For two decades China has repeatedly built more industrial plant than its own households and builders could absorb in a normal year. Steel, cement, aluminium, chemicals, shipyards and later solar panels, batteries and electric cars all show versions of the same pattern. Factories go up because credit is cheap, local officials want GDP, and state firms face soft budget constraints. When domestic demand falters, the plant does not disappear. Output either sits idle, which is expensive, or finds buyers abroad, which other countries experience as a flood of low priced goods. The hinge that made this logic permanent was the global financial crisis of 2008 and 2009. Export orders collapsed. The property market, the main domestic sink for heavy industry, froze. Beijing’s answer was a vast stimulus that refilled furnaces at home and left China with even more capacity. After that episode, keeping capacity busy meant using domestic stimulus and foreign markets together.
What excess capacity means in China
Excess capacity is not simply a few quiet months at a mill. Every industry needs spare plant for maintenance, seasonal peaks and unexpected orders. Steel analysts often treat utilisation near 80 percent as a workable normal, not a failure. The problem begins when capacity keeps rising faster than final demand for years, so that utilisation falls well below that band and prices stay depressed even in good years. That is structural overbuilding rather than ordinary cyclical slack.
In China the drivers of that overbuilding are familiar. Local governments competed on growth. For a long period cadre promotion rewarded visible investment: roads, towers, industrial parks, furnaces. State owned enterprises in heavy industry could borrow from state banks at favourable rates and were rarely forced into bankruptcy. Land, power and water were often underpriced for favoured projects. Household consumption stayed relatively low as a share of national income, while investment stayed high. Illness, schooling and old age still push families to save. Between 2000 and 2024, according to the Bertelsmann Stiftung, investment averaged just over 42 percent of Chinese GDP, roughly double the share in the United States or Germany. World Bank series for gross fixed capital formation show the same intensity: about 38 percent of GDP in 2007 and about 44 percent by 2009 after the stimulus began. The International Monetary Fund has likewise flagged China’s savings rate as far above the levels of peer industrial economies.
The result was a set of industries that Chinese policy documents themselves repeatedly named as overcrowded: steel, cement, flat glass, primary aluminium and shipbuilding. Later the same pattern appeared in solar modules, lithium ion batteries and electric vehicles. The products change. The incentives that produce too much plant do not. Officials and scholars have sometimes split the problem into types, such as weak link surplus in steel and cement, local surplus in aluminium, and more cyclical surplus in shipbuilding. The practical point for outsiders is simpler. When utilisation stays low for years, and firms keep producing anyway, surplus has to go somewhere.
It is important to separate two stories that get mixed up in foreign debate. One is that Chinese firms became highly competitive through scale, learning and ruthless cost control. That is real. The other is that capacity decisions were shaped by local GDP contests and soft finance, so that supply could outrun profitable domestic demand for long stretches. That is also real. Explaining excess capacity does not require pretending Chinese industry is weak, nor does it require treating every Chinese export as illegitimate. It requires looking at who pays when plant is built and who bears the cost when it is not used.
Capacity before 2008
China’s entry to the World Trade Organization in December 2001 opened a long boom in exports of manufactures. Coastal provinces specialised in assembly and processing trade. At the same time, urbanisation and rising incomes pulled tens of millions of people into cities that needed flats, offices, roads, bridges and power lines. Property construction and public infrastructure became the great domestic absorbers of steel and cement. Mills and kilns were built to feed that boom. Because investment creates both demand for materials and new supply of materials, capacity could race ahead of ordinary household consumption of finished goods. A tonne of steel used to build another steel mill temporarily absorbs surplus metal while adding to future surplus. Yongding Yu and other Chinese economists have long described that circular logic as a structural feature of investment heavy growth.
