Every other thinker in this book is a contribution. Marx is a critique. The chapter walks the economic apparatus, the social-historical theory, the predictive program that did not land, the three live readings of what the project was actually for, and what survived.
That asymmetry is why a section in chapter 6 would not have done. A critique that worked through the classical school’s own vocabulary, took its categories seriously, and then turned them against the school’s conclusions cannot be summarized as one more position inside the lineage; it is the lineage interrogating itself.
Karl Marx (1818–1883) is the only figure in this book whose project, as he understood it, was not to add a theorem to the discipline he found but to expose what the discipline could not see about its own subject. He worked through classical political economy from the inside, reading Smith, Ricardo, Mill, Quesnay closely, taking their categories seriously, often defending them against their later critics, and then turned those categories against the conclusions the school had drawn. He treated capital as a historical formation that produces specific kinds of social relations and then takes those relations as natural, eternal, given. And he proposed that economic doctrines themselves arise from the conditions in which they are written, so that the discipline's blind spots are not accidents of method but symptoms of where it stands.
Three intertwined operations, then. The first is immanent critique — the method of using a position's own categories and premises against its conclusions, rather than imposing an external standard. Marx does not refute the labor theory of value; he pushes it through to consequences Smith and Ricardo did not draw. The second is the analysis of capital as a self-undermining historical system: not a stable allocative mechanism but a configuration that, by its own logic of accumulation, periodically negates the conditions that made it function. The third is historical materialism as method: the claim that the economic structure of a society conditions its legal, political, and ideological forms, and that doctrines about the economy are not exempt from this conditioning. The three operations are not separable. The critique of classical political economy needs the historical-system analysis to explain why the classicals' static categories deformed what they tried to describe; the historical-system analysis needs historical materialism to explain why the school could not see this from where it stood.
The working position is stated upfront so the reader knows what the rest of the prose is arguing. Marx's economics failed as prediction. Capitalism did not collapse from internal contradictions; the reserve army was domesticated by welfare-state institutions and union density rather than eliminated by revolutionary action; the rate of profit's empirical trajectory has not behaved as the falling-rate theorem requires. Those failures are specific and the chapter walks each. But the diagnostic vocabulary — alienation, commodity fetishism, exploitation as a structural relation rather than a moral charge, the analytical category of class as a relation rather than an income stratum — has more analytical life than its textbook treatment recognizes, and economists who reject Marx politically use these categories without the labels. Failed prediction, productive diagnostics. The chapter walks both halves at depth and takes a position on what the split means.
That position is not the only one available. The Marxian revival, which has been visible in waves since the 1960s and acutely after 2008, divides into three readings of what Marx's project actually was. Analytical Marxism (Roemer, Cohen, Elster, Wright) ran the claims through neoclassical and game-theoretic machinery, accepted that the labor theory of value as a price theory was wrong, and reformulated exploitation as differential property rights over productive assets to preserve the explanatory core on rebuilt foundations. Mainstream dismissal, the academic discipline's revealed verdict — holds that the labor theory is an analytical dead end, the falling-rate theorem is mathematically defective per Okishio, the predictions failed, and what is usable in Marx has been absorbed without the framework. The new reading (Heinrich, Backhaus, Reichelt, Postone, working out of the Neue Marx-Lektüre developed in West Germany from the 1960s) argues that the standard interpretation misread Capital by treating the labor theory as a model of how prices form rather than as a critical analysis of the value form itself. On this reading, the empirical failure of the labor theory as a price theory is irrelevant because it was never a price theory. Section 7.5 walks the three at strongest form and takes the chapter's call.
A note on usage. Marxian in this chapter names the analytical tradition that takes Marx's categories as objects of intellectual work, including writers who reject Marx politically. Marxist names a political-doctrinal commitment with party-historical content. The distinction matters because much of what survives Marx analytically is held by Marxians who are not Marxists, and a chapter that conflates the two cannot describe what survived.
Chapter 3 treats Marx as the terminal figure of classical political economy: the point where the school reaches its self-critique and its limit. That treatment is correct as far as it goes. This chapter takes the critique outward: not Marx the school's endpoint, but Marx the inaugurator of a body of analysis that exits the school. The value-theoretic lineage that brought Marx to socially necessary labor time is walked in §2 below and runs through the book's value-lineage walkthrough. What Marx built on top of the labor theory is where this chapter begins its substantive work. Where Marx sits relationally is on the marx node of the intellectual-history timeline.
The value-lineage walkthrough traces the descent that brought Marx to socially necessary labor time (Smith's toil-and-trouble, Ricardo's labor embodied, Marx's adjustment for prevailing technique). That genealogy does not need redoing here. The reader who has not followed it needs only the result: under Marx's account, the value of a commodity is the labor time required to produce it under the technique normal for the society and at the average intensity prevailing in that society, and exchange-value (the ratio at which one commodity exchanges for another) tracks this socially necessary labor time. What Marx built on top of that foundation is what this section walks.
The first move is the surplus value category. The worker sells not labor but labor power, the capacity to work for a defined period. Labor power is a commodity like any other; its value is the labor time socially necessary to reproduce it (food, shelter, clothing, training, the costs of raising the next generation of workers). The worker is paid the value of labor power, in the form of wages, and the contract is honored. But labor power, when used, produces value in excess of its own cost. The capitalist who buys labor power for a day owns whatever the worker produces during that day, and what the worker produces is worth more than what the worker is paid. The difference — the value labor produces minus the value labor receives as wages — is surplus value. It is the form in which the social product is appropriated by owners of productive assets, and it is the category that turns the labor theory from a price theory into an account of how the social product is distributed.
The second move is the structural reading of exploitation. The category, in Marx's technical sense, is not a moral charge that workers are paid too little. It is a structural property of the relation of production: under capitalism, surplus value is appropriated by those who own productive assets, and the relation holds whether wages are at subsistence or substantially above. A unionized worker in a high-wage industry, paid five times the cost of reproduction, stands in the same structural relation to the firm's owners as a sweatshop worker paid at subsistence. The two cases differ in distribution; they are identical in form. The point is the relation, not the wage level. Conflating the two collapses Marx's category into a moral complaint that wage rises can dissolve, which is precisely what wage rises did not do to the structural relation through the second half of the twentieth century.
