We recently submitted two comment letters to the offsets registry Verra on the topic of using ton-year accounting to issue carbon credits. The first comment letter responds to Verra’s proposal to introduce a ton-year accounting option across its program — a change that would represent a profound shift in the way that Verra manages the permanence of carbon offset claims. The second comment letter responds to a methodology developed by Verra and the forest carbon offsets company NCX that uses ton-year accounting to credit forest harvest delays as short as one year.
Our letter about Verra’s program-wide reforms focuses on two overarching problems with the registry’s planned use of ton-year accounting to issue carbon credits.
First, we document how ton-year accounting is fundamentally inconsistent with net-zero climate goals that seek to stabilize planetary temperatures. Ton-year accounting asserts that temporary carbon storage is equivalent to the permanent effects of CO₂ emissions on the basis of a very limited consideration that ignores temperature outcomes. Using ton-year accounting to justify ongoing emissions with short-term carbon storage leads to higher warming over time, not climate stabilization. As a result, we argue that any credits issued on the basis of ton-year accounting should be labeled as inconsistent with net-zero climate goals.
Second, ton-year accounting introduces novel additionality risks that must be explicitly addressed. As proposed by Verra and NCX, ton-year accounting effectively gives projects an open option to exit from carbon storage commitments as short as a single year. This flexibility enables strategic, non-additional enrollment behaviors. For example, a forest landowner that is planning to harvest a nearly mature forest parcel five years from now could enroll in a ton-year offsets program and, in the meantime, earn credit while waiting for a business-as-usual harvest outcome. In light of these vulnerabilities, we believe that ton-year accounting should only be used with methodologies that explicitly account for its novel additionality risks.
Although a handful of other registries have authorized the use of ton-year accounting, to our knowledge, Verra's proposal is far and away the most expansive to date. The proposal not only allows the development of new methodologies that include ton-year accounting — like the methodology developed by NCX — but it also allows any project to petition Verra to use ton-year accounting under any methodology.
We urge researchers and market analysts to pay close attention to what Verra decides here. Ton-year methods are physically inconsistent with net-zero temperature stabilization targets and create substantial additionality risks. And because Verra’s proposal contemplates letting any project replace conventional permanence protections with ton-year methods, the consequences of erring on any of these dimensions could be significant.
Our second comment letter addresses a proposed ton-year methodology developed by NCX for crediting short-duration harvest deferrals — including projects that delay harvest for as little as one year. We focused on the novel additionality risks that ensue.
Critically, NCX’s proposed methodology defines any carbon stored above the baseline scenario as additional — but doesn’t specify a particular model or method. Instead, it outsources baseline model development and application to financially interested project proponents like NCX and does not require full public disclosure of those methods — a practice with a striking resemblance to a policy adopted by the Climate Action Reserve on which we’ve previously commented.
So while we know NCX has a model that predicts the likelihood of harvest for any given forest, we’re provided few details about model accuracy or validation benchmarks and are left guessing about how NCX’s approach actually works. The failure to fully disclose the model makes it impossible to assess the soundness of NCX’s proposed methodology and, as a consequence, calls into question its additionality claims.
Although we have serious concerns about any effort to sell offset credits on the basis of black-box modeling methods, the issue is more than philosophical here. We also describe what appears to be a gaping loophole in the proposed methods that makes clear what’s at stake.
The issue arises when landowners sell a timber harvest option to a third party, which is a common practice motivated both by tax law and the on-the-ground realities of harvest operations. The NCX methodology suggests that landowners who have already sold a harvest option to a third party can assume a 100 percent likelihood of harvest. The methodology then suggests that any resulting harvest deferral should be considered fully additional.
What makes this outcome so problematic is that once a landowner sells a timber harvest option, they no longer control if and when that harvest occurs. That decision now rests entirely with the option holder. In such a scenario, paying the landowner to defer harvest makes no sense: the credits generated under the methodology would be non-additional because they cannot change the landowner’s behavior.
Beyond these critical additionality concerns, we also document a pair of issues with the proposed method’s treatment of emissions leakage — that is, how to account for the fact that deferred timber harvests don’t necessarily change timber demand, which could shift (or “leak”) to forests that don’t enroll in offset programs. We spotted a potentially significant loophole in how large timber owners could define a key methodological term to shift harvest around within their own lands without accounting for leakage effects, and argue that any timber harvest prediction models that are used to credit deferred harvest under the methodology should also be used to stress test the thin assumptions underlying what few leakage protections apply.
The issues we raise in our NCX comment letter underscore our overarching concerns about Verra’s program-wide proposal. NCX’s proposed methodology is specifically designed around the use of ton-year accounting — a best-case scenario for managing the novel risks introduced by Verra’s ton-year accounting option. But even here we found major red flags. The use of a proprietary baseline model with methods and results no one else can see makes it impossible to tell if the methodology mitigates ton-year accounting’s unique additionality risks, and the presence of explicit loopholes for crediting non-additional behavior undermines our confidence in the methodology’s rigor.
Overall, we’re left with major questions about how any of this will work — and, if critical methods remain proprietary, whether we’ll be able to tell at all.