The Surplus Has No Author

Beijing's defense of its industrial surplus is built from truths made individually innocent.

Eleven days separated two official descriptions of the same industrial machine. On July 17, 2026, Luo Wen, director of China's State Administration for Market Regulation, published "Deepening the Rectification of “Involution-Style” Competition and Creating a Sound Market Environment" on page ten of the People's Daily. Some industries, he wrote, had become trapped in a crude struggle for scale and market share, leaving “excess capacity and weak innovation coexisting.” Local governments, public investment funds and local finance were all implicated, leaving industrial parks to copy one another while firms expanded blindly.

Luo then followed the goods out of China where some firms had extended homogeneous production and price competition into foreign markets, compressing their own profits and provoking trade friction. He called the process 内卷外化 (nèijuǎn wàihuà), involution externalized.

On July 28, the Ministry of Commerce released "China's Position on the So-Called Excess Capacity Issue" in Chinese and English, addressing foreign governments that accused China of overwhelming their industries. The paper tried to cleave industrial subsidies from excess capacity, trade surpluses from overproduction, and economic imbalance from either. Its title had already placed the accusation at a distance through the use of 所谓 (suǒwèi), “so-called.”

The documents do not quite contradict each other. Luo wrote about pathologies in particular sectors, while the ministry answered a much broader foreign charge against Chinese manufacturing. Their institutional purposes also differed. A regulator repairing industrial competition must identify mechanisms, while a commerce ministry resisting foreign restrictions must prevent those mechanisms from becoming a national confession. Even so, the overlap is large enough to matter with Luo assembling the machine, and the ministry returning eleven days later to remove the embarrassing joinery.

The ministry's treatment of subsidies shows the general method at its most concise. Industrial subsidies and excess capacity, the paper says, have no 必然联系 (bìrán liánxì), no necessary connection. But the foreign charge is narrower, holding that subsidies can preserve capacity that prices and profits would otherwise force to contract, particularly where local governments reward scale and banks delay exit. The ministry therefore converts a claim about contribution into one about inevitability, then defeats the stronger claim before declaring the cause absent.

That maneuver, as transparent as it is, gains force from the carelessness of some foreign criticism. Few foreign critics acknowledge that overcapacity has no universal threshold. Moreover, utilization differs across industries, spare capacity can anticipate demand still to come, and a country has no obligation to consume everything it makes. Boeing sells most of its commercial aircraft abroad, which says very little on its own about whether America has too many aircraft factories.

Neither can China's advantages be said to be artificial. Dense supplier clusters shorten the passage from design to mass production, while cheap logistics and capable engineers allow firms to iterate quickly under competitive pressure. Chinese batteries, for instance, improved as their prices fell, while Chinese solar equipment lowered the global cost of decarbonization. A February 2026 comparison by the Rhodium Group attributed only about 5 percent of BYD's estimated $4,700-per-vehicle cost advantage over Tesla in China to subsidies, with vertical integration and lower overhead costs accounting for considerably more. Consumers elsewhere received better goods for less because Chinese companies became exceptionally good at making them.

Foreign governments enter the argument with evasions of their own. The United States and Europe subsidize certain industries while presenting Chinese industrial policy as a departure from normal commerce. In March 2026, Washington opened Section 301 investigations into structural excess capacity across sixteen economies, including the European Union and Japan, making excess capacity a category broad enough to describe a genuine economic pathology, a fear of strategic dependence, and a case for protecting domestic producers, without requiring governments to say where one ended and the next began.

And that ambiguity compromises the accusation without resolving the condition it names. Chinese competitiveness explains why its firms win orders, whereas the existence of structural excess depends on whether losing firms inside China are allowed to exit. Such excess need not announce itself through idle machinery. Finance or political protection can keep production running beyond remunerative demand, shifting the adjustment into losses, falling prices, extended credit or foreign markets even as factories earn too little to sustain the capital and suppliers behind them.

A superior Chinese firm can therefore displace an inferior foreign rival through ordinary competition even while its industrial system at home preserves too many producers, presses the cost of survival down the supply chain, and asks foreign markets to clear the result. Superior efficiency and deferred adjustment can enter the same battery. Wherever production outruns remunerative demand, someone must carry the difference, and lower prices settle consumers' bills while helping determine which factories close and where the technical skills and supplier networks required for the next generation remain. Efficiency explains who wins the order without deciding who carries the loss.

The present episode also differs from earlier rounds of Chinese export expansion. As David Lubin of Chatham House notes, Chinese export growth has often outrun world import growth before, but the gap usually narrowed again as Chinese domestic demand recovered. Since early 2023 it has remained open, with Chinese export volumes growing faster than world imports without interruption. The divergence does not by itself prove excess capacity, but it shows foreign markets being asked to absorb Chinese expansion faster than their own demand is growing.

The ministry avoids that allocation question by giving each possible cause a separate room in which to establish an alibi. Subsidies do not invariably produce surplus; a trade surplus does not invariably prove excess capacity; domestic demand contributes most Chinese growth; competition should remove producers whose goods cannot be sold profitably. Each proposition contains truth, while responsibility disappears through the cracks made by their separation.