Stylize a working day as a way to see the category in operation. A worker enters the factory at eight in the morning. By noon, the worker has produced enough output to cover the value of labor power expended that day: the food, shelter, and replacement costs that wages will pay for. From noon onward, the worker continues producing, and what is produced now is surplus. The structure does not change if the wage is two times subsistence rather than equal to it; what changes is when noon arrives. With high wages, the worker covers the value of labor power by ten in the morning rather than noon, and the surplus-producing portion of the day is six hours rather than four. The structural relation between paid time and unpaid time does not vanish; it shifts. The category survives wage compression that the moral reading of exploitation does not survive.
With surplus value pinned, the apparatus's accounting categories follow. Constant capital (denoted c) is the value invested in means of production: machinery, raw materials, factory buildings, the inputs that transfer their value to the product without creating new value. Constant capital is conserved, not augmented. Variable capital (denoted v) is the value invested in labor power: the wage bill. It is variable because labor power, when used, creates new value rather than transferring stored value. The total value of a commodity decomposes into three components: c (the value of means of production used up), v (the wages paid), and s (surplus value). The bookkeeping is c + v + s. The labor theory provides the substance behind this; the decomposition is what the labor theory becomes once surplus value is named.
The organic composition of capital is the ratio of constant to variable capital, c/v, in value terms. As capitalist development proceeds (mechanization, larger plants, more dead labor per unit of living labor), this ratio rises. Marx's claim was secular: across the long run of capitalist accumulation, the organic composition rises because each round of competition rewards firms that substitute machinery for labor, and the firms that do not follow are eliminated. This is the empirical premise of the falling-rate theorem.
The Marxian rate of profit is surplus value divided by total capital advanced: s/(c + v). It differs from the marginalist rate of return on investment in two respects. The denominator includes constant capital as the basis on which capitalists compute return; the numerator is computed in value terms over the full circuit of capital, not as a one-period yield on a financial position. The rate of exploitation, by contrast, is s/v: surplus over the wage bill, an internal feature of the production process rather than a return on investment. The two ratios behave differently as the organic composition rises.
From this geometry, the falling-rate-of-profit theorem follows. If the rate of exploitation s/v is roughly constant and the organic composition c/v rises, then s/(c + v) falls. The intuition: a capitalist who invests more constant capital per unit of labor is multiplying the denominator faster than the numerator can grow, because surplus value comes from labor and not from machinery. Marx's tendency of the rate of profit to fall (usually abbreviated TRPF) is the claim that this geometry obtains under the secular rise of the organic composition. He named it a tendency, not a law: he listed counter-tendencies that work in the opposite direction. Raising the rate of exploitation through longer hours or intensified work increases s/v and offsets the rising c/v. Cheapening the elements of constant capital, through productivity gains in capital-goods sectors, slows the denominator's growth. Foreign trade can supply inputs at lower cost than domestic production. The reserve army of labor, by suppressing wages, raises s/v across the economy. The tendency is the resultant of these forces, and Marx did not commit to a monotonic decline.
What the discipline did with the theorem is the second half of the story, and two formal results carry the verdict together. Nobuo Okishio’s 1961 result, Okishio’s theorem, showed that under viable cost-reducing technical change with constant real wages, the equilibrium rate of profit cannot fall: if a capitalist adopts a new technique only when it raises the rate of profit at existing prices and wages, then once all firms have adopted it the new equilibrium rate of profit will be at least as high as before. Ian Steedman’s Marx After Sraffa (1977) extended the verdict by reformulating the value-theoretic apparatus in physical-quantity terms (after Sraffa) and showing the falling-rate result does not survive that reformulation either. The transformation problem in Capital’s third volume is the related embarrassment: Marx’s procedure for transforming values into prices of production has been shown, since Bortkiewicz’s 1907 critique, to be mathematically inconsistent — it cannot simultaneously preserve total values equal total prices, total surplus value equal total profits, and equalized sectoral rates of profit. The Bortkiewicz tradition produced reconstructions that retain a recognizably Marxian structure, but the procedure as Marx wrote it does not work. Both verdicts are reported with the formal apparatus forwarded to the technical literature; even the analytical Marxians (Roemer, Steedman themselves) ratify the results, which is part of why the chapter does not relitigate them.
Drive Marx's value bookkeeping. Push the organic composition (c/v) up with the rate of exploitation (s/v) held fixed, and the rate of profit falls — the falling-rate tendency. Then switch on Okishio's constraint (a technique is adopted only when it raises the profit rate at current prices and a constant real wage) and the equilibrium rate stops falling. That contrast is the §7.2 verdict, made drivable rather than asserted.
Figure 7.3 (interactive). The Marxian rate of profit s/(c+v) as the organic composition rises at fixed exploitation. Marx's tendency falls; Okishio's constraint keeps the equilibrium rate from falling. Drag the sliders; flip the scenario.
More machines per worker, with the same squeeze per worker, spreads the surplus thinner over a larger capital — so the rate of profit falls. Unless the new machines also cheapen everything, including the wage goods that fix the value of labor power: then the squeeze per worker rises and the fall is arrested. Okishio's theorem generalizes that second case — a technique a capitalist would actually adopt, because it raises profit at current prices and wages, cannot lower the economy-wide equilibrium rate.
The rate of profit is surplus over total capital advanced, $r = \dfrac{s}{c+v} = \dfrac{s/v}{(c/v)+1}$. Holding $s/v$ fixed and raising $c/v$ sends $r$ down.
Okishio's condition: if a technique is adopted only when it raises the profit rate at ruling prices and a constant real wage, the post-adoption equilibrium rate satisfies $r' \ge r$. The counter-tendency slider raises $s/v$ as $c/v$ rises — the offset Marx himself listed; Okishio shows viable change makes it binding.
The transformation problem is the volume-III embarrassment of the same bookkeeping. Two sectors with different organic compositions have different profit rates in value terms, yet competition equalizes profit rates. Run Marx's procedure and the two aggregate identities (total price = total value; total profit = total surplus) both hold — but the uniform rate Marx uses is computed on capital valued at values, not at the prices commodities actually sell for. Flip to Bortkiewicz's correction, which prices inputs and outputs consistently: the profit rates genuinely equalize, but one of Marx's two aggregate identities now breaks. The inconsistency is shown, not derived.
