Section XXII: XXIIWhy China Is Different

This section carries the weight of this work’s second half. The argument is that the People’s Republic represents a structurally durable alternative architecture. The argument is contested by every major analytical tradition: liberal modernization theory, Western Marxism in its Trotskyist and council-communist forms, the post-1991 social-democratic synthesis, and the right-developmentalist tradition that admires China without granting it socialist content. The case has to be built against all of them. The test I’ll apply has three parts: First, what does the historical record of the Bolshevik tradition tell us about which failure modes a socialist project can encounter, and which of those modes the CCP has institutional answers to? Second, does the contemporary Chinese formation satisfy Lenin’s specific structural definition of imperialism, in which case its socialist content would be substantially compromised? Third, what does the empirical record of Beijing’s exercise of political authority over capital, particularly across the property-sector and tech-platform demolitions of 2020–2025, tell us about whether the transition is moving toward the lower phase Marx described or has settled into state capitalism as a destination? If China passes all three tests, the structural-durability claim holds. If it fails any one, the claim is in trouble.

One more tradition goes on the list, because it supplies the control case. Arrighi’s reading of the early 1990s, quoted in the pattern section, described East Asia as a capitalist archipelago: islands of centralized profit rising above a sea of low-cost labor into which they thrust their roots “without providing them with the means needed to rise to or above ‘sea level,’” every island dependent on the United States for military protection, energy, food, and the profitable disposal of its manufactures. Huntington’s triad for Japan, energy, food, and military security, named the dependency exactly, and the 1985-1987 harvest documented in Section XVIII showed what the dependency was worth when the hegemon called it in. The People’s Republic is that problem solved term by term: the security triad internalized, the strategic reserve demonstrated at half-rate import endurance through a closed strait, the food program, the deterrent, the laboring sea raised rather than tapped, and the islands’ export dependency replaced by the continental formation the archipelago could never assemble. In the archipelago frame, the durability claim this section defends is the claim that the sea learned to stand. The three tests that follow ask whether the standing formation is what it says it is, and the second of them, Lenin’s, is run in Section XXV.

One datum from the summer leads them, because it operationalizes the question. The class-character debate was never whether capitalists exist in China, the billionaires demonstrably survive their fines, but which direction the discipline runs, and the Western archive supplies the mechanism by which capital disciplines states: exit. Mitterrand was broken by flight, Syriza by a liquidity switch. Capital rules constitutional states through its option to leave. In May the securities commission fined the largest offshore brokerage $271 million and closed the channel itself, the travel-quota loophole declared off limits, the offshore trusts taxed as of late July, mainland accounts reduced to sell-only, the surplus rerouted into licensed domestic channels that return profits in yuan. The measure was aimed at the channel where earlier ones had been aimed at men. If it holds, a capitalist class that cannot flee cannot rule by the threat of flight, and the weapon that broke every Western subordination of capital has no ammunition. Whether it does is observable: the errors-and-omissions line of the balance of payments, where Chinese flight has always shown, through the next season of pressure on the yuan. This settles nothing by itself, and the strongest counter-reading, the bureaucracy as collective capitalist, consolidating itself by breaking individuals, is answered only by where the trapped surplus is then directed, which is what the tests below examine. But if the door holds the structural veto is gone, and its absence would be the strongest single datum the durability claim yet has.²⁶⁹

The exit was closed against a deal that had run for 40 years. Since 1978 the state had let private wealth accumulate almost untaxed in exchange for growth and employment: no tax on gains from mainland shares, no estate tax, no property tax outside two pilot cities: the machinery an American fortune has to build for itself, buy, borrow, and die, supplied in China as the default. The wealth left anyway, because there was nowhere at home for it to grow after 2007 and because a socialist state’s deal never promised that the tax would not one day arrive. Global Financial Integrity put illicit outflows at $3.8 trillion for 2000 to 2011; a trillion dollars left in 2015 alone; more than 15,000 millionaires emigrated in 2024; half of Singapore’s 10 richest were born on the mainland; and the Hong Kong trusts held $670 billion, more than half of it mainland and Hong Kong money.²⁷⁰

What ended the deal was visibility and need. China joined the Common Reporting Standard in 2015 and the account data began flowing in 2018, from 52 jurisdictions that year and 80 by 2024, Hong Kong, Singapore, Switzerland, the Virgin Islands, and the Caymans among them, with the United States the one hole in the net. At home the property registry, the digital tax system, and the traceable currency made what is held inside the country as visible as what is held outside. The land-sale revenue that had funded local government fell with the property sector, near half a trillion dollars a year of it, and personal income tax rose 13% in the first half of 2025 against growth of 5 on the tax read’s count of the ministry’s monthly data, almost all of it from high earners.

