Insights · Methodologies
Additionality for carbon credits: how tree-planting projects prove it
Additionality is the test behind every carbon credit: would this have happened without the carbon money? For tree planting the question has a twist, because trees sometimes come back on their own. This post explains how additionality for carbon credits works under VM0047, how it differs from older tools, what evidence a project needs, and what a feasibility screening can and cannot tell you.
What additionality means
A carbon credit is supposed to represent one tonne of CO2 that was removed or avoided because of the project. If the same tonne would have been removed anyway, selling it as a credit lets a buyer claim a benefit that never happened. That is the core of additionality.
The ICVCM, the body that sets the Core Carbon Principles for the voluntary market, puts it as Principle 5: emission reductions or removals must be additional, meaning they would not have occurred without the incentive created by carbon credit revenues.
For tree-planting projects (afforestation, reforestation and revegetation, or ARR), there are two ways a project can fail the test:
- The trees would have been planted anyway. A timber plantation that pays for itself, a government programme that already funds planting, or a law that requires it.
- The trees would have grown back anyway. Abandoned farmland in a wet climate can regrow forest by itself. Crediting that regrowth would pay for nature doing what it was already doing.
The second failure mode is specific to land projects, and it is why VM0047 handles additionality differently from methodologies for, say, energy projects. Many past problems in the market came from weak additionality, as covered in why carbon projects fail on additionality.
How VM0047 tests additionality
VM0047 has two quantification approaches, and each has its own additionality route. The details below are from version 1.1 of the methodology (Section 7).
| Area-based approach | Census-based approach | |
|---|---|---|
| Typical projects | Block planting and assisted natural regeneration that turn non-forest into forest, or add carbon to forest not managed for wood products in the past ten years | Dispersed direct planting with a full count of trees, for example on farms (at most 50 planting units per hectare) |
| Method type | Performance method | Project method |
| Regulatory surplus | Required | Required |
| Performance benchmark | Required, dynamic, re-tested at each verification | Not used |
| Investment analysis | Required, unless the project has no revenue or incentives other than carbon credits | Always required |
| Common practice | Covered by the benchmark | Required: adoption below 15% among similar landowners |
Regulatory surplus is the simplest test. The project must not be something the law already requires. If a national rule obliges landowners to replant, or a permit condition requires restoration, those trees are not additional. VM0047 refers to the VCS Methodology Requirements for the detail.
Investment analysis asks whether the project would be financially attractive without carbon revenue. VM0047 points to Step 3 of Verra's VT0008 Additionality Assessment tool, which allows a benchmark analysis or an investment comparison analysis. For area-based projects it is not required where the project earns nothing and receives no incentives other than carbon credit sales.
Common practice applies to the census-based approach. The project surveys similar landowners in a comparable area. If 15% or more of them already do the same planting without carbon finance, the activity is common practice and not additional. Census-based projects are also exempt from a leakage deduction, because they must not change the existing land use: the farm stays a farm, with trees added.
Which approach applies is not a free choice. VM0047 excludes area-based projects that plant fewer than 50 planting units per hectare and could use the census-based approach, and the census-based approach is limited to at most 50 planting units per hectare.
The performance benchmark: additionality measured, not argued
The area-based approach replaces much of the old argument with measurement. The project picks control plots outside any registered project that match its own land on biophysical, social and political conditions and on their historical stock trend. Using remote sensing, it then tracks a vegetation stocking index on both.
If the project plots gain significantly more than the control plots (VM0047 uses a Z test), the project shows additionality for that period. If the surrounding land is regrowing just as fast, the project does not. The same benchmark sets the crediting baseline, so regrowth that would have happened anyway is netted out of the credits.
- Measured withRemote sensing, control plots
- Re-testedEach verification, at least every 5 years
- Also setsThe crediting baseline
At validation, the project only has a forecast: an ex-ante calculation that compares its modelled growth with a benchmark derived from the control plots' history over the ten years before the start. The real test comes later. That has a direct consequence for anyone forecasting credits: the deduction for business-as-usual regrowth is not known in advance. We explain what that does to a forecast in the dynamic baseline in VM0047.