Figure 7.4 (interactive). A two-department illustration of the transformation problem (Sector I produces the capital good, Sector II the wage good; rate of exploitation fixed at 100%). The bars are sector profit rates. Value scheme: unequal. Marx's procedure: equal by construction, but on value-priced capital. Bortkiewicz: equal on consistently-priced capital, at the cost of one aggregate identity. A simplified exposition of the Bortkiewicz (1907) result; the full treatment is Steedman, Marx After Sraffa (1977).
Marx tried to add up profits in value terms and redistribute them across sectors as a single average rate — but the redistribution changes the prices of the inputs too, and the procedure doesn't close. The same commodity is an input at one price and an output at another. Once you price inputs and outputs consistently, you can equalize profit rates or preserve both of Marx's aggregate totals, but not both at once.
In value terms, sector $i$'s profit rate is $s_i/(c_i+v_i)$, which differs across sectors when organic compositions differ. Marx sets a single rate $R=\sum s_i/\sum(c_i+v_i)$ and prices of production $p_i=(c_i+v_i)(1+R)$; then $\sum p_i=\sum W_i$ and $\sum \text{profit}=\sum s_i$ both hold.
Bortkiewicz prices the capital good at $x$ and the wage good at $y$ and solves $W_I x=(c_I x+v_I y)(1+\rho)$ and $W_{II} y=(c_{II} x+v_{II} y)(1+\rho)$ for a uniform $\rho$. With the normalization total price = total value, total profit no longer equals total surplus — the residual the readout reports.
Why bother with the apparatus at all if its central predictive theorem is defective? Because surplus value, the structural reading of exploitation, and the analytical category of class as a relation are not what Okishio defeated. What Okishio defeated is the claim that under capitalist development the rate of profit secularly falls toward a crisis-inducing minimum. The claim that surplus value is appropriated by owners of productive assets — the structural fact about the wage relation — is not a claim about the rate of profit's trajectory. It is a claim about the form of the relation, and that claim survives. Marx's position and Capital as a work sit relationally in the timeline. The non-quantitative apparatus that survived the predictive failure is what the next section walks.
You just drove the apparatus the verdict rests on — surplus value, the falling rate, Okishio's defeat of it. "Was Marx right?" doesn't take one answer; it takes five.
The question fractures into five separate verdicts, and they don't point the same way. The labor theory of value as a price theory: wrong. The falling rate of profit: defective in standard form (Okishio — you just watched the line refuse to fall). The immiseration prediction: refuted by the wage record. Exploitation as a structural relation: a live analytical category. The diagnostic vocabulary — alienation, fetishism, fictitious capital: live. The apparatus this section walks is the spine all five verdicts rest on; the full accounting is the walkthrough.
Marx's structural reading of exploitation is the for-voice this walkthrough opens with, before the welfare-economics apparatus answers back.
On Marx's account, the wage-profit split is a structural consequence of who owns productive assets, not a market-clearing return to factors. Capital I.25 — the general law of capitalist accumulation — is the for-voice this walkthrough mounts against the marginal-productivity defense (the worker is paid the marginal product of labor). Distribution as a property relation, or distribution as factor pricing: the walkthrough holds the two against each other across the welfare apparatus.
What survived the predictions' failure was a vocabulary. Four categories, each making a structural claim about a specific kind of social relation that capitalism produces, and each (as the chapter will argue across §7.4 and §7.5) doing analytical work that did not depend on the predictions and that no other tradition produced an equivalent for. Alienation, commodity fetishism, primitive accumulation, the reserve army of labor. Each is a structural category, not a moral category and not a psychological category. The section walks the three diagnostic clusters and shows what each is built to do.
Alienation in Marx's technical sense is a four-fold structural estrangement of the worker under capitalist production, not a description of how workers feel about their jobs. The category appears in the Economic and Philosophic Manuscripts of 1844 and persists in altered form through Capital; the structure is continuous though the terminology shifts. The four-fold breakdown follows.
From the product. The worker produces an object that becomes the property of the capitalist. The shoe the worker stitched goes to the warehouse; the worker has no claim on it, may not be able to afford to buy it, will not see it again. The structural fact is not that the worker dislikes this; it is that the productive activity terminates in something that belongs to someone else, by the form of the relation, regardless of how the parties feel about it.
From the act of production. The labor process itself is structured by the requirements of capital accumulation, not by the worker's intentions. Pace, intensity, sequence, the use of tools: these are determined by the firm's competitive position, not by the worker's understanding of the work. The worker executes a process whose structure was chosen elsewhere. The estrangement here is not that work is unpleasant but that productive activity is not the worker's own activity in the sense that craft work was the artisan's own activity.
From species-being. Marx's term for what he took to be the human species's distinguishing feature: the capacity to produce creatively and intentionally, transforming nature according to a conception held in mind before execution. Wage labor, by structuring production around capital's requirements, alienates the worker from this productive-creative capacity in its general human form. The worker remains capable of it (outside work, in unalienated activity), but inside production, the activity that is humanly distinctive has been organized so that it expresses the firm's logic rather than the worker's.
From other workers. The wage relation atomizes. Workers compete with each other for jobs, for piece-rate bonuses, for promotion within the firm's hierarchy. The shop floor is structured so that cooperation occurs through the coordination of capital, not through the workers' own organization of their activity. Solidarity, where it forms, forms against the structure of the wage relation, not as a natural extension of it.
All four estrangements name the form, not the affect. They describe how production is organized under the wage relation; they do not describe how any particular worker feels about that organization. A satisfied worker in a well-paid job stands in the same alienated relation as a dissatisfied worker in a precarious one, in the same way that a high-wage worker and a low-wage worker stand in the same exploitative relation. The pop-Marxist reading that maps alienation onto job dissatisfaction is what this category was built to refuse. The structural framing is what makes alienation an analytical concept rather than an emotional report.
Commodity fetishism is the highest-load category in the section. Marx introduces it in the first chapter of Capital, in a passage that has resisted easy reading for a hundred and fifty years. The set-up: under generalized commodity production, social relations between producers (who works for whom, on what terms, with what reward) appear to producers as relations between things (commodities, money, capital). Producers experience their own products as autonomous powers that dominate them. Marx's example is a wooden table. As an ordinary object, the table is wood, joined and shaped by labor, useful for setting things on. Once it enters the market as a commodity, Marx writes, it “stands on its head, and evolves out of its wooden brain grotesque ideas.” It acquires properties (a price, a market position, an exchange-relation to other commodities) that are not properties of the wood but properties the wood seems to possess in virtue of being a commodity.