Then the doors closed in order: gains on American shares bought through Hong Kong brokers taxed, Hong Kong insurance policies taxed on their investment income, offshore trusts taxed, gifts out of the trusts taxed on the gain, a foreign passport no exit for anyone whose home, family, and business remain in China, all at 20% and with the gain taxed or carried to the heir at death, so that the one loophole the American fortune lives inside, the gain that vanishes when its owner dies, does not exist for the Chinese one. What remains untaxed is domestic, gains on mainland shares, estates, property outside the pilots, left alone while the markets that carry them are fragile. The sequence is the one the direction test needs: the exit closed first, the tax followed, and the fortune that stayed is visible to the state that let it form. The 25 richest Americans paid a true rate near 3.4% of their wealth growth on ProPublica’s count. The class that got rich first in China is being billed for the 40 years, and the bill is the deal ending.²⁷¹

The argument that what Beijing has built is structurally durable runs against confident prediction, made since 1949 in both liberal and Marxist traditions, that the project must fail. The predictions have been continuously wrong about timing while sometimes correct about underlying fragilities. The traditions making them have not revised their frameworks. What follows builds the case that those frameworks are the problem, not the project.

China itself faces two failure modes that have to be overcome. The first is ossification, the trajectory the Soviet Union exhausted, in which a project calcifies into catechism, cannot perceive changed conditions, and becomes incapable of self-correction. The second is capture, the trajectory in which a project becomes functionally identical to authoritarian capitalism, where the party-state apparatus gets captured by the domestic capitalist class it was supposed to direct, and the resulting formation is America or Russia 2.0. Both failure modes are live, but both are at least partially addressed by the institutional response Beijing has been constructing across the past decade. Whether the response is sufficient is what will be tested.

The history that produced the institution is in the section on the bubble on the sea, from the Bolshevik problem through the Cultural Revolution, and the liquidation section carries Reform and Opening as the other side of the offshoring it answered. What follows here is the adjudication: the tests the present CCP is running and the evidence for each.

The CCP is recovering a gradualist alternative the Bolshevik tradition had identified, tested, and lost, not inventing one. Bukharin’s path was a serious analytical position, not a deviation. Stalin’s destruction of it was forced less by doctrine than by an encirclement no gradual path could then survive, which Section XXIII takes up, and that is why the alternative could be recovered half a century later and not before. The question for the China argument is whether the lost alternative can be recovered under different conditions, and whether the conditions of the recovery prevent the failure mode that destroyed the original.

The Gotha Programme test

The CCP’s legitimacy through this period has come from the explicit acknowledgment that the system is not yet at the higher phase Marx specified. The primary-stage formulation names the present as socialism’s first stage, and this work reads that name as a transition still to be won, the status the name claims being the thing in contest. The CCP maintains the doctrinal trajectory toward the higher phase while permitting the developmental work the transition requires. The Critique of Gotha Programme passage, which Marx wrote in 1875, anticipates exactly this structure. Marx specified that the lower phase of communism retains many features of capitalist society: distribution according to work performed, labor-time vouchers as proxy for currency, the state as transitional apparatus, forms of bourgeois right in distribution. Only in the higher phase, after the productive forces have developed sufficiently and the antithesis between mental and physical labor has dissolved, does the principle “from each according to ability, to each according to needs” become operative.

The framing is double-edged. Used carelessly, it can rationalize indefinite preservation of capitalist features under the heading “we’re still in the lower phase.” That is the failure mode Trotskyist and other left criticism correctly identifies. The framing itself does not authorize indefinite capitalist preservation since a transition is a passage by definition, and the trajectory matters. A formation that retains capitalist features and is moving toward their abolition is doing what a transition is supposed to do. A formation that retains capitalist features and is not moving anywhere is ossified, antithetical to socialist development.

Between capitalist and communist society, Marx specified in the same text, lies the period of the revolutionary transformation of the one into the other, to which corresponds a political transition period whose state can be nothing but the revolutionary dictatorship of the proletariat. That is the transition. The lower phase of communism comes after it and presupposes the means of production already in common ownership, so that what survives of the old society is bourgeois right in distribution, the state reduced, in Lenin’s reading of the same passage, to enforcing that right. This matters for placing China, because Marx’s phrase for a society “stamped with the birthmarks of the old society from whose womb it emerges” describes the lower phase, and the lower phase presupposes value production already abolished. China is not there. China is in the transition, which is exactly where Lenin located what he named state capitalism under the dictatorship of the proletariat: commodity production, private enterprise, foreign concessions, and a national bourgeoisie, all retained to build the productive base while the commanding heights are held and political power is closed to capital.

Lenin called it a step forward toward socialism and marked it off from its Western namesake in one line, that the state in this case is represented by the proletariat and not the bourgeoisie. The objection that China still runs markets, private firms, and a bourgeoisie is then a category error rather than a fact the framing must absorb, lower-phase criteria applied to a transition formation, where commingled forms are the definition of the stage rather than evidence against it. The discriminator is not the presence of the forms. It is the one Lenin drew and Deng restated as markets being tools: whether capital is barred from political decision-making, and whether the class that holds the heights and directs accumulation is the party of the proletariat or the bourgeoisie. He had set the test eight months before, in September 1917: a state monopoly directed in the interest of the landowners and capitalists makes a reactionary-bureaucratic state, an imperialist republic, and directed in the interest of the whole people it is a step towards socialism, socialism being “merely state-capitalist monopoly which is made to serve the interests of the whole people”, with no middle course. The form decides nothing. Whom the monopoly serves decides it, which is what the household and labor tests below are written to measure.²⁷² The surviving test is still trajectory, now specified as a transition moving toward the lower phase rather than the internal motion of a lower phase already reached.

