Widening global imbalances draw attention to policies shaping international trade. Using firm-level data, the ECB blog compares government support in China, the United States and the euro area. We find that subsidies drive Chinese exports in strategically important sectors.
Concerns about imbalances in the flow of goods and money have recently resurfaced.[1] The current account broadly measures whether a country earns more from the rest of the world than it spends, through trade in goods and services and cross-border income flows. The latter are divided into primary income (from labour and capital) and secondary income (mainly remittances, foreign aid and government transfers). Countries that earn more than they spend run surpluses, while those that spend more run deficits.
A country's current account balance reflects the difference between national saving and investment. When domestic saving exceeds domestic investment, the excess funds are invested abroad, resulting in a current account surplus. Conversely, when investment exceeds saving, the country typically runs a current account deficit.
But the debate has also turned to policies that shape trade flows, including the growing use of industrial subsidies.[3]
For a recent discussion of the debate surrounding Chinese industrial subsidies, their sectoral targeting and their implications for export competitiveness, see Tett, G. (2026), "China is smarter about subsidies than everybody else", Financial Times, 31 July.
To examine this issue, we use the fresh OECD dataset on Manufacturing Groups and Industrial Corporations (MAGIC), which provides internationally comparable firm-level data on government support. We compare the public support received by large firms in the United States, China and the euro area. We also examine how subsidies relate to firms' export performance and the potential implications this has for global trade imbalances.[4] The OECD MAGIC database compiles information from firms' audited financial statements, annual reports and other public corporate disclosures, supplemented by government sources where available. By relying primarily on company-level disclosures and applying a harmonised methodology across countries, including China, the database improves cross-country comparability and reduces reliance on official government reporting. Estimates are conservative, as the database currently covers only selected forms of government support and the largest manufacturing firms.
Here are our findings in a nutshell: Government support for large firms is rising globally. Chinese firms, however, stand out for both the amount of support they receive and the number of sectors receiving support - with those receiving more support also tending to export more. This link is much weaker in the United States and the euro area. And while subsidies may not explain a country's overall trade surplus or deficit, they can influence trade imbalances in key industries.
Yet further research is needed to establish a robust causal link between subsidies and exports. We argue that examining this link at sector level is important, as trade imbalances in strategically important industries such as solar technology, semiconductors and automobiles can put employment and industrial capacity in trading partners under pressure.
Why are industrial subsidies back in focus?
Let's take a step back and look at the wider picture. Global imbalances are commonly defined as the sum (regardless of whether they are positive or negative) of individual countries' current account balances as a percentage of world GDP. This captures the overall size of external surpluses and deficits rather than the net global balance.
Global imbalances climbed to 3.7% of world GDP in 2025 (IMF, 2026) after falling to 2.8% in 2019.[5] See IMF (2026), External Sector Report: Global Imbalances in a Shifting World, July.
What's behind those imbalances? With the exception of the euro area, most of the current account balances of the United States and China come from the goods trade balance (Chart 1, panel b).[6] See footnote 1 for a definition of the current account balance and its components as shown in Chart 1, panel b).
Chart 1
Global imbalances are increasing again, primarily driven by the three largest economies
(percentages of world GDP and percentage points)
a) Global imbalances and country contributions since 1990 |
b) Composition of current account balance |
 |
 |
Sources: IMF, World Economic Outlook and ECB staff calculations.
Notes: Panel a) sums the absolute changes in current account balances across economies. As both widening surpluses and widening deficits enter in absolute value, the measure captures the size of global imbalances rather than a net global balance. The latest observations are for 2025.
Discussions among G7 and G20 countries focus on large and persistent external imbalances and the risk of trade tensions and protectionist responses.[7] These discussions have reaffirmed that global imbalances mainly result from unbalanced growth models and are better understood through the lens of national saving and investment gaps, where high domestic savings relative to domestic investment lead to the accumulation of current account surpluses and vice versa. These gaps largely reflect structural factors (demographics, income levels and resource endowments), cyclical conditions (business cycle) and policy choices (fiscal, health and exchange rate policies) (IMF, 2025). See G7 Academic Expert Group (2026), Recommendations of the G7 Academic Expert Group to the G7 Leaders on Global Economic Imbalances, French G7 Presidency, June and IMF (2025), External Sector Report: Global Imbalances in a Shifting World, July.
See Evenett et al. (2025), Industrial Policy Since the Great Financial Crisis, IMF Working Paper, No 2025/222 and Millot, V. and. Rawdanowicz, Ł. (2024), The Return of Industrial Policies: Policy Considerations in the Current Context, OECD Economic Policy Papers, No 34.