The image is making a precise structural claim. Under commodity production, the social act of allocating labor across kinds of work is performed indirectly, through the market: producers do not coordinate face to face; they consign their products to the market and discover, after the fact, what their labor was worth. The market mediates the social allocation of labor, and producers experience this mediation as a property of the commodities themselves. The price of bread is the social labor of bread production, expressed through the form of a market relation; producers see only the price, and the social-allocative content of that price is invisible to them. Fetishism is the analytical name for this misperception — the structural form by which generalized commodity production presents itself to its participants, rather than a metaphor for consumerism or a moral complaint about materialism.
The vocabulary that the new reading in §7.5 will deploy lives here. Use-value is the qualitative usefulness of a thing, what the table is good for. Exchange-value is the quantitative ratio at which one commodity exchanges for another. Value is the social substance (abstract human labor under conditions of generalized commodity production) of which exchange-value is the form of appearance. The value-form is the form that value takes as it moves through commodity, money, and capital. Heinrich's argument, which §7.5 walks at length, is that Capital is an analysis of the value-form's operation, not a model of how prices in particular markets are formed. The structural reading of fetishism is what makes that argument possible. If fetishism were a metaphor about consumerist culture, no amount of textual reconstruction would license the value-form reading. Because fetishism is a structural account of social mediation under commodity production, the reading is licensed.
Primitive accumulation (Marx's term ursprüngliche Akkumulation, where “primitive” means original or originary, not crude) is the historical precondition of capitalist relations. For wage labor to exist, two things must be true: there must be capitalists who own productive assets, and a population of workers who own none and must sell their labor power to live. The second class is produced, not given. Primitive accumulation is the analytical name for the historical processes (enclosure, colonial dispossession, the destruction of common rights, the violent expropriation of peasantries) that strip a population of independent subsistence and create the propertyless worker as a category. The English enclosures, the Highland clearances, the Atlantic slave trade as a labor regime feeding the plantation economies of the Americas: these are not pre-capitalist episodes that ended once capitalism began. They are the form by which capitalist relations expand into new territories, and they continue to operate.
The Highland clearances will do. Tenant farmers who had grazed sheep and cultivated small plots on the laird's land for generations were evicted, sometimes burned out of their cottages, when the landlord's calculation favored sheep walks or sport hunting over rents. The dispossessed emigrated or moved to the industrial Lowlands to sell their labor in the new factories. Both options fed the British industrial expansion. The point is not the moral character of the dispossessions, though that is real. The point is that the “free labor” classical political economy took as its starting condition was not a natural state. It was produced by specific processes of expropriation, and the analytical category names that production. The factory system hist ch. 7 walks required this prior labor-supply formation. Atlantic slavery, treated in hist ch. 9, is the paradigm case of primitive accumulation through a different legal form.
Where the industrial revolution sits in the timeline as the productive context the analytical category was built to read appears on the timeline node.
The reserve army of labor is the equilibrating mechanism. The unemployed and underemployed population, by competing with employed workers for jobs, suppresses wages toward the value of labor power. When capital expansion pulls workers in, the reserve army shrinks, wages rise, and the rate of exploitation falls. When capital expansion slows or technical change displaces workers, the reserve army grows, wages fall, and the rate of exploitation recovers. The mechanism stabilizes the relation. Marx took it to imply that capitalism could not abolish unemployment without abolishing itself, because the wage relation requires the reserve army's discipline to operate.
What happened to the reserve army in the second half of the twentieth century is the test the chapter will return to. Welfare-state institutions (unemployment insurance, public health, retirement systems), high union density, and full-employment policy regimes produced postwar labor markets in which workers had reservation wages well above the value of labor power and in which the disciplinary force of unemployment was substantially attenuated. The reserve army was domesticated. The structural framing of the category survives this; what was domesticated were the parameters of the mechanism's operation, not the analytical fact that wage formation under capitalism is conditioned by a population that can substitute for employed workers. Whether the parameter shift counts as refutation or accommodation is the question §7.4 takes up.
Is the postwar wage rise a refutation of Marx's mechanism, or a parameter shift within it? Drag the reserve army down (what the welfare state, union density, and full-employment policy did) and the wage rises above the value of labor power — with the mechanism itself unchanged. Switch to “mechanism refuted” and wages decouple from slack entirely. The chapter's claim is the first picture, not the second.
Figure 7.5 (interactive). Wage relative to the value of labor power as labor-market slack varies. “Mechanism intact”: wages track slack (Marx's relation), so shrinking the reserve army raises wages without changing the relation. The shaded bands mark the high-slack pre-welfare-state regime and the low-slack postwar regime. Illustrative; a live wage-share series would replace the stylized bands.
Unemployment is the thing that holds wages down; the welfare state and full-employment policy shrank it, so wages rose. But that slack sets wages is exactly Marx's mechanism — it didn't stop being true, its input just moved. Reading the postwar wage rise as a refutation confuses a changed parameter with a changed law.
With these vocabulary categories pinned, the chapter has the diagnostic half of its central evaluative claim. Section 7.4 walks the methodological frame and the predictive program that did not land.
Historical materialism, in Marx's mature formulation, is a methodological claim about how to explain societies. The economic structure of a society, namely the productive forces (the technical capacities for production) and the relations of production (the legal and customary forms by which productive assets are owned and labor is organized), constitutes the foundation on which the legal, political, and ideological superstructures arise. Changes in the foundation, driven by the development of productive forces, drive changes in the superstructure. The claim is methodological: where to look first when explaining a doctrine, an institution, a political form.
The textbook caricature treats this as a one-way determinism: economy causes politics, politics is epiphenomenal, ideas are reflexes of class interest. Engels's letters to Joseph Bloch (1890) and Conrad Schmidt (1890–94) addressed this caricature directly. Engels insisted that the relation between foundation and superstructure is reciprocal: political forms react back on the economic structure, ideologies have their own internal logics that develop semi-autonomously, and the determination of the superstructure by the foundation is “ultimately” rather than “immediately.” The qualifications are not retreat from the method but the method's correct statement; the caricature is what gets attacked and what serious users of the framework do not deploy.
