Section XXIV: XXIVThe Falling Rate in Shenzhen

The law the first half established does not stop at the Pacific. China is a commodity producer under competition, and the rate of profit falls in Shenzhen by the same arithmetic it falls in Santa Clara: more constant capital against less variable capital, the price pulled to the new value as each advantage generalizes, the margin gone once everyone has the machine. Beijing’s name for the condition is involution, and it has to be run against Beijing with the discipline the first half ran it against Washington, because no formation escapes the law. The test of a successor is what happens when the law arrives.

It has arrived. Overproduction, price wars, and the longest stretch of deflation since the 1990s. The GDP deflator down nine consecutive quarters. Auto-sector profits down 33% between 2017 and 2024 while sales rose 21%. EV margins from 7.8% in 2017 to 4.3% in 2024 and 4.4% in 2025, the two lowest years on record. Some 400 EV producers gone since 2018 and some 50 more expected to cut back in 2026. Producer prices negative from late 2022 into early 2026 and turned positive in March only by the war’s oil price. Every entry is the mechanism of the dilemma section running on the other side of the trade. The flood that destroys the incumbent’s rent destroys the challenger’s margin at the same time, because it is one flood, and the moral depreciation that strands Washington’s constant capital is Beijing’s own falling rate seen from the flooder’s side. A country that gives away the weights and sells the panel below the incumbent’s cost has no rent of its own to collect on either.²⁹⁹

What differs is what validates the constant capital, and who holds the claims. In the American case the constant capital is validated by profit through private claims, the three layers, and when validation fails the claims are destroyed and the household’s deferred wages are the collateral. In the Chinese case the claims sit substantially with the state: the banks, the local financing vehicles, the state enterprises, and a party that let private developers fail without socializing their losses. A formation that holds its own claims can validate constant capital by fiat, roll it, write it down, recapitalize it, for as long as the physical surplus is produced and distributed, which means it can run on the surplus product where the incumbent has to run on surplus value.

That is the value-form discriminator, and it is the term the Preobrazhensky test the liquidation section set was missing. Preobrazhensky asks which sector accumulates at the other’s expense. The discriminator asks what the socialized sector needs in order to grow. A plan target denominated in use-values needs no rate of profit. A private balance sheet needs one or it is written down. Between those two sentences sits the whole difference between a formation that survives the law and one that does not. The expansion since 2012 is financed from inside, by state banks lending against a household savings rate near a third of income and a current account the export machine keeps in surplus,³⁰⁰ which is the difference between a formation that runs on the surplus product and one that runs on surplus value, and it is why the recycling into the incumbent’s debt stopped when it stopped being needed.

The difference is only live if the surplus product can be realized somewhere, and the war has closed most of the somewhere. China’s surplus realizes through three channels: the core’s consumption, the periphery financed by Chinese credit, and its own households. The first channel is the household the reproduction section documented and the Europe the cracks section documented, the borrower kept on credit until not kept at all, the continent’s storage and fertilizer, tariffs on the goods that used to be cheap, and now Beijing’s own cancellation of photovoltaic export rebates and its price undertakings in EVs, a strategy its own commentators describe as treating internal ailments through external means, compelling exporters toward higher-margin models because the volume game has closed. The second channel is the yuan bridge, and the war lands hardest exactly where the bridge leads: Pakistan’s grid, Egypt’s wheat, Ethiopia’s nitrogen. China can sell into that periphery only by financing the buyer, which is vendor financing at state scale, the shunt again, with a state owner and state-held claims, and a shunt is a shunt whoever owns it. The one thing the periphery buys with its own dollars is the panel that cuts its dollar bill: Pakistan has taken more than 30 GW of Chinese panels since 2019 against a 35 GW grid, paid for in reserves it could not spare, to stop paying for gas it could not afford, and shut the net-metering door in February 2026 to protect the capacity charges its grid still owes the fossil plants.³⁰¹ The third channel is domestic, and it is the number the critics hold up.

