whose spending is not drawn from the value the circuit has so far distributed. Where does that demand come from?

The classical Marxist answer, built from the reproduction schemas discussed below, is that the demand comes from capital itself in its investing capacity: producers buying means of production from one another, expanding plant, hiring additional labour whose wages then support additional consumption. On this answer the system can, in principle, realise its surplus through its own growth, production creating the market for production in an ascending spiral, provided the proportions between branches are the right ones. Luxemburg judged the answer a swindle of accounting, a circulation of paper claims that postpones rather than solves the difficulty, and demanded to know for whom the ever-expanding product is ultimately produced if consumption is not its end. Her critics replied that she had mistaken a question about proportions for a question about an external buyer, that the schemas demonstrate the possibility of balanced expanded reproduction without any purchaser outside the system, and that possibility is all a theory of crisis needs, since crises then arise from the failure of the conditions, not from a logical bar on growth itself. The dispute has run for a century because both sides hold a real point. The schemas do prove possibility in principle. They do not prove that the conditions of balance are ever secured by a system of independent producers deciding in the dark, and the gap between possibility and assurance is where crisis theory lives.

Applied through the relative-to-what test, the realisation problem is an underconsumption claim with its denominator finally made precise: output measured against the demand generated within the circuit itself, where the suspect term is the demand for the accumulating residue. That precision is what elevates it above the crude doctrine. It does not say that workers consume too little in some moral sense. It says that the arithmetic of the circuit leaves a specific tranche of value hunting for a buyer, and it predicts trouble at the point of sale for that tranche, bunching in investment goods and in the industries furthest from final consumption. Whether the prediction survives contact with the schemas and with the historical record is a further question, and it is exactly the kind of question a serious theory should expose itself to. Vague underconsumption can never be wrong, which is why Marx called its simplest form a tautology. The realisation version can be wrong, in identifiable ways, and thereby earns the standing of a theory.

The reproduction schemas and what they settle

Marx’s answer to the realisation question, and the technical foundation for most later argument, is the pair of reproduction schemas in the second volume of Capital, examined in their own right in the guide to the marx reproduction schemas. The schemas divide social production into two great departments: one producing means of production, machinery, materials and plant, and one producing means of consumption, the goods that enter individual use. Each department’s output is analysed into the value of the means it used up, the value of the labour power it employed, and the surplus it generated. Simple reproduction, the abstract case in which the whole surplus is consumed and production repeats at the same scale, turns on a single elegant condition: the portion of the first department’s output that must be exchanged against consumption goods must equal the second department’s demand for means of production. When the equality holds, every commodity finds its buyer, every department replaces its inputs and feeds its people, and the circuit closes without residue. Expanded reproduction complicates the flows with accumulation in both departments, but the logical structure survives: there exist proportions at which the system buys its own product entire, surplus included, with no external market required.

What the schemas settle, then, is the question of principle, and they settle it against the strongest form of the realisation objection. A closed capitalist system is not arithmetically condemned to choke on its surplus; the demand for the accumulating residue is supplied by accumulation itself, by capitals in the first department selling to capitals in both departments that are expanding. Luxemburg’s demand for an outside buyer, as a logical necessity, does not survive the schemas, and most later Marxist economists, including many sympathetic to her politics, have judged that her proof fails at this point. The schemas also deliver, almost as a by-product, the analytical basis of disproportionality theory, since their central lesson is that balance is a set of exact proportions among magnitudes that no agency is charged with maintaining. The conditions of smooth reproduction are precise, multiple, and hostage to every harvest, invention and shift of taste, which is a very different picture from either classical harmony or underconsumptionist doom.

How can balanced departments still produce a breakdown?

Because the schemas state proportions that must hold, not a mechanism that holds them; the system reaches its proportions, if at all, through adjustments of price, profit and bankruptcy after the fact. Investment decisions rest with capitals acting on expectations, with long lived plant that cannot be redirected, so overshooting is discovered when output fails to exchange on required terms.

What the schemas do not settle is equally important, and the underconsumptionist tradition has always taken refuge in the unsettled remainder. Demonstration that balanced growth is possible says nothing about its probability, nothing about the role of money and credit in bridging or widening the gaps, nothing about the time structure of investment, in which plant ordered in optimism is delivered into pessimism, and nothing about the distribution of income between classes, which the schemas treat as a parameter rather than as a variable with its own dynamics. A critic can therefore accept the schemas entire and still maintain that the system’s demand problem is chronic, displaced into credit expansion, into the search for foreign markets, into state expenditure, rather than solved. That is, in substance, where the serious debate has stood for decades: not over whether realisation is possible, but over what the system must do, and what deformities it must acquire, in order to keep realising. The schemas converted a fake logical scandal into a real research programme, and both the defenders and the critics of Marx have worked inside it ever since.

There is a pedagogical moral here that the persona sections below will draw out, but it belongs to the argument as well. The schemas are the point at which this whole debate becomes technical, where slogans about gluts and shortages give way to flows that can be written down and checked. Students who meet the realisation problem only through polemical summaries inherit a quarrel of adjectives. Students who work through a two-department table with actual numbers discover that the quarrel has a skeleton of arithmetic, that each school can be located by the condition it expects to fail, and that the relative-to-what test, applied to the schemas, resolves into a question about which equality breaks first. No other part of Marx’s work rewards the effort of slow reading more directly.

Disproportionality: the third explanation

Between the claim that consumption is too small and the claim that capital is too large stands a third possibility, quieter than either and in some ways harder to refute: that nothing is too large or too small in aggregate, and the failure lies in the proportions among branches. Production of means of production has grown faster than production of consumption goods can support, or the reverse; raw materials lag behind the plant built to process them; one great industry, railways in the nineteenth century, construction or electronics in later ones, has drawn in a share of social labour that the rest of the structure cannot validate. The crisis then expresses an imbalance, and its remedy within the system is the forcible reallocation performed by devaluation, bankruptcy and unemployment in the swollen branches. This is disproportionality theory, associated within the Marxist tradition above all with the Russian economist Mikhail Tugan-Baranovsky, who built on the reproduction schemas the startling conclusion that capitalism could in principle expand with consumption stagnant or even shrinking, provided proportions were maintained, and that crises arise precisely when they are not.

Tugan-Baranovsky’s conclusion scandalised contemporaries because it severed crisis from the workers’ poverty altogether, making the whole drama a matter of allocation among departments, and later critics accused it of proving too much: a system that can grow by producing machines to make machines, with consumption as a dwindling sideshow, has purchased its arithmetic consistency at the price of describing an economy unmoored from any human purpose, and strains in demand, in credit and in the labour market would break its proportions long before the arithmetic did. Yet the challenge he posed remains standing. If proportions alone can generate crisis, then evidence of crisis is not evidence of underconsumption, and every appeal to unsold goods must first exclude the possibility that the goods were simply the wrong goods, produced in the wrong branch in the wrong volume. Disproportionality functions in this debate as the great alternative hypothesis, the explanation that must be eliminated before either of the grander theories can claim a case.

Disproportionality receives its denominator from the test without strain: each branch’s output is measured against the requirements of the other branches, as fixed by the reproduction proportions, and the excess is internal and relational rather than aggregate. That is also why disproportionality blends so readily with the overproduction of capital in concrete accounts of actual crises. A department that has overshot its proportion has, from the standpoint of the capitals trapped inside it, overproduced capital in exactly Marx’s sense: plant and stocks that cannot be valorised, because the complementary demand from other departments is not there. The theories are rivals at the level of first cause and partners at the level of mechanism, and sophisticated crisis writing moves between them accordingly, asking whether disproportion is the trigger, the transmission line, or the whole story. The one move the test forbids is the lazy one: citing the existence of disproportion, which is permanent, as if it were itself the explanation of a crisis, which is periodic. Like restricted consumption, disproportion is a standing condition of the system. What requires explanation is its periodic conversion into general breakdown, and that conversion usually runs through credit, through the bunching of investment in long-lived plant, and through the profit collapse that turns a branch’s embarrassment into a general refusal to lend, to order and to hire.

The three explanations compared

The argument so far can be gathered into the findable form that this series reserves for the point of maximum confusion. The three explanations compared, set side by side with their denominators exposed, show at a glance why the schools talk past one another: each row measures a different magnitude against a different standard, so evidence that confirms one row may leave the others untouched.

Explanation	What is in excess	Relative to what	Mechanism of crisis	Remedy the logic implies	What Marx said of it
Overaccumulation	Capital seeking profitable employment	The rate of profit its owners require	Valorisation fails, investment halts, capital is devalued	None internal; the crisis itself destroys the excess value	His own usage: overproduction means overproduction of capital
Underconsumption	Finished goods awaiting sale	The purchasing power of the mass of consumers	Realisation fails at the point of sale	Redistribution toward wage earners and consumers	Rejected as a pure tautology in its crude form
Disproportionality	Output of particular branches	The requirements of the other branches	Proportions break, and the break spreads through credit	Reallocation between branches by devaluation and bankruptcy	Built into the reproduction schemas as their standing risk

Read down the third column and the article’s discipline explains itself. Overaccumulation is a claim about capital and profit. Underconsumption is a claim about goods and purchasing power. Disproportionality is a claim about branches and their mutual requirements. A controversy in which one disputant produces evidence about profit rates and the other answers with evidence about wage shares is not a controversy at all until the columns are aligned, and a surprising share of the published literature consists of exactly that misalignment, conducted at high temperature. The table also exposes the asymmetry that the textual sections established: the rows are not three neutral options among which Marx chose impartially. He authored the first, supplied the apparatus from which the third is built, and attacked the second by name, while later Marxists rebuilt the second into stronger forms he did not live to answer. Any account that flattens that asymmetry, presenting underconsumption as simply Marx’s theory of crisis, has failed the most elementary test the texts impose.

A table, however, is a map and not the territory, and two warnings belong with this one. The first is that actual crises are compound events, in which an originating failure in one row propagates into the others within months: a valorisation failure halts investment, the halt throws workers out of employment, consumption falls, goods pile unsold, and the finished scene displays every row at once. Sequence, not surface, is where the theories differ, and sequence is what a table cannot show. The second warning is that the rows differ in their falsifiable content. Overaccumulation predicts that crisis gathers where profitability has been deteriorating under the boom’s surface. Underconsumption in its strong form predicts chronic pressure at the point of sale, relieved by identifiable external or state-supplied outlets. Disproportionality predicts sectoral origins that can be traced through input and output relations. A reader who knows what each row predicts can interrogate a narrative of any given crisis, asking which predictions the narrative confirms and which it quietly omits. That interrogation is the test in action.

Wages before crises: the evidence and its limits

Return to the sharpest empirical claim in the whole debate, Marx’s observation that crises are prepared by periods of generally rising wages, because its career illustrates how evidence functions in this literature. The observation was aimed at a doctrine of his time and it drew on the visible rhythm of the nineteenth century trade cycle, in which the boom’s later stages brought labour scarcity, successful wage demands, rising costs in the wage goods industries and a squeeze on margins, followed by collapse. Later scholarship has complicated the picture without dissolving it. Wage movements differ across countries and across cycles; some crises do follow periods of wage stagnation and deteriorating household balance sheets, a configuration the modern descendants of underconsumptionism are built to describe; and the profit squeeze literature of the 1970s revived Marx’s observation from the opposite political direction, treating rising wages not as the refutation of a crisis theory but as a crisis mechanism in their own right, through the compression of profits.

The honest summary is that the wages observation defeats the doctrine it was aimed at and does not defeat the tradition that doctrine belonged to. Crude underconsumption predicted crisis from a falling wage share; the recurring spectacle of crises detonating near the peak of wage gains was a genuine anomaly it could not process. The stronger realisation theories make no such prediction, locating the shortfall in the accumulating residue rather than in the wage bill, and they can absorb rising wages as comfortably as falling ones. What the episode teaches is the value of asking, for any crisis theory, which observations would embarrass it, and of noticing when a theory has been quietly rebuilt so that nothing can embarrass it any longer. A doctrine that accommodates every wage movement, every profit movement and every sectoral pattern with equal ease has drifted back toward the tautological condition Marx diagnosed, whatever its author’s intentions. The relative-to-what test, with its demand for a stated denominator and a stated mechanism, is the prophylactic against that drift.

