The September 2026 evidence strengthens China Has the Floor more than the earlier version suggested because the strength is visible not only at the state and industrial level but also across important civilian measures. The previous formulation treated slow retail growth, weak mortgage borrowing, and the property downturn too broadly as evidence that China’s household economy itself was substantially weak. Current household data show real income growth, real consumption growth, rising business income, extremely high business formation, and continued participation in an expanding productive economy. China’s civilian economy is not behaving like an American consumption-and-credit economy, but that is not the same thing as economic failure. From a Civilian Intelligence perspective, China currently shows a strong execution floor combined with a civilian population that is earning more, spending more in real terms, saving heavily, forming businesses, and participating in an increasingly technological production system.
The household numbers are concrete. In the first half of 2026, nationwide per-capita disposable income reached 22,981 yuan, up 5.2% nominally and 4.2% after inflation, while median disposable income rose 4.7% to 19,036 yuan. Urban disposable income increased 4.4% nominally and 3.4% in real terms, while rural disposable income increased 6.4% nominally and 5.5% in real terms, meaning rural incomes were actually rising faster than urban incomes. Nationwide per-capita consumption expenditure reached 14,836 yuan, increasing 3.7% nominally and 2.7% after inflation, with rural consumption expenditure rising 4.6%. Those figures do not describe a civilian population broadly losing purchasing power; they describe households whose real income and real consumption were still increasing during the first half of 2026.
The composition of that income is even more important for Civilian Intelligence because Chinese households are not participating only through wages. Wage and salary income increased 5.3%, transfer income increased 5.8%, and per-capita net business income increased 6.5%, faster than overall disposable income. Business income accounted for 15.8% of nationwide disposable income, which means household exposure to enterprise and self-employment is economically meaningful rather than marginal. Chinese households also maintain an unusually high savings rate, which the IMF estimates at roughly 20% of GDP, about twice the OECD-country average. That savings behavior can support investment and self-insurance rather than appearing immediately as retail consumption, so judging Chinese civilians exclusively by Western measures such as mortgages, credit expansion, and shopping activity gives an incomplete picture.
Business creation strengthens that conclusion. China recorded 25.74 million newly established business entities in 2025, with an average of roughly 26,000 enterprises being established every day, according to the National Bureau of Statistics. In the first quarter of 2026 alone, another 5.098 million business entities were created, including 2.074 million enterprises and 3.014 million individual businesses. Civilian Intelligence rrcognizes not every one of those registrations as evidence of China’s physical floor because some belong to retail, services, digital commerce, and the global internet economy. The correct method is to separate the global or digital business layer from the businesses that physically manufacture, move, inspect, build, process, and operate things inside China.
Once that separation is made, the physical entrepreneurial layer remains strong. During the first half of 2026, China added 28,000 new high-end equipment-manufacturing enterprises and 15,000 new high-tech manufacturing enterprises, bringing the total number of high-tech manufacturers to approximately 330,000. New integrated-circuit manufacturing businesses increased 24.2% year over year, while spacecraft and launch-vehicle manufacturing registrations increased 185.7% and optical-fiber and cable manufacturing registrations increased 129.4%. In the first quarter alone, new intelligent unmanned-aircraft manufacturing entities increased 15.7%, while equipment manufacturing produced about 70,000 new business entities. These are not merely online storefronts or financial registrations; they represent entrepreneurial formation inside semiconductors, aerospace, drones, machinery, transportation equipment, and other parts of the physical execution floor.
The operating factories show the same pattern. In August 2026, Chinese manufacturing output increased 6.1% year over year, equipment manufacturing increased 12.1%, and high-technology manufacturing increased 16.7%. Lithium-ion battery production increased 57.2%, industrial-robot production increased 34.6%, and 3D-printing-device production increased 29.9%. Private-enterprise industrial output still increased 3.7%, and profits across large industrial enterprises rose 17.6% during the first seven months of the year. The significance is that China is not merely adding factories; the industrial system is increasingly producing the batteries, robots, electronics, and advanced machinery that improve the productive capacity of the system itself.
The battery and electric-vehicle system demonstrates how far that execution floor now reaches. China produced approximately 70% to 75% of the world’s electric cars in 2025, more than 80% of global battery cells, about 85% of cathode active material, and more than 90% of anode active material used in electric-car batteries. Chinese producers supplied almost 75% of global EV battery deployment, while China itself accounted for about 60% of worldwide EV battery deployment. China also exported more than 2.5 million electric cars in 2025, and its EV exports more than doubled again year over year during the first quarter of 2026. China is therefore not simply selling automobiles abroad; it is exporting batteries, electrical architecture, motors, electronics, manufacturing relationships, charging requirements, replacement components, and portions of an entire transportation ecosystem.
Below the battery sits an even deeper floor: material processing. The International Energy Agency estimates that China processes more than 70% of global lithium, cobalt, graphite, and rare earths, while China also holds approximately 80% of lithium-ion battery supply-chain production capacity, about 85% of solar supply-chain production capacity, 95% of photovoltaic-wafer capacity, and 97% of battery-anode-material capacity. This matters because owning mineral deposits is not the same thing as possessing usable industrial material; lithium, cobalt, graphite, and rare earths must be refined, separated, purified, and transformed before they become batteries, magnets, motors, electronics, drones, or defense equipment. China’s export controls have already demonstrated that concentration in these middle layers can interrupt downstream manufacturers in other countries even when those countries own sophisticated final-product companies. The physical floor therefore includes not only mines and factories but the processing knowledge, specialized equipment, chemistry, and industrial discipline that connect raw matter to advanced technology.