One inversion is an open question beside the test: the Preobrazhensky test read forward suggests that both a ruptural socialism in the imperial core and a China that takes Andropov’s turn, the disciplinarian’s reform from inside the party, are live routes to a resurrected task of democratic planning beyond state capitalism, which would make the present the transition Lenin described with the NEP, on a horizon measured in decades rather than years and at global rather than national scale. Stated as an open question beside the Gotha test, since this work’s discipline is to state such things before the tape decides them.

The successor points the same machines in a different direction, and the difference is what the Preobrazhensky test should read. Its anti-graft system of the last decade, Zero Trust, cross-referenced bank, property and construction records against the declared lives of officials across some 30 counties and flagged 8,721 of them before the officials it watched refused it their data and shut it down in most places, and the party’s discipline commission made a digital supervision system the center of its 2026 program. Pointed upward at the cadre and at the flows capital must report, the machine is the informational half of the override, the state able to see what it intends to direct, and the Andropov turn in digital form. Pointed by the cadre at the worker, it is guard labor automated, and the rollback shows the apparatus defending itself as a stratum does. The test has to read both, whether the supervision runs from above alone or is opened to the people it serves, which is the difference between discipline and democratic planning.²⁷³

Lenin’s pamphlet of 1917 gives the measures by which to read both states. He asked for the banks united under the state, the great syndicates taken over, commercial secrecy abolished, the employers compelled into associations, and over all of it accounting and control by the organizations of the workers. The successor has most of the forms. Its largest banks are the state’s, firms with some state ownership hold about two thirds of all registered capital, its tax system now reads invoices, bank flows and payrolls together, its platforms carry special state shares and its firms party committees, and its machines are pointed at its own officials. What it lacks is the last term: the books are open to the state and its discipline commission, and not to the workers’ organizations Lenin meant to do the reading. The incumbent runs the measures in reverse. Its banks are private and rescued. Its stakes in Intel and US Steel serve the firms that receive them, and the subsidy it gave the memory maker restrained that firm’s buybacks for two years and no longer. Its builders move their spending off the lines the market reads, and its machines are pointed down, at migrants and at the guards’ own capacity to refuse, which is his reverse control in practice, the worker’s life an open book and the capitalist’s accounts a secret. By his criterion the forms decide nothing and whom they serve decides everything, and the question of whom the successor’s forms serve waits on the opening of the books.²⁷⁴

The workers’-state question

The Trotskyist critique of the Bolshevik experience produces an additional sharper observation. Actually existing socialist states have suppressed workers and workers’ movements. Kronstadt 1921. Tiananmen 1989. Various labor strikes suppressed by socialist states. Polish Solidarity. If a workers’ state suppresses workers, the argument runs, how can it meaningfully be a workers’ state?

The observation has empirical content. Engels’s On Authority, written in 1872, provides one part of the analytical apparatus for engaging it. Engels argued that complex social production requires authority. A railway cannot run without authority. A factory cannot run without authority. Most fundamentally, “a revolution is certainly the most authoritarian thing there is; it is the act whereby one part of the population imposes its will upon the other part by means of rifles, bayonets and cannon.” The anti-authoritarian socialist either does not understand revolution at all or, by demanding the abolition of authority before the conditions for its withering have been met, ends up as objectively liquidationist in function.

Lenin extended this in State and Revolution. The dictatorship of the proletariat is a workers’ state with coercive capacity, defending the revolution and dismantling class society. The withering happens through gradual reduction as the conditions for class antagonism dissolve. Lenin acknowledged that suppression of counter-revolutionary elements is necessary, but argued that, over time, the apparatus of suppression becomes simpler and more diffuse, eventually performed by the armed people themselves rather than by a specialized state apparatus. This did not occur in the Soviet Union before dissolution.

The bare fact that a workers’ state can suppress workers’ movements does not by itself answer the question. Counter-revolution can wear and has worn workers’ clothing. Trade union consciousness and class conflict, and the intermingling of the two with a Communist Party, exist even after a revolution. Not every movement that names itself “workers” carries the working class’s historical interest. The same principle that the Trotskyist tradition applies to its critiques of existing Communist Parties should apply equally to its readings of other movements that have appeared as workers’ movements.

This does not justify any specific suppression. Engels’s argument legitimates the necessity of authority in general but does not legitimate the particular use of authority in any given case. Authority that defends the revolution against counter-revolution is justified. Authority that defends bureaucratic privilege against workers attempting to deepen the workers’ revolution might not. The same passage in On Authority makes clear that authority is justified by what it accomplishes, by whose interests it serves, by whether it advances or impedes the dissolution of class society writ large. Plekhanov set the condition in 1883: the conscious participation of the producers in organizing production cannot be replaced by any skill or daring of those who would organize it for them, and production managed by a caste would be patriarchal and authoritarian communism, under which the people would lose the capacity for further progress. The successor is the first formation to break through with much of the base already built, and whether its organization of production is opened to the producers or kept by the caste he warned of is the test this section has set.²⁷⁵

Marx’s analysis of the Paris Commune in The Civil War in France, written in 1871, shows the same balanced judgment in operation. The Commune was a workers’ state in formation. Marx praised its democratic forms: the elected and recallable officials, the abolition of the standing army in favor of the armed people, the direct accountability of all functionaries, the destruction of the parasitic state apparatus. He simultaneously identified its tactical errors: the failure to march on Versailles immediately, the respect for the Bank of France, the refusal to seize centralized power decisively when conditions required it. The same framework that defended the Commune’s democratic content noted that the Commune’s failure to exercise sufficient revolutionary authority at the right moments contributed to its destruction. Authority and democracy are coupled in Marx’s framework, not opposed.