There is also a matter of fairness in how the observation is quoted. In popular Marxist apologetics the rising wages point is sometimes deployed as a conversation stopper, as if it disposed of every argument connecting distribution to crisis. It does not, and Marx’s own system gives distribution a deep role in conditioning the system’s limits, since the rate of surplus value, which is a distributional magnitude, governs the mass of profit available for valorisation. Distribution enters his crisis theory through the profit side rather than the consumption side, through what capital extracts rather than through what workers can spend. Critics who wish to argue that inequality destabilises capitalism are therefore not refuted by the passage; they are invited to specify their channel, and if their channel runs through household debt, asset prices and the fragility of consumption financed by borrowing, they are describing mechanisms Marx did not analyse in a financial system unlike his own, which must be judged on their evidence rather than excommunicated by quotation.

Inequality, demand and the Keynesian neighbour

No question in this territory draws more traffic than the claim that inequality causes crises, and it deserves an answer with the denominators showing. In one sense the claim is a commonplace of the strong underconsumption tradition: a distribution that concentrates income in hands with a low propensity to spend it, while restricting the mass market, tilts the system toward a chronic gap between capacity and consumption. In another sense the claim belongs to a different theory altogether, the Keynesian theory of effective demand, in which deficient aggregate demand is a general possibility of a monetary economy, arising from the psychology of saving and investment decisions rather than from the class structure of production. The two theories share a policy neighbourhood, redistribution and public expenditure recommend themselves to both, and they are constantly conflated in political argument, but their mechanisms part company under the test. Underconsumption in the Marxist lineage is a claim about the circuit of capital and the realisation of surplus value. Effective demand failure in Keynes is a claim about the determinants of investment and liquidity preference, available to economies of any class description. The full comparison is reserved for the guide to marx vs keynes on crisis; the precision point here is that calling an inequality-and-demand argument Marxist does not make it so, and calling it Keynesian does not refute it.

Judged on its own evidence, the inequality channel has genuine cases in its favour and genuine difficulties. Its favour is strongest where a crisis has been preceded by stagnant mass incomes bridged by household borrowing, so that consumption was sustained by debt against a distributional shortfall, and the crisis arrived as the debt structure failed; the sequence fits the strong underconsumption template of a realisation problem postponed by credit. Its difficulties are those of timing and universality: inequality in many economies has risen through booms and crises alike, as a slow variable, and slow variables explain the terrain of crisis proneness better than the moment of outbreak; and crises have occurred in economies and eras of comparatively compressed distribution, which the theory must treat as exceptions requiring auxiliary mechanisms. The measured verdict is that inequality operates in this debate as a conditioning variable of the first importance and a trigger of the second rank, and that arguments promoting it to sole cause repeat, in modern dress, the error of the crude doctrine Marx attacked: promoting a permanent feature of the system into the explanation of its periodic accidents.

Is an inequality account of crisis automatically a Marxist one?

No, because the Marxist member of this family is defined by its mechanism, the realisation of surplus value within the circuit of capital, and an inequality argument earns the label only when it works through that mechanism. An argument running from unequal income to deficient spending to recession may be Keynesian, or generically demand side, without engaging anything distinctively Marxist.

The converse error deserves equal notice. Because Marx attacked underconsumptionism, some of his defenders treat every distributional argument as foreign matter, as if the master’s system assigned no role to the division of the product between classes. That is false to the system, in which the wage share and the profit share are the central distributional magnitudes and their ratio drives accumulation, the reserve of unemployed labour, and the mass of surplus available for investment. What separates the Marxist framework from its neighbours is not indifference to distribution but the location of distribution’s effects: on the supply of surplus and the requirements of valorisation, first of all, and on the composition of demand between departments, rather than on a simple aggregate of consumer spending. A reader equipped with that distinction can place any inequality argument on the map within a paragraph, identifying its mechanism and therefore its evidence, instead of sorting arguments by the political company they keep.

Credit: the mechanism that postpones and transforms

Every party to this debate needs credit, because credit is the device by which the system’s demand gaps are bridged in fact, whatever the theories say in principle. Credit lets investment proceed ahead of realised profit, lets consumption proceed ahead of wages, lets governments and foreigners and future incomes stand in the market as buyers of the accumulating residue. For the strong underconsumptionist, credit is the great postponement: the realisation problem is chronic, and the history of the system’s survival is the history of expedients, borrowing by households, by firms, by states, by whole nations, that manufacture the missing demand and convert a crisis of sale into a later crisis of debt. For the theorist of overaccumulation, credit is the accelerant that lets capital overshoot its valorisation limits furthest before the correction, inflating the excess that the crisis must then destroy. For the disproportionality account, credit is the transmission line along which a branch’s imbalance becomes a general stoppage, since the refusal of lenders to renew advances is the moment when private troubles synchronise into public collapse.

Where does credit enter the realisation story?

Credit enters as the standing substitute for demand the circuit has not distributed, advancing purchasing power from the future into the present so that goods embodying surplus value can be sold against incomes not yet earned. It becomes decisive when the volume of obligations grows faster than the profits available to service them.

Marx’s own treatment of credit is extensive and double-edged, and it resists recruitment to any single row of the table. He analysed the credit system as a form in which the reproduction process is both liberated from its immediate monetary limits and driven beyond its real ones, so that overproduction, in his sense, reaches its maximum extension under credit’s stimulus: production overshoots, speculation detaches prices from values, and the eventual crisis presents itself first as a monetary and banking panic, whatever its deeper composition. The surface of a crisis is therefore a poor guide to its theory. A banking collapse may crown an overaccumulation of capital, may unmask a postponed realisation failure, may transmit a disproportion, and in the wreckage the three are thoroughly mixed. The discipline the relative-to-what test imposes is to treat the financial surface as the last act and to ask, of any historical crisis, what had been accumulating beneath it: unsold goods against stagnant purchasing power, capital against inadequate profit, or branches against one another in broken proportion.

The strongest objection: semantics, and the sophisticated reply

Two objections to this article’s whole construction must now be heard at full strength, because the construction is worth nothing if it survives only against weak opposition. The first attacks the relative-to-what test itself. The test, the objection runs, is a semantic sieve that dissolves a real and massive phenomenon into a quibble about denominators. Workers thrown out of employment while factories stand stocked and shops stand full do not experience a relation or a comparison; they experience a system that produced too much and consumed too little, and any vocabulary that fits the scene that exactly has earned its place. To insist that the excess be specified is to substitute the pleasures of taxonomy for the work of explanation, and worse, it is not innocent: the specification habit systematically favours the overaccumulation row, because that row’s denominator, profitability, is the one capital itself uses, so the test smuggles the standpoint of the system’s managers into what advertises itself as neutral method. Ordinary language, on this objection, knew what it was doing when it fused the two words. The fusion records the unity of the event, in which overproduction and underconsumption are the same failure seen from the two ends of the circuit, and splitting them is the scholastic error.

The second objection defends underconsumption in the form its best proponents actually hold. Marx’s tautology passage, on this reading, hits a target of straw: no serious underconsumptionist claims that crises are caused by workers’ consumption being low in some absolute sense, and the sophisticated theory is untouched by the wages observation, because its claim is dynamic and relational. Productive capacity grows through accumulation at a pace set by competition, while the distribution of income confines mass consumption within limits set by the wage relation; the gap between capacity and consumption therefore widens structurally, and the system survives only by finding ever larger supplementary outlets, investment booms, luxury expenditure, armaments, foreign markets, consumer credit. Those outlets are real but self-undermining: investment enlarges the very capacity that needed absorbing, credit accumulates obligations against future income, foreign markets develop into competitors. Crisis, on this account, is the periodic exhaustion of the postponements, and it is entirely consistent with wages rising in booms, since rising wages are one of the boom’s own transitory supports. This version predicts not that crisis strikes when consumption falls, but that the system’s demand problem deepens across its history and its expedients grow more desperate, a prediction its proponents support with the growth of state expenditure, of the sales effort, of household debt, across the long twentieth century. Nothing in the tautology passage answers it. Marx died before its classic statements were written, and the passage’s authority cannot be stretched across a century of revised theory without turning scholarship into scripture.

Both objections score real points, and the reply begins with concessions. The tautology passage does not dispose of the sophisticated version; this article has said so from the start, and any use of the passage as a universal solvent of demand-side argument is an abuse of the text. The sophisticated theory is a genuine theory, with a mechanism, a dynamics and evidence that must be met on its merits, and where its predictions fit a case better than the rivals, the case belongs to it. The semantics objection also lands a partial hit: the circuit’s failure is indeed one event, and the test would deserve the scholastic charge if it claimed that naming denominators replaces the analysis of how the failure propagates. It claims nothing of the kind. What the test dissolves is not the phenomenon but a specific confusion, the inference from unsold goods to a consumption cause, and that inference is not ordinary language’s wisdom but a fallacy that ordinary language makes easy, exactly as it makes the confusion of heat and temperature easy. Experience of a crisis is evidence about the scene; explanation requires the order of events, and the order is precisely what the fusion of the words conceals. As for the charge that the test favours the profitability denominator: it favours whichever denominator the evidence supports in a given case, and it has been applied in these pages to convict Marx’s own side of denominator drift, in the apologetic use of the wages observation, as readily as anyone. A sieve that catches all parties is not a party weapon.

What the sophisticated underconsumptionist must then accept, in return for surviving the tautology passage, is the discipline of its own predictions. If the demand gap is structural and widening, the theory owes an account of the long stretches in which realisation proceeds without breakdown, and of crises that erupt at the height of investment booms with consumption buoyant and order books full, where the visible failure runs from profitability outward rather than from consumption inward. If the supplementary outlets are doing the explanatory work, the theory must identify them case by case and show their exhaustion, rather than invoking them as a universal reserve. And where the sophisticated version retreats, under pressure, to the claim that in the end everything is a realisation problem because everything must finally be sold, it has circled back to the tautology in modern dress, since a condition true of every moment of the system’s life cannot explain the moments at which it breaks. The strongest form of underconsumptionism stands. It stands as one row of the table, not as the table’s owner, and the controversies worth having are the ones that proceed row by row through the evidence of actual crises.

Reading the record: what the cases show

Theories of this generality are not settled by single cases, but cases discipline them, and a brief tour of the canonical episodes shows the test at work. The crisis of the mid 1820s, the first recognisably general commercial crisis of the industrial era, combined speculative investment in foreign loans and mining shares with an expansion of productive capacity in the cotton districts; its anatomy mixes rows, with overextended capital and broken proportions legible throughout and any consumption shortfall playing, at most, a supporting part. The mid century crises that Marx studied most closely followed the rhythm his wages observation describes, gathering through booms of railway construction and rising employment before breaking in monetary panic. The long deflationary decades after the crisis of the early 1870s displayed a different configuration, of capacity built for expanding markets meeting price falls and profit compression across heavy industry, an overaccumulation landscape in which the complaint of the age, voiced by businessmen rather than theorists, was of production outrunning profitable sale rather than of consumers too poor to buy. The great contraction after the crash at the end of the 1920s remains the most contested case of all: mass consumption had been sustained through the preceding boom partly by instalment credit and a speculative redistribution toward the wealthy, investment had overshot in construction and automobiles, and the collapse propagated through banking failure into a general refusal to spend, so that each school can find its mechanism in the wreckage and the discriminating work lies in the sequence of the first eighteen months.

What would count as evidence for each of the three explanations?

For underconsumption, a shortfall of mass purchasing power relative to installed capacity in the run up, with consumption weakening before investment fails. For overaccumulation, deteriorating profitability beneath a prosperous surface, with investment sustained by credit against shrinking returns. For disproportionality, a traceable sectoral origin whose overshoot transmits the disturbance through input relations and shared lenders.