Marx's most subtle deployment of the method is The Eighteenth Brumaire of Louis Bonaparte (1852), the analysis of how Louis-Napoleon Bonaparte rose to power in the years after 1848 and what the rise revealed about the class structure of French society. The book reads its political subject through the configuration of class interests (bourgeoisie, peasantry, lumpenproletariat, financial faction versus industrial faction) without reducing political agency to economic interest. Bonaparte exploits real political openings created by the irreconcilability of class fractions; he is not the puppet of any of them; the book treats his agency as effective and surprising. The methodological lesson is that historical materialism does not assert that politics is reducible to economics. It asserts that political forms are intelligible in light of the class structure they negotiate, and that doctrines about politics are intelligible in light of the class positions they articulate.
Apply the method to economic doctrines themselves and a striking consequence follows. If the marginalist revolution of the 1870s, which displaced the labor theory of value just as Marx's Capital was beginning to be read, is not refuted by historical materialism but is instead an instance of doctrinal change that the method should be able to account for, then a partial explanation becomes available: the labor theory had a Marx-shaped exit at the end of it, and the discipline that adopted marginal utility instead of pursuing the labor-theoretic line was, among other things, declining the political payload the labor theory had begun to carry. This is the mirror image of the historiographical claim chapter 5 takes up from the marginalists' side: the displacing program had partly political work to do, regardless of whether Jevons, Menger, or Walras was conscious of doing it. Ch. 7 names that mirror; ch. 8 carries the displacing program in its own voice.
Now turn to what the framework predicted, and what happened. Four predictions matter, each made by Marx's framework, each tested by the historical record across the twentieth century, and each falsified in a specific way that the chapter will name.
| Prediction | Marx's framework | Historical record |
|---|---|---|
| Capitalism's trajectory | Collapse from internal contradictions | Welfare-state, regulatory capitalism, postwar growth domesticated the contradictions |
| Reserve army of labor | Eliminated only by revolutionary abolition of capitalism | Domesticated by welfare-state institutions and union density |
| Middle classes | Proletarianization → revolutionary unity with industrial workers | Postwar middle-class formation; professional-managerial stratum; no unifying revolutionary subject |
| Rate of profit | Secular fall per the falling-rate theorem | No clean falling trend; Okishio result defeats the theorem in standard form |
Failure 1: capitalism did not collapse from internal contradictions. Marx's framework expected the falling rate of profit, the secular concentration of capital, and the deepening immiseration of the working class to produce contradictions the system could not contain: periodic crises of growing severity, ending in collapse and proletarian revolution. The twentieth century did not produce that trajectory. Industrialized capitalism passed through major crises (1873, 1929, 1973, 2008) and did not collapse. Welfare-state institutions absorbed the social cost of unemployment; regulatory regimes constrained financial speculation, however imperfectly; postwar growth (1945-1973) raised real wages substantially across most of the industrialized world, weakening the immiseration premise. Each crisis was followed by institutional reconfiguration that addressed the proximate failure mode without abolishing the system. Hist ch. 14 walks the postwar institutional regime that did the absorbing.
Failure 2: the reserve army was domesticated, not eliminated. Marx's framework treated the reserve army as the disciplinary mechanism wage labor required, with full employment achievable only through the abolition of capitalist relations. The postwar full-employment regime (high union density, sectoral wage bargaining, social insurance, Phillips-curve demand management) produced labor markets in which the disciplinary force was attenuated for several decades without abolishing the wage relation. Union members in West Germany, Sweden, the UK, and the US in the 1950s and 1960s had reservation wages well above the value of labor power, and the cost of refusing a particular job was low. The wage relation continued; the disciplinary mechanism was institutionally constrained, and the constraint lasted long enough to count as evidence that what Marx took to be a necessary feature of capitalist labor markets was instead a parameter that could be dialed across a wide range. Hist ch. 12 and hist ch. 14 carry the institutional history.
Failure 3: the middle classes did not unify with industrial workers. Marx's framework predicted that capitalist development would proletarianize the middle strata, meaning that the small business owners, the artisans, the clerks, the professionals would be ground down into the mass of wage labor, producing a unified revolutionary subject. Twentieth-century capitalism produced a different outcome. Wage labor expanded enormously, but the wage-earning population stratified rather than homogenized. A professional-managerial stratum developed, distinguished from the industrial working class by salary form, education credential, employment security, and identification with the firm rather than against it. White-collar work expanded faster than blue-collar, and the white-collar workforce did not see itself as proletariat. The middle classes were not eliminated; they were restructured. The unifying revolutionary subject the framework expected did not assemble. The failure is specific: proletarianization in the sense of rising wage dependence did happen; proletarianization in the sense of class unification with industrial workers did not.
Failure 4: the rate of profit's empirical trajectory has not behaved as the falling-rate theorem requires. The Okishio verdict from §7.2 was already half this argument on the formal side. The empirical record carries the other half. Long-run estimates of the rate of profit in advanced capitalist economies do not show the secular decline the theorem requires. There are episodes of falling profitability (the 1965-1982 decline in the US; the post-2008 deceleration) and episodes of recovery, but the trajectory is not a downward path toward a crisis-inducing minimum. Brenner's work on the post-1973 long downturn argues for a structural explanation of the 1970s decline from international competitive pressure on installed capacity, not from the standard falling-rate theorem. The standard theorem, in the form Marx proposed it, does not survive contact with the empirical record any better than it survives Okishio.
Marx's crisis mechanism — the falling rate of profit, overproduction, capital concentration — is the crisis-theorist slice of this chapter. Set against Schumpeter's creative destruction, the two predicted capitalism's self-undermining by opposite mechanisms: Marx by a profitability squeeze, Schumpeter by the innovation that renews the system while destroying its incumbents.
Calibration here sits between two failure modes. Naive falsification (the predictions failed; therefore the framework is dead) is wrong because it conflates the predictive program with the diagnostic vocabulary. The diagnostic categories of §7.3 are structural-analytical claims about kinds of social relation, and what failed in §7.4 does not touch them directly. Naive vindication (the framework is unfalsifiable; therefore intact) is wrong in the opposite direction, because it pretends the predictive program was never the test. The predictive program was substantive: capitalism would collapse, the reserve army would be eliminated, the middle classes would proletarianize, the rate of profit would fall. The historical record falsified each. To pretend Marx's claims were always purely structural is to evacuate the economics of its substance to save it from its empirical defeat. The honest position is that part of the program was substantive prediction, and that part was falsified. The diagnostic vocabulary survived because it was doing different work, and the next section walks the survival inside the contested historiography of how to read the project as a whole.
