How to find the key insights

Data-driven Chapters

Subsidies to steel firms are widespread, persistent, and delivered through multiple instruments. Using the OECD MAGIC database (2005–2022), the report finds that nearly all observed steel firms receive government support in some form—only 6% of firm-year observations show zero subsidies, while about 90% of firms receive subsidies in at least half of their observed years. Main subsidy instruments include: cash grants and transfers (direct financial support for projects, equipment, upgrades, or cost reimbursements); below-market borrowings (BMB) (preferential or implicitly guaranteed loans from state-controlled or influenced lenders); and corporate income-tax concessions (tax exemptions, reduced rates, credits, or benefits linked to investment or performance).
Subsidies fluctuate substantially, increasing sharply during crises. In the 2009 global financial crisis and China’s 2015–2016 slowdown, BMB surged. Subsidy intensity (subsidy size relative to firm size) varies widely: China’s subsidy intensity is five times higher than that of other non-OECD countries and 10 times higher than OECD countries. Key firm characteristics influencing the size and likelihood of subsidization include home jurisdiction, firm size and the degree of state ownership. The percentage of state ownership is more informative than a binary state-owned enterprise (SOE) vs. non-SOE classification, and higher government ownership strongly correlates with receiving larger subsidies relative to firm size.
Steel firms with higher state ownership receive significantly more subsidies across all instruments: Firms with >25% state ownership receive twice as much in grants and tax concessions (relative to asset size) as firms with <10% state ownership. They receive three times more BMB. This reflects both direct support (grants, tax benefits) and indirect support (cheaper or guaranteed loans).
Higher state ownership also implies higher indebtedness and lower profitability. State-owned or state-influenced firms tend to be more indebted, with debt-to-asset ratios around 30% higher. Such firms are less profitable, with return on equity and return on assets almost three times lower for majority-owned SOEs. In China, SOEs' debt-to-capacity ratio is 2.5× that of private firms, showing deep structural imbalances driven by state support.
State ownership is the primary driver of subsidy allocation. Once ownership share is controlled for, profitability indicators generally do not predict subsidies. In OECD countries, higher debt results in fewer subsidies. However, in non-OECD countries the data shows that more debt can result in mores subsidies (no negative effect). Firm size also matters, with larger asset bases likely to attract more subsidies in non-OECD countries. Global steel demand is usually not significant in subsidy allocation. China is the exception; Chinese SOEs maintained very high utilization rates (84% vs. national averages ~65%), implying government support sustained production even in weak markets.
There is a correlation between subsidies and overcapacity. Cash grants increase steel production capacity, but only in non-OECD countries, not in OECD countries. In non-OECD countries, a sustained USD 1 million annual grant is linked to a 5,000–15,000 ton capacity increase. No measurable impact is found from grants in OECD members. BMB’s effects are complicated by endogeneity: governments often provide BMB during crises, making rising BMB appear negatively correlated with capacity due to reverse causality. BMB likely helps maintain existing capacity during downturns rather than expand it. Using instrumental variables to isolate the exogenous component of BMB, a USD 1 million increase in BMB is linked to a ~1,000 ton capacity increase. This is smaller than the grant effect but reflects a structural (non-crisis) contribution to capacity growth.