A Claim About a World That Did Not Happen

The offset is the most comfortable idea in travel: a small payment at checkout that converts a flight into a neutral act. Two decades of evidence have been fairly unkind to it. The mechanism is not fraudulent in general, but the specific product sold to air passengers has failed on its own terms often enough that the honest position is to stop treating the checkbox as a solution and start treating it as a donation.

Why the accounting is hard

An offset claims that a payment caused a reduction in emissions that would not otherwise have happened. That claim requires four things to hold simultaneously.

Additionality: the project would not have gone ahead anyway. Renewable installations that were already commercially viable, and forests that were never going to be felled, generate credits for doing nothing. A correct baseline: the counterfactual against which the saving is measured is honest rather than inflated, which is the failure mode investigators have found most often. Permanence: carbon stored stays stored, which for forestry means for a very long time in a world with fire, drought, pests and changes of government in it. And no leakage: protecting one forest did not simply move the logging to the next valley.

Investigations across the major registries have found large fractions of credits failing at least one of these, most commonly the baseline — projects credited for protecting forest that was not under meaningful threat. The problem is not that verification is absent; verification is extensive and expensive. It is that the counterfactual is unobservable, and an unobservable quantity with money attached moves in the direction of the money.

An offset is a claim about a world that did not happen. Nobody can audit a counterfactual, which is why the market kept discovering it had paid for one.

The aviation-specific problem

Even a perfect offset addresses only the carbon dioxide. A substantial share of aviation’s warming effect comes from non-CO₂ effects at altitude — contrail cirrus above all, plus nitrogen oxides and water vapour in the upper troposphere — and those are not priced in any passenger offset scheme. A payment calibrated to the CO₂ alone is, by construction, paying for a fraction of the effect, and the multipliers proposed to correct it disagree with each other badly enough that no single figure commands consensus.

The timing mismatch is the second aviation-specific issue. The emission happens today; forestry credits deliver their sequestration over decades, if the trees survive. Paying now for a removal in 2060 is not the same transaction as not emitting, and the difference matters most for exactly the emissions that persist longest.

The defence of the market, fairly put

It exists. Some project types have much better integrity than forestry credits — landfill and industrial gas capture with metered results, and direct removal with geological storage. (Cookstove programmes, long sold as the safe choice, have their own documented over-crediting problem, for exactly the baseline reasons forestry does.) Standards bodies have tightened methodologies repeatedly in response to the investigations, and a credit issued today under a revised methodology is not the same product as one issued a decade ago.

And carbon finance has genuinely funded conservation and rural energy projects that no other mechanism was going to fund. The reasonable criticism is not that the money did nothing; it is that the claim attached to the money — that a specific flight has been cancelled out — was not supportable.

What has better evidence

Flying less, and this is not a moral point but an arithmetic one: the largest single lever available to a frequent traveller is one long-haul return not taken, and no purchasing decision within a trip comes close to it. Flying direct, since climb burns fuel fastest and a connection adds both a second climb and extra distance. Choosing the operator and aircraft type where a real difference exists on a route, which is published and checkable. Taking the train where the journey is under roughly six hours and the network exists.

Staying longer is the underrated one. The emissions of a trip are dominated by the flight, so a fortnight in one country has a much better ratio than two long weekends, and the destination gets more of the benefit as well.

And if you are going to pay something, pay it as what it is. Direct removal with durable storage is far more expensive per tonne and far more likely to be real. A contribution to an organisation working on the problem, made without claiming it cancelled the flight, is intellectually cleaner than a checkbox that lets the flight be described as neutral. The flight was not neutral. Saying so is the only version of this that survives scrutiny.