Since September, based on integrated information from those recently released by relevant departments, China’s economic data for August 2026 reveals a notable contrast. The manufacturing Purchasing Managers' Index (PMI) stood at 49.8%, up 0.6 percentage points from July, but remained below the boom-bust threshold for the second consecutive month. The non-manufacturing business activity index for the services sector was 49.3%, and the business activity index for the construction industry dropped further to 46.9%. This, of course, does not mean that the Chinese economy is deteriorating comprehensively. In August, the manufacturing production and new orders indices rebounded to 50.4% and 50.6%, respectively. Among 21 manufacturing industries, 16 saw their PMIs rise month-on-month, with equipment manufacturing and high-tech manufacturing PMIs reaching 51.4% and 52.9%, respectively. What truly deserves attention is that while manufacturing and certain new industries maintain their resilience, real estate, infrastructure, state-owned investment, and construction, which are the sectors that traditionally functioned as economic shock absorbers during past downturns, have failed to experience a simultaneous rebound. In the first seven months of 2026, national fixed-asset investment of the country fell by 6.7%, among which state-held investment decreased by 3.3%, infrastructure investment dropped by 3.6%, real estate development investment fell by 19.2%, and construction and installation engineering investment declined by 9.2%. This implies that current issues facing China may no longer be merely a routine cycle of insufficient demand, but rather a shift in the set of economic floor-support mechanisms that have operated over the long term in the past.
Over the past three decades, China has developed a distinct counter-cyclical transmission mechanism. During economic downturns, fiscal and financial resources flowed through local governments, state-owned enterprises (SOEs), and financing platforms into real estate, infrastructure, and industrial projects. This generated orders for construction enterprises and further translated into industrial demand for steel, cement, and mechanical equipment, while simultaneously creating land revenue, employment, and household income, ultimately diffusing into consumption and service sectors. Consequently, real estate, infrastructure, and construction were not merely massive industries, but long constituted a vital transmission chain for macroeconomic counter-cyclical regulation. This mechanism operated efficiently during periods of rapid industrialization and urbanization. However, its viability relied on sustained growth in land and housing demand, the expansion capacity of local government finance, and the ability of new infrastructure to continue yielding high economic returns.
These conditions for China are now all changing. In the first seven months of 2026, real estate development investment decreased by 19.2%; the floor space of newly built commercial housing sold was 450.21 million square meters, a decline of 11.8%. Sales revenue reached RMB 4,271.8 billion, down 13.1%. Meanwhile, floor space under construction by real estate developers fell by 12.7%, and at the same time, newly started floor space dropped by 24.0%. In addition, funds available to development enterprises decreased by 20.3%. Concurrently, fixed-asset investment fell by 6.7%, and construction and installation engineering declined by 9.2%. Infrastructure investment dropped by 3.6%, investment in the road transport industry fell by 7.5%, and investment in water conservancy, environmental, and public facilities management dropped by 9.1%. The construction business activity index stood at only 46.9% in August, with new orders and employment also remaining clearly in contraction. Extreme weather can account for some short-term slowdowns in construction, but it is no longer sufficient to explain the medium-term trend wherein real estate, infrastructure, construction installation, and construction industry prosperity are declining simultaneously.
Even more noteworthy than the contraction in real estate and construction is the change now underway in the government balance sheet that previously supported this floor-support mechanism. The latest government debt report submitted by the State Council to the Standing Committee of the National People’s Congress shows that, by the end of 2025, the national government debt balance had reached RMB 102.5 trillion. This comprised RMB 41.2 trillion in treasury bond debt, RMB 54.8 trillion in local government statutory debt, and RMB 6.5 trillion in existing local government hidden debt. The national statutory government debt balance stood at RMB 96 trillion, corresponding to a statutory liability ratio of 68.5%. When existing local hidden debt is taken into account, the national government liability ratio reached 73.2%. This does not mean that the Chinese government has lost its borrowing capacity, but it does indicate that the debt foundation underlying the government’s past balance-sheet expansion to sustain investment differs significantly from that of more than a decade ago.
What is especially important among these developments is that local governments are shifting from expanding their balance sheets to simultaneously repairing them. At the end of 2023, the balance of local government hidden debt still stood at RMB 14,300 billion. To address this, the central government increased the local government debt limit by RMB 6,000 billion to replace existing hidden debt and, starting in 2024, allocated RMB 8,000 billion annually from newly issued special bonds for five consecutive years to help resolve the debt. By the end of 2025, existing local government hidden debt had declined to RMB 6,500 billion. While such debt replacement helps reduce financing costs and mitigate risks, it also means that new government financing can no longer be understood simply as funding for new investment. An increasing share of fiscal and financial resources must instead be devoted to replacing existing debt, repaying maturing obligations, and repairing local government balance sheets. In 2025, local governments issued a total of RMB 10,300 billion in bonds, including RMB 5,400 billion in new bonds and RMB 4,900 billion in refinancing bonds.
A deeper layer of change lies in the fact that local fiscal pressure is increasingly transmitting to the central fiscal authorities. In 2026, the central general public budget revenue is budgeted at RMB 9,567 billion, central general public budget expenditure is projected to reach 15,007 billion yuan, and the central fiscal deficit is set at RMB 5,090 billion, an increase of RMB 230 billion compared to 2025, which would be financed through the issuance of treasury bonds. The national fiscal deficit is budgeted at RMB 5,890 billion, with the entire RMB 230 billion increase allocated to the central government. Meanwhile, central transfer payments to local governments have reached RMB 10,415 billion, remaining above RMB 10 trillion for four consecutive years. The Ministry of Finance has also explicitly stated that it will temporarily raise the central government's burden ratio in areas such as childcare subsidies and the exemption of preschool childcare and education fees, so as to alleviate local fiscal expenditure pressures.