The ICVCM approved VM0047 as meeting the Core Carbon Principles in a Governing Board decision dated 11 December 2024.
Older tools vs the benchmark approach
Before VM0047, most ARR projects under Verra demonstrated additionality with VT0001, the tool for AFOLU project activities. VM0047 lists VT0001 among its sources. That style of test is argued at the start: identify realistic alternatives for the land, show through an investment analysis or a barrier analysis that the project is not the most attractive option without carbon revenue, and show that it is not common practice.
The weakness of that approach is that much of it rests on the project's own narrative and financial assumptions, judged once. Barrier arguments in particular were hard to falsify. In 2024 Verra released VT0008 and VT0009 for non-AFOLU projects, which tightened this: barriers now need verifiable evidence that carbon revenue is decisive, technological barriers were dropped, and the “first of its kind” argument was removed. Verra said at the time that the new tools do not yet replace VT0001 for AFOLU, and may absorb it later.
VM0047's area-based approach keeps regulatory surplus and an investment analysis, but adds a test that keeps running. A project that looked additional on paper and turns out to sit in a landscape that is regreening anyway will see it in its monitoring results.
For the bigger picture of the methodology, read VM0047 explained.
The evidence a project needs
Under VM0047 a project developer should expect to assemble:
- A legal review showing that no law, regulation or permit condition requires the planting.
- A financial model for the investment analysis: costs, any timber, fruit or other revenue, subsidies and grants, with assumptions that match what financiers are told.
- Land history showing what the land looked like and how it was used before the start date, and that it was not managed forest at any point in the ten years before.
- Control plot design for area-based projects: matched plots, a stocking index from remote sensing and the historical trend used for the ex-ante benchmark.
- A common practice survey for census-based projects: a representative sample of similar landowners within five years of the start date, or qualifying public statistics.
What a feasibility screening can and cannot say
A desktop carbon feasibility study cannot prove additionality. It can surface the signals that make it hard, early and cheaply:
- Natural regrowth pressure. Land history from satellite data shows whether tree cover around the site is already rising. Wet, abandoned land next to forest is a warning sign for the performance benchmark.
- Current land use. The intake asks what the land is used for today. That answer shapes leakage, and in our screenings it was the answer that moved net figures most on farmed land.
- Revenue other than carbon. A plan with substantial timber or crop revenue will need a convincing investment analysis.
That is why our reports list additionality under next steps, not as a pass or fail. More on how the screening fits the wider path is on the reforestation feasibility page.
Frequently asked questions
A project is additional if the emission reductions or removals would not have happened without the incentive of carbon credit revenue. For tree planting, the question is whether the trees would have been planted, or would have regrown by themselves, anyway. If they would, the credits do not represent a real climate benefit.
It depends on the approach. Area-based projects must show regulatory surplus, pass a dynamic performance benchmark that compares project plots with matched control plots, and usually complete an investment analysis. Census-based projects, which count individual planted trees, must show regulatory surplus, complete an investment analysis and pass a common practice test.
Not under the VM0047 area-based approach. The project shows expected additionality at validation with an ex-ante forecast, then additionality is reassessed at each verification by testing whether project plots grew significantly more than control plots. That is different from older approaches, where the test was mostly done once, at the start.
No. A desktop screening can flag obvious risks: land that is regrowing on its own, planting that is already common practice, or a project that would pay for itself without credits. The proof comes from a financial model, legal review and, under VM0047, monitoring results against control plots over the life of the project.
Sources
- Verra: VM0047 Afforestation, Reforestation and Revegetation, v1.1 (Section 7 Additionality; Appendix 1 Performance Benchmark)
- Verra: VM0047 methodology page and related documents
- Verra (2024): Verra releases new VCS additionality tools (VT0008 and VT0009)
- ICVCM: the Core Carbon Principles (Principle 5, additionality)
- ICVCM Governing Board decision VCS_VM0047_v1_2024 (11 December 2024)