However, their impact on the current account balance is not clear a priori. Gourinchas et al. (2026) argue that micro industrial policies such as subsidies and tariffs have ambiguous effects on current account balances and typically require complementary macro policies to have a meaningful impact on external balances.[9]
See Gourinchas et al. (2026), Global Imbalances, Industrial Policy and Tariffs, IMF Working Paper, No 2026/067.
By contrast, macro industrial policies operate at the country level and influence broader economic conditions, making them more likely to affect the current account. For example, policies that encourage savings or keep the exchange rate undervalued can generate current account surpluses by holding back domestic consumption.
This blog post looks at micro industrial policies in the three largest economies. Focusing first on the intensity and sectoral scope of government intervention, we then investigate whether these subsidies are linked to trade imbalances, particularly in strategic industries.
Four facts about global subsidies
To get a better idea of where support is going, we use a new OECD database on Manufacturing Groups and Industrial Corporations (MAGIC), which is based on publicly accessible information. It records support received by 525 large manufacturing firms across 15 sectors in 52 countries from 2005 to 2024. Government support is classified into below-market-rate borrowing, government grants and tax concessions.[10] Below-market-rate borrowing includes loans, guarantees and other forms of financing provided on terms more favourable than those available on the market. Government grants are direct transfers of public funds that do not require repayment. Tax concessions reduce firms' tax liabilities through measures such as tax credits, exemptions, deductions and preferential tax rates (OECD (2021), Measuring distortions in international markets: Below-market finance, OECD Trade Policy Papers, No 247.
First, industrial subsidies mainly support a few strategic industries. Since the global financial crisis, government support as a percentage of global firms' costs (i.e. the sum of costs across sectors and countries) has almost doubled, reaching USD 108 billion in 2024. The largest increase can be observed in China (Chart 2, panel a). Strategic industries such as automobiles and semiconductors received the largest share (Chart 2, panel b). Across sectors, subsidy intensity (i.e. subsidies relative to costs) is larger for semiconductors and solar panels. Here, government support in the last five years averaged around 3.9% and 3.6% of firms' costs, compared with 0.8% in the automotive industry. Subsidies tend to be awarded relatively more to export-oriented companies. And state-owned companies get a substantially bigger share than private companies, especially in Asia. These patterns suggest that government support tends to prioritise strategically important sectors.
Chart 2
Global subsidies
(percentages of global firms' costs)
a) Subsidies by region |
b) Subsidies by sector |
 |
 |
Sources: OECD MAGIC dataset and ECB staff calculations.
Notes: ASEAN = Association of Southeast Asian Nations. Values are calculated as total subsidies received by firms in each country/sector divided by the total costs of goods sold (cogs) by all firms in the global sample, using five-year averages. The sample is restricted to firms that have been included in the database since 2010, ensuring a constant composition of the sample over time. Euro area, US and Chinese firms account for 15%, 18% and 27% of the sample respectively. Total costs of goods sold includes the cost of intermediate goods and services consumed by firms in their production process, as well as their labour costs, depreciation and amortisation. The latest observations are for 2024.
Second, while economic theory suggests that subsidies should be targeted to address specific market failures, OECD data reveal a much more heterogeneous pattern. Companies in China stand out in terms of the scale and sectoral scope of the support received, when measured as a percentage of domestic firms' costs. In 2024 subsidies accounted for about 2% of domestic firms' costs in China, compared with 1.4% and 0.6% in the United States and the euro area respectively (Chart 3, panel a). Moreover, China in particular targets a large number of sectors, as opposed to the United States (Chart 3, panel b). Meanwhile, in the euro area, government support has become broader-based in recent years. Between 2010 and 2024, Chinese companies received mainly below-market-rate borrowing, while companies in the United States received mostly tax concessions and those in the euro area mainly government grants.
Chart 3
Scope and size of industrial subsidies vary across countries
(percentages of domestic firms' costs and HHI index, three-year moving average)
a) Evolution of subsidies across major economies |
b) Sectoral concentration of subsidies |
 |
 |
Sources: OECD MAGIC database and ECB staff calculations.
Notes: Panel b) depicts an index for which higher values indicate greater concentration of subsidies in specific sectors. The latest observations are for 2024.
Third, the sectoral allocation of subsidies has shifted over time, particularly among Chinese companies. Between 2010 and 2024 subsidies granted to sectors linked to real estate (e.g. aluminium, cement) or more mature industries decreased. At the same time, strategic sectors such as semiconductors, transport, fertilisers and chemicals saw an increase in subsidies (Chart 4). For the United States and the euro area, the pattern is less clear-cut. Here, the sectoral allocation of subsidies has been relatively stable over time. The only exception are the solar and telecommunication sectors, where we observe shifts in the opposite direction in the United States and the euro area.