Figure 7.1 names the four failures; here you can hold each prediction against the historical record and see where the paths diverge. The point is not “Marx was wrong” but the specific shape of each divergence — which is what telling a failed prediction from an evacuated framework requires.
Figure 7.6 (interactive). Each dimension: Marx's framework's predicted trajectory against the twentieth-century record (indexed). Series are stylized illustrations of the qualitative divergence, not a sourced dataset; a live B-sourced series would replace them.
A prediction can fail in different ways. “Collapse” failed because institutions absorbed each crisis; “reserve army” failed because a parameter moved, not the mechanism (see Fig 7.5); “unity” failed because the working population stratified rather than homogenized; “rate of profit” failed because no secular fall appears (and Okishio shows why — Fig 7.3). Naming the specific divergence is what separates a falsified prediction from a diagnostic category that survives it.
What Marx's project actually was depends on who is reading it. The Marxian revival, as it has run since the 1960s and acutely after 2008, divides into three readings, each constituting a coherent research program with a substantive position on what the project was for. The chapter walks the three at strongest form before taking its call.
| Dimension | Analytical Marxism | Mainstream dismissal | New reading (Heinrich et al.) |
|---|---|---|---|
| Exponents | Roemer, Cohen, Elster, Wright | The discipline's revealed verdict; Okishio, Steedman as formal moments | Heinrich, Backhaus, Reichelt, Postone |
| What is accepted | Labor theory as price theory is wrong; predictive program failed | Transaction-cost reasoning, class as voting bloc, capital concentration — absorbable without framework | Labor theory is not a price theory; standard interpretation misread the text |
| What is rejected | Labor-theoretic apparatus; falling-rate theorem; immiseration prediction | Labor theory of value; falling-rate theorem; historical-materialist predictions | Standard interpretation as price-theory misreading; Marxist-Leninist textbook |
| What is preserved | Class, exploitation, historical change — on rebuilt foundations | What has been absorbed without the framework's vocabulary | Capital as critique of the value form |
| Central insight | Exploitation as differential property rights over productive assets | Marx's framework is not the live frontier of economics | Value-form analysis is what Capital was always about |
Hold one axis fixed and swap the reading. The three readings re-interpret the same apparatus, so the only way to see “same labor theory, three verdicts on it” is to fix the axis and switch the reading. The chapter's call across the three is in the prose (§7.5 close and §7.6), not here — this widget presents the positions at strongest form; the narrator takes the position.
Figure 7.7 (interactive). Each reading's treatment of three axes — the labor theory of value, exploitation, and what 2008 did or didn't vindicate — each cell at its strongest form. Switch readings on a fixed axis to compare. The chapter's verdict is in the prose, by design.
Analytical Marxism is the reformulation that ran the framework through the methodology of contemporary analytical philosophy and neoclassical economics. John Roemer's A General Theory of Exploitation and Class (1982) is the program's central economic statement: a game-theoretic reformulation of exploitation in which the structural fact Marx identified is preserved without the labor theory of value as its analytical foundation. Roemer's argument is that exploitation can be defined directly in terms of differential property rights over productive assets. A coalition of agents is exploited if it would be better off, and another coalition would be worse off, under a counterfactual redistribution of the assets the second coalition owns. The definition does not require labor as the unique source of value; it requires only that there is a structure of asset ownership that, when applied through normal market processes, produces a class of agents whose labor sustains a class of agents whose property is sustained.
G. A. Cohen's Karl Marx's Theory of History: A Defence (1978) was the methodological companion: a reconstruction of historical materialism using analytical philosophy's apparatus, defending the central claim that productive forces explain relations of production. Jon Elster's Making Sense of Marx (1985) ran the framework through methodological individualism. Erik Olin Wright's Class Counts (1997) built empirical class analysis on contradictory class locations adapted to the contemporary occupational structure.
What analytical Marxism produces, taken together, is the most defensible reconstruction of Marx's framework available to a working economist. The price theory is gone. The falling-rate theorem is gone. The substantive predictive content is gone. What remains is a coherent theory of class as differential property relations, a theory of exploitation as a structural feature of those relations, and an empirical research program that uses these categories to study contemporary economies. The reformulation is internally rigorous, it accepts the discipline's verdict on the labor theory, and it preserves what was distinctive about Marx's analytical contribution. The cost is that what is preserved is no longer recognizably Marx's; the rebuilt foundation does substantially different conceptual work from the labor theory it replaced.
Which Marxian apparatus survives a coherent-left reconstruction? Analytical Marxism keeps exploitation-as-property-rights and discards the labor theory as a price theory; the transformation-problem constraint (Bortkiewicz 1907 — Fig 7.4) bounds any sub-strand that tries to keep it. This chapter is that walkthrough's transformation-problem home.
Mainstream dismissal is the academic discipline's revealed verdict, expressed less in single texts than in what economics departments teach and journals publish. The discipline does not refute Marx; it does not engage Marx; in most graduate programs Marx is not on the syllabus. The dismissal is not casual. It is the considered position that the framework is not the live frontier of the field and that engaging it has substantial opportunity cost.
The dismissal rests on four substantive moves. First: the labor theory of value as a price theory is wrong. Prices in actual markets are not determined by socially necessary labor time but by supply, demand, and competitive equilibration with marginal cost as the binding constraint. The Salamanca-Menger subjective-value lineage walked in chapter 5 (the marginalist revolution) got this right. Analytical Marxism accepts this verdict, which means the discipline and the analytical Marxians agree on the price-theory question.
Second: the falling-rate-of-profit theorem is mathematically defective per Okishio (back to §7.2). The result is robust, the formal apparatus has been examined for sixty years, and the theorem in standard form does not survive the examination. The discipline has not relitigated this since the 1980s.
Third: the historical-materialist predictions failed. The four failures of §7.4 are read as decisive. Capitalism did not collapse, the reserve army was domesticated, the middle classes did not unify, the rate of profit did not fall. Each is empirical and each is settled by the historical record. The framework that made these predictions does not have a recovery route that preserves the predictions; recovery routes that abandon the predictions are abandoning what the framework substantively claimed.
