Carbon starts to be taxed: How green compliance moved from "PR slogan" to "cost line item"

The Analysis

The date to hold on to is 1 January 2026, and the number is 70%. Under Phase 2 of the EU Emissions Trading System for maritime transport, from that year the shipping company must surrender allowances covering 70% of verified emissions, against 40% in 2025, and the scope now reaches methane and nitrous oxide. That is a rule about the carrier's obligation. Standing alone, it is not a rule about your freight bill.

The gap between those two statements is where most of the confusion sits. A regulated obligation and a cost passed through by contract are different legal objects, and the second does not follow automatically from the first. A surcharge on your invoice is a contract term. It is not a tax, it does not acquire the authority of a tax because the cost behind it is regulatory, and whether it can be levied is decided by the clause rather than by the directive.

Three instruments are being discussed as though they were one: EU ETS, FuelEU Maritime and the CII. Different obligors, different triggers, different consequences. Merging them produces an exposure map that looks complete and is not.

For ETS maritime the obligated party is the company responsible for the ship. Your importing entity is not that party. It holds no allowance obligation, needs no registry account for this purpose, and faces no penalty under the directive when the ship comes up short. Its exposure is contractual: whether the carrier may levy a carbon or ETS surcharge, on what basis, supported by what evidence, and subject to what cap. The regulation is silent on all four, deliberately, because it regulates the ship and not the sale.

Seventy per cent is a surrender ratio, not a rate. The cost is verified emissions multiplied by the allowance price multiplied by that ratio. Two of the three variables sit outside the regulation: the allowance price moves with the market, and verified emissions move with the voyage. A per-container carbon figure is therefore a derived number, not a published one, and any quotation that states it without stating its basis has hidden the two variables that matter most.

Which voyages count, and at what share, is settled in the implementing provisions rather than in the headline. The material I am working from gives the 70% ratio and the inclusion of methane and nitrous oxide, and does not give the voyage attribution rules. I am not going to reconstruct them. If a quotation arrives with a carbon line, the question to ask is which voyage basis it uses, because two suppliers can be arithmetically correct and a factor of three apart.

Bringing methane into scope has a specific consequence that is easy to miss. Methane slip from LNG propulsion is now counted, which turns a fuel choice that was defensible on particulates and sulphur into a cost variable. That is a design constraint operating on the carrier's fleet, and it reaches you either as a surcharge or as a change in which ships are deployed on your lane. It never appears on your invoice as a methane line.

The three regimes should be kept apart. ETS is an EU obligation to surrender allowances, enforced through the registry. FuelEU Maritime sets a greenhouse-gas intensity limit per unit of energy, with its own penalty structure, and enters its monitoring phase in 2026. The CII is an IMO rating regime, where a poor rating forces slow steaming or withdrawal, and the consequence runs through charter terms and port state attention rather than through a duty. A vessel can be comfortable under one and exposed under another. Calling all three a carbon tax is convenient, and it puts the risk in the wrong place.

The capacity effect is the half that quotations do not show. Where older tonnage carries a poor CII rating, effective slots on Europe-bound routes are removed by regulation rather than by demand, and that removal stacks on top of Red Sea rerouting. The result is structural tightness in compliant slots specifically, which is a different market from the one a spot index describes. Carbon then appears twice: as a line item, and as scarcity of the compliant kind of space.

Exposure separates by contract type more than by company size. A large buyer with an annual tender, a carbon questionnaire and a scored environmental criterion has already moved this into procurement, which is where the source material locates the threshold behaviour of large clients. A smaller shipper booking spot sees the same cost with none of the visibility, because the surcharge arrives inside a number it cannot decompose. Both pay. Only one of them can evidence what it paid.

The 2026 obligation is already running, and the step from 40% to 70% is already inside the base. What arrives later is contractual: the next tender cycle, and the next round of annual freight agreements, where carbon reporting capability stops being a differentiator and becomes an entry condition. That is the transmission point, not the directive's commencement date.

Here is a proportional calculation, deliberately narrow. Assume a lane's carbon component was US$100 per FEU in 2025, which is my assumption and not a figure from the source, and that the same voyage basis and the same allowance price apply. Moving the surrender ratio from 40% to 70% multiplies that component by 1.75, so US$175 per FEU, a difference of US$75. At 500 FEU a year, another assumption, the delta is US$37,500. The calculation fails if the allowance price moves, if the voyage basis changes, if the carrier absorbs part of it, or if a contractual cap applies. Four conditions, any one of which breaks it. The industry estimate the source lists is a figure at carrier-fleet level, aggregated across ships; the distance between that number and your per-container cost is precisely the contract.

Without a carbon clause the cost does not vanish. It arrives inside an all-in rate, unlabelled, unverifiable and uncapped. A shipper in that position cannot reconcile a surcharge against any published basis, cannot pass it downstream with evidence, and cannot tell a regulatory increase from a commercial one. The absence of a clause is not protection. It is a decision to accept an unverifiable number.

Two mistakes recur. One is treating the surcharge as a tax, which makes it feel mandatory and non-negotiable; legally it is a price term, and price terms are negotiable by definition. The other is assuming the surcharge equals the ETS cost, when nothing requires reconciliation unless the contract does. The corollary is that a low-carbon service tier is a product and not a regulation, so being offered one says nothing about where you stand.

Where does this hold, and where does it break? It breaks if your cargo never calls at an EU port, because the obligation attaches to the voyage and not to the cargo's destination. It breaks if you buy on delivered terms and the seller holds the freight contract, because the surcharge is then someone else's line and your recourse runs through the sale contract. It breaks if the carrier absorbs the cost as a commercial decision, which some will do on competitive lanes. It breaks if the implementing provisions attribute your leg at zero. And it breaks at the level of my own source: I am reasoning from a summary that gives a ratio and a scope, not from the consolidated text. Until that text is read, the correct posture is to record what you were charged and on what stated basis, and to leave the characterisation open.

↑ Back to the key points

— By Dr. Ingrid Voss