The drivers and impacts of subsidies to steel firms

4. The impact of subsidies on crude steelmaking capacity

The impact of cash grants on capacity

Cash grants have a clear-cut and positive impact on capacity, but only in partner economies, and have no statistically significant impact in OECD Member countries. A grant worth USD 1 million annually, sustained over a number of years,1 is associated with an increase of 5 000 to 15 000 metric tonnes2 in steel production capacity in partner economies, whereas grants show no correlation with capacity increases in OECD Member countries, a result similar to those of a previous study (Mercier, 2024[1]) which features prominently in the OECD Steel Outlook (OECD, 2025[2]). Interestingly, though, a similar decrease in grant amounts shows no statistically significant impact on capacity. This can be explained by how grants are recorded, and how increases (and not decreases) in grants serve as clear signalling devices that additional cash transfers have been provided by the government or some government-related agencies.3

The impact of BMB on capacity

The role of BMB is not straightforward to disentangle, as it is often used in a counter-cyclical fashion to rescue firms during crises in steel, as explained in Box 3 and illustrated in Figure 3. Lagging the variable by one year is likely insufficient to avoid endogeneity issues. If governments in partner economies choose to deliver BMB to firms in difficulty during stressed times, there is an opposing estimated impact (stemming from reverse causality) that makes BMB increases negatively correlated with capacity. This is a similar to those what is observed in a horizontal, cross-industrial sectors study (OECD, 2025[17]).

In (Mercier, 2024[1]), adding debt as an explanatory variable made BMB’s effect irrelevant, suggesting BMB’s effect was not impacting capacity once total debt was taken into account (with indebtedness levels clearly being correlated with steel crises and thus incorporating the endogeneity). In the sense that BMB served to rescue or alleviate financial stress on steel firms, it could be argued that although BMB did not result in a net increase of capacity at those times, BMB helped maintain capacity levels through crises, thus preventing a capacity reduction. This countercyclical aspect of BMB is very well illustrated in Figure 3.

Although the countercyclical effect of BMB during steel crises cannot be directly estimated due to its probable endogenous nature, it is still possible to identify the effect of BMB outside these crises’ episodes. This is the approach taken in ‎Annex G, where instruments are used to estimate first the exogenous (independent of government response to crises) component of BMB, which is then used as an explanatory variable. The estimated impact of this exogenous BMB is statistically significant: a USD 1 million increase in BMB support is associated with an increase in capacity of approximately 1 000 metric tonnes. While this effect is roughly ten times smaller than that estimated for grants, it is important to note that grant amounts recorded in the data often understate the actual value received by steel firms (see Box 8 in Annex B), whereas BMB figures reflect the full annual value of below-market financing actually benefitting the firm.

The coefficient on BMB should therefore be interpreted as capturing its structural contribution to capacity growth, rather than its countercyclical role during crises. Indeed, by instrumenting BMB, the IV strategy effectively removes the crisis-driven component of both BMB allocation and capacity variation. In other words, the coefficient reflects how BMB affects capacity in normal times, outside of its function as a crisis-response tool. While this means the stabilising effect of BMB during steel downturns is not captured, the estimate provides valuable insight into the long-term, policy-driven impact of BMB on capacity.

Other factors

Steel product demand

Other variables correlated with capacity changes are the changes in steel product demand in money terms (an indicator of global demand): the impact is negative, indicating that decreasing global demand somehow still slows increases in capacity (‎Annex F).

Firm size

Firm size also seems to play a role, with increases in a steel firm’s revenue increasing its capacity in the following years.

Profitability

Profitability, as captured by the profit margin, is negatively correlated with capacity changes in partner economies, even when controlling for all fixed firm individual effects and the usual controls (‎Annex F).

In OECD Member countries, profitability, as captured by EBITDA, is positively correlated with capacity changes in subsequent years, but not in the current year.4

Notes

← 1.  Due to the accounting treatment of some asset-linked grants in steel firms’ financial statements, the actual financial inflow is often greater than what is initially reported. These grants are recorded progressively over several years, following the amortisation schedule of the related asset. This explains the use of the phrase 'sustained over a number of years'.

← 2.  The bounds of the interval reflect variations across model specifications, with significant rounding applied to avoid overstating the precision of the results. Further estimations are provided in ‎Annex G.

← 3.  The main results of our revisited impact study are presented in ‎Annex F and can corroborate the findings of [DSTI-SC-2024-7].

← 4.  A proper inspection of Table A ‎F.2 indicates that EBITDA is negatively correlated with current capacity changes, but this is mostly due to revenue becoming significant and taking part of the EBITDA coefficient. Re-estimating by dropping the variable “revenue” in those two equations yields a non-statistically significant EBITDA variable (for explaining contemporary changes in capacity), consistent with the absence of a link between current EBITDA and current capacity.