The empirical evidence on that channel as of 2026 is mixed. Household consumption share of GDP has remained stubbornly anchored in the high 30s to low 40s, 22 years after rebalancing toward consumption was first declared the explicit policy priority in 2004. The labor share of national income, which the ILO put at 51.6% in 2019, has shown only modest motion since the 20th Party Congress called for raising it in 2022. Per-capita disposable income and GDP each grew 5.0% in real terms in 2025, income keeping pace with output, which says nothing about the household share either way. Capital-biased technical change continues to compress the labor share. Most damaging to the phase-two narrative, Li Shi, Dean of the Institute for Common Prosperity and Development at Zhejiang University, the officially designated Common Prosperity demonstration province, presented research in 2026 showing China’s wealth Gini coefficient rose from 0.45 in 1995 to above 0.7 by 2023. The 14th Five-Year Plan’s stated target of narrowing gaps between individual incomes by 2025 was not met. The Zhejiang demonstration zone’s own Action Plan explicitly avoided “adjusting high incomes” to prevent capital flight, which measures how seriously the redistribution component is being pursued. The labor-share stagnation is real and is the place where the reading of a transition moving toward the lower phase has to demonstrate trajectory rather than residue. The Gini increase is the strongest single argument for trapped-transition. It is the strongest because the productive achievement has not yet reached the household on the scale the thesis requires, and whether it does, with who controls investment and how durable the concessions are, is the joint test set below.

These data points are real and they are stress points. They are not yet the structural failure of the project. Productive substrate dominance does not depend solely or primarily on Chinese household consumption, but the escape route usually taken here, global electrification demand as an independent variable, cannot be taken by this work, because this work has spent its first half demolishing the independence. Run Section XV’s weapon at world scale. China is the world economy’s Department I, and its surplus currently realizes through three channels: the imperial core’s consumer demand, which the realization-circuit argument says is failing. Global South demand, which Section XI documents being crushed in the dollar-scarcity phase, the smallholder priced out of fertilizer being also the marginal buyer of Chinese goods. And internal demand, anchored by the numbers above in the high thirties for 22 years.

A Department I whose realization rides on a credit bridge is stable exactly as long as the bridge, and the bridge under the present configuration is dollar-denominated. What converts this from refutation into test is the substitution already underway in Section XVIII’s settlement data. The yuan credit bridge is replacing the dollar bridge beneath the same trade: CIPS above 1.22 trillion renminbi daily, the oil-for-goods architecture, 53% of China’s cross-border flows in its own unit. Whether it replaces the dollar bridge fast enough is an empirical race, and the race takes an observable now: renminbi-denominated trade finance extended to the Global South, the series that measures whether Beijing is financing its own realization or waiting for a dying system to do it.³⁰²

The economists Michael Pettis and Matthew Klein put the objection in its sharpest form: a formation that consumes 39% of its output and invests at diminishing returns has a surplus product it must place abroad, and if the core closes its borders and Europe rations and the periphery is short of dollars, no owner of the losses can prevent a debt deflation at home. Two answers, one on each side of the bridge. On the home side, the saving that suppresses consumption is precautionary rather than cultural: households save against illness, age, and housing that the state does not yet pay for. The campaigns of 2025 and 2026 are the beginning of the state paying for them. The childcare subsidy of July 2025, 3,600 yuan a year for every child under three, the first universal cash payment the state has made since 1949 on the health commission’s own description, with 90 billion yuan budgeted for its first year and 20 million families in it. The consumption plan of March 2025 that raised pensions and put 67 billion yuan into employment subsidies. The sums are small against the gap. The childcare payment and the employment subsidies together come to 157 billion yuan, about a ninth of one point of output, where one point of the consumption share is 1.4 trillion. They establish a direction and an instrument at a scale that cannot yet move the number, which is why the consumption share over the next four quarters is the test and these measures are only its first entries.