Distinguishing requires analytical work, not categorical fiat. The same act can be defended on different grounds by different traditions, and the disagreement may be irreducible because it depends on judgments about what counter-revolution is or what advances or impedes the dissolution of class society in a specific case.

Take Kronstadt 1921. The Bolshevik defense holds that the rebellion was driven by counter-revolutionary Whites with peasant support, that it threatened the survival of the revolution at a moment of extreme weakness, and that suppression was a tragic necessity. Trotsky himself defended the suppression for the rest of his life, including from exile, even after he had become the most prominent critic of Stalinist repressive tactics that were explicitly built on the example of Kronstadt. The anarchist and council-communist counter-position holds that Kronstadt was a workers’ rebellion against Bolshevik bureaucratic-centralizer deformations, that the suppression was the beginning of the regime’s degeneration, that the organic soviets, the organs of workers’ democracy, were already being absorbed and crushed into the party-state apparatus. Both positions have evidence. The disagreement is partly about what counter-revolution looks like when its forms include workers acting against the workers’ state.

Take Tiananmen 1989. The CCP defense holds that the protest movement was being captured by capitalist-restorationist forces, that workers’ movement elements were a minority, that the suppression preserved the revolution against a deeper counter-revolutionary moment that would have ended the project’s continuity. The opposition position holds that the protests included substantial workers’ movement content (the autonomous workers’ federation), that the suppression was the bureaucracy defending its accumulating privilege against pressure for democratic deepening, that the CCP’s subsequent rightward drift in some dimensions confirms this reading. Both positions have evidence.

The record supports a mixed verdict: suppression of workers’ movements by workers’ states is sometimes a necessary defense of revolution, sometimes betrayal of revolution by bureaucracy, and sometimes both at once. The same act can have multiple components running simultaneously. A revolutionary state can suppress a workers’ movement in self-defense and in service of bureaucratic privilege at the same time. The analysis required to disentangle these is harder than declaring categorically that all suppression is betrayal or all suppression is necessary.

The Tiananmen question does not admit of a one-sentence verdict, and a reader who demands that it does will find no historical or future formation in the actually-existing socialist tradition that qualifies.

The deeper point against the Trotskyist and left demand that workers’ democracy in council form is the criterion for socialism: workers’ democracy is not a single monolithic form. The question is how authority gets exercised, by whom, accountable to what, whether the conditions for workers’ political agency are maintained over time. A council form that exercises authority democratically can ossify too. A party-state form that exercises authority through cadre democracy can preserve workers’ agency through other mechanisms. The demand for the organic sovereign council form specifically, applied as the test of legitimacy, makes it impossible for any actually existing workers’ state to qualify, because the pure council form has never been militarily and economically sustainable as a sovereign state form for more than a few months to a few years against sustained military-political-economic challenge from capitalism.

The state-capitalism question

The dominant relation is being tested at the transition stage Lenin specified, where state capitalism under proletarian power is the expected form, and the party’s own name for the present, socialism’s primary stage, is the claim in contest. Braudel’s sentence, which Arrighi made the hinge of his own book, is the one to invert: capitalism triumphs, in Braudel’s words, “when it is the state.” In the successor the identification runs the other way, the state over capital, and the direction of the transfer is the test of which it is.²⁷⁶ The form is the weak link’s inheritance, told in the challengers’ century: a state that had to make its own proletariat operated wage labor to make it, and a charge that reads the necessity as restoration rules out the revolution that produced the state. Against the related Trotskyist position that the USSR and by proxy China are state capitalism, the argument is more substantive. Tony Cliff’s analysis specifically argued that Soviet state enterprises extracted surplus value, that the bureaucracy functioned as a collective capitalist allocating surplus, and that production was for accumulation through plan fulfillment, which functioned analogously to capital self-expansion.

This is a specific structural claim, not an unfalsifiable categorical move. The structural counter-arguments in the Soviet case are specific too. Production was not dominantly for exchange. Most output was for use within the planning system. The competition element of capitalism, among enterprises at home, was structurally absent.²⁷⁷ Capital accumulation as an automatic drive was mostly absent. Surplus was allocated through political-administrative processes rather than extracted as profit through M-C-M’.