Fourth: what is usable in Marx has been absorbed without the framework. Transaction-cost analysis of why firms exist (Coase, Williamson) captures part of what Marx's analysis of capital and the firm aimed at. Class as voting bloc, in modern political economy, captures part of the structural analysis of class. Capital concentration is studied within industrial organization. The discipline has the analytical tools the framework reaches for, in vocabularies that do not require the labor theory. From this, the dismissal concludes, there is little reason for a working economist to begin with Marx.
This is not a foil. It is a serious position, held by serious economists, on the basis of analytical considerations that the chapter has just walked. The hardest test of the chapter's evenhandedness is to render it at its strongest form, and the chapter is taking the test rather than ducking it.
The new reading is the Neue Marx-Lektüre, developed in West Germany from the 1960s by Backhaus and Reichelt and brought into the Anglophone literature by Michael Heinrich's An Introduction to the Three Volumes of Karl Marx's Capital (German 2004; English 2012) and Moishe Postone's Time, Labor, and Social Domination (1993). It argues that the standard interpretation of Marx, the interpretation that analytical Marxism and mainstream dismissal both share at the level of what the labor theory was trying to do, has been a misreading of Capital from the start.
Heinrich's central claim: the labor theory of value is not a theory of how prices in particular markets are determined, and treating it as one is to misread Capital's subject. Capital is not a book about price formation. It is a book about the value form: about what kind of social relation produces a society in which products take the form of value-bearing commodities, money appears as a self-moving substance, and capital appears as a power that constrains and dominates its producers. The analysis is of the form, not the magnitudes. The prediction that prices will track socially necessary labor time, which the labor theory as a price theory makes and which empirical evidence falsifies, is not a prediction Capital was making. It is a prediction the standard interpretation imputed to Capital by reading the book as continuous with Ricardian price theory rather than as a critique of the categories Ricardo took for granted.
The textual basis is the first chapter of Capital: the analysis of the commodity, the value-form, and commodity fetishism. On the standard reading, this chapter is preliminary scaffolding for the labor-theoretic price model the rest of the book builds. On the new reading, the first chapter is the book. It states what Capital analyzes (the value-form as the form social labor takes under generalized commodity production), and the rest of Capital investigates how this form develops through commodity, money, and capital. Whether prices in any given market track values is, on this reading, not the question.
The structural reading of commodity fetishism (§7.3) is what makes this argument coherent. If fetishism were a metaphor, the value-form analysis would be metaphysical superstructure built on a metaphor. Because fetishism is the structural account of how social mediation appears under commodity production, the value-form analysis is its analytical extension. Postone adds that the labor producing value is an abstracted social form (“abstract labor”) specific to capitalism, and Marx's critique is of the social form, not of any particular distribution of its products. The political conclusion is not that workers should seize the surplus they produce, which would leave the value-form intact; the conclusion is that the value-form itself is what the critique points beyond.
On the new reading, the empirical failure of the labor theory as a price theory is irrelevant because the labor theory was never a price theory. The Okishio verdict is correct against the standard interpretation and beside the point against the new reading. The four failures of §7.4 are real falsifications of the framework as standardly read, and the new reading accepts each. But the diagnostic categories — alienation, fetishism, exploitation as a structural relation, the value-form as social mediation — survive because they were the analysis all along. The predictions were a misuse of categories the value-form analysis did not authorize. Whether this is reading Capital faithfully or imaginatively is contested; the philological case is substantial, and the position now dominates continental Marx scholarship.
With the three readings laid down, 2008 becomes legible as a specific kind of event. Sales of Capital spiked through 2009; reading groups proliferated; David Harvey's online lecture series accumulated millions of views; the Verso editions sold out in successive printings. The surge tapered through 2014 as the financial system stabilized. It was the largest spike of public engagement with Marx's economics since the 1970s.
What was vindicated, and what was not, depends on what is meant. Not vindicated: the falling-rate-of-profit theorem in its standard form. The 2008 crisis did not look like a falling-rate crisis. It originated in the financial sector, in the collapse of mortgage-backed asset values and the seizure of interbank lending markets, not in a secular decline of profitability across productive sectors. The macro literature on the crisis, in both mainstream and heterodox traditions, does not invoke the falling-rate theorem. Whoever claims that 2008 vindicated Marx in this sense is overstating what the crisis showed.
Vindicated: the diagnostic vocabulary that treats crisis as endogenous to capital accumulation rather than as exogenous shock. Going into 2008, mainstream macro had treated financial fragility as something the post-Volcker regime had retired; the “Great Moderation” literature read the post-1985 period as evidence that policy had stabilized the macroeconomy. The crisis falsified that reading. Capital's own logic of accumulation produced the configuration that broke: internal dynamics of leverage, asset-price formation, and shadow banking that built fragility through ordinary expansion. The Marxian picture of capital as a system that periodically negates its own conditions of valorization fits this much better than the Great Moderation's picture did. Fictitious capital, Marx's category for claims on future surplus value that circulate as if they were capital, gained renewed analytical life as a name for the financial-asset collapse the crisis turned on. Minsky's financial-instability hypothesis, treated in Walkthrough 08, sits adjacent to the Marxian tradition without belonging to it; Brenner's long-downturn thesis and Minsky's hypothesis are read together in much of the post-2008 literature.
Welfare economics, briefly. Marx's challenge is that the welfare theorems' equilibrium-as-Pareto-efficient result presumes the legitimacy of the existing distribution of property in productive assets, which is precisely what Marx's category of class is built to interrogate. Pareto efficiency is silent on whether the starting distribution is itself defensible. The substantive machinery lives in economics ch. 3; what Marx adds is the question of whether the distribution the framework takes as given is the question.
Sraffa is sometimes claimed for the Marxian revival; the Sraffian system, including the Cambridge capital controversy, is a separate program treated in chapter 5 (the marginalist revolution) and in the book's distribution-thread walkthrough. Brenner's long-downturn thesis frames post-1973 stagnation as international competitive pressure on installed capacity. The dependency and world-systems traditions (Prebisch, Frank, Wallerstein) descend from Marxian categories through development economics; their substantive treatment is forwarded to chapter 17.