Steel makes the arithmetic clear. World Steel Association figures put Chinese crude steel production near 129 million tonnes in 2000. By 2007 it was approaching 490 million tonnes. In 2008 it was about 512 million tonnes. China had flipped from net importer toward net exporter for many products. Utilisation rates before the crisis were still high by later standards, around the mid 80s percent according to later Peterson Institute reconstructions. The system looked tight because construction demand was enormous. The plant base that would later look oversized was already being laid down, financed by banks and encouraged by localities that counted tonnage as proof of development.
Cement and aluminium tell parallel stories. Between 2011 and 2013, according to figures from China’s National Bureau of Statistics and the United States Geological Survey cited by CKGSB, China produced about 6.6 billion tonnes of cement, more than the United States produced across the whole twentieth century. By the mid 2010s Chinese aluminium capacity was reported near 40 million tonnes, exceeding global consumption by roughly 9 million tonnes on Antaike estimates. Those later snapshots show how far the investment race could go once the post crisis stimulus had run. The underlying tendency, though, was already visible before Lehman Brothers failed: build first, find demand later. Shipbuilding yards expanded on the same credit tide, only to face a global freight slump when trade volumes stalled.
Before 2008, foreign markets still acted as a partial outlet when domestic construction paused. American and European demand for Chinese manufactures was strong. That outlet did not remove the domestic bias toward overbuilding. It merely delayed the moment when surplus capacity would have nowhere convenient to go.
The 2008 and 2009 shock
The global financial crisis hit China first through trade. Research associated with Bai, Chen and Song notes that China’s annualised GDP growth dropped from about 9.5 percent in the third quarter of 2008 to about 6.4 percent in the first quarter of 2009, as total exports roughly more than halved between September 2008 and February 2009. Full year merchandise exports in 2009 fell by around 16 percent from 2008. Factories along the coast cut shifts. Migrant workers went home early for the Spring Festival and many did not return on time. Orders for metal intensive machinery and construction related manufactures vanished overseas just as domestic confidence cracked.
Housing was the more dangerous domestic channel. Construction and property sales are the main way China turns steel, cement, glass and copper into domestic final demand. When that channel freezes, heavy industry has nowhere to go. Through late 2008 the freeze was severe. National statistics for January to November showed commodity housing sales area down 18.3 percent year on year and sales value down 19.8 percent. Vacant floor space rose 15.3 percent by the end of November. Mortgage lending for the year fell nearly 30 percent. In December 2008, prices in the official 70 city index fell 0.4 percent from a year earlier and 0.5 percent from the previous month. Fifty cities recorded month on month declines in new home prices. Shenzhen new home prices were down 18.1 percent year on year; Guangzhou was down 9.4 percent. Industry reports described the first nationwide decline in house prices in about a decade, with the fall gathering force from August into the fourth quarter. Developers’ willingness to start new projects turned negative from mid year.
That is not the same as saying the Chinese property market was destroyed. Prices did not collapse to zero. Most developers did not vanish overnight. National averages still mixed rising inland cities with crashing coastal ones. But sales volumes, credit and prices moved together in a way that threatened the cash flow of builders and the order books of mills. For an economy that absorbed heavy industry output mainly through building, late 2008 was a near miss. If housing had stayed frozen while exports stayed dead, furnaces and cement kilns would have faced a demand hole large enough to force mass closures, bank losses and local fiscal stress. Beijing treated the risk as existential for growth and employment, which is why the response was so large and so fast.
The four trillion yuan response
On 9 November 2008 the State Council announced a stimulus package estimated at 4 trillion yuan, about 570 to 586 billion dollars at the time, to be spent over roughly two years. The list covered low income housing, rural infrastructure, water, electricity, transport, environmental projects, technological upgrading and reconstruction after the May 2008 Wenchuan earthquake. Academic reconstructions of the allocation put about 1.5 trillion yuan into railways, roads, airports, water conservancy and urban power grids, about 1 trillion into post disaster rebuilding, about 1.14 trillion into housing and rural livelihood projects, and smaller sums into environment and education. Credit ceilings on commercial banks were lifted. Fiscal policy was described as active and monetary policy as moderately easy. The package was large relative to GDP by international standards, and it was designed to work through construction rather than through household cash transfers.