The liberal version of the same objection is the economist Branko Milanovic’s: what this section calls a state over capital is an autonomous bureaucratic elite that uses markets to maximize national power and its own discretion, and its crackdowns, the Ant listing, the tutoring industry, the developers, are self-preservation against a rival power base rather than anything a class did. The answer is the direction of the surplus, the test this section runs and the one the objection has to meet. In its crude form a bureaucratic elite that governs for itself loots the state to enrich its own and parks the proceeds abroad. The record of that formation is Suharto’s Indonesia, Mobutu’s Zaire, and Yeltsin’s Russia, and its signature is the offshore account. The party-state closed the offshore exit in May and taxed the trusts in July, liquidated the largest private fortunes of the property era without socializing their losses, and put the state banks into unprofitable, long-horizon physical assets, the ultra-high-voltage grid, the terawatts, the fabs, the hulls, at returns no rent-seeking elite would accept. A looting elite does not do those things. Milanovic’s elite might, since growth and national power are its only title to rule, and his bureaucracy disciplines fortunes while it tolerates the corruption it cannot end. What his model cannot do without is a private capitalist class left in place as the engine of that growth, and an inequality accepted as its price. A formation carrying a class into production can dispense with both, whatever its cadre think of themselves. The test is not settled by the crackdowns. It is settled by what the discipline leaves behind: a private capitalist class still reproducing itself at the center of accumulation, which is Milanovic’s China, or a class whose command over investment and the surplus has been taken from it. The strategic-industry class getting the treatment the property class got is the one move his elite cannot make without killing its engine, provided the treatment takes command over investment from the class as a class and does not hand it to a new private winner. The consumption share, the social wage and the wealth Gini bear on the question together, and none of them alone settles it, since a richer household is what his elite rules by while the growth lasts, and capitalism itself has conceded the household a larger share whenever it had to, the 30 years after 1945 the longest such concession. What settles it is the concession that is not taken back when the return on capital falls, and who is left to take it back. Capital has taken its concessions back whenever the rate demanded it, the incumbent last and now, its top tenth carrying half the spending while the rate is defended. A formation whose private class no longer commands investment has no one who must take the concession back, only a plan that can choose arms or accumulation over the household instead, and that choice is what the tests below watch for. Those are the falsifiers the closing section names.²⁷⁸

What is unfalsifiable is the stricter criterion that says any system with markets, money, state, or wage labor is capitalism. That criterion makes Lenin’s transition into capitalism, which is anti-Marx on its own terms. Cliff’s analysis was not operating that criterion. Cliff’s analysis was operating a substantive structural claim about whether the dominant relation in the USSR was capitalist accumulation through bureaucratic-collective form. That claim is debatable on its merits.

Beijing today operates structurally differently from Western capitalism in ways that produce different outputs and threaten the existing global ordering principle. The CCP’s own answer to the Cliff position was articulated by Deng during the 1992 Southern Tour. “Capitalism may have markets, socialism may also have markets.” Markets are tools, not modes of production. The mode of production is determined by the dominant social relation, by who owns the productive forces, by who directs accumulation, by which class exercises political authority over the strategic project. Beijing operates markets as tools deployed within socialist construction. Cliff’s position in its strongest form puts the competition outside: the arms race compels the state to accumulate as a single capitalist whatever its domestic markets do, so the question it puts to Beijing is whether the world market and the war economy set the pace of accumulation, whatever is deployed at home. The answer this work gives is the household test: a state accumulating under the arms race’s compulsion could not sustain the household’s share against the return on capital when the two press against each other, and whether Beijing does, across the 2029 tests and not in one year’s print, is what they are written to catch. Cliff’s own criterion gives a second test beside it: whether labor power is priced against accumulation’s needs or against the household’s, which the record of hours, arrears, the labor share and the migrant’s status will show by the same date.²⁷⁹ The protocol difference, with state ownership of commanding heights, party political control, refusal to socialize losses to private capitalists, and capacity to subordinate specific factions when their accumulation threatens the strategic project, is the empirical answer.

A libertarian looks at the present administration’s equity stakes and state direction and calls it communism, and he is wrong for the whole criterion: the direction serves capital, secures bourgeois accumulation, enriches the family and the firms, so it is bourgeois despite the form. The reading cuts both ways at once, which is the part to hold. The orthodox critic cannot read bourgeois off the presence of markets any more than the libertarian reads communist off the presence of state ownership, and this work cannot read transitional off Beijing’s discipline of capital, for the same reason. The form settles nothing on either side. What settles it is who the state serves and how, and who serves in the state, the composition of the cadre and how those served interact with those who serve. State direction is universal now. Nothing but the regulator question is left to decide it, which is where the discriminator and the tests stand.

The discriminator underneath the protocol difference can be stated in value-form terms, which is where the strongest contemporary version of the state-capitalism argument lives. Postone and Heinrich, the value-form theorists, read capitalism as impersonal domination by the value form rather than as a fact about ownership: production is validated through exchange, the law of value passes the verdicts, and no owner, private or state, can refuse the verdicts without ceasing to accumulate. The question is then whether the value form governs reproduction in the PRC or is instrumented by it. The record shows instrumentation.

The law of value operates in bounded domains and is overridden at the commanding heights, where plan targets are denominated in use-values, grid capacity, housing floor space, fabrication nodes, hull tonnage, and met in use-values whether or not the exchange verdict approves. Strategic sectors run losses for decades without dying. Evergrande shows the limit of the verdict and the limit of the override. The value form passed a verdict of collapse on the largest developer in the system and the verdict was carried out on the firm, which was liquidated. What the state refused was the cascade the verdict compels in a capitalist credit system, and it set the order of the losses politically: the unfinished homes first, the offshore bondholder last. A social formation in which the value form’s verdicts are politically interruptible at the heights is not governed by the value form at those heights, whatever operates in its retail domains. The criterion is falsifiable and should be written as such: the day the party cannot override a market verdict at the commanding heights without systemic collapse, the value form governs, the instrumentation has inverted, and the state-capitalism reading wins.