Later episodes extend the test rather than retire it. The stagflationary breakdown of the 1970s fits none of the rows cleanly in its textbook form, and its literature quarrelled profit squeeze against monopoly pricing against external shock, a quarrel that the denominator discipline clarifies even where it does not close it. The financial crisis at the end of the first decade of the twenty first century gave the strong underconsumption template one of its best modern cases in its distributional chapters, stagnant household incomes bridged by mortgage credit until the bridge failed, while simultaneously displaying an overaccumulation of capital in housing and its financial superstructure so literal that the excess could be photographed in empty subdivisions. A framework that can hold both halves of that event without confusing them has justified itself. A framework that insists one half is an illusion has merely chosen its row in advance.

Misreadings this article corrects

Certain errors recur so reliably in coverage of this subject that naming them together has its own value. The first error is that overproduction means goods nobody needs, heaps of unwanted product in a world of satisfied people. Nothing in Marx corresponds to this picture; his overproduction is relative to profitability, and it coexists, as a matter of definition, with unmet need on a vast scale. The second error is that Marx was an underconsumptionist, that his theory of crisis reduces to the underpayment of workers. His explicit text says the opposite, in the plainest polemical language he used on any economic doctrine, and the attribution survives only because the picture is intuitive and the passage is unread. The third error treats the two words as two descriptions of one event, interchangeable labels for a slump. They are rival mechanisms with different predictions, and treating them as synonyms makes the whole subsequent literature, in which able people dispute real questions, unintelligible.

The fourth error holds that the realisation problem has been settled, in either direction, by the reproduction schemas. The schemas settle the possibility of balanced realisation and leave open everything about its likelihood, its props and its pathologies, which is where the live argument continues. The fifth error infers crisis directly from rising inequality on Marx’s account, skipping the mechanism. On his account distribution works on crisis through the mass and rate of profit, and any modern argument from inequality must build its own bridge, through debt, demand composition or asset markets, and be judged as the construction it is rather than as a quotation it is not. Behind all five errors stands the single habit this article exists to break: the use of either word without its denominator. Correct the habit and the errors lose their root system, since each of them feeds on the ambiguity the test removes.

The glut controversy in the classical schoolroom

The quarrel behind this article is two centuries old, and its classical phase repays attention because the positions were stated there with a nakedness later writers learned to disguise. Sismondi opened it in his New Principles of Political Economy of 1819, written in the distress that followed the long wars, when machinery was displacing hand workers across Europe and the displaced were visibly unable to buy the products of the machines that replaced them. His mechanism ran through revenue: production expands by competition’s compulsion, each producer racing to cheapen and enlarge output, while the revenue of the consuming population, above all the wages of workers separated from any property in the product, grows more slowly or shrinks. Glut is then not an accident but a rhythm, the recurring fate of a system in which production is guided by the rivalry of producers rather than by the known dimensions of consumption. Sismondi drew conservative remedies from the diagnosis, the protection of small producers and the slowing of machinery, and posterity has often discarded the diagnosis with the remedies. That is an error the relative-to-what test prevents, because his denominator is plainly declared: output against the revenue of those who must buy it, a first-measure claim in its earliest systematic dress.

Against Sismondi stood the doctrine of markets in its rigorous Ricardian form. Ricardo held that demand is limited only by production, that no man produces but with a view to consume or to sell, and that he sells with a view to buy, so that the circuit of exchange cannot jam in aggregate, however badly it may jam in particular branches. James Mill and then his son gave the doctrine its crispest statements, and it entered the textbooks as the refutation of all talk of general gluts. Yet the son, John Stuart Mill, in an essay on the influence of consumption on production written in the 1830s and published in 1844, conceded the crucial exception with a clarity the fathers lacked. During commercial crises, he allowed, there can be a real excess of all commodities, because money is not a mere counter: in a panic everyone wishes to postpone purchases and hold money itself, so that goods in general exceed the demand for goods in general, temporarily and violently. The concession is the thin end of the wedge that Marx would later drive home, since it admits the monetary separation of sale and purchase as the abstract form of the general glut, while still denying any permanent tendency to excess. The classicals’ own best mind had granted the phenomenon and confined it to the panic; the subsequent argument is entirely about what fills the interval between panic and underlying condition.

Why did Mill concede a temporary general excess of commodities?

Because his father’s doctrine assumed exchange against goods directly, and Mill saw that in a monetary economy a general desire to hold money and defer buying can leave all commodities in excess at once during a loss of confidence. The concession preserves the law of markets for settled times while admitting the universal glut of the crisis itself.

The schoolroom phase matters for a further reason: it fixed the polemical map on which Marx later worked. When he attacked underconsumption, he was attacking the Sismondian inheritance as it had passed into the socialist literature of his century, where it justified schemes of redistribution as the sovereign cure. When he defended the possibility of general crisis, he was building on ground the Mills had partly cleared, against the Ricardian veto. And when he constructed the reproduction schemas, he was answering the question the whole controversy had posed, how the product finds its purchasers, with an instrument nobody in the schoolroom possessed. A reader who knows the older quarrel can see Marx’s crisis writings as interventions at three distinct points of it, rather than as a single doctrine. That is the difference between knowing a thinker’s position and knowing the conversation the position belongs to, and it is the difference this series exists to teach.

One more classical voice deserves rescue from the margins, because it shows the denominator problem operating before any Marxist touched it. Malthus, in his Principles of 1820, argued against Ricardo that effectual demand could fail in general, that saving could outrun the profitable fields for investment, and that a class of unproductive consumers was therefore a functional necessity of a wealthy economy. His mechanism mixed several denominators in a way the test finds instructive: sometimes the excess is of goods against wants, sometimes of capital against the rate of profit that would induce its employment, and the argument gains plausibility by moving between them. Ricardo’s replies pressed the confusion hard, and on points of logic Ricardo usually won the letters while losing the century. Keynes would later salute Malthus as the ancestor who saw what Ricardo’s victory concealed. The Marxist tradition, which had its own reasons to fight Ricardo’s system, nonetheless inherited Ricardo’s side of this particular argument more often than it noticed, since Marx’s attack on underconsumption is an attack on the Malthusian mechanism in socialist dress. Lineage in this subject is never a guide to allegiance, which is why the article has insisted on mechanisms from its first page.

The plethora of capital: money, interest and idle balances

A form of overproduction that rarely reaches the headlines, because it has no warehouse to photograph, is the glut of money capital: periods in which loanable funds accumulate in the banking system faster than borrowers of substance will take them up, interest rates sag, and money lies idle in balances or chases speculative assets for want of productive employment. Marx treated this plethora of capital as the mirror image of the overproduction of productive capital, and the two appear together in his analysis of the cycle’s turning point. When valorisation falters, capitals withdraw from production and present themselves as money seeking placement; the same excess that stood as plant and stock now stands as idle deposits and depressed rates, and commentators who meet it in this form often fail to recognise the phenomenon they would readily name in its industrial dress. The denominator has not changed. What has changed is the costume, from commodities that cannot be sold at a profit to money that cannot be lent at an acceptable risk and return.

The coexistence of idle money and idle labour is the combination that most scandalises common sense, and it scandalised Marx’s contemporaries into some of their best questions. If funds are superabundant and workers are unemployed and needs are unmet, why does the system not join the three together? The answer the framework gives is austere: because the joining must pass through the form of profitable investment, and it is exactly that form which has failed. Money capital is not a social fund awaiting allocation to needs; it is private value awaiting valorisation, and where valorisation cannot be assured it will wait, or speculate, or emigrate, in perfect indifference to the needs across the street. The scandal is therefore real but diagnostic rather than explanatory in itself. It demonstrates that the barrier is the profit form, which the overaccumulation account asserts, more directly than it demonstrates any shortfall of consumption, since the idle money is itself unsold capacity to demand. A critic of the framework may reply that this merely restates the mystery in the system’s own vocabulary, and the reply has force wherever the framework’s defenders stop at the restatement. Pressed further, the account must show why valorisation failed, in costs, in markets, in proportions, which returns the argument to the rows of the table and their rival triggers, with the plethora of money capital registered as symptom and amplifier rather than as first cause.

There is a modern echo here that the article notes without adjudicating. Discussions of a savings glut among the world’s creditor economies, and of persistently low interest rates amid corporate cash piles, conduct in central bank vocabulary an argument whose structure the nineteenth century would recognise: accumulations seeking placement faster than placement can be found, pressing returns down and inflating the prices of existing assets. Whether those accumulations are best read as capital overproduced against profitable outlets, or as the counterpart of deficient demand elsewhere in the world system, is a live dispute among economists who mostly do not use this article’s vocabulary. Settling their dispute is not the test’s office. It identifies the dispute, showing that the two readings differ in denominator and therefore in remedy, the one pointing toward the destruction or socialisation of claims, the other toward the manufacture of demand. Readers who can perform that identification on an argument whose authors never mention Marx are using the instrument as it was meant to be used, on material its maker never saw.

Consumption inside the system Marx built

Because Marx attacked underconsumptionism, careless readers conclude that consumption plays a minor part in his economics, and the conclusion is worth a section of correction. Consumption in his system is not a sideshow; it is the destination of one whole department of production, the livelihood of the working class whose labour sets the rest in motion, and a magnitude whose restricted basis he names among the deep conditions of crisis. In the third volume he writes of the contradiction in terms of an antagonism between the conditions of production and the conditions of realisation: production expands as if the whole consuming power of society stood behind it, while that consuming power rests on foundations that hold the mass of producers to a fraction of the product. The passage is the closest Marx comes to the underconsumptionists’ territory, and its logic is worth precision. The restricted consumption of the masses figures here as the terrain on which the drama occurs, the permanent narrowness that production perpetually overshoots, and precisely because it is permanent it cannot serve as the trigger of the periodic outbreak. Terrain and trigger are different explanatory objects, and most misreadings of Marx on crisis are, at root, a confusion of the two.

The architecture of the reproduction schemas gives consumption its exact place. Workers’ consumption is the reflux that returns labour power to the market, renewed and ready for another period of employment, and in the schemas it appears as a flow determined by the wage bill, not as an independent force that might rescue or ruin the whole. Capitalists’ consumption, including the luxury production that Marx separates as a subdivision of the consumption goods department, absorbs a share of the surplus whose size varies with the phase of accumulation: in booms the share devoted to accumulation swells, in the aftermath of crisis the survivors feast on devalued assets, and the luxury trades feel both movements with a lag. Individual consumption also performs a quiet disciplinary function in the system, since the workers’ standard of life is at once a cost to be minimised by each employer and the substance of the demand that all employers collectively require, a contradiction at the level of the class that no individual capital can resolve. That formulation, class against individual, is Marx’s signature move in this territory, and it explains why he could hold simultaneously that consumption is structurally restricted and that its restriction explains crises only in the mouth of a tautologist. The restriction is the water in which the system swims. Crises are the moments it strikes its head, and what it strikes is the accumulated consequence of its own expansion.

A further refinement separates Marx’s treatment from every simple demand story. Consumption goods are themselves divided by destination, into necessaries that enter the reproduction of labour power and luxuries that do not, and the two subdivisions behave differently across the cycle, luxury output swinging more violently because its demand is concentrated in revenue that evaporates when profits fall. An analyst armed with that subdivision can read a crisis in the consumption industries themselves with more resolution than the aggregate word consumption allows, asking whether the failure sits in wage goods, where a fall registers employment already lost, or in luxuries, where it registers revenue already vanished. Sequence again: the subdivisions turn the consumption side of a crisis from a single blurred magnitude into an ordered set of consequences, and ordered consequences are what distinguish a mechanism from a mood.

