Author : Marguerite Fauroux
Many companies want recognition for actions that reduce emissions outside their value chain. These actions do not show up in their carbon footprint as defined by the GHG Protocol: that inventory measures the emissions of an organisation and its value chain, and the effect of its decisions on other parties’ emissions is not part of it. Distribution system operators (DSOs) are a case in point. They invest in infrastructure that lowers emissions across their territory, yet their own carbon footprint does not reflect it. It may even go up.
It is therefore legitimate for these players to seek recognition for this effort. The GHG Protocol is working on a framework that could address this: the forthcoming Actions and Market Instruments (AMI) standard. It sets out where and how such actions should be reported. The trickiest question has yet to be settled: against which baseline scenario should an action be judged to have actually reduced emissions?
Counting what was avoided means imagining what would have happened
An avoided emission is the difference between two scenarios: one in which the action took place, and a baseline scenario in which it did not. That second scenario does not exist. It has to be constructed.
What would have been chosen without the DSO’s investment? Heating oil, gas, or a heat pump that would have been installed a few years later anyway? Each option gives a different result.
Methods exist to frame these choices. In France, Carbone 4’s Net Zero Initiative has, since 2020, called for a clear separation between an organisation’s own emissions (its carbon footprint) and the emissions it helps others avoid. The initiative also distinguishes avoided emissions that correspond to an actual reduction from those that only reflect a smaller increase. In 2023, the WBCSD published guidance on avoided emissions, developed in collaboration with the Net Zero Initiative. The Climate Dividends association, founded in 2022, offers its own third-party-verified protocol for calculating avoided emissions, in which one tonne of CO₂e equals one “climate dividend”.
These frameworks converge on several principles: report avoided emissions separately, never subtract them from the carbon footprint, and document the chosen baseline scenario transparently. The WBCSD adds that they must not be used to claim net zero. On the baseline scenario, they agree that it must be justified. The choice itself is still made case by case. The WBCSD and the Net Zero Initiative also agree on one important point: the baseline should reflect what would have happened anyway. If regulation requires an installation to be replaced, the baseline is the average market solution that complies with that regulation. For a new need, it is the average solution sold that same year. A solution that merely meets the standard, or performs no better than the market average, therefore avoids no emissions, even if it is low-carbon. The GHG Protocol is considering a similar criterion for the AMI but has not yet decided on it.
None of these texts amounts to a common standard with the international recognition the GHG Protocol enjoys for carbon footprints.
The forthcoming AMI standard: several GHG statements, kept strictly separate
The AMI keeps the physical inventory as it is and proposes adding up to three separate statements alongside it, each with its own calculation method. The GHG Protocol presents them separately, with no netting between them, and leaves it to programmes such as the Science Based Targets initiative (SBTi) to decide what can be combined.
| Statement | What it shows | Example |
| 1. Physical inventory | Actual value chain emissions, scopes 1, 2 and 3 | Unchanged from today |
| 2. Contractual inventory (market-based) | What is attributed to the company through contracts or certificates | A company buys certificates for low-carbon steel produced somewhere other than at its own supplier |
| 3. Impact statement | What the action changed compared with a baseline scenario | A DSO funds infrastructure that reduces emissions across its territory |
| 4. Non-GHG indicators | Transition metrics not expressed in tonnes of CO₂ | The share of investment dedicated to decarbonisation |
The contractual inventory already exists for electricity, through the scope 2 market-based method. The AMI is considering extending it to scopes 1 and 3. The impact statement, for its part, would bring together the effects of a company’s actions, calculated against a baseline scenario: emissions avoided, reduced or removed, both within and beyond the value chain. Emissions avoided by products sold would form a separate category, and carbon credits would also be reported there.
The consultation held over recent months received more than 400 responses. According to the GHG Protocol, 84% of respondents support this structure. Other submissions, such as EFRAG’s, urge caution: address the gaps in existing inventories first, and build in safeguards against greenwashing. Many also ask that the statements remain strictly separate and that contractual reporting should not mask physical emissions.
