Clement Fuest has recently has shared a critique of the EU’s CBAM. This landmark pillar of the EU’s climate policy toolkit is under heavy attack from many sides. The narratives and attacks that are used follow simple arguments that are pervasive. But I fear that there is a high degree of simplicity in the arguments. His original article can be found here
If we zoom out, the goal of the CBAM is to not levy a border tax, because other countries adopt carbon pricing that is deemed equivalent. And in fact, as I highlighted elsewhere, this seems to be working. There are many other dimensions that are crucial to discuss, of course. We are in the process of a fundamental economic rebalancing and restructuring of many aspects of how our economies function, not only in terms of the energy consumption dimension, but more broadly, in robustifying societies that are subject to demographic change.
Now, let me dissect some of his comments to highlight some of the fallacies or challenges that I see in his thinking or argument.
1. Static vs. dynamic world: he assumes a frozen low-transparency equilibrium
Fuest treats today’s opaque, paper-based supply chains as if that’s the permanent state of the world. He assumes high monitoring and verification costs for CBAM are structural, not temporary. In reality, we’re in the middle of a technology shock that makes carbon accounting cheaper and easier over time, not harder. I shared an example of how AI reasoning model can very quickly and accurately estimate CO2 emissions. Its just math!
2. Carbon content is becoming easy to estimate
New tools (digital product passports, automated LCAs, IoT, AI on procurement data, etc.) enable sharp and relatively cheap estimation of carbon content for many goods. This means the compliance burden is endogenous: as these tools spread, the unit cost of compliance falls. Our work on the AI generated production network highlights how this can, in principle, work. Fuest’s claim that broad CBAM coverage is “administrativ kaum machbar” rests on yesterday’s technology, not on where we are heading.
3. Invoice-level VAT and e-invoicing: the new information layer
Within the G20, there is a clear shift towards invoice-level VAT / real-time e-invoicing. There is an obvious tension with the US on this dimension. In any case, a growing public data layer can help creating the information signature that is necessary to obtain default carbon intensity estimates and these can be supplied to the various annotator software solutions that carry out the estimation in absence of granular accounting. This creates a dense, transaction-level public information layer that can be used to: In such an environment, verification and enforcement can be largely automated. His picture of CBAM as manually checking every shipment is outdated.
4. Smart supply chains: compliance cost inside countries will fall
Payment systems and supply-chain software are moving towards “smart” supply chains with embedded metadata. This becomes most obvious when we think of invoicing software, or invoice metadata being attached to financial transaction data. This has the added benefit of reducing structural tax evasion, fraud, and other forms of evasion. Compliance with domestic regulation becomes a by-product of the infrastructure. So while Fuest argues CBAM will impose prohibitive bureaucracy, the direction of travel is the opposite: compliance cost per transaction is going down, not up.
5. Europe’s real problem: institutional, not technical
I fear that the real problem is an EU-institutional one. Similar as with competition policy preventing the completion of the single market through cross-border merger and acquisitons in the domains that is my understanding of what Enrico Letta suggests: banking and telecommunications, to create EU champions, I don’t think that the main challenge for Europe is not “CBAM is inherently not deliverable”. The real bottleneck is institutional: the EU still lacks a strong pan-European fiscal and data capacity. Right now it seems that the OECD is starting to become a knowledge hub for such data integration in Europe. Not the European Union.
6. The Kronospan/Turkey example is structurally misleading
In the piece, Clemens highlights a compelling example of a company that may face relocation pressures due to the CBAM. The case is a company making plywood. The production of which involves urea to make glue. The urea is mostly imported and is energy intensive or CO2 intensive under the current paradigm. Fuest’s relocation story (Kronospan → Turkey → duty-free access with dirty inputs) assumes CBAM’s product coverage stays narrow. But CBAM is designed for gradual expansion of coverage to more goods and more embedded emissions over time. A relocation strategy that exploits a temporary loophole does not define the long-run equilibrium: Using this as a central example is therefore misleading: it freezes a transitional design and treats it as a permanent structural flaw.
7. The Kronospan/Turkey example is structurally misleading (2)
An alternative to producing urea from natural gas (through production of ammonia) would involve working with hydrogen H₂ produced from water electrolysis powered by renewables (wind, solar, hydro), combined with N₂ from air. This produces green ammonia NH₃. The green ammonia can then be combined CO₂ that is potentially captured with firms being compensated if the urea enters a carbon sink (plywood). For example, one could capture CO₂ from industrial flue gases (cement, steel, power plants) that will in forseeable future be hard to avoid.
This would provide a double dividend as it would provide a valuable onward use for CO₂ and establish a carbon sink. This would be in the spirit of the circular economy. This cuts most of the upstream CO₂ emissions and offers a win-win. Of course, such investments in adapting alternative production processes will not happen, in a world where dirty production processes have a structural trade-induced comparative advantage. Many pilot projects and commercial plants are being built around the world focusing on green ammonia for fertilizers and there is a ton of innovation to effectively replace synthetic fertilizer and re-establish soil as a carbon sink.
8. CBAM is already working as intended: it induces foreign carbon pricing
In his analysis, other countries’ climate policies are treated as fixed and exogenous. But the core design logic of CBAM is to encourage other countries to expand on their climate policy tookit. And this process is underway. The CBAM comes with a dynamic shadow of increasing coverage across different products. This creates a powerful incentive for trading partners to adopt or tighten their own carbon pricing schemes. Evidence already points to this dynamic as more and more countries are adopting carbon pricing schemes, and this includes Turkey, the example that Clemens draws upon in his example of Kronspan. So the bleak scenario “EU decarbonises alone and loses competitiveness” ignores the central mechanism that CBAM is meant to trigger – and which is already in motion.