Fixed capital and the calendar of breakdown

Why should breakdowns arrive in a rhythm at all, rather than as a steady drizzle of failures? Marx’s most concrete answer runs through fixed capital, the plant and machinery whose long life imposes a calendar on the system. A machine once installed will demand replacement in a run of years determined by its physical durability and by the faster moral depreciation that better machines impose, and because investment bunches, whole cohorts of plant approach renewal together. In the first volume he had already seized on this renewal cycle as the material basis for the decennial rhythm of the industrial cycle, the periodic foundation on which booms built and broke with a rough regularity his contemporaries could set their watches by. The connection to this article’s subject runs through the first department. Replacement demand for means of production is the steadiest large component of the market for Department I output, and its bunching means that department’s expansion and contraction are driven by a metronome built into the capital stock itself, largely indifferent to the state of consumption in any given year.

The metronome explains features of crisis that consumption based theories handle awkwardly. It explains why the capital goods industries, steel, machinery, construction, swing with an amplitude far beyond the industries that serve the consumer, since their market is other capitals’ decisions to expand, decisions that evaporate together when profitability clouds. It explains why a crisis can detonate while shops are still busy and wage packets still rising, because the first refusals occur in the ordering of plant whose delivery lies years ahead, invisible to any observer counting shoppers. And it explains the curious phenomenon of the recovery’s mechanics, in which the forced renewal of worn plant during the slump, purchased cheaply amid devaluation, lays the cost basis for the next expansion’s profitability. Marx’s cycle is thus a machine with moving parts, in which proportions, profitability and the physical calendar of capital interlock. Underconsumption, where it enters, enters as the narrow market into which the machine’s output must finally be sold, the condition that turns each mechanical oscillation into a crisis of the whole rather than a sectoral adjustment. Condition and mechanism, terrain and trigger: the distinction organises even the parts of the theory where the rival vocabulary seems most at home.

Later students of the cycle, working inside and outside the Marxist tradition, have repeatedly returned to this calendar with better statistics than Marx commanded. Studies of construction cycles, of equipment replacement bunching, of the clustering of innovations that call whole families of plant into simultaneous being, all elaborate the insight that capital’s physical form imposes periodicities no demand management abolishes. The elaborations do not vindicate every use Marx made of the idea, and the rough decennial rhythm belonged to a world of shorter lived equipment and smaller states. What survives is the structural point: a system whose productive apparatus is embodied in long lived, lumpy, jointly installed masses will generate endogenous waves of over and under investment, and at the crest of each wave there will exist, in the most literal sense, an overproduction of capital relative to the profits the wave can deliver. The calendar gives the overaccumulation row its clock, and a theory with a clock will always out predict a theory with only a level.

Luxemburg’s question pressed to its limit

Rosa Luxemburg’s intervention deserves a slower hearing than the lineage section allowed, because it is the point where underconsumption in its sophisticated dress came closest to capturing the Marxist tradition outright. Writing in the years before the First World War, she subjected the reproduction schemas of the second volume to a hostile audit and announced that they concealed the decisive question under their arithmetic. Expanded reproduction, as Marx modelled it, requires that part of the surplus be capitalised, embodied in additional means of production and additional wage goods for additional workers. But the model, she argued, simply assumes the additional demand those goods require; in a society of capitalists and workers alone, the demand for the growing surplus product has no visible buyer, since the workers’ wages are fixed by the value of labour power and the capitalists’ own consumption is the part of the surplus not destined for accumulation. The schemas show the proportions a solution would have to satisfy. They do not show the money, or the market. Her own solution reached outside the model: capitalism lives by trading with and devouring the non-capitalist strata and territories around it, peasants, artisans, colonies, and its history of expansion is the history of that feeding; when the outside is consumed, the system’s realisation problem becomes absolute and its end begins. With that argument she fused crisis theory to a theory of imperialism and gave underconsumption its most dramatic historical form, in which the denominator is nothing less than the unconsumed remainder of the pre-capitalist world.

The counterattack came from within the tradition and was led, after her death, by Nikolai Bukharin, whose critique accused her of misunderstanding the schemas’ purpose. The schemas, on his reading, are an analysis of proportions under abstraction, like the analysis of supply and demand in equilibrium: they no more need to name an external buyer than a proof that a see-saw can balance needs to name the children. Demand for the additional product is supplied by the expansion itself, since accumulation in one branch is simultaneously demand for the products of the others, and the money question dissolves into the credit and gold mechanisms the second volume handles elsewhere. The dispute was never merely technical. Beneath it stood two readings of what ails the system: Luxemburg’s, in which its limits are external and historical, a frontier that closes; and her critics’, in which its limits are internal and periodic, a valorisation that fails and is restored by destruction. The relative-to-what test locates the quarrel precisely: her denominator is the closed circuit’s own demand, indicted for insufficiency, while her critics answer that the circuit is not closed in that way, and that the insufficiency she measured was an artefact of holding one flow, consumption, in view while the investing flow ran offstage. Judged strictly as logic, the critics had the better of the exchange about the schemas. Judged as historical intuition, her insistence that the system reaches compulsively outward, and that the reaching shapes its politics, has worn better than the serenity of the see-saw. A balanced article lets both judgements stand, since they concern different objects: a proof, and a prognosis.

Her case also exposes a fault line in the strong underconsumption tradition that its heirs rarely acknowledge. If realisation requires an outside, then the theory’s crisis is not periodic but terminal and secular, a single long crisis of saturation approached asymptotically; yet the tradition also wants the theory to explain the ordinary periodic crises of the trade cycle, which arrive and depart while the outside still exists. The two assignments pull against each other, and her own writing oscillates between them, the accumulation book arguing the secular case while her political journalism addressed crises of the moment with disproportion and profitability arguments borrowed from the common stock. The oscillation is not hypocrisy but the mark of a real tension inside any theory that makes a structural gap do the work of a trigger. It is the same tension this article met in the wages passage, between condition and cause, and it recurs wherever a permanent feature of the system is asked to explain its periodic accidents. Naming the tension does not refute the tradition. It tells its defenders where their theory still owes work, which is the most useful thing a critical summary can do for any school.

How Marx became an underconsumptionist in the popular tradition

If Marx’s text is as plain as the tautology passage suggests, why does the attribution persist that he explained crises by the poverty of the workers? The reception history supplies the answer, and it begins with Engels. In his polemical and popular writings, above all in the work against Dühring, Engels presented the contradiction between social production and private appropriation as the system’s master key, and described crises as the revolt of the productive forces against the relations that confine them, with the restriction of mass consumption rhetorically prominent among the confining conditions. The presentation is not false to Marx, since the condition is genuinely Marx’s, but its compression for polemical use flattened the condition trigger distinction his partner’s manuscripts preserved. Popularisation then did what popularisation does: the Second International’s catechisms and party textbooks needed a crisis theory that could be taught in an evening, and the formula that production outruns the workers’ purchasing power taught beautifully, fitted the visible facts of every slump, and carried the correct political moral. Bernstein’s revisionism attacked the breakdown expectations built on such formulas; the orthodox defence of those expectations leaned harder on the formulas; and through the quarrel the simplified doctrine sank deeper into the movement’s common sense, acquiring the authority of the founders by repetition rather than by citation.

The twentieth century’s great popularisations completed the process. Textbooks of Marxist economics written for mass readerships, some of them works of real distinction within their limits, organised crisis theory around the realisation difficulty and the narrow market, because those were the themes the monopoly capital school was simultaneously raising to scholarly eminence. A student formed by such textbooks, meeting the tautology passage for the first time, experiences a small shock of disorientation, since the passage appears to attack the doctrine the textbooks taught as the master’s own. That shock is pedagogically valuable and this article is built to produce it under controlled conditions. It marks the exact point where the received tradition and the text part company, and a reader who has felt it will never again accept an attribution in this field without asking for chapter and verse. The guide to the marxism glossary key terms exists to fix the corrected usage, and the correction, once made, reorganises the student’s whole map: the master appears less as a prophet of demand failure and more as the analyst of a system whose expansion undermines the profitability that expansion requires.

Scholarly recovery of the textual position has been the work of the last half century, as the second volume’s manuscripts and the third volume’s drafts became available in critical editions and as historians of economics reconstructed the glut controversy behind the texts. The recovery has its own excesses, including a tendency in some quarters to treat the underconsumptionist tradition within Marxism as mere error, a lapse from an orthodoxy the master defined complete. That tendency should be resisted for the reasons the objection section gave: the tradition contains theories Marx never faced, answering questions his century had not yet posed in their modern form, and its best work stands or falls on evidence rather than on pedigree. The point of the reception history is not to police the inheritance but to explain a paradox of attribution, how the most famous sentence Marx wrote on consumption came to be the one his movement’s common sense forgot. Traditions remember what they can use. The tautology passage was unusable as a slogan, and so the tradition, needing a slogan, remembered something else and signed his name to it.

The destruction of goods: scandal, evidence and its limits

Among the images that fix this subject in the public mind, the most potent is the deliberate destruction of usable goods amid want: coffee burned as locomotive fuel, crops ploughed under, milk poured away, fish landed and dumped, while unemployment queues lengthened past the warehouses. The 1930s supplied the canonical instances on a grand scale, in the Americas and in Europe, under schemes designed to raise prices by restricting supply, and the images retain their power to outrage because they seem to strip the system to an essence: here, undeniably, is overproduction, and here are the needy it refused. Any honest treatment must grant the outrage its object. The destruction was real, the want was real, and the policy that joined them was a choice made within institutions that preferred scarcity prices to distribution. What the analysis must refuse is the inference the images invite, that they identify the mechanism of the crisis itself.

Does the destruction of goods prove overproduction relative to need?

It proves the opposite relation: goods were destroyed because they could not be sold at prices covering costs and sustaining the incomes built upon them, which is excess relative to solvent demand and profitability in a population whose need stood demonstrably unmet. Destruction is administered devaluation, undertaken when free sale threatens to ruin the producers the policy protects.

Read through the test, the destroyed coffee teaches a precise lesson. Brazil’s coffee in the years after the great crash was overproduced relative to the world market’s demand at the valorisation price the growers and their creditors required, and trivially underproduced relative to the world’s desire for coffee, which no one measured and no institution served. The state’s destruction programme was the overaccumulation crisis managed by hand, an attempt to destroy the excess capital embodied in the stocks before its free sale destroyed the capital of the growers instead. Analogous logic governs the quieter destructions of ordinary times, the fashion stock pulped, the previous model withdrawn, the harvest left unpicked when the price will not cover the picking. In each case the denominator is price and profit, and the act of destruction is the confession of it, since goods genuinely excessive relative to need would not need destroying; they would need only to be given away, and it is exactly the giving away that the price system, defending the value of the remaining stock, forbids. The scandal of destruction is thus the strongest popular evidence for this article’s central claim, once it is read with the denominators attached: a system that must burn food beside hunger is not suffering from abundance. It is suffering from the rule that goods may move only where profit calls them, which is a statement about the organisation of production, not about its volume.

Two cases worked through the test

Method is best verified on material, so consider two episodes in the detail the test demands. The first is the Lancashire cotton industry in the middle decades of the nineteenth century, the industry’s world workshop and Marx’s standing exhibit. Cotton manufacturing expanded under steam power at rates no final market could permanently validate; periodic crises, in the late 1830s, in the 1840s, above all in the overproduction crisis that preceded the American Civil War’s cotton famine, left the districts with unsold yarn and cloth, short time working and operatives on relief beside mills that could have clothed continents. Apply the questions in order. The excess, at the moment of breakdown, was of finished goods and of the fixed capital embodied in mills and machinery. The denominator divides the case: relative to need, in a world short of cheap clothing, there was no excess worth the name; relative to solvent demand, the glutted markets of India and the home districts took what their incomes allowed and no more; relative to profitability, the mills built in the speculative phase could not earn their keep at the prices the glutted market offered, and it is in this third register that the episode’s casualties were decided, by bankruptcy, consolidation and the devaluation of mill property into stronger hands. The trigger sits in the competitive expansion of capacity, each firm enlarging against the others under credit’s encouragement, which is the overaccumulation mechanism wearing industry dress; restricted working class consumption is present throughout as terrain, in Lancashire and in the export markets, and explains the narrowness of the ledge on which the expansion danced rather than the moment it fell.