Where the 2008 crisis and Minsky sit relationally appears on the timeline.
The chapter's call. The new reading is most faithful to Capital's text; the philological case is substantial, and the structural reading of fetishism in §7.3 makes the case coherent rather than ad hoc. Analytical Marxism produced the most defensible reconstruction for the working economist who wants Marx's explanatory core (class as relation, exploitation as structural fact, historical change as endogenous to social-property structures) without the predictive program. Mainstream dismissal is correct about most of what it dismisses: the labor theory as price theory is wrong; the falling-rate theorem is defective; the four predictive failures are real. Where the dismissal goes wrong is on what is left. The structural reading of fetishism, of exploitation, and the materialist explanation of doctrines are not absorbed by transaction-cost economics or political-economy treatments of class as voting bloc. Those vocabularies cover their regions but do not reach the social-form analysis Capital was always doing. The dismissal evacuates the analytical residue at the moment it correctly retires the predictive program.
Marx's economics failed as prediction and succeeded as diagnostics. This split is interesting because it tells us something about what kinds of intellectual work survive failed predictions and what kinds do not.
What survived, specifically: alienation as four-fold structural estrangement, doing analytical work in the regions where wage labor's structural form is the question, regardless of any particular worker's psychological response. Commodity fetishism as the structural account of how social mediation under generalized commodity production appears to its participants, an analysis that no other tradition reproduces without recovering Marx's apparatus to do it. Exploitation as a structural relation between owners and non-owners of productive assets, holding regardless of wage level: a category that locates an analytical fact about the form of capitalist labor markets that the marginal-product picture of factor pricing cannot reach. Fictitious capital as the analytical name for claims on future surplus that circulate as if they were capital, recovered in the post-2008 literature on financial-system fragility. Each of these is a structural-analytical claim about a kind of social relation, and each survives parameter shifts — the rise of the welfare state, the post-1973 turn, the financialization of the post-1980 economy — because the categories pin the form of the relation rather than the trajectory of the system.
The reason the diagnostic categories survived is internal to what kind of work they do. Predictive claims of the form “capitalism will collapse,” “the reserve army will be eliminated only by revolution,” “the rate of profit will secularly fall” depend on the parameters of the system continuing to operate as the framework's mechanism requires. When welfare-state institutions, union density, and full-employment policy regimes shift the parameters far enough, the predictive claims fail because the mechanism that produced them has been institutionally constrained. Structural claims of the form “under generalized commodity production, social labor is mediated through the value-form” do not have this dependence. They describe the form of the relation, and the form is what the parameters operate on. Welfare-state Sweden in 1970 and financialized Britain in 2010 differed in the parameters, and the form (wage labor, commodity exchange, capital as self-valorizing value) was the same form in both cases. The diagnostic categories survive because they pin the form, and parameter shifts leave the form intact.
The three-readings call from §7.5 follows from this. The new reading is most faithful to Capital's text because the structural-analytical work is what the book was always doing; the standard interpretation read the book as continuous with Ricardian price theory and inherited a misreading from that continuity. Analytical Marxism produced the most defensible reconstruction available to a working economist who wants the explanatory core without the predictive program; the rebuilt foundations do not preserve everything Marx's value-form analysis did, but they preserve enough to study contemporary class structures with the framework's distinctive moves. Mainstream dismissal is correct about what it dismisses: the labor theory as price theory, the falling-rate theorem in standard form, the historical-materialist predictions in the senses they were tested. It is wrong about what is left, because the categories that the value-form analysis of Capital centered on (fetishism as social mediation, exploitation as structural relation, the materialist explanation of why doctrines arise as they do) are not what transaction-cost economics or modern political economy reproduce. The dismissal evacuates the analytical residue at the same moment it correctly retires the predictive program, and the residue is what the chapter has been arguing matters.
There is a more general lesson. Frameworks whose analytical load is structural (that pin the form of a relation, the type of an object, the kind of social mediation in operation) can survive predictive failure because their work is not predictive. Frameworks whose analytical load is predictive (the rate falls, the system collapses, the classes unify) cannot survive predictive failure, because the predictions are what the framework offered. Marx's economics carried both kinds of work, in a configuration that the standard interpretation made hard to disentangle. What this chapter has done is disentangle them, accept the predictive failures cleanly, and show that the structural work continues to have analytical purchase. The lesson generalizes: when a framework is being evaluated, the first question is what kind of work it is doing, and the answer is rarely uniform across the framework.
The forward connections are several. Chapter 5 carries the displacing program from the marginalists' side: the mirror image of the historical-materialist explanation in §7.4. Chapter 17 places the modern Marxian descendants (Piketty, Brenner, Harvey, Minsky's financial-instability inheritance) within the broader modern-pluralist landscape. The book's distribution-thread walkthrough threads the distribution lineage from Ricardo's rent through Marx's surplus value through marginal-product distribution through Piketty's r > g, with this chapter's account of surplus value as one station. Walkthrough 09 carries the modern inequality debate, which inherits Marx's framing of distribution as a question about the structure of property in productive assets. Walkthrough 08 carries the modern recession debate, where Minsky's financial-instability hypothesis lives.
What happened when Marxism became state doctrine (Soviet planning, Maoist development, the Cuban experiment, the planning tradition's collapse in 1989-1991) is hist ch. 15's territory. This chapter has covered Marxism as intellectual project, not as state doctrine. The two are not the same, and this chapter's evaluation does not extend to the planning tradition's economic record. Piketty sits on the timeline as the modern descendant the distribution-lineage thread runs forward to.
Marx, Economic and Philosophic Manuscripts of 1844; The Eighteenth Brumaire of Louis Bonaparte (1852); Capital vol. 1 (1867), vol. 3 (Engels, ed., 1894). Engels to Bloch and Schmidt (1890–94). Böhm-Bawerk, Karl Marx and the Close of His System (1896). Okishio (1961). Steedman, Marx After Sraffa (1977). Cohen, Karl Marx's Theory of History (1978). Roemer, General Theory of Exploitation and Class (1982). Elster, Making Sense of Marx (1985). Wright, Class Counts (1997). Postone, Time, Labor, and Social Domination (1993). Heinrich, An Introduction to the Three Volumes of Capital (2004; English 2012). Brenner, The Economics of Global Turbulence (2006). Harvey, A Companion to Marx's Capital (2010).