The urbanization plan of 2024 removes the settlement restrictions of the hukou in cities under 3 million and lets the migrant working class carry its entitlements to the city, which begins to end the arrangement Section VIII describes, a class that financed the accumulation from 1978 by working in the city without the city’s social wage. A state that pays for illness, age and housing raises the propensity to consume without a foreign buyer.³⁰³ On the bridge’s side, the vendor financing is on state banks, behind capital controls, and moving into the successor’s currency as the yuan bridge is built, and the borrower that fails, Pakistan or Zambia or Sri Lanka, leaves a loss that a sovereign can recapitalize by fiat with no foreign creditor, and with the currency mismatch moved off its own books once the loan is in yuan, the borrower still earning rupees or kwacha against a yuan bill, at a cost paid in domestic inflation risk that the formation has so far been running the other way, nine quarters of deflation. A shunt is a shunt whoever owns it. The difference is that this one’s owner can absorb the shunt’s failure without a run, and the incumbent’s cannot.

What the data establishes is that empirical indicators in 2026 cannot resolve the trajectory-versus-plateau question by themselves. Continued stagnation through 2029 is consistent with either awaiting the strategic-industry-class subordination test, or failure with the test never arriving. If labor share, consumption share, and wealth Gini continue stagnating through 2029 without the new productive-industrial class facing the institutional treatment the property class faced, the trapped-transition reading is empirically confirmed and the structural-durability claim is wrong. If the apparatus exercises against CATL or BYD by the end of 2029, an act comparable to the three red lines whose terms take command over investment from the class rather than hand it to another private winner, the trajectory reads as Phase 2 operating. The framework commits to that falsification structure rather than retreating into open conditional.

So the war does to China’s export route what the exhaustion of offshoring did to America’s. It closes the outside. That is the compression thesis applied to Beijing, and it forces the branch decision onto the domestic distribution of the surplus product inside the same window as the core’s cascade.

The periphery comes before the domestic question, because the yuan bridge is where the class content of Chinese accumulation is most often misread in both directions. Lenin did not define the export of capital by keeping the periphery agrarian. He said the opposite: the export of capital influences and greatly accelerates the development of capitalism in the countries to which it is exported, and Russia was such a country, industrialized on French and British money. The permanent strip mine is the dependency school’s addition, and it is contested even on its own ground. What makes capital export imperialist in Lenin is control and destination: finance capital of the exporting country holds the assets, the super-profits flow home and fund a labor aristocracy there, and the monopolist blocs divide the world among themselves.

So whether the periphery industrializes does not discriminate by itself. Lenin expected it to. What discriminates is who owns the capacity, where the surplus goes, whether the technology and the productive base transfer, and whether the relation subordinates the recipient state or is contracted between sovereigns with the recipient setting the terms. The socialist precedent for the last two is the one China itself received. The Soviet Union’s 156 projects of the 1950s transferred turnkey plants, blueprints, and cadres on credit repaid in goods, which is what building another country’s productive base looks like. China’s own second stage, the 1980s and 1990s, forced Western capital into joint ventures, local content, and technology transfer as the price of its market, which is what a recipient extracting industrialization from a capital exporter looks like.

That second stage is now being run against China, and it is the new structural fact. A capital exporter with overcapacity and no core outlet must export, which puts the recipient in a buyer’s market for capital, the reverse of Lenin’s era, when capital was scarce and the periphery competed for it. Indonesia is the case. It banned raw nickel ore exports in January 2020. Western trading partners filed disputes at the WTO and predicted failure. Its nickel export value went from $1.1 billion in 2014 to $36.2 billion in 2024, it accounts for roughly 58% of global production, and its nickel-related industries grew fivefold between 2013 and 2022.