The second case is housing in the first decade of the twenty first century in the United States and parts of Europe, chosen because its surface cries underconsumption and its depths answer otherwise. Construction boomed on credit extended to households whose incomes were stagnant, so that demand for houses was, for a time, manufactured by the lending itself; when the credit structure failed, the surface presented unsold and foreclosed houses beside households in acute housing need, the exact tableau of the naive picture. The test reads it differently. The excess was of capital embodied in residential structures and in the securities built upon their mortgages, accumulated beyond what the underlying incomes could valorise; the denominator of need was never met and never approached, as the foreclosure queues themselves attested; the denominator of solvent demand had been faked by the credit mechanism and collapsed when the fakery was recognised; and the denominator of profitability, returns on the mortgage assets at their marked values, was the trigger line along which the failure propagated into the banking system. A sophisticated underconsumptionist reading claims the case through the stagnant incomes, and the claim deserves its hearing, since without the income stagnation the credit bridge would not have been needed. The overaccumulation reading answers that the stagnation was decades old while the crisis was an event of specific years, dated by the overshoot of construction and leverage. The case therefore ends, as honest cases should, with the condition conceded to one row and the trigger to another, and with the relative-to-what test vindicated not by crowning a winner but by making the division of explanatory labour visible.

The afterlife of the veto: from Say to the supply side

The classical veto on general gluts did not die with the classical school; it changed its clothes and governs common sense still. In its marginalist dress it became the proposition that a rational economy tends toward the full employment of its resources, with unemployment and unsold capacity treated as frictions or as the wages of interference. In its journalistic dress it became the assurance, heard in every generation’s boom, that production cannot outrun consumption because production is consumption’s source, an assurance that somehow never prevents the crisis it declares impossible. Marx’s monetary point remains the standing reply to every version: where sale and purchase are separated by money, and where production is undertaken for profit by independent capitals, the balancing of aggregate supply and demand is an outcome to be achieved through prices that may ruin the producers, not an identity guaranteed by logic. The veto’s persistence is instructive about the sociology of doctrine. It survives because it is comforting to producers, flattering to policy and simple to state, and its refutation requires a distinction, between sale at some price and sale at a reproducing price, that takes a paragraph to explain. Doctrines are selected for their fitness in argument as much as for their truth, and the relative-to-what test is, among other things, an attempt to give the refutation a form as portable as the error.

A subtler descendant of the veto deserves separate notice, because it recruits the language of equilibrium to concede the phenomenon while denying the theory. On this view a general glut is a disequilibrium, a temporary mismatch between the price structure and the underlying data, corrected as prices adjust, and calling it overproduction mistakes a transition for a condition. There is a real distinction buried here, and Marx’s framework has its own version of it, since his crises too are corrections that restore proportions and profitability. The difference lies in what does the correcting and at whose expense. In the equilibrium story, correction is the market’s information system healing itself, and its costs are the tuition of adjustment. In Marx’s story, correction operates through the destruction of capital values, the bankruptcy of the weaker capitals, the devaluation of plant and the immiseration of the dismissed, and it restores not an optimum but the conditions for another round of the same accumulation that produced the breakdown. The two stories can agree about the sequence of prices in a given quarter and disagree about everything that matters: whether the recurrence is pathological or constitutional, and whether the restored state is health or merely the interval between fevers. A reader who sees that the quarrel is about the character of the correction, not about the existence of gluts, has found the level at which the classical controversy is still alive.

The world market: outlets, exports and the export of capital

No account of this subject is complete without the spatial dimension, because the search for outlets beyond the home market is where underconsumptionist intuition and historical fact most plausibly meet. Foreign trade relieves a glut at home in the most literal way, by finding the buyers elsewhere, and the nineteenth century’s great exporting economies visibly breathed through their foreign markets, with whole districts, Lancashire preeminently, producing for populations on other continents whose incomes they did not control. Theories of imperialism built on this observation their grandest structures: Hobson traced overseas expansion to the surplus capital and surplus goods that a maldistributed home income could not absorb, and Lenin, adapting Hobson, made the export of capital the signature of the monopoly stage, the vent by which the advanced economies exported their overaccumulation along with their goods. These theories gave the outlet idea a geopolitical body, and whatever their fate as complete accounts of imperialism, they record a genuine mechanism: expansion outward postpones the reckoning inward, and the reckoning, when the outside is organised into rival capitalist economies of its own, returns in the intensified form of competition for markets and the crises of a world system rather than a national one.

How does the export of capital postpone an overaccumulation crisis?

By relocating the excess to where returns remain high, since capital exported to economies with cheaper labour, untapped resources or scarce infrastructure can valorise itself at rates the saturated home market no longer offers. The postponement is genuine but self limiting, because exported capital builds the recipient economy into a new seat of accumulation with its own surplus.

Marx’s own remarks on foreign trade, scattered through the third volume’s discussion of the counteracting influences on the falling rate of profit, treat the world market with characteristic doubleness: trade cheapens the elements of production and expands the field of accumulation, raising the rate of profit, while simultaneously drawing the home economy into the crises of every market it touches and generalising overproduction to the scale of the world. He did not build from these remarks an outlet theory of crisis, and the difference matters for attribution. The outlet theories belong to Hobson and to the imperialism literature, standing in the underconsumptionist lineage by their mechanism, home demand insufficient, foreign fields required, and they should be credited there rather than retrofitted onto the founder. What Marx supplies to the discussion is the frame in which the outlet appears as a counteracting influence, real, effective and temporary, one entry in a ledger of forces that delay the tendencies they cannot abolish. The relative-to-what test, applied to the export of capital, asks its usual question and gets a revealing answer: the excess is of capital, the denominator is the home rate of profit, and the geography is the mechanism’s costume. An outlet theory that begins in unsold goods ends, on inspection, in the migration of capital toward valorisation, which is the overaccumulation row travelling under a passport.

The genuinely underconsumptionist residue in the world market story should not be argued away. Export markets for finished goods do fail for want of purchasing power among the populations they serve, colonial and postcolonial demand was chronically restricted by the poverty of the producers drawn into it, and the crises transmitted along trade routes often detonated in the weakest consuming regions first. The framework’s claim is not that such failures are imaginary but that they are instances of the general circuit failing at its points of least resistance, and that the systematic question, why the circuit keeps generating product and capital beyond what its markets can valorise, is answered at the centre of accumulation and not at the periphery of consumption. Centre and periphery, trigger and terrain: the same grammar, applied to the map.

A disproportion traced slowly

Abstract proportions persuade few readers, so follow one imbalance through its career in slow motion. Imagine, in the manner of the schemas, an economy in which the department producing means of production has been expanding rapidly on the strength of a new technique, say a generation of plant built to make the machines that make consumer goods. Its output must be bought by capitals in both departments that are themselves expanding; its prosperity is therefore a claim on the future growth of the whole structure. Now let the consumer goods department falter, not from any fall in the population’s needs, which are unchanged, but because the wages it pays and the revenues it generates no longer grow fast enough to validate the prices at which its enlarged output must sell, or because a shift of fashion or a failed harvest redirects what demand exists. Its capitals curtail orders for new equipment. The first department discovers, with a lag measured by its order books, that its market was a mirage of extrapolation; its own expansion halts mid gesture, with plant half built and specialised steel unsaleable to anyone else. Its workers are dismissed, and their lost wages strike the consumer goods department in a second wave, deepening the original faltering. Banks that financed both departments find their advances frozen in unsaleable product and half finished plant, and call loans from borrowers uninvolved in the original imbalance, spreading the stoppage to branches whose proportions were sound. At the end of the sequence the economy displays unsold goods everywhere, idle capacity everywhere and unmet needs everywhere, the full tableau from which the naive observer concludes underconsumption or overproduction in the lump. The slow motion shows instead a disproportion, kindled in the relation between two departments, propagated by the derived character of demand for means of production, amplified by credit, and resolved only when devaluation has written the swollen department’s capital down to the scale its partners can sustain.

The parable’s moral is methodological. Every magnitude in it behaved reasonably at the moment of decision; the catastrophe arose from the relations among magnitudes, each set in ignorance of the others. That is why the schemas, which are nothing but those relations written down, occupy the centre of Marx’s answer to the realisation question, and why the claim that the system lacks a buyer for its surplus, while suggestive as journalism, dissolves under analysis into claims about proportions, timing and profitability that can be separately examined. It is also why disproportionality, though it wins few headlines, keeps winning the technical argument as the necessary form in which deeper causes must express themselves. Whatever starts a crisis, a failing rate of profit or a genuine shortfall of consumption, it must pass through the departmental relations to become a general breakdown, because the economy has no other wiring. The wiring is not the current, but no current travels without it, and confusing the two is the characteristic error of single cause theories in both directions.

Crisis and its impostors: panics, harvests and shocks

A theory of crisis is not a theory of every bad year, and part of precision is knowing what the theory declines to explain. Economies suffer contractions from harvest failure, from war, from epidemic, from the sudden repricing of a dominant commodity, from policy error of monumental scale, and none of these requires any of the three mechanisms to be at work. Marx, who followed commercial history closely, knew that the immediate occasion of a crisis was frequently external to the accumulation process: the crisis of the late 1840s was entangled with harvest failures and the bullion drain they caused, and he treated such occasions the way a physician treats the draught that gives a weakened patient pneumonia, as the trigger of a condition prepared by the constitution. The distinction between the periodic crisis of the system and the accidental contraction is therefore internal to his method, and it protects the theory from the lazy falsification that points at any downturn with an external cause and declares the mechanism absent. Equally, it protects the reader from the theory’s imperialism, the temptation to conscript every downturn as evidence. The test that separates the genuine case is the presence of the prepared excess: capital accumulated beyond valorisation, proportions stretched, realisation dependent on postponement, waiting for its occasion. Where the excess is absent, a shock is only a shock, and its study belongs to economic history rather than to crisis theory.

Why do crises of overproduction so often announce themselves as banking panics?

Because credit is the connective tissue of the whole circuit, the point where every department’s promises to every other are concentrated and can be called in at once. When valorisation falters anywhere important, the refusal to renew credit synchronises the discovery everywhere, and the crisis surfaces as a monetary panic whatever its deeper composition among the three explanations.

The impostor problem has a mirror image in the financial crisis that is genuinely primary, where the excess accumulated inside the credit system itself, in claims layered upon claims whose underlying valorisation was fictitious from the start. Marx’s category of fictitious capital, developed in the third volume’s treatment of credit, covers precisely this: paper wealth that represents not capital but capitalised expectations, capable of expanding in serene independence from the production beneath it and of collapsing without any prior glut of goods at all. A crisis of fictitious capital fits the overaccumulation row with the denominator made explicit in an unusual way, since the excess is of claims relative to the profits that must service them, and it warns the analyst that the absence of warehouses full of goods proves nothing against an overproduction diagnosis. The warehouses of a credit crisis are the balance sheets, and they are fuller than any photograph can show.

Cousins, not twins: overaccumulation and the falling rate of profit

The overproduction of capital is often confused with the tendency of the rate of profit to fall, and the confusion is natural since both are denominated in profitability and both culminate in crisis. They are nonetheless distinct arguments, and conflating them muddles two debates that the literature has kept separate for good reason. The falling rate argument, treated fully in the guide to the tendency rate of profit to fall, is a claim about a secular tendency generated by the rising composition of capital, the growing mass of means of production set in motion by each worker, pressing down on the rate of return across the system’s history unless counteracting influences intervene. Overaccumulation is a claim about the cycle: whatever the secular trend, accumulation in the boom runs ahead of the valorisation available at its peak, and the crisis is the correction of that cyclical excess. An economy could exhibit cyclical overaccumulation with a stable long run profit rate, and a falling long run rate could in principle be accommodated without crisis if accumulation adjusted smoothly to it, though Marx’s framework gives reasons to doubt the smoothness. The distinction matters practically because the two arguments have different evidence and different critics: the falling rate stands or falls on long run data about composition and profitability, while overaccumulation stands or falls on the anatomy of booms and their turning points. A writer who merges them inherits the refutations aimed at both and the evidence proper to neither, a fate the relative-to-what test, with its insistence on one denominator and one mechanism at a time, is designed to prevent.