Subsidized entry can multiply firms in a favored industry while weak household consumption limits the market available to them. Local governments can then delay closure, allowing competition to punish margins without reducing aggregate production, before exports absorb what domestic demand does not. Once the walls between the ministry's propositions are removed, its separate alibis become a single causal sequence.

Beijing aggregates achievement and disaggregates responsibility. When the subject is industrial success, the state appears whole, and the ministry describes an “enterprise-led, government-guided” investment system backed by research facilities and infrastructure, with public power carrying innovation from laboratory to production line. When the subject turns to imbalance, however, support becomes one influence among many, local protection is framed as an aberration, and the final surplus becomes an outcome nobody chose.

The asymmetry is sharpest in the paper's appeal to market competition as the most effective mechanism for preventing disorderly expansion, not least because its sequence is clean. Goods fail to sell, firms lose money and weak capacity leaves the market. That ending, however, depends on loss becoming exit, and when weak producers remain financed and dominant manufacturers pass the cost of survival down the supply chain, competition can drive prices and profits down without forcing capacity from the market. In this situation, every producer may be punished while the system still fails to contract.

No conspiracy is required to make this hang together. Different institutions see different parts of the machine because they answer to different political tasks, and their documents become revealing only when placed together. What emerges is a gradient of permissions and returns, with no central order book — the center may name desirable structures, but it's local governments, firms and banks that decide how vigorously to reproduce them.

The gradient begins with an explicit preference for manufacturing. The Communist Party's Third Plenum in 2024 instructed the state to “establish an investment mechanism for maintaining a reasonable share of manufacturing,” paired with lower costs and tax burdens as official explanations directed labor and capital toward manufacturing through land, infrastructure and technical knowledge. Technological self-reliance raised the political value of domestic capacity even where no current order book could justify it.

That preference is defensible. A large country may value industrial depth, and China has watched Western economies regret the disappearance of capabilities their markets once classified as redundant, while security places a value on capacity that a quarterly return cannot capture. The difficulty begins when every locality reads the same strategic signal.

A city that lands a battery plant gains construction, employment, tax receipts and political prestige while bearing only a fraction of the national cost when dozens of other cities reach the same conclusion. Within this ecosystem guidance funds attract private capital, development zones copy successful neighbors, and local banks hesitate to recognize losses in firms tied to the surrounding economy, leading to a result that Luo described with unusual clarity: investment funds became disguised subsidies, while industrial parks acquired “a thousand parks with one face,” and zombie firms remained upright through local finance.

The center can officially oppose waste while preserving the incentives that generate it. The Ministry of Commerce insists that “China never deliberately pursues a trade surplus,” a claim that remains plausible in the narrow sense that the surplus itself need never appear as a policy target. Deliberate pursuit, however, sets too exacting a test of authorship. The state chooses the scale of industrial production it wishes to preserve, lowers the cost of producing it, rewards self-reliance and maintains an economic structure in which households absorb a modest share of national output.

Localities compete within those choices, producing a trade balance that is neither commanded nor accidental. It is the least disruptive means of reconciling the capacity Beijing wishes to preserve with the domestic absorption it has been unwilling to enlarge at the necessary political cost. As the same production establishes Chinese firms at technological chokepoints, the balance may acquire strategic value without ever appearing as an explicit target, allowing the center to disclaim the surplus while continuing to choose, and benefit from, the conditions that produce it.

The battery industry gives this distributed authorship its hardest test because it was an achievement before it began to resemble a problem. Chinese companies improved energy density, reduced manufacturing costs and supplied the world's largest electric-vehicle market, while demand for grid storage rose sharply as renewable generation expanded. During the first half of 2026, sales of power and storage batteries reached 979 gigawatt-hours, almost half again as much as a year earlier, and storage-cell sales rose by 83 percent. Demand in China and abroad genuinely absorbed the output.

Production still rose faster, reaching 1,069 gigawatt-hours during the same six months — up 53 percent — while exports absorbed 181 gigawatt-hours, or 18.5 percent of sales. Production and sales are flows measured at different points, however, so the figures alone cannot establish structural excess, while exports can reflect efficiency as readily as pressure. Their narrower significance lies in an industry expanding at extraordinary speed as foreign markets became an important part of its clearing mechanism.

Those figures acquire meaning beside the state's own response. In January, the industry ministry warned battery makers about the risk of overcapacity. On April 9, officials from that ministry, the National Development and Reform Commission, the market regulator and the energy administration met sixteen leading battery companies. The official account placed capacity warnings beside price competition, supplier payment periods, local investment recruitment and involution externalized, thereby putting the same mechanism Luo later described into the same room.

The meeting also showed where adjustment had gone before reaching the border. Some manufacturers protected their own cash by extending payment periods and compressing supplier margins. In June, eleven leading battery makers backed a sixty-day limit for paying smaller suppliers after the industry ministry warned that long payment cycles strained their cash flow and weakened innovation. A large manufacturer that delays payment turns the smaller supplier into its involuntary bank, preserving its own liquidity via somebody else's balance sheet.