The headline fiscal number understates the real expansion. Much of the money was raised and spent through local government financing vehicles borrowing from banks. The Bank for International Settlements recorded that Chinese banks extended 7.4 trillion yuan in new loans in the first half of 2009 alone, more than the 4.2 trillion yuan of new loans for the whole of 2008. Work by Bai and co authors estimates about 4.7 trillion yuan of extra lending in 2009 relative to a normal path, of which roughly 2.3 trillion yuan went to local financing vehicles. About nine tenths of local government investment in that year was financed by bank loans. Investment as a share of GDP jumped further. Equity markets and housing transactions revived through 2009 as credit flooded in. Mortgage terms were eased and purchase restrictions loosened. The property market that had seized up in late 2008 began to clear stock and restart starts.
For heavy industry the immediate effect was relief. Chinese crude steel output rose from about 512 million tonnes in 2008 to about 577 million tonnes in 2009 while many other countries cut production. Cement kilns and aluminium smelters found orders again in railways, urban works and a rebounding property market. The furnaces were refilled at home. That was the point of the programme, and by the narrow test of avoiding a hard landing it worked.
The hangover was the other half of the story. Stimulus did not only use existing plant. It encouraged new plant. Local governments, flush with loan backed projects, raced to add capacity. CKGSB later summarised the investment share of GDP as moving from around 40 percent before the crisis toward 50 percent, with total debt relative to GDP rising sharply over the following five years. By the end of 2015 Chinese crude steel capacity stood near 1,127 million tonnes against production of about 799 million tonnes, a utilisation rate near 71 percent on the official capacity series used by the Peterson Institute. Conventional arithmetic put excess capacity near 328 million tonnes; even after allowing for normal spare capacity, PIIE still counted around 128 million tonnes of true surplus. Empty floors and underused infrastructure accumulated in weaker cities. China had saved growth in 2009 by doubling down on the investment model that creates excess capacity.
Exporting the surplus
Once domestic absorption cooled in later cycles, Chinese firms looked outward again. Steel is the clearest case. United States Department of Commerce steel export reports put Chinese steel exports near 110 million tonnes in 2015 and about 107 million tonnes in 2016. Those volumes met a thicket of anti dumping and countervailing duties in the United States and the European Union. European industry groups argued that cheap imports had contributed to large job losses after 2008. Similar fights appeared in aluminium, flat glass and solar panels. Pre crisis exports had served, in the European Chamber of Commerce’s phrase cited by CKGSB, as a safety valve on a pressure cooker. After the crisis, the valve still worked, but partners were less willing to absorb the steam without tariffs and political protest.
The pattern did not end with the mid 2010s. When China’s property sector weakened again in the early 2020s, steel demand at home softened and exports rose. MERICS reported that China produced just over one billion tonnes of steel in 2023 and exported around 90 million tonnes, with 2024 export volumes rising even as unit values fell. The OECD’s steel outlook for 2026 put global excess capacity near 640 million tonnes in 2025, with China accounting for about half of the global capacity demand gap. World Steel Association data show national crude steel output falling to 960.8 million tonnes in 2025, a seven year low, while trade reports described record or near record steel export volumes in the same period. Weak domestic construction and strong export shipments are two sides of one capacity story. Importers in Southeast Asia, the Middle East and elsewhere took more Chinese metal even as rich country trade barriers thickened.
Solar panels, batteries and electric vehicles are later chapters of the same book. Deloitte estimates for 2023 put Chinese solar panel exports near 56 billion dollars, lithium ion battery exports near 70 billion dollars and electric vehicle exports near 34 billion dollars, with total motor vehicle exports rising from under 9 billion dollars in 2019 to about 78 billion dollars in 2023. China held about three quarters of world lithium ion manufacturing capacity in 2022 and more than 80 percent of capacity across solar manufacturing stages. Reuters reported in 2024 that Chinese solar module capacity had reached around 861 gigawatts, with utilisation in some readings near a quarter, which helps explain rock bottom panel prices worldwide. Foreign governments answered with tariffs and subsidy probes. Chinese firms answered by seeking new markets and by shifting some assembly abroad. The politics differ by product. The underlying pressure does not: plant built for rapid domestic growth seeks buyers wherever prices clear.