Within the couplet there is also a division of labour in explaining the boom’s peculiar psychology. Overaccumulation accounts for the moment when prosperity itself becomes the problem, when every indicator the participants watch, sales, orders, employment, prices, signals health while the foundation, the rate of return on the ever growing capital, quietly thins. The participants are not fools; they are reading flow variables in a system whose decisive variable is a ratio between a stock and a flow, capital and profit, and ratios deteriorate silently in a way flows do not. That is why crises surprise their victims with a regularity that mocks the idea of learning from experience, and why each generation’s assurance that the cycle has been mastered is followed by a rediscovery of the old anatomy under new instruments. The assurance usually rests on demand, on the observation that consumption is strong and credit available, which is to say it rests on exactly the variables that the overaccumulation account predicts will look strongest at the edge of the break.

Neighbouring terms sorted

Precision compounds when the near neighbours are sorted as well, since half the confusion in this field is smuggled in through adjacent vocabulary. Effective demand, the Keynesian term, names the level of demand at which entrepreneurs’ expectations of proceeds justify the employment they offer; it belongs to a theory of employment determination and should not be used as a synonym for consumption, since investment demand is half its substance. Aggregate demand, the national accounting term, is a statistical sum of spending categories, useful for measuring a shortfall after the fact and neutral among all theories of why the shortfall came. Surplus absorption, the monopoly capital school’s term, names the problem of finding outlets for a rising economic surplus and is the strong underconsumption position in its own vocabulary, to be engaged as such and not mistaken for a neutral description. Realisation names the conversion of commodity capital into money through sale, the moment at which surplus value becomes available for distribution and renewed investment; it is a stage in a circuit, not a synonym for selling in general. Valorisation names the expansion of value in production, the process for whose sake the circuit runs, and its failure is the overaccumulation denominator stated as a verb. A glut, finally, is the vernacular word for a localised excess of a commodity against its market, honest in its place and misleading only when promoted to a theory of the whole. A reader who keeps these six neighbours in their separate compartments will find that many celebrated disagreements shrink to their real size, and that the disagreements which survive the sorting are the ones worth the years they have consumed.

There is a corresponding discipline for the word crisis itself, which popular usage stretches to cover any difficulty with a headline. In the strict usage of this literature a crisis is a general and violent interruption of reproduction, in which the circuit’s phases, production, sale, renewed production, cease to mesh across the economy at once, and the restoration of meshing requires the destruction of values on a mass scale. Mere recession, sectoral decline, financial embarrassment and slow stagnation are neighbouring conditions with their own anatomies, and the precision matters because the three explanations compared in this article are theories of crisis in the strict sense. Stagnation, in particular, has its own literature and its own strongest theory in the surplus absorption account, and arguments that migrate between crisis and stagnation without signalling the move reproduce, at the level of the event, the same denominator drift this article has tracked at the level of the word.

Where Marx treats the question: a concordance in prose

Because attribution is this article’s recurring battleground, a reader deserves a map of the texts themselves, in the order a student might work through them. The discussion begins from the treatment of Malthus and Ricardo in the manuscripts on the history of economic theory, where the young science’s glut controversy is fought through page by page and Marx’s allegiances are formed: with Ricardo against the Malthusian theory of gluts on most points of logic, against Ricardo on the complacency of the law of markets, and against Sismondi on remedies while absorbing his question. The Grundrisse notebooks of the late 1850s then give the first full statement of the mature architecture, production driven beyond every given limit by capital’s nature, the market perpetually too narrow for the expansion it must absorb, and overproduction named as the moment the barrier is struck. The first volume contributes the abstract form, the separation of sale and purchase in the metamorphosis of commodities, and with it the bare possibility of general crisis, together with the analysis of the industrial cycle’s rhythm through the renewal of fixed capital. The second volume supplies the schemas and the tautology passage, the technical apparatus and the polemic, side by side in the work Marx left least finished and posterity read least carefully. The third volume completes the construction with the absolute overproduction of capital, the plethora of money capital, fictitious capital and the credit system, the materials from which the overaccumulation row of this article’s table is assembled. Read in that sequence, the texts show a mind returning to one problem across thirty years and refining its answer, from the schoolroom controversy of his student reading to an account in which consumption, proportions and profitability each have an assigned and limited office.

Where in Marx should a reader look for the fullest treatment of overproduction?

The fullest single discussion stands in the third volume of Capital, in the chapters on the law of the falling rate of profit and its internal contradictions, where the absolute overproduction of capital is defined and the plethora of money capital beside it. The indispensable companion is the second volume’s twentieth chapter, where the underconsumptionist rival is named and answered.

The concordance also corrects a lazy assumption about the unfinished state of the work. Because the second and third volumes were assembled by Engels from manuscripts of very different dates, readers sometimes treat their crisis materials as strata from incompatible periods, the schemas belonging to one Marx and the profit chapters to another. The manuscripts do span decades, and the critical editions let scholars date the layers precisely, but the architecture described above is stable across the layers: the possibility of crisis from the monetary form, the necessity of proportions from the schemas, the trigger in valorisation, the condition in restricted consumption. What changes across the decades is emphasis and evidence, not the assignment of offices. That stability is itself a finding about the theory, since it means the corrections this article has pressed, about denominators and about attribution, apply to the whole mature corpus and not to a cherry picked volume. A critic who would overturn the reading must overturn it in all five places at once, and a defender who would quote it must be prepared to be checked in all five.

Testing the rows: what the evidence would look like

Theories that cannot in principle be tested against records deserve suspicion, so it is worth specifying, row by row, the empirical signatures each explanation would leave in statistics a researcher can actually gather. Capacity utilisation is the most direct witness for overaccumulation in its physical dress: an economy entering crisis with utilisation rates that have peaked and begun to slip, after a period of heavy investment in plant, displays the anatomy of capacity outrunning its profitable employment, especially where the slip begins in the capital goods industries. Inventory behaviour tells a related story with finer grain: an involuntary accumulation of stocks, finished goods produced for a market that does not arrive, is the realisation failure made visible in warehouse ledgers, and its position in the sequence, before or after the investment turn, discriminates between consumption led and valorisation led accounts. Profitability data, the share and the rate of profit as national accountants and historians reconstruct them, test the overaccumulation row at its denominator: a crisis preceded by several years of thinning margins beneath booming sales is its signature case, while a crisis arriving with margins fat and consumption collapsing is the signature of its rivals. Wage and income distribution data test the underconsumption rows, but only when handled with the timing discipline the wages passage imposes: the question is never whether the wage share is low, which it nearly always is, but whether its movement, or the debt structure erected upon its stagnation, tightens measurably as the break approaches. Input and output tables, finally, give disproportionality its laboratory, since a genuine disproportion should be legible as specific sectoral mismatches, orders cancelled along identifiable chains, rather than as a uniform grey failure.

Handled together, these instruments convert the controversies of the schools into a research practice with agreed rules of evidence, and the best modern work on historical crises does exactly this, whatever banner its authors march under. The long running debates over the great contraction of the 1930s, over the postwar golden age and its ending, over the financial crisis of the early twenty first century, are at their most illuminating precisely where they behave as contests among rows, each side producing the signature its theory predicts and auditing the signatures its rivals claim. They are at their least illuminating where they behave as contests among attributions, what Marx really meant deployed as a substitute for what the data really show. This article’s insistence on the textual questions is not a preference for exegesis over evidence. It is the recognition that the textual questions decide which theory a given body of evidence is evidence for, since a profit rate series confirms or refutes the overaccumulation account only if overaccumulation is the account Marx’s denominator defines. Precision about words is the toll gate through which evidence must pass to reach the theories at all.

There remain obstacles that no refinement of method removes, and candour requires naming the largest. The relevant magnitudes are not directly observable: the rate of profit depends on the valuation of the capital stock, a constructed quantity over which competent accountants quarrel in different centuries’ prices; capacity is a judgement about sustainable intensity, not a physical constant; and the counterfactual against which any excess is measured, the output or investment that would have been validated, is by construction invisible. Every empirical verdict in this field is therefore an interval rather than a point, and the schools’ persistence owes something to that honest uncertainty. What method can do, and what this article has attempted, is to narrow the intervals by fixing the definitions, so that the residual disagreement concerns the world and not the words.

Open questions: the state, the form of crisis and the unfinished argument

Two developments since Marx’s century press on all three rows and keep the argument unfinished in identifiable places. The first is the growth of the state as an economic actor of the first magnitude, spender, employer, borrower and guarantor, whose operations supply exactly the kind of demand from outside the immediate circuit that the strong underconsumptionists once sought in colonies and armaments. Whether state expenditure represents the system’s discovery of a permanent internal outside, displacing its realisation problem into public debt and inflation, or merely a new set of proportions to be mismanaged, divides the modern schools along lines the table predicts: surplus absorption theorists claim the state as their mechanism institutionalised, while their critics point to crises that migrated into fiscal and currency forms as evidence that the underlying valorisation problem was postponed rather than solved. The second development is the corresponding migration of crisis form itself. Where the classical crisis announced itself in glutted warehouses and stopped mills, its successors have arrived as inflations, as debt deflations in asset markets, as currency collapses at the periphery, as long stagnations without a cathartic break. A framework that defines crisis by its surface will declare the old theories obsolete at each migration. A framework that defines it by mechanism, as this article has done, expects the surface to migrate, since the circuit’s failure will present at whichever joint the period’s institutions have made weakest, and asks of each new form the old questions about excess, denominator and trigger.

The unfinished business can be stated as three open problems, offered here as the honest edge of the subject. The world economy problem: the denominators of all three theories were framed for national economies or for an abstract closed system, while the actual circuit is planetary, with consumption, production and valorisation distributed across jurisdictions whose statistics and policies do not compose into a single set of books, so that an excess measured in one frame may be a shortage in another and the choice of frame is partly political. The measurement problem: the capital stock and its rate of return resist the precision the theories invite, and until historical national accounting matures further, several famous verdicts in this literature rest on series too fragile for the weight placed on them. The agency problem: each row implies a different bearer of remedy, redistribution, reallocation, or the supersession of the valorisation form itself, and the passage from mechanism to remedy crosses from analysis into politics, where the theories’ adherents dispute not what the machine does but what should be done about it. On that crossing this article, bound by the neutrality its series requires, takes no side. Its business has been with the machine’s description, and with the discipline that keeps descriptions honest: name the excess, state the denominator, show the trigger, and let every school, including the founder’s, submit to the same three questions.

Wages in their double character: cost to one, demand to all

A compact way to feel the whole debate in miniature is to follow the wage bill through its double character. To each individual employer, wages are a cost, and the rational employer presses them down without limit of principle; to employers as a class, the same wages are the largest single fund from which their products are bought, and pressing them down without limit would saw off the branch of demand on which the class sits. The contradiction is a fallacy of composition built into the system’s structure, and every participant in this debate uses it somewhere. The underconsumptionist builds the whole theory on it: the class cannot help restricting the consumption on which its own realisation depends, and crisis is the bill arriving. Marx accepts the contradiction completely and denies the inference built on it, because the class has other outlets for its product than the workers’ consumption, in its own investment and luxury above all, and because the wage bill’s movements across the cycle are governed by the labour market’s phases in ways that invert the simple story, rising as crisis approaches and collapsing only after it has arrived. Popular folklore supplies the contradiction’s favourite illustration in the story of the five dollar day, the doubling of wages at the Ford works in the second decade of the twentieth century, endlessly retold as the moment a far sighted capitalist grasped that his workers must be able to buy his cars. Historians treat the story’s motives as mixed at best, turnover and discipline prominent among them, and the economics of one firm’s wage bill are in any case a homeopathic dose of the class wide demand in question. The folklore survives because it dramatises the fallacy of composition in a single gesture, and the article’s discipline suggests the correct use of it: as an illustration of the contradiction’s existence, never as evidence for the mechanism that turns the contradiction into crisis.