The restoration question carries an evidentiary asymmetry that the debate rarely states. Restoration in every observed case arrived as a rupture with a date. The Soviet sequence has one: the property laws of 1991 and 1992, the voucher privatizations, the nomenklatura converting stewardship into title, the enterprises of the plan re-chartered as the estates of a class. Restoration is a property revolution, and property revolutions leave records, statutes, share registries, oligarchs. The Chinese record contains no such rupture. A restoration could arrive without one, by the slow conversion of stewardship into command that the rival reading would have to document decade by decade, and the burden below is written for that case too. The commanding heights were never mass-privatized, cadre stewardship was never converted into heritable title across the system, and capital’s political organs were never chartered. The absence cuts both ways and honesty requires saying so.

It means the restoration the state-capitalism school dates to 1978 or 1992 cannot produce its rupture and has not yet documented the cumulative transfer that would replace one, and it means the contest the primary-stage doctrine names has not been finally resolved, because the same continuity that blocks the restoration verdict blocks the consolidation verdict. The asymmetry is in the burden each reading carries. The restoration case must point to the transfer of title and power, and has not. The construction case must show only that the contest is live and the trajectory since 2012 runs toward the heights, which is what the paragraphs below document. What remains between the two positions after the categorical claims are cleared away is time-scale, whether four decades of transition is a phase or a destination, and time-scale is a question the falsification structure in the closing section keeps open rather than settles by fiat.

This apparatus also settles an account the framework has carried since 2010. The strongest critics of Arrighi’s China book, Leo Panitch first among them, found its central weakness precisely here: markets conceived as instruments of states with the theory of the state left undeveloped, and China rendered non-capitalist nearly by definition because its development ran outside the core. The criticism was correct about the book. The answer is the apparatus above, which replaces the definitional move with tests: the transition-formation placement, the value-form instrumentation criterion with its stated falsifier, the closed exit, the restoration-rupture asymmetry. Arrighi’s own criterion in that book, whether the capitalist class controls the state or the state controls the capitalist class, is what the exit closure and the Evergrande override are evidence entries in. The claim survives or fails on the tests, not on the definition, which is the difference between this section and the book whose question it inherits.

The configuration has an exact image from the last time a revolution took the old order’s instruments into its own hands. In December 1804 Napoleon took the crown from the pope’s hands and set it on his own head, then crowned his empress, with the papacy present and conferring nothing. David painted the second gesture because the first was too naked to hang. The WTO accession of 2001 would fill the same canvas: the party entering the market’s own cathedral, taking the regalia of exchanges, billionaires, and export rankings into its hands in front of the priesthood of the order it was joining, and never once conceding that the priesthood confers legitimacy, which the mandate had already conferred on itself. The bourgeois revolution universalized its content by betraying its form, the dynast outside the dynasties spreading the Code at bayonet point across a continent that had refused it. The form-content split is the exact question this section’s tests answer from the other side: whether the crown is carrying the Code or only the Empire, whether the instruments are wearing the state or the state the instruments.

The institutional response

What carries through this dense historical material to the present moment is the structural picture of how the CCP has constructed its institutional capacity to address both failure modes. Ossification was the Soviet failure mode. The response Deng built after the Cultural Revolution was procedural: term limits, collective leadership, retirement by age. The Xi consolidation set much of that aside, and the challengers’ century says what that risks, since concentration is how the Soviet apparatus hardened. What it bought is told below. What it put at risk has a test with a date, the succession, and the congress of 2027 is the first place it prints.

Capture is the second failure mode, the one that became most visible across the Jiang Zemin years of the 1990s and 2000s when the Shanghai faction was associated with coastal market reform, business-elite alignment, and the structural risk that the CCP would become functionally captured by the capitalist class it was permitting to emerge. The response to this failure mode has also been the Xi consolidation, which has included an extensive anti-corruption campaign that took down Bo Xilai and substantial elements of the Jiang network, the property-sector controlled non-rescue, the tech-platform crackdown, and the Common Prosperity campaign.

The property-sector demolition is real. Real estate’s direct contribution to GDP fell from roughly 7.3% in 2020 to under 6% in 2023 on the value-added measure, with the broader real-estate-and-infrastructure share falling from a 2015 peak near 34% to roughly 25% in 2025. Evergrande liquidated with over $300 billion in debt. Country Garden defaulted. Vanke required state intervention by January 2025. At least 50 major developers entered restructuring. State-owned developers now take the large majority of land purchases. The state cut the sector’s credit with the three red lines and then declined the rescue. It declined to socialize the private developers’ losses to preserve them, declined to bail out their creditors, and declined to extend the implicit guarantee that had underwritten the prior expansion. The support that came later went to finishing the buyers’ homes and to the sector’s function, while the developers’ equity and their offshore creditors stayed where the refusal had put them.²⁸⁰ Refusal of rescue at that scale is itself a doctrinal move, the party-state that claims political authority over capital exercising that authority against a specific capitalist faction whose accumulation had begun to threaten the strategic project.²⁸¹

