This paper did not begin with steel, or even with the European Union’s Carbon Border Adjustment Mechanism. It began while we were working together on two Horizon Europe proposals.
Our collaboration first developed around BEYOND EPICENTERS, a project on disaster resilience, which was subsequently funded with our team as coordinator. We then worked together on MATRIX, this time turning to climate resilience. MATRIX received very strong evaluations, and we are now preparing a resubmission, again as coordinator.
Writing these proposals meant spending a great deal of time thinking together about resilience from different disciplinary perspectives. What makes an economy, an industry, or a community vulnerable to an external shock? What determines whether that shock produces lasting damage or instead accelerates adaptation? And what kinds of investments actually increase resilience?
Those conversations gradually took us beyond the proposals themselves.
One issue that kept drawing our attention was the European Union’s Carbon Border Adjustment Mechanism, or CBAM. For an economist, CBAM raises questions about trade, competitiveness, market access, and adjustment costs. For an engineer, it raises a different set of questions: how much carbon is embedded in a ton of steel, where do those emissions come from, and which technological changes can reduce them?
We realized that neither perspective was sufficient on its own.
That insight became the starting point for this paper. We chose steel because it sits directly at the intersection of these questions. It is one of the world’s most carbon-intensive industries, it is deeply integrated into international trade, and it is central to the industrial economies of several countries across the Middle East and North Africa. At the same time, the production processes and electricity generation mixes used across these countries influence the carbon intensity of their steel.
So the question that interested us was not simply: How much will CBAM cost MENA steel exporters?
It was more practical: What would they actually have to change to reduce CBAM cost?
Answering that required us to combine pieces of evidence that are usually studied separately. We brought together life-cycle assessment of steel production, emissions from electricity generation and their implications for carbon costs, production technologies, exports to the European Union, carbon prices, and alternative decarbonization pathways through 2035.
This interdisciplinary approach was central to the project. Export volumes tell us something about economic exposure, but not about the carbon embodied in those exports. Emissions data tell us something about environmental performance, but not what those emissions mean once carbon acquires a price at the border.
Bringing the two together changed the policy question.
We examined the main steelmaking routes used across the region: the traditional blast furnace-basic oxygen furnace route, electric arc furnaces, and direct reduced iron combined with electric arc furnaces. Their carbon footprints differ substantially. But production technology is only part of the story. Electric steelmaking can be relatively low carbon when electricity is clean, and much less so when the national grid remains heavily dependent on fossil fuels.
This became one of the central insights of the paper: sometimes the electricity system is the steel policy.
The importance of that distinction became clearer as the country results emerged.
We had started with a regional question, but what we found was striking heterogeneity. There is no single MENA exposure to CBAM and, correspondingly, no single MENA solution.
Türkiye faces the largest absolute exposure because of the scale of its steel exports to the European market. Egypt is particularly dependent on that market. Saudi Arabia stands out because of the carbon intensity associated with its production and electricity system.
More importantly, the most effective route to reducing exposure differs across countries.
For Egypt and Morocco, where electric arc furnace production is already dominant, cleaner electricity does most of the work. The critical constraint is not primarily the steelmaking technology itself, but the carbon intensity of the grid.
For Algeria and Türkiye, the picture is different. Both the production route and the electricity system matter. Reducing exposure therefore requires coordinated change across industrial technology and power generation.
For countries with substantial direct-reduced-iron capacity, another possibility lies further ahead: replacing natural gas with green hydrogen. That pathway could eventually alter the competitive position of parts of the region, particularly given the renewable-energy potential of several MENA economies, although its economics remain challenging.
The scenarios also produced an important result for the policy debate.
Under high carbon prices, regional carbon-border liabilities increase substantially through 2035. Yet ambitious decarbonization can reduce those liabilities by as much as 56 percent compared with business as usual.
There is, of course, another way for exporters to pay less at the border: they can export less.
Our trade scenarios show exactly that. As carbon costs rise, exports can contract, lowering the total CBAM bill. But this apparent improvement comes at the price of lost export revenues. A smaller carbon-border payment caused by shrinking trade is not the same thing as becoming more competitive.
That is why decarbonization, rather than export contraction, is the more meaningful adjustment mechanism.
This also changed the way we came to think about CBAM itself. It is easy to describe it as a new cost imposed at the European border. But its implications go further. CBAM effectively makes the carbon content of production one component of international competitiveness.
For carbon-intensive industries, comparative advantage will increasingly depend not only on wages, energy prices, productivity, logistics, or proximity to markets, but also on how goods are produced and on the carbon intensity of the energy systems behind them.
Seen this way, CBAM is simultaneously a trade challenge and an industrial-policy signal.
For MENA economies this creates a difficult adjustment problem, but potentially also an opportunity. Several countries in the region possess substantial renewable-energy resources, experience with direct-reduced-iron steelmaking, or both. Investments that might otherwise be viewed primarily as climate policy can therefore also become investments in future export competitiveness.
The analysis also points beyond technology. Measurement matters. Exporters need credible information about the emissions embedded in their products. Domestic carbon pricing may matter as well, because carbon prices paid in the exporting country can be taken into account under CBAM. This creates the possibility of retaining some carbon revenues domestically and using them to finance the transition.
Our study inevitably has limitations. We use country-level electricity mixes rather than plant-level emissions data, and our life-cycle assessment captures a somewhat broader emissions boundary than the one ultimately used for CBAM compliance. National renewable-energy targets are also targets, not guarantees. The scenarios should therefore be read as pathways rather than forecasts.
But these limitations also point to the next questions. Plant-level emissions data would allow us to examine competitiveness within countries, not only between them. Future models could allow CBAM itself to influence investment and technology adoption. And the economics of green-hydrogen-based steelmaking, particularly in countries with strong renewable resources, deserves much closer attention.
Looking back, the route from our Horizon Europe work on disaster and climate resilience to steel and carbon-border adjustment no longer seems accidental.
Resilience is ultimately about the capacity to absorb a shock without losing the ability to adapt and compete. CBAM provides a particularly concrete test of that capacity for carbon-intensive industries.
And perhaps that is the broader lesson we took from this research. The most useful question is not simply how large the next external shock will be. It is whether we can identify, before it fully arrives, which structural changes will make economies better able to respond to it.