Therefore, it is no longer sufficient to characterize the current situation in China simply as “local governments are short of funds, while the central government still has money to spend”. The country’s central government has monetary sovereignty, stronger tax-raising capacity, and greater access to government bond financing, giving it significantly greater fiscal capacity than local governments. It would therefore be inaccurate to suggest that the central government has “run out of money”. At the same time, however, the central government is using transfer payments to stabilize local government finances while assuming an increasing share of the responsibilities required to provide a fiscal backstop, including local debt resolution, social security, sci-tech and industrial policies, and major projects. Its own fiscal spending is also becoming increasingly dependent on deficit financing and government bond issuance. This indicates that some of the fiscal responsibilities left by the traditional growth model are gradually being shifted to the central government. The central government remains the ultimate backstop, but its balance sheet also has limits and cannot provide an unlimited source of financing without costs or constraints. As a result, the challenges facing the traditional backstop mechanism are gradually shifting from whether local governments can continue to sustain investment to how quickly, and to what extent, the central government is willing and able to expand its own balance sheet.
China is not simply finding itself unable to invest. Rather, capital flows are shifting. In the first seven months of 2026, investment in construction and installation engineering fell by 9.2%, but investment in the purchase of equipment and instruments grew by 9.0%, and investment in intellectual property products increased by 9.1%. Investment in railway, ship, aerospace, and other transport equipment manufacturing grew by 18.7%, while computer, communication, and other electronic equipment manufacturing rose by 7.8%. Capital is migrating away from real estate, civil engineering, and certain traditional infrastructure toward equipment, technology, and advanced manufacturing.
The issue is that while new growth engines can succeed old industries in generating output value, they may not necessarily succeed them simultaneously across all of their macroeconomic functions. Real estate and infrastructure are characterized by large investment scales, long industrial chains, high local correlation, and extensive labor inputs, allowing a single round of expansion to concurrently generate land revenue, construction orders, building material and mechanical demand, and substantial local employment. Advanced manufacturing, artificial intelligence, new energy, and aerospace possess higher technological and capital intensity, enabling them to create higher-tech output, but they may not generate local fiscal revenue, construction demand, and widespread service consumption demand in the same way. Consequently, the continuous expansion of new industries and persistently weak traditional demand can very well coexist.
The August manufacturing data precisely reflects this state. The production index and new orders index have reached 50.4% and 50.6%, respectively, indicating that improvement is not entirely devoid of genuine demand. However, orders in hand still stood at only 46.7%, raw material inventories at 48.1%, finished product inventories at 48.4%, and employment at 48.7%. The PMI for large enterprises has reached 50.6%, that for medium enterprises stands at 49.4%, and small enterprises register only 47.9%. It is not that orders do not exist, but that newly added orders have not yet fully translated into the accumulation of orders in hand, corporate restocking, and expanded employment. Manufacturing is repairing itself, but this recovery has not yet formed the widespread economic diffusion seen during past real estate and infrastructure boom periods.
This insufficient transmission ultimately reflects on the services sector. In August, the services business activity index stood at only 49.3%, flat compared to July, while new orders dropped further to 44.5% and employment declined to 45.8%. At the same time, business activity indices for industries such as postal services, telecommunications, radio, television and satellite transmission services, and internet software and information technology services all exceeded 55%, whereas traditional service sectors such as wholesale and retail remained below the boom-bust threshold. The coexistence of high prosperity in the new economy and weak broad-spectrum consumer demand is becoming an increasingly distinct structural characteristic.
This also makes this year’s economic data particularly worth observing through the traditional market observation of July, August, and September as a cycle of slump, recovery, and eventual turnaround. Originally a piece of market wisdom reflecting the transition from the slow summer period to the autumn peak rather than a strict economic rule, and in any case applied here to already seasonally adjusted PMI data, the latest figures point to a clear departure from past patterns. After holding at 50.0% in July 2024 before rising to 50.2% in August, and following a similar pattern in 2025, when the index rose from 50.0% to 50.5%, this August’s reading remained sluggish at 49.3%, with new orders continuing to decline. With July marking a pronounced slump and August failing to deliver the usual recovery, whether September can bring about a genuine turnaround now depends far more on employment, income, and consumer expectations than on a simple seasonal rebound.
What the August data truly reveals is perhaps not that the Chinese economy has completely lost its growth momentum, but that the floor-support mechanism underpinning growth is continuing to undergo an unsynchronized transition. The old mechanism composed of real estate, land finance, local financing, and traditional infrastructure is gradually weakening, with local governments and certain SOEs increasingly bearing the tasks of stock debt and balance-sheet repair. The central fiscal authorities are taking over more stabilization responsibilities, but the center itself is increasingly relying on deficits and treasury bond expansion to maintain the intensity of fiscal expenditure. Meanwhile, advanced manufacturing and new quality productive forces, while forming new growth drivers, have not yet fully taken over the macroeconomic functions previously borne by the old mechanism in local finance, employment, income, and consumption.
Final analysis conclusion:
China has not lost its macroeconomic floor-support capacity. What has truly changed is the cost, space, and efficiency of traditional support methods. Whether the economy can establish a new endogenous growth cycle in the future may no longer depend solely on how much debt the government can add or how many projects it can build, but on whether the growth generated by advanced manufacturing and technological progress can more fully translate into employment and household income, and further transform into consumer demand via social security, public services, and income distribution improvements. Only when this new transmission chain gradually takes shape can the new growth engines be considered to have truly succeeded the old ones. Else, even if production and exports continue to grow, the economy may face for a relatively long period a structural divergence characterized by a relatively robust industrial end and weak broad-based domestic demand.
______________
Zhou Chao is a Research Fellow for Geopolitical Strategy programme at ANBOUND, an independent think tank.