The rest of the ledger is on the same scale and this work owes it the same standard. The local government financing vehicles, the off-budget borrowers that built the infrastructure the plan ordered and the land-sale revenue then stopped paying for, carry a debt that China’s executive director at the Fund put at 44 trillion yuan and that market estimates put above 60 trillion. The augmented government debt that includes them stood at 124% of output on the Fund’s 2024 count, against a general-government figure of 60 and a total non-financial debt of 312%, and the 12 trillion yuan swap of November 2024 covered about a quarter of the hidden load on Fitch’s estimate. Household consumption has sat between 38 and 40% of output for two decades, the lowest share of any large economy, and the labor share Section XXIV gives has moved only modestly. August’s releases printed the divergence again, industrial value added up 5.2% on the year and retail sales up 0.4%, fixed investment down 7.2% over January to August, and the Bureau’s own release naming an imbalance between strong supply and weak demand.²⁸²

The 2022 mortgage boycotts, buyers on more than 300 stalled projects refusing to pay for apartments that were not being built, and the Henan village-bank freeze the same summer, several billion dollars of deposits locked and depositors met by plainclothes police, were runs on the base of the system, and the state ended both by administrative order rather than by market clearing. The population has fallen every year since 2022, and the youth unemployment series was suspended at 21.3% in June 2023 and redefined before it resumed. None of this is hidden and all of it is the falling rate’s ledger on the other side. The difference this work claims is stated in Section XXIV. The losses are held by an owner that can write them down without a run, and what matters is where the surplus goes while they are held.²⁸³

The tech-platform crackdown was also substantive. The Jack Ma humiliation of late 2020. The Ant Group IPO suspension. The ride-hailing platform Didi’s delisting. The claim these facts carry does not move with the stock cycle: the platform, property, tutoring, and fintech crackdowns happened, imposed sector-wide losses, left no victorious private faction, and the regime remains in force, with Ant still broken up, the three red lines’ monthly reporting dropped by the regulators in January 2026, and the commanding heights still state-owned. PIIE’s tracker puts the private sector at 41% of the market value of the top 100 listed Chinese companies by mid-2026, 27 of the 100 new to the list and 26 of those upstream makers of chips, components and equipment, recovering from the 33.5% low of mid-2024 and still far below the 55.4% peak of mid-2021. By ownership as against market value, firms with some state ownership held about 68% of the registered capital of all Chinese firms in 2017 and the state’s own capital 31% of the whole, on the Stanford center’s reading of the registration data.²⁸⁴ The measures answer different questions, what share of the market’s value private owners hold, what share of registered capital sits in firms the state has a stake in, and what the state itself owns, and none of them alone shows who controls investment, which is why the class test below is joint.

The rebound was driven by technology firms in the AI boom, with Tencent, Alibaba, CATL, Xiaomi, and BYD among the private names in the top 100. Recovery within bounds under continued structural crackdowns is subordination rather than abolition, the disciplined-and-allowed-back pattern a restoring state would not produce and a captured bourgeoisie would not survive. Capital is being channeled instead into the strategic emerging industries through what the Brookings Institution calls the “new national system.” Strategic emerging industries rose from 7.6% of GDP in 2014 to over 13% in 2022, with a target of 17%-plus by 2025. The 2025 figure has not been published in the series the target was set in, so the target stands unscored. The crackdown and the reallocation read as one operation, accumulation redirected by plan.

What the allocation discipline demonstrates is the political subordination of capital, and that part is already in hand. The property demolitions, the three red lines, the refusal to socialize the developers’ losses, the freed capital pushed into strategic industry against the return it could have earned in the inflated asset, are present-tense facts, and they establish that the party stands above capital and directs it. The deeper question the Marxist economist Michael Roberts presses is whether the law of value is being subordinated as the general regulator of investment across the whole economy, or whether the plan commands the strategic core while value still allocates the rest. That is the discriminator. Command over capital is demonstrated and in hand. Value displaced as the general regulator is on the horizon and not yet shown. Accumulation enlarges the material possibility of labor’s emancipation. Whether the possibility is exercised is read in labor’s position, and a higher household share, a larger social wage, wider authority over the labor process and shorter working time are related outcomes and not one, which is why the household test and the labor test are set apart.²⁸⁵ The falsification structure already committed to is what tests the second.

Whether this institutional response is sufficient to prevent capture is the open question the next years will test. The test is whether Beijing subordinates the new strategic-industry class as it did the property class, and it has a clock. The property class stood in 2018 where this class stands now, the three red lines came two years later, and the largest developer defaulted the year after. If by the end of 2029 the fortunes of this class have stayed intact through whatever successions have come, its owners sit in the organs that set policy beyond the consultative bodies they already sit in, and no comparable act has come against it, the capture reading has won. The new productive-industrial billionaires (Zhang Yiming of ByteDance, Robin Zeng of CATL, Wang Chuanfu of BYD, the new AI billionaires at MiniMax and Zhipu) are the class in question. Beijing’s defenders have to commit to the test.

The answer to that test depends in part on whether the apparatus has cadre-base pressure under it for the act. The pattern documented by Yinhao Zhang in “The Material Basis of a Spectre: Why China’s Youth Are Rediscovering Mao” (Monthly Review, Vol. 77, No. 11, April 2026) is what would be expected if the pressure were forming.