For the materials producer, the mechanism is felt in the widening interval between its own obligations and its customer's payment, as wages, energy bills and loan installments fall due while the battery maker leaves its invoice unpaid. The supplier is made to finance the continued operation of its customer's assembly line, transferring the loss away from the producer whose exit would actually remove capacity. Prices and profits have imposed discipline, but not the exit on which the ministry's defense depends.

Rapid growth in grid storage may eventually vindicate much of today's investment, and Chinese firms may continue taking global share because they remain better and cheaper. But both possibilities leave the big institutional question intact: before demand arrives and weak producers leave, someone must finance the interval. In batteries, the burden has fallen partly on suppliers and local lenders, with overseas buyers helping to absorb the output, yet the distinction between an efficient export and exported adjustment cannot always be found inside the container.

The July paper answers that ambiguity with the scale of domestic demand. Between 2013 and 2024, the ministry says, domestic demand supplied an average of 93 percent of Chinese growth, split between 55 percentage points from consumption and 38 from investment. A country whose growth is overwhelmingly domestic, the figures suggest, cannot plausibly suffer from inadequate demand.

The arithmetic can be accepted while leaving the conclusion unproven. Contribution to growth measures how changes in consumption and capital formation contribute to changes in GDP; it leaves household absorption of existing industrial output unmeasured. A trade surplus that remains unchanged contributes nothing to current growth under this calculation because the measure records movement and leaves the level invisible.

Investment further complicates the accounting because expenditure on new capacity is recorded as demand before that capacity begins adding to supply. Building a battery factory raises present expenditure, employment and GDP, but once the factory opens it enlarges the volume of batteries that customers at home or abroad must absorb. The same investment can therefore make demand appear stronger today while deepening the imbalance it is invoked to disprove tomorrow.

The ministry enlarges the category again through retail sales. China's total reached 50 trillion yuan in 2025 and, converted using purchasing-power parity, was said to equal 1.7 times the American figure. But this proves scale while leaving absorption unresolved. Official retail sales include purchases by government organs, the military, institutions and enterprises, and principally record goods and catering while omitting much service consumption. A huge market can remain unbalanced between household consumption and investment-led supply.

Documents by Chinese officials have already supplied the narrower diagnosis that these aggregate measures evade. A Qiushi essay published in July 2025 described effective domestic demand as conspicuously insufficient, connecting weak household willingness and ability to consume with excess capacity in some industries before identifying local intervention and incomplete bankruptcy arrangements as obstacles to clearing. Some firms, it concluded, were 活不好、死不了 (huó bù hǎo, sǐ bù liǎo), unable to live well or die.

Xi Jinping had stated the causal chain a decade earlier. Capacity was “excessive overall,” he warned in 2015, and further investment-led expansion offered diminishing returns. Final consumption, he continued, was “the lasting force of economic growth.” The remark belongs to another historical moment, yet its provenance leaves the July defense with a harder problem. The connection between investment-led expansion, weak final demand and excess capacity had formed part of China's own diagnosis, at its highest level, for more than a decade before the ministry dismissed it as foreign confusion.

The July defense escapes that diagnosis by folding investment into demand, so expenditure on new capacity becomes evidence of the market required to justify it. Investment is demand in the national accounts, but insufficient final consumption cannot be remedied by allowing the factory to appear as its own customer.

Accounting, however, is only able to postpone rather than abolish the adjustment. Once production exceeds what can be sold at remunerative prices, China can allow firms to close, leaving suppliers, shareholders and creditors to carry the losses as capacity disappears; it can raise household incomes and strengthen social provision so that Chinese consumers absorb more of what the country makes; or foreign markets can take the additional output, transferring the pressure to producers elsewhere until their governments intervene through tariffs and procurement rules.

There is no frictionless world market in which an abstract unit of inefficiency simply vanishes. When a factory closes, skilled workers disperse, supplier relationships dissolve and research capacity may disappear along with the ability to manufacture strategically important goods in the future. Governments that once welcomed inexpensive Chinese products therefore resist becoming dependent on China for batteries and ships, just as Beijing once resisted dependence on foreign semiconductors.

The strategic case for resistance to China's exports can readily become a shelter for comfortable incumbents, raising the cost of the green transition and denying developing countries equipment they could not otherwise afford. Chinese machinery may lower the cost of building a foreign factory even as Chinese finished goods narrow the market in which that factory must survive. Europe and the United States cannot demand cheap decarbonization while excluding every producer capable of delivering it, and no country needs to reproduce every stage of every supply chain.

By September, that resistance had produced an unusual alignment. At the G20 finance ministers' meeting, China alone rejected language calling for the elimination of “non-market policies” associated with global imbalances, preventing agreement on a communiqué. The phrase remained capacious enough to contain economic diagnosis, strategic anxiety and domestic protection, but Beijing's isolation showed how many governments had come to regard China's domestic adjustment as a cost being transferred into their own economies.

Behind the argument over excess capacity lies a conflict over where adjustment should occur, which is a polite, abstract way of asking who should carry its losses. Beijing claims national credit for the achievements of its industrial system, then invokes the world market when the resulting production must be absorbed. Other governments increasingly reject that division of labor because they can see their own factories disappearing beneath it.