Nothing has been finalised yet. The AMI will be integrated into a future corporate standard developed jointly by the GHG Protocol and ISO. A public consultation is planned for the second quarter of 2027, with final publication in the fourth quarter of 2028. The GHG Protocol also makes clear that it will not decide whether these figures count towards climate targets. That decision will rest with the SBTi, regulators and other programmes.
Emissions, contracts, impacts: three different things
The physical inventory, the contractual inventory and the impact statement answer three different questions. The first tells you what the company actually emits. The second tells you what is attributed to it by contract, for example through a renewable electricity certificate. The third tells you what its actions changed compared with a baseline situation. The risk lies in mixing these three answers: presenting a certificate purchase as a drop in the physical footprint, or as an emissions reduction the company brought about. The GHG Protocol is clear on this point: a reduction claim must rest on a method designed to measure reductions.
Guarantees of origin, used for years in market-based scope 2 accounting, are a good illustration of this drift. A study published in 2024 in the Journal of Industrial Ecology shows that the guarantees of origin purchased by companies, in Norway as in Germany, do not lead to new renewable capacity. The purchase only shifts where the low-carbon attribute is claimed.
The AMI working group is aware of the problem. Its working documents point out that the 2015 market-based method does not allow a company to claim it has caused an increase in renewable generation. Extending contractual logic to scopes 1 and 3 without resolving these issues would multiply the risks already seen in scope 2.
Meanwhile, global emissions keep rising
According to the Global Carbon Project, fossil CO₂ emissions are projected to reach a record 38.1 billion tonnes in 2025, 1.1% more than in 2024. The remaining carbon budget for limiting warming to 1.5 °C amounts to about four years of emissions at the current rate.
An avoided emission is calculated as the difference from a scenario that never happened. It therefore cannot be observed directly: how real it is depends on how credible that scenario is. It can represent a genuine climate benefit without emissions actually falling. The Net Zero Initiative makes this point by distinguishing an actual reduction from a smaller increase: in the second case, emissions grow more slowly, but they still grow. Recognising these efforts is still worthwhile. Presenting them as a fall in emissions would be misleading.
This is why the strict separation between the physical inventory and the impact statement, at the heart of the AMI proposal, is a step in the right direction, provided the statements remain truly separate and the baseline scenario is justified transparently. The question is how to keep this link with physical reality once the reporting framework is in place.
Keeping the link with reality by anchoring claims in a territory
Emissions are physical flows. They occur in a specific place, on a specific grid, in a specific installation. A single rule applied everywhere in the world will always end up drifting away from that reality.
At the scale of a territory, the link with reality becomes observable again. A DSO can track how consumption changes on its network. It can tie its investments to the climate strategies of the local authorities it serves. And it can check, year after year, whether the territory’s physical emissions are going down.
Several recent developments point towards stricter geographic anchoring. Version 2 of the SBTi Corporate Net-Zero Standard now limits market instruments for electricity to regions where the electricity can physically be delivered. The GHG Protocol’s revision of its Scope 2 Guidance is moving in the same direction, although the subject remains hotly debated: a market-based method requiring hourly matching and delivery on the same grid is being considered alongside the current method. On the AMI side, working documents propose that an impact may only be presented as linked to the value chain if it occurs in the same sourcing or use region.
Anchoring avoided emissions in a territory does not solve everything. The choice of baseline scenario is still a choice. But it becomes easier to discuss, to test against data and to correct.
Recognising action without losing sight of reality
The AMI responds to a legitimate expectation. Companies are acting beyond their value chain, and today’s carbon footprint gives them no way to show it. Clearly separating what is emitted, what is attributed by contract and what has been avoided could be a step forward, provided that separation comes with robust safeguards against greenwashing.
These new forms of reporting will only be valuable if they stay connected to physical reality. Before recognising an action, it must be possible to say what it is reducing emissions against, and then to check, wherever possible, that emissions are actually falling. The territory offers that point of verification. It is at this scale that part of the answer can be built.
These questions are at the heart of several of our current projects with energy-sector players. If you are thinking about them too, we would be glad to discuss them with you.