Chart 4
Sectoral allocation of subsidies: visible strategic shift in China
(percentage point changes)
a) Change in subsidies in China |
b) Change in subsidies in the euro area and United States |
 |
 |
Sources: OECD MAGIC database and ECB staff calculations.
Note: The two panels refer to the change in subsidies as share of domestic firms' costs between 2024-2020 and 2010-2014. The latest observations are for 2024.
Fourth, the Chinese industries that receive more support also tend to export more, although this does not prove that subsidies caused the export gains.[11] The recent widening of China's trade surplus has renewed interest in the nexus between the use of industrial policies, overcapacity and export growth. Using a text-based indicator of industrial policies based on firms' earning calls, Cesa- Bianchi et al. (2026) find that they are not first order drivers of current account surpluses unless combined with policies that suppress domestic demand, such as tight fiscal policy, foreign reserve accumulation and capital flow controls. Similarly, Gourinchas et al. (2026) argue that industrial policies are generally not expected to have aggregate effects on current account balances because most measures are sector-specific and too limited in scope to materially affect national saving or investment decisions. See Cesa-Bianchi et al. (2026), Industrial policies, global imbalances and technological hegemony, CEPR Discussion Paper, No 21253 and Gourinchas et al. (2026), Global Imbalances, Industrial Policy and Tariffs, IMF Working Paper, No 2026/067.
Chart 5
Foreign revenues vs government support
(percentages of domestic firms' costs and percentage changes)
 |
 |
Sources: OECD MAGIC database and ECB staff calculations.
But the link may also run the other way: governments may support sectors that were already strong exporters. To better isolate the role of subsidies in driving (our proxy for) exports, we estimate an econometric model for the Chinese sectors covered in the MAGIC database.[12] The SVAR is estimated over the period 2012 to 2024 at annual frequency for a sample of 14 Chinese sectors. The system comprises five endogenous sectoral variables in log differences: domestic revenue, foreign revenue, government support received, producer price indices (PPI) and total factor productivity (measured as revenue per employee). All variables but PPI are based on OECD MAGIC and are computed as sector averages. PPI is sourced from the National Bureau of Statistics of China via Haver analytics. Using sign and zero restrictions, we identify five key shocks: a domestic demand shock, a foreign demand shock, a subsidy shock, a TFP shock and a cost-push shock. The model is estimated with two lags and includes sectoral fixed effects.
First, increases in government support are followed by higher domestic revenues and exports (proxied by foreign revenues) for the subsidised Chinese companies. The effects are strongest in the first two to three years after a subsidy is employed and gradually fade thereafter (Chart 6, panel a).
Second, while subsidies appear to have made only a modest contribution to Chinese firms' overall export growth, in relative terms their impact is much larger for exports by strategic industries. In sectors such as automotive, solar panels, wind turbines and semiconductors, subsidies are estimated to have contributed between four and fourteen times more to export growth compared with the overall sample (Chart 6, panel b).
Chart 6
Contribution of a subsidy shock to foreign revenue growth
(percentage point change to one standard deviation subsidy shock, ratio of sector-specific vs aggregate subsidy-shock contribution to foreign revenue growth)
a) Impulse response function of foreign revenue growth to a subsidy shock |
b) Relative subsidy-shock contribution to foreign revenue growth in selected sectors |
 |
 |
Sources: OECD MAGIC database and ECB staff calculations.
Notes: The bars show the ratio between the contribution of subsidies to foreign revenue growth in a given sector and the contribution to aggregate foreign revenue. Contributions are obtained from a SVAR-implied historical decomposition of cumulative foreign revenue growth between 2014 and 2024.
Subsidies can have significant effects on strategic sectors' exports
Industrial subsidies alone are unlikely to drive aggregate global imbalances, which primarily reflect broader saving and investment patterns. But our findings suggest that they may meaningfully reshape trade within strategically important sectors. In particular, if subsidies boost exports beyond what underlying economic conditions would suggest, they may entail significant adjustment costs for trading partners, including job losses or de-industrialisation, ultimately causing trade tensions. This is a topic for future research and the results presented here should be interpreted with caution, as they are based on a relatively small sample of sectors and may not fully separate the effects of subsidies from other factors affecting Chinese firms.
The views expressed in each blog entry are those of the author(s) and do not necessarily represent the views of the European Central Bank and the Eurosystem.