China’s energy system reinforces the same execution strategy. By the end of June 2026, renewable generating capacity had reached 2.455 billion kilowatts, equal to 60.7% of total installed Chinese generating capacity, while wind and solar together represented roughly 48.3% of the national total. Renewables produced 41.2% of China’s electricity during the first half of 2026, while coal-fired generation’s share fell below 50% for the first time over a half-year period. China has not abandoned coal, oil, nuclear, hydro, or natural gas; it is building an energy stack in which multiple sources support an increasingly electrified transportation, manufacturing, robotics, and city system. From the Civilian Intelligence perspective, this creates optionality because an electric bus, robot, factory machine, or EV can consume electricity generated from multiple domestic sources rather than being permanently locked to one globally traded liquid fuel.
Shenzhen shows what happens when the physical floor becomes infrastructure rather than a collection of products. By 2026, Shenzhen had built more than 1,200 low-altitude takeoff and landing facilities and opened 310 drone logistics routes, while cargo drones completed more than one million flights in 2025. The city produces roughly 70% of China’s consumer drones and 50% of its industrial drones, placing manufacturing and deployment inside the same urban ecosystem. The important innovation is not that China invented the drone, because it did not; the important innovation is that Shenzhen is surrounding drones with landing facilities, flight routes, logistics demand, batteries, software, permissions, maintenance, mapping, emergency applications, and commercial customers until the drone becomes infrastructure. That is the core of China’s execution mentality: take an existing technology, industrialize it, reduce its cost, connect it to physical systems, and make it ordinary enough to become part of daily economic life.
China is now trying to apply that same mentality to artificial intelligence. DeepSeek’s September 2026 V4.1-Flash release uses a 552-billion-parameter mixture-of-experts architecture while activating only 8 billion parameters for input and 16 billion for output, and DeepSeek reports that its new cache architecture requires only one-quarter of the previous generation’s high-bandwidth memory. That approach reflects an emphasis on extracting greater capability from constrained compute and lowering the cost of large-scale deployment rather than treating model size alone as the objective. China still has a hard physical bottleneck in advanced high-bandwidth memory, and Chinese AI-chip companies raised prices in September because HBM shortages and U.S. export restrictions increased procurement costs. Civilian Intelligence therefore records both facts simultaneously: China is pushing AI efficiency and industrial deployment aggressively, while the incomplete semiconductor stack remains one of the clearest physical dependencies underneath its intelligent floor.
China’s monetary position remains much weaker than its physical position, which reinforces rather than weakens the floor-versus-ceiling model. In the first quarter of 2026, the U.S. dollar still represented 57.13% of world foreign-exchange reserves, while the renminbi represented only 1.99%. America therefore retains extraordinary command-layer power through currency, capital markets, finance, software, intellectual property, advanced chip design, and other upper layers of the global system. China’s strength is different because it is concentrated more heavily in the physical systems required to manufacture, electrify, transport, automate, refine, and deploy. China has not taken America’s ceiling, but it has built a much broader physical floor beneath a technological world that increasingly depends on batteries, minerals, robotics, electrical systems, manufacturing, logistics, and material processing.
The property downturn and youth labor market remain real liabilities, but they no longer justify describing the entire Chinese civilian economy as broadly weak. Property investment was down 19.9% through August, and the unemployment rate among urban 16-to-24-year-olds excluding students reached 17.9% in July, while overall urban unemployment remained much lower. Those figures identify specific pressure points—housing wealth and younger labor-market entry—rather than overturning the simultaneous evidence of rising real household income, rising real consumption, rising household business income, millions of new businesses, and fast growth in physical manufacturing. Retail sales growing only 0.4% in August also shows restrained consumer demand, but restrained consumption is not identical to declining civilian purchasing power when first-half real disposable income was rising 4.2%. Civilian Intelligence therefore classifies property and youth employment as significant liabilities inside an otherwise expanding civilian-production system, rather than treating them as proof that Chinese households as a whole are economically failing.
The corrected September 2026 thesis is therefore stronger and simpler: China’s floor is strong, and the civilian layer participating in that floor is stronger than the earlier version gave it credit for. Chinese households are earning more in real terms, consuming more in real terms, generating more business income, saving heavily, starting millions of businesses, and participating in a production system expanding through semiconductors, equipment manufacturing, drones, batteries, robots, electricity, electric vehicles, logistics, and advanced industrial technology. China’s civilian economy should not be judged solely by the American expectation that strength appears through more mortgages, more consumer debt, faster retail spending, and higher household leverage. China’s model converts a much larger share of national activity through saving, business formation, manufacturing, physical investment, infrastructure, and productive capacity, while maintaining real household income growth despite serious problems in property and youth employment. From the Civilian Intelligence perspective, the central September 2026 finding is no longer merely that China has the floor; it is that China’s civilians are actively embedded in, earning from, forming businesses inside, and increasingly operating the floor that China is building.