The double character has a second edge that cuts the other way, against an easy politics of wage led recovery drawn from the underconsumptionist row. If wages are raised as a class wide matter, the demand for consumption goods rises, and so does the cost structure of every employer, compressing the profit share from which investment is funded; the net effect on the circuit depends on the balance between the realisation gained and the valorisation lost, a balance that varies with the phase of the cycle, the openness of the economy and the state of profitability at the start. Marx’s framework, which assigns wages their place in the reproduction flows rather than treating them as a policy dial, predicts exactly this ambiguity, and the historical record of wage led and profit led episodes in different economies has confirmed ambiguity often enough to embarrass both simple doctrines. Precision here, as everywhere in this subject, is not the enemy of commitment but its precondition: a politics that knows which denominator its remedy addresses can argue for it honestly, while a politics that promises demand without cost, or profit without demand, is trading in the fused vocabulary this article set out to split.

Horizons: cycle, stagnation and the long run

Many quarrels in this literature are secretly quarrels about the clock. The three explanations operate most naturally at different horizons, and a claim that is false of the cycle may be true of the epoch, and vice versa. Overaccumulation is a creature of the cycle’s turning point, a condition assembled in a handful of prosperous years and discharged in the break. Disproportionality belongs to the same horizon, since proportions are stretched and snapped within the same span. Underconsumption in its crude form failed precisely as a theory of this horizon, since its variable moves too slowly and in the wrong direction to date the break. In its strong form, however, underconsumption changed horizons and became a theory of the epoch: a chronic tendency toward stagnation, toward surplus unabsorbed and capacity underused across decades, punctuated by the expedients that postpone it. Judged as a theory of the epoch, its evidence is the long deflationary stretches, the reliance on armaments and credit creation, the secular sagging of investment opportunities that several generations of economists, by no means all Marxists, have diagnosed under different names. The founder’s tautology passage, aimed at a cyclical doctrine, does not reach the epochal one, and the epochal one’s defenders weaken their case when they borrow cyclical evidence, just as its critics weaken theirs when they answer an epochal claim with a cyclical text. To the three questions a fourth should therefore be added, at what horizon, and many apparently irreconcilable authorities reconcile themselves once their clocks are displayed side by side: Marx explaining the break of the cycle, the monopoly capital school explaining the drag of the epoch, each trespassing on the other’s ground only when their expositors grow careless.

The long run holds one further possibility that this article has deliberately left at the edge of its frame, the question of breakdown, whether the system’s crisis tendencies intensify across its history toward a limit at which recovery no longer follows. The founders’ texts support arguments in several directions on that question, and the tradition’s most famous quarrels, between revisionists who saw adaptation and orthodox who saw deepening contradiction, were fought over it with the crisis theories as their weapons. What can be said within this article’s discipline is that breakdown claims inherit the obligations of every other claim: a stated excess, a stated denominator, a stated mechanism, now joined to a stated horizon and a stated trend in the evidence. A breakdown asserted from the mere recurrence of crises is the tautology error at the largest scale, since recurrence is the phenomenon and not its intensification. A breakdown argued from measurable trends in profitability, in the scale of postponement required, or in the closing of outlets, is a theory entitled to its hearing. The distinction, once more, is not between loyalty and heresy. It is between claims that expose themselves to the record and claims that consume it, and on that distinction the whole of this long argument between overproduction and underconsumption finally rests.

Production for profit and production for use: the oldest framing

Beneath every technical version of this debate runs an older and simpler contrast, between production for profit and production for use, and the contrast explains the subject’s persistent popular energy. A system that produced for use, the intuition runs, could not suffer overproduction in any sense that mattered, since goods would be made as they were needed and stopped when needs were met; a system that produces for profit can make too much and too little at once, too much of what pays and too little of what is needed, and its crises are the periodic disclosure of that inversion. The intuition is not wrong, but it operates at a level above the mechanisms, and the article’s discipline applies to it as to everything else. Production for use names a destination, not a denominator, and the sentence that a profit system overproduces must still be completed before it can be tested: overproduces against the demand its own distribution creates, against the returns its own accumulation requires, or against the proportions its own departments impose. The oldest framing is the horizon within which the three explanations compete, the reason the competition matters beyond the seminar, rather than a fourth explanation that spares its holder the work of the other three. Writers who rest in the framing have located the moral of the subject and mistaken it for the analysis, and the tradition’s internal critics, from the schemas’ defenders onward, have always pushed past the moral to the mechanism, not because the moral is false but because it is inert until the mechanism tells it where to bite.

The contrast also guards against a misreading in the opposite direction, in which the technical apparatus of departments and rates is taken to have dissolved the human content of the theory into engineering. Marx’s own writing never permits that dissolution for long. The third volume’s great summary sentence on the barrier of capitalist production being capital itself gathers the whole analysis back into the contrast: a mode of production that develops the productive forces without limit, and collides periodically with the narrow limits of the valorisation that is its only permitted purpose, is a mode in which production for use is present as a permanent reproach inside production for profit. The reproach is the condition; the collisions are the crises; and the three explanations are the competing accounts of the collisions’ immediate mechanics. Held in that order, the subject’s levels compose into a single intelligible structure, and the reader can move between the moral framing and the technical rows without confusing them, which is as good a definition as this series can offer of what it means to understand a doctrine rather than to recite it.

Prices in the crisis: what a general fall testifies

Every general crisis speaks first in prices, and the testimony of prices is worth a hearing of its own, because a general fall of prices is the point at which the three explanations meet in a single observable fact and part company in its interpretation. For the classical tradition a general fall was the self correcting mechanism itself, the means by which glutted markets clear and proportions restore, painful to debtors and merchants caught mid contract but healthy in its destination. For Marx the fall was the devaluation of capital in its monetary form, the process by which values that could not be sustained are written down together, and its violence measured the scale of the excess accumulated in the boom. For the underconsumptionist a general fall testifies to the demand gap directly, prices dropping because buyers are absent. One fact, three readings, and the readings can be discriminated by what accompanies the fall: whether quantities contract before prices move, which is the signature of a demand failure arriving first, or whether margins and asset values collapse while physical trade still bustles, which is the signature of valorisation failing underneath a prosperous surface. Price data alone decide nothing, since prices are the circuit’s common language. Price data read in sequence, against orders, margins and credit conditions, decide a great deal, and the historiography of the great crises is at its best exactly there, in the patient reconstruction of which magnitude moved first.

Debt gives the price fall its peculiar cruelty, and the mechanism deserves statement because it connects the crisis of overproduction to the long deflations that have scarred whole decades. Obligations are fixed in money while incomes and asset values are not, so a general fall of prices raises the real burden of every debt contracted in the boom, transferring wealth from debtors to creditors at the moment debtors can least bear it and converting illiquidity into insolvency across whole classes of producers. Irving Fisher, writing out of the wreckage of the great contraction, formalised the spiral as debt deflation, over indebtedness liquidated into falling prices that swell the remaining debt, and his formalisation has been absorbed by traditions far from his own. Within the framework of this article the spiral is the overaccumulation crisis working itself out through the credit system: the excess capital of the boom survives into the slump as excess debt, claims on valorisation that valorisation cannot honour, and the price fall is the slow, socially ruinous form of the devaluation that a sharper crisis performs quickly. Read so, deflation is not a separate monetary accident that befalls a healthy real economy. It is one of the costumes in which the correction of overproduced capital appears when the correction is stretched across years by policy, by gold standard discipline in one era or by creditor protection in another, and the choice between sharp crisis and long deflation is a choice between tempos of the same reckoning.

The price testimony also disciplines a popular error about inflationary episodes, which are sometimes recruited as evidence that overproduction has been abolished by monetary management. Rising prices can mask a valorisation failure for a time, since nominal revenues swell while real returns stagnate, and the masking is one reason the crisis of the 1970s presented as stagflation rather than as glut: the excess capital of the postwar boom was devalued partly through the currency, in a slow confiscation of creditors and wage earners whose earnings lagged the price level, rather than through the classical sequence of panic and bankruptcy. Whether such masking represents a permanent escape from the old anatomy or a translation of it into a chronic form is among the live questions dividing the schools, and it is, at bottom, a denominator question once more. Measured in money of the day, the excess vanishes into the price level. Measured in real claims on profit, it persists through the inflation that conceals it, and the analyst’s task is to keep the real measure in view while the nominal one performs its vanishing trick on the front pages.

A working method for reading crisis claims

The practical yield of the argument can be stated as a method with a handful of moves, offered to any reader who meets these words in a textbook, a newspaper or a pamphlet. First, find the excess: goods, capacity, capital, or a branch’s output, and refuse to proceed until the claimant’s noun is identified. Second, find the denominator: solvent demand, profitable sale, the requirements of other branches, or human need, and watch for the moment, which comes in most popular arguments, when the denominator changes mid-paragraph. Third, ask for the trigger: the standing conditions are always present, restricted consumption among them, so an explanation of a dated event must name what changed. Fourth, ask what the claim predicts that its rivals do not, and whether the case at hand shows it. Fifth, check the attribution: when a doctrine is assigned to Marx, ask for the volume and chapter, since on this subject the famous attributions are wrong in both directions, with underconsumption assigned to him against his text and his own usage of overproduction quietly replaced by the warehouse picture he rejected.

Which denominator does a crisis headline usually conceal?

Profitability is the denominator most often hidden, because headlines prefer the visible arithmetic of goods and shoppers to the invisible arithmetic of returns on capital. A report of glutted markets and failing retailers invites the underconsumption completion by default, while the question whether the industry’s capital could ever have earned its required return goes unasked in the coverage.

For the reader who wishes to keep this method in working order, the instruments are modest: a notebook in which crisis claims are logged with their denominators filled in, a habit of returning to the second volume’s chapters on reproduction when the polemics heat up, and a willingness to let a case stay unassigned when the sequence of events is genuinely mixed. The study notes collected in the marxism study notes companion are built for exactly that kind of logging, since a claim recorded with its denominator exposed can be revisited when the evidence arrives. What the method promises is not certainty. It promises that disagreements will be about evidence and sequence rather than about words, which is the condition under which this old debate makes progress, and the condition its founder, for all his polemical ferocity, observed in the passage with which this article began its work.

For students and exam candidates

The mark carrying distinction in this topic is the relative-to-what test, and answers that deploy it early outperform answers that narrate the theories in sequence. When a question asks whether Marx explained crises by underconsumption, the high scoring response states the denial with its text, the second volume’s tautology passage and the rising wages observation, before conceding ground to the stronger realisation versions and locating them in later writers rather than in Marx. When a question asks what overproduction means, the strong answer refuses the warehouse picture, gives the denominator of profitability and the phrase overproduction of capital, and only then mentions unsold goods as the phenomenon’s surface. Examiners reward the sequence of the two departments in the schemas because it can be checked, so be able to reproduce the simple reproduction condition in words and to say what its failure would look like. Credit is the most commonly mishandled element: treat it as postponement and amplifier, never as a fourth theory. Finally, keep the attributions sorted under pressure. Sismondi and Malthus open the lineage, Luxemburg sharpens the realisation question, Tugan-Baranovsky builds disproportionality, Baran and Sweezy rebuild underconsumption as surplus absorption, and Marx stands against the crude doctrine that textbooks keep assigning to him. An answer that keeps those names in their places has done most of what the topic demands.

For teachers and seminar leaders

The pedagogical hazard in this topic is the intuitive picture, since students arrive with the warehouse and the empty purse already fused into a single explanation and experience the correction as pedantry unless the correction is staged as a discovery. Stage it with the tautology passage read aloud before any summary is given: let the seminar sit with the fact that the founder’s most quoted sentence on crisis attacks the doctrine most students were about to attribute to him, and let the discussion earn the distinction between condition and trigger out of that discomfort. A two department exercise with small invented numbers repays the hour it costs, because students who have made a schema balance, and then broken it by shifting one magnitude, stop asking whether realisation is possible and start asking the productive question, which condition fails first and who notices. Separate three things relentlessly in discussion: gluts of goods, which are events in markets; gluts of capital, which are events in valorisation; and the literary afterlife in which the two were confused by the tradition itself. Assessment should probe the separation directly, asking students to assign a described episode to a row of the comparison table and defend the denominator they used. The strongest seminars end with the sophisticated objection unanswered in part, the realisation problem standing as a live research question rather than a catechism with the replies printed in advance.