At Tsinghua University, the Selected Works of Mao Zedong went from outside the top 50 library loans in 2016 to the number one position by 2019, a position held every year through 2024. The pattern repeats at Peking, Fudan, and Shanghai Jiao Tong, the four most prestigious universities in the country and the source of the next generation of senior cadre, academic, and business leadership. A 2020 MyCOS survey found the pattern across the top 10 loan lists of 13 of 80 universities surveyed, weighted toward top-tier institutions. Beihang’s 2020 list, posted on the university’s official Douyin account, showed the Selected Works first and Volume V second. Volume V was compiled during the Cultural Revolution, published in 1977, and effectively suppressed after 1978 for its radical content and direct criticism of Deng Xiaoping. Most university libraries do not carry it. The active pursuit of the most politically charged Mao text by students at one of China’s top engineering institutions is the signal.

The platform pattern confirms the character of the movement. On Zhihu, a 2017 question asking who was the greatest figure in Chinese history was flooded with pro-Mao answers from high-voted users. The platform’s pro-Western liberal owners closed the thread under a procedural pretext. A 2020 attempt at the same question produced more than 5 thousand responses, with the highest-voted answers naming Mao, and the platform deleted many of them. The thread survives in heavily censored form with its top-voted answer reading: “The people say it is him; he says it is the people.” The April 2025 tariff confrontation produced a peak WeChat index of 139 million for “Jiaoyuan,” the term that emerged among educated Chinese youth as workaround for the platform censorship of “Mao Zedong.” Articles and videos on On Protracted War circulated across the platforms at scales the state did not produce. The fever survives suppression from both directions: the state’s wariness of uncontrolled Mao discussion, particularly on the Cultural Revolution and the post-1949 radical period, and the pro-Western liberal media elites who own the private platforms and treat pro-Mao content as politically sensitive.

The material substrate is the contradiction between the post-1978 social contract’s promise of mobility through education and the actual labor market the educated youth face. Youth unemployment hit a 21.3% official rate in June 2023 before the government suspended the release.²⁸⁶ The “involution” and lying-flat movements have named the condition in popular vocabulary. The class stratification documented in the red princeling scandals (the Mercedes in the Forbidden City, the Beiji Nianyu affair, the 4+4 medical track) has made the formation Mao called “bureaucratic capitalists” legible to the demographic that should have been its natural cadre. The same generation that has experienced the failure of the market-era promise is also the most patriotic and anti-imperialist cohort since the reform era began. The two registers are coupled: as Zhang frames it, the new domestic business elite is read as both exploitative and comprador, aligned ideologically and sometimes economically with the Western interests Beijing is contesting at the geopolitical level. The struggle for social justice at home and the struggle for national sovereignty abroad are the same struggle in the same vocabulary.

This is the Chinese-side correlate of the imperial core’s own political-economic articulation contest. The same principle that material conditions force consciousness operates bilaterally. The educated youth at Tsinghua and Peking are not the demographic the CCP necessarily needs to convert, they are the demographic whose material conditions have already produced the recognition the structural-durability argument requires. The doctrinal commitment to phase-two consolidation now has cadre-base pressure forming under it from the demographic that could be running the apparatus in 15 years.

Zhang’s own framing of the limits stands. The movement is concentrated in the digital realm. It lacks formal organization. The same students reading the Selected Works as revolutionary text are also reading it as a manual for office politics and individual advancement in the hypercompetitive labor market. The revolutionary content sits alongside the self-help reading and can be defanged by it. Zhang describes the current configuration as “a loose, sometimes confused mix of feelings rather than an organized movement.” That is what an unorganized but theoretically energetic ferment looks like in its pre-formal phase, on the pattern the articulation section names for the imperial core, a party in formation that is still drawing its lines. Class consciousness has formed under material pressure on both sides of the inverted topology. The political vehicle that absorbs it into organizational form is the work the next phase has to do.

The difference between the two ferments is not their size or their degree of organization, which are comparable, but the position of the party each faces. In the core the ferment’s text is the enemy’s text for both parties: the apparatus spent $65 million against its candidate in Michigan and lost, and the coercion layer met its movement in Minnesota. In China the ferment’s text is the ruling party’s own canon. Volume V is the party’s compilation of the transition years, and the demographic reading it is the demographic the party recruits from: the cadre are taken from the universities where the loans are counted and run through the party school on the same texts. Since 2021 the readers have been answered in policy with what they wanted, the platform crackdown, the tutoring ban, the ruling against the 996 week, common prosperity, the anti-involution campaign.

In the same years the ferment’s organizational expression outside the party’s channels has been policed, the Jasic affair of 2018 the case, when the Marxist student societies that organized with Shenzhen workers were suppressed. That is the pattern: absorb the doctrine, recruit the readers, answer the demand in policy, forbid the independent organization. It is the trapped transition’s risk, an apparatus renewing itself from a ferment it will not let stand on its own, and it is also the capacity this section claims, a party that can turn the pressure into cadre selection where the core’s parties can only fight it or be replaced by it. The cadre-base pressure the class test needs is not a movement outside the party. It is the generation inside it, the thing the Soviet apparatus never had.²⁸⁷

What can be said with confidence is that the CCP has built the institutional capacity to act against capture in ways the Soviet framework never did. Whether it has kept its capacity against ossification is what the succession will show. The history of the years since 1949 is the history of that construction.