The capital that built the smelters was compelled: more than $30 billion of foreign investment, primarily Chinese, into domestic refining and HPAL plants, because the ore could no longer leave the country unprocessed. Thailand ran the same play on autos, cutting EV import tariffs on the condition that every imported vehicle be matched by domestic assembly by 2025, and BYD, Great Wall, and Changan answered with more than $3.5 billion of greenfield factories. Africa’s version is the market side: on May 1, 2026 China implemented zero tariffs on 100% of tariff lines for all 53 African countries with diplomatic ties, unilaterally, with no reciprocal reductions required, the first major economy to do so. That is the periphery ending the strip-mine role by leverage, and Beijing complying because its overcapacity needs somewhere to go. The flood that destroys the incumbent’s rent is, from the periphery’s side, an industrialization subsidy: the panels, machinery, and vehicles a poor country could never have bought at the incumbent’s prices, sold at the challenger’s, and installed on the recipient’s terms. It is structural rather than charitable, and that is what makes it durable.

The other side of the same record is the test, and the record is honest about it. In Indonesia, Chinese firms control approximately 75% of refining capacity, China supplies 80 to 90% of the refining machinery, and China took 82% of Indonesia’s nickel exports in 2024. The figure that would settle the question, what share of the export value is captured by Indonesian capital rather than by majority-Chinese processors, does not appear in any public dataset. Local-content targets are set at 30% and audited outcomes have lagged the announcements. In Africa, the zero tariff arrives on top of a structure that is still raw for manufactured: 2025 trade of $348 billion, Chinese exports up 25.8% to $225 billion, imports up 5.4% to $123 billion, and a record African deficit of $102 billion, with Africa exporting oil, ores, and agricultural products and importing textiles, machinery, and electronics, and Afreximbank’s own read is that the measure needs broader reforms to be captured.³⁰⁴

On default, Beijing’s banks extend maturities and adjust financing costs rather than accept haircuts on principal. Zambia’s deal cut the coupon to 1% until 2037 and pushed maturities on $6.3 billion of bilateral debt to 2043, an average extension of more than 12 years. That is the deferral mechanism of the creditor section at state scale, insolvency converted into duration, and it cuts both ways at once: a 1% coupon for a generation is a loss taken through time rather than through principal, close to a grant in present value, and it is also a creditor protecting its claim. Against it stands the captive discount on sanctioned oil, Iranian and Russian barrels bought below the world price because the seller has no other buyer, which is rent extracted from a partner.³⁰⁵ The debt-trap reading, meanwhile, fails on its own exhibit: Hambantota was a 70% stake Sri Lanka sold to raise cash, with the proceeds used to service Chinese and other debt, a privatization rather than a seizure.

So the ownership is substantially Chinese, the surplus’s destination is substantially China, the terms are softer than the last century’s and harder than a gift, and the industrialization is real. All of it at once. The developmental content is being extracted by the recipient state’s leverage rather than granted by the exporter’s intent, which is exactly how China got its own, and the same act carries two class contents decided by the terms. A socialist formation exporting overcapacity to build a periphery that can absorb it is Branch 1’s world-scale move, and it is self-interested, because it creates the demand the capacity needs. A capitalist formation doing the same for interest and collateral is Lenin’s capital export.

The incumbent’s law has already ruled on the case, and the ruling goes in the ledger. The WTO panel found Indonesia’s ban inconsistent with its obligations in November 2022. Indonesia appealed into an appellate body the incumbent had itself emptied by blocking its appointments, and as of 2024 the appeal sat there with nowhere to go. The capital built the smelters anyway. The form used against a periphery state’s sovereignty over its own ore, the form’s enforcement organ dead by the incumbent’s hand, and the challenger’s capital constructing what the form said should not exist: that is the guarantor test of Section XXIII with a case attached, and it is the case the periphery will cite.