For researchers, historians, and journalists

For researchers, the reproduction schemas remain the technical entry point into every live question in this field, since the realisation dispute, the disproportionality tradition and the modern literature on sectoral imbalance all conduct their arguments in the schemas’ vocabulary of departments, proportions and flows. Work on the historical cases rewards the sequence discipline this article has pressed: date the turn in orders, in profitability, in credit and in consumption separately before assigning a case to a row, and treat single indicator narratives, however distinguished their source, as hypotheses about sequence rather than as findings. For historians, the reception story is a research field in its own right, the process by which a movement’s textbooks overwrote its founder’s text on a point this central, and the critical editions now make the overwriting traceable edition by edition. For journalists and analysts, the practical yield is a question that can be put to any crisis commentary, including the commentary of central bankers and fund managers who use none of this vocabulary: relative to what is the alleged excess measured, and what changed to make it bind now? Copy that answers the question can be argued with; copy that cannot is decoration. The naming of the denominator is the smallest unit of rigour in economic writing, and this subject, where the founder attacked his own followers’ favourite formula, is the standing proof that rigour begins at home.

Frequently Asked Questions

Q: What is the difference between overproduction and underconsumption?

The two terms name rival explanations of crisis that measure opposite sides of the market against different standards. Overproduction, in Marx’s usage, means that capital has accumulated beyond the opportunities to employ it at an adequate rate of profit, so the excess is of capital and its measure is valorisation. Underconsumption means that the purchasing power of the mass of consumers falls short of the output brought to market, so the shortfall is of demand and its measure is consumption relative to production. Because the denominators differ, the two are not descriptions of one event viewed from two ends; they imply different mechanisms, different sequences within a crisis, and different remedies, redistribution in the one case and the destruction or devaluation of excess capital in the other.

Q: Did Marx reject underconsumption theory?

Yes, in its crude form, and in writing that leaves little room for doubt. In the twentieth chapter of the second volume of Capital he argues that explaining crises by a scarcity of effective consumption explains nothing, since the unsold goods that are the phenomenon are simply renamed as their own cause, and he supports the charge with an observation: crises are regularly prepared by periods in which wages generally rise and workers receive a larger share of consumption. The rejection is aimed at the doctrine that workers’ restricted consumption triggers crises. It does not address the stronger realisation theories developed later by Luxemburg and the monopoly capital school, which were built to survive his objection and must be judged on their own evidence.

Q: What does overproduction mean in Marx?

In the passages that carry his theory, overproduction means overproduction of capital: an accumulation of productive capacity, of stocks and of money seeking placement beyond what can be set to work at the rate of profit capital requires. A portion of the accumulated value must then stand idle, accept a reduced return, or be destroyed through devaluation and bankruptcy, and the crisis is the process that accomplishes that correction. The term does not mean that society has produced more goods than its members need; in Marx’s analysis the crisis unfolds amid unmet need, and the unsold goods are the surface appearance of a failure of valorisation underneath. His fullest statement is the discussion of the absolute overproduction of capital in the third volume, where additional capital produces no additional profit and the excess is competed away.

Q: What is the realisation problem?

The realisation problem is the question whether the surplus value produced in a capitalist economy can be converted into money through sale, given who holds the purchasing power. Wages cover only part of the product’s value, and capitalists do not personally consume the whole surplus, so the portion of output embodying surplus destined for accumulation appears to lack a buyer. Rosa Luxemburg pressed this question hardest, arguing that a closed capitalist system cannot solve it and must feed on markets outside itself. The standard reply, grounded in Marx’s reproduction schemas, is that accumulation itself supplies the demand, as expanding capitals buy means of production and hire additional workers, so realisation is possible in principle when departmental proportions hold. The dispute therefore concerns not arithmetic possibility but whether the system’s actual arrangements secure the conditions, and it remains the technical core of the underconsumption debate.

Q: What is disproportionality theory?

Disproportionality theory explains crisis by imbalance among the branches of production rather than by any aggregate excess or shortfall. Social production must divide itself between the department making means of production and the department making consumption goods, and between industries within each, in proportions that allow every output to be bought as an input or as consumption. Since those proportions are set by thousands of independent investment decisions taken in ignorance of one another, they are periodically violated: one branch overshoots, its output cannot be exchanged on the required terms, and the failure propagates through orders, employment and credit into a general breakdown. The theory builds directly on Marx’s reproduction schemas and was developed most boldly by Tugan-Baranovsky, who held that proportions, not consumption, are the system’s weak point. It functions in the debate as the standing alternative that any underconsumption or overaccumulation claim must first exclude.

Q: Do wages rise before crises?

In the pattern Marx observed and cited against the underconsumptionists, yes: crises are typically prepared by a phase of the boom in which employment is plentiful, labour is scarce relative to demand, and wages generally rise, giving workers a greater share of the annual consumption product than in quieter times. The observation matters because crude underconsumption theory requires the opposite sequence, a squeezing of workers’ purchasing power as the trigger of collapse. Later evidence complicates the pattern without restoring the crude doctrine: wage behaviour differs across countries and cycles, and some downturns do follow wage stagnation financed by household debt. The sophisticated realisation theories are constructed so as not to depend on the wage sequence at all, locating the shortfall in the demand for the accumulating surplus rather than in the wage bill.

Q: Is inequality a cause of economic crisis?

It can be a contributing condition, but the answer depends on the channel specified, and the channel decides which theory the claim belongs to. Through the underconsumption channel, concentrated income restricts the mass market while capacity expands, leaving realisation dependent on credit, investment booms or state spending; the modern version points to crises preceded by stagnant household incomes bridged by borrowing. Through Marx’s own framework, distribution acts mainly on the profit side, since the division between wages and surplus governs the mass of value available for valorisation. What the evidence does not support is the direct inference from rising inequality to crisis without a mechanism: inequality is a slow variable that rises through booms and slumps alike, so it explains crisis proneness better than the timing of any particular outbreak, which requires a trigger in profitability, proportions or credit.

Q: Who were the main underconsumptionists?

The lineage begins with Sismondi, who argued after the Napoleonic wars that machine production expands faster than the revenue of the consumers who must buy its product, and, from a very different standpoint, with Malthus, who held that deficient effectual demand could produce general gluts and defended unproductive consumption as a remedy. In the later nineteenth century J. A. Hobson traced gluts to oversaving by the rich and connected the surplus to imperial expansion. Within the Marxist tradition the decisive figure is Rosa Luxemburg, whose accumulation theory made the realisation of surplus value dependent on non-capitalist markets, and the tradition’s modern form is the monopoly capital school of Paul Baran and Paul Sweezy, who recast the doctrine as a chronic problem of absorbing a rising economic surplus.

Q: Does overproduction mean too many goods for people’s needs?

No, and the distinction is the entry point to the whole subject. Needs that nobody can pay to satisfy do not register in a market economy as demand, so goods can be unsold, and producers ruined, in the presence of intense unmet need for the very articles standing in the warehouses. When Marx wrote of overproduction he meant an excess measured against profitable sale, and in its deepest form an excess of capital against the rate of profit its employment requires. The spectacle of destruction of food beside hunger, a recurring feature of great crises, is the clearest demonstration: goods destroyed because they cannot be sold at sustaining prices are excessive relative to purchasing power and profitability, and by the standard of need they are not excessive at all.

Q: Is underconsumption the same as a demand shortfall?

Only loosely, and the looseness hides the theory. A demand shortfall is a description: spending in some period falls below output or below the level that keeps resources employed, and every crisis theory, including those that reject underconsumption, acknowledges such shortfalls as part of the event. Underconsumption is a specific explanation of the shortfall, locating its cause in the restricted consumption of the mass of the population relative to expanding production, and in its strong forms in the difficulty of realising the surplus destined for accumulation. Keynesian effective demand failure is a different explanation again, running through investment decisions and liquidity preference rather than through the class distribution of consumption. Calling every shortfall underconsumption empties the term of its mechanism and reproduces, in miniature, the tautology Marx complained of: the failure to sell is renamed as its own cause.

Q: How does the relative-to-what test sort an overproduction claim?

It asks three questions of the claim in sequence and sorts it by the answers. First, what is said to be in excess: finished goods, installed capacity, a branch’s output, or capital itself? Second, against what standard is the excess measured: the demand backed by money at ruling prices, the sale that yields the required rate of profit, the requirements of other branches, or human need? Third, by what mechanism does that measured gap produce a breakdown, and in what order do sales, profits and orders fail? A claim about goods against purchasing power sorts as underconsumption; a claim about capital against profitability sorts as overaccumulation; a claim about branches against each other sorts as disproportionality. Claims that change standards midway, or answer the third question by pointing at the crisis itself, fail the test whichever school utters them.

Q: What separates underconsumption from effective demand in this debate?

The mechanism and the location of the failure. Underconsumption, in the Marxist lineage, is a claim about the circuit of capital: the value produced includes a surplus that must be realised through sale, and the distribution of income between classes confines the consumption fund in ways that make that realisation chronically precarious. Effective demand, in Keynes’s construction, is a claim about the determinants of employment in a monetary economy: investment, driven by expectations and the rate of interest, and the propensity to consume together fix the level at which the economy settles, which may be below full employment without any reference to surplus value or class distribution. The two frameworks share policy conclusions about redistribution and public spending, which is why they are conflated in political argument, but a claim is sorted by the machinery it runs on, not by the remedies it recommends.

Q: Why did Marx reject crude underconsumptionism?

For two connected reasons, one logical and one empirical. The logical reason is that the doctrine, as commonly stated, is circular: it explains the failure to sell goods by a scarcity of consumption, when the scarcity is inferred from the failure itself, so the crisis is redescribed rather than explained, and a permanent feature of the system, the restriction of workers’ consumption, is offered as the cause of a periodic event. The empirical reason is the wage sequence: crises characteristically arrive after periods in which wages have generally risen and workers’ share of consumption has grown, the reverse of the sequence the doctrine requires. Marx did not deny that mass consumption is restricted or that the restriction conditions the whole system; he denied that the restriction, being constant and moving perversely, can serve as the trigger that dates a crisis.

Q: How do department schemes show a disproportionality crisis?

The reproduction schemas divide production into a department making means of production and a department making means of consumption, and they specify the proportions in which the two must exchange for reproduction to proceed smoothly: the first department’s demand for consumption goods must match the second department’s demand for equipment and materials, with matching conditions inside each department when accumulation is added. A disproportionality crisis appears in the scheme as the violation of those equalities: if the first department expands beyond what the second can absorb and equip, its output cannot be exchanged, investment halts, and the failure transmits to employment and consumption in the other department. Because no agency enforces the equalities in an actual economy, which reaches them only through price movements and bankruptcies after the fact, the schemes display both the possibility of balanced growth and the standing mechanism of its breakdown.

Q: What should a teacher separate between gluts of goods and gluts of capital?

A glut of goods is a market event: a stock of finished commodities that cannot be sold at the prices their producers require, cleared by price falls, losses and the writing down of the capital embodied in it. A glut of capital is a valorisation event: an accumulated mass of value, whether embodied in plant, in stocks or in money balances seeking placement, that cannot find employment at the rate of profit its owners require, so that part of it stands idle or loses value outright. The two regularly appear together in a crisis, the second usually preceding and causing the first, but they answer to different evidence: inventories and price cutting for the one, margins, investment intentions and the movement of interest rates for the other. Students who can state which glut a given passage or headline describes, and what it is measured against, have mastered the distinction this topic exists to teach.