Why surplus keeps leaving the country
The mechanism is ordinary industrial economics under Chinese political constraints. A blast furnace, a cement kiln or a solar cell line is a sunk cost. Once the capital is spent, the extra cost of producing another tonne or another panel is often low, especially when energy and credit are soft. Idling the plant still leaves debt service, local tax shortfalls and unemployed workers. Local governments therefore resist closures. Banks prefer to roll loans to zombie firms rather than recognise losses. Exporting at thin margins, or even at prices that look like dumping to foreign investigators, can still beat a cold furnace. Cash flow, utilisation statistics and employment all improve, at least for a while, even if long run returns on capital do not.
From the partner country’s side, the same shipments look like unfair competition. Workers and firms abroad face price wars they cannot match if their own capital must earn a market return and their governments do not socialise losses. Trade defence cases multiply. Some importing countries welcome cheap inputs for their own builders and consumers. Exporting countries with competing industries do not. That tension is what commentators mean by a second China shock or by overcapacity diplomacy: the surplus is managed by moving it across borders rather than by writing it off at home. Belt and Road infrastructure talk sometimes framed the same problem as a development opportunity for capital scarce countries, though those markets are usually too small to absorb China’s steel or cement overhang on their own.
None of this requires a conspiracy theory. It follows from high investment, soft local budgets and the political cost of mass layoffs. Beijing has at times tried to force cuts. It has also at times prioritised growth and employment over rapid closure. Firms and localities then do what the incentives tell them to do. Partners who only complain about dumping without noticing the domestic political economy miss half the mechanism. Partners who excuse every surplus shipment as natural comparative advantage miss the other half.
From supply side cuts to the new three
By late 2015 overcapacity in steel and coal had become an official crisis. Producer prices had fallen for years. Losses mounted across major mills. In February 2016 the State Council set a target of cutting 100 to 150 million tonnes of crude steel capacity between 2016 and 2020. Officials later reported more than 65 million tonnes of iron and steel capacity closed in 2016, ahead of that year’s 45 million tonne target. A central fund of about 100 billion yuan was created to resettle workers. Hundreds of thousands of steel and coal employees were moved or compensated. Consolidation produced champions such as China Baowu from the Baosteel and Wuhan merger. Prices and profits recovered for a time. Outside analysts warned that gross closures could overstate net shrinkage if new efficient furnaces replaced old ones, but the campaign showed that Beijing could force capacity discipline when political attention was high.
The 2020s surplus in green technology does not erase that history. It repeats the investment race under a new industrial policy banner. Property no longer absorbs materials as it did in 2009 to 2013. Policymakers lean on advanced manufacturing and the so called new three. Capacity again outruns domestic take up. Exports and trade friction follow. The products are cleaner and more sophisticated. The capacity arithmetic is recognisable. Subsidies, local competition for battery and car plants, and national targets for renewables all recreate the old habit of building first and sorting demand later.
The lasting lesson of 2008 and 2009 sits underneath both the steel fights of the 2010s and the electric car fights of the 2020s. When export demand and property sales failed together, China used a credit and infrastructure surge to keep capacity busy at home. That worked as emergency medicine and left a larger industrial plant. In later downturns, when domestic sinks weakened again, the same plant sought foreign buyers. Excess capacity in China is therefore not only a domestic planning problem. It is a system that, once built, tends to export its surplus unless someone pays the political and financial cost of closing it. The crisis years taught Beijing and Chinese firms that domestic stimulus and foreign markets are twin ways to keep the furnaces lit. The world has been living with that lesson ever since.