At home, the state’s response to the falling rate is on the record, and it points both ways in the same quarter. The anti-involution campaign launched in July 2025, with more than 50 measures across industries. Amendments to the national Pricing Law barring companies from selling below cost. Two-year plans for 10 key industries with 2025 and 2026 output growth targets set lower than 2024’s. And, in the other direction, EV purchase subsidies renewed into 2026 amid deflationary pressure. On one reading that is the plan restoring proportionality between departments. On the other it is the collective capitalist managing the rate of profit on behalf of its firms, banning the competition that lowers it, and exporting the overcapacity, which is the counteracting tendency on Marx’s list and the whole of the orthodox critique. The state managing the rate of profit and the state putting money in households at once is the branch question stated as policy, and where the margin the campaign preserves goes decides it: to the wage share and the social wage, Branch 1. To retained margins and export prices, the recomposition.

The trajectories of Korea, Taiwan, and Singapore, and of Eastern Europe after 1991, are the recomposition’s comparators: developmental states that industrialized under direction and resolved into subordinate nodes of the incumbent’s architecture, their planned forms absorbed into the accumulating purpose, their bourgeoisies competent for a generation and then rentier. The clause that distinguishes China from every one of them is the party in command, without the Russian rupture, and without the subordination that made the tigers’ competence temporary. Whether that clause holds is what the tests measure.

The exits are few, and the war has closed the historical ones. A capitalist formation with overcapacity and no external market has, in the record, three ways out: war, which destroys capital and seizes markets. Empire, which exports capital to a solvent outside, and redistribution, which raises the wage share until the population absorbs the surplus. The solvent blocks war as a settlement but not the war economy as demand, which is the Taiwan buildup read as military Keynesianism, and the air-naval section carries the failure mode it leads to. Empire needs a solvent outside, and the reason this crisis is different is that there is none. Financing the outside yourself is the shunt. It buys time and no more, and the time is spent well only if the periphery it builds can pay. Japan’s route, 30 years of zombie balance sheets, is unavailable, because Japan stagnated inside the American absorber and China has no absorber.

What remains is redistribution: the surplus product routed to the household as wages, public goods, housing, health, and pensions until domestic demand carries the capacity the world will not. That is Branch 1’s content arrived at by necessity rather than doctrine, and a capitalist China that takes it starts becoming what this work says it is, while one that refuses it goes to the war economy or to breakdown. The materialist claim is that the party’s consciousness follows the conditions, and the conditions are already forcing the question, which is what the campaign and the subsidies are. The war economy is the one exit that also builds the standing capacity to regenerate, which is why a state that takes it is hard to distinguish from one preparing to inherit.

The Grundrisse can now be read from the other side. The reduction of necessary labor to a minimum is the catastrophe the first half described, and it is the precondition of everything else for a formation that does not need profit to reproduce itself. Only that kind of formation survives the success horn, because only it can distribute use-values without a wage to buy them. The open weights are the value-form move already on the tape: giving away what the incumbent encloses, because the formation needs the capability and not the rent. A state that can distribute use-values by plan does not need the enclosure.

The tests, stated before the outcome so that either outcome counts. The household consumption share of GDP, and its direction over the next four quarters. The social wage: pensions, health coverage, and housing provision as a share of the surplus. Where the margin the anti-involution campaign preserves goes, wages or retained earnings, readable in the industrial-profit series against the wage series. Whether the rebate cancellations raise export prices or household prices. The terms of the next restructurings, Zambia’s and Sri Lanka’s successors, settled by haircut or by duration. Local content in the periphery audited rather than announced, the Indonesian value-capture share published or still absent, and the African deficit narrowing or widening under the tariff. The yuan-credit exposure against domestic absorption. And what a capitalist China looks like on the same tape, so that the reader knows it when it prints: Japan in 1990 at five times the scale without the absorber, deflation with the banks holding the claims and the households holding the bill, a fiscal crisis of the collective capitalist, or the war economy as its way out. This work’s claim is that the formation that inherits is the one that does not need profit to reproduce itself. Beijing is the only candidate, the law has reached it, and the next eight quarters are the test.