ncx-gitbot
commented
2 years ago NCX response: Projects are additional when the carbon stocks in the project scenario are greater than the carbon stocks expected under the baseline scenario–this is the basis for any carbon project verified against any standard. Because additionality, and therefore, creditable carbon is dependent on an accurate baseline, eligibility is limited to forests that are truly at risk of being harvested in the next year. Deferring that harvest results in additional carbon in the landscape. Longer project terms do not give higher confidence in model results as modeling forest management behavior over very long periods is in fact more difficult and uncertain than modeling over short periods. Furthermore, where the methodology is not explicitly prescriptive, it is expected that project developers will implement appropriate safeguards to avoid adverse selection. NCX signs a legal agreement with landowners that affirms their willingness to harvest the volume they are instead credited with deferring. We look forward to working with other developers and academic researchers to explore methods of measuring adverse selection directly in the future.
Commenter Organization: Berkeley Carbon Trading Project, Environmental Center, Goldman School of Public Policy University of California, Berkeley
Commenter: Barbara Haya
2021 Deferred Harvest Methodology Section: Comment on overarching approach - Additionality: The proposed methodology is vulnerable to non-additional crediting from adverse selection.
Comment: Additionality is trickier with tonne-year accounting than with land use methodologies requiring longer-term storage. If we knew perfectly what each participating forestland owner would do each year without offsets we could accurately measure the effect of offsets on on-site forest carbon stocks and credit appropriately. In practice, baseline stocks are uncertain in a long time frame and are even more uncertain in any particular year. Forest management differs between parcels with similar characteristics because of a myriad of factors. This is especially true for small-scale landowners, who are the main focus of this protocol. Compared to large industrial timberlands which can have established harvesting schedules, small landowner harvesting decisions are commonly affected by less predictable and less modelable circumstances such as the financial needs and forest management goals of forestland owners. Models can statistically estimate what a landowner is likely to do by comparing with other similar lands using dynamic baselines and taking into account the landowner’s past practice. But it is not possible to predict with confidence what would happen on all plots in any particular year. This means that carbon offsets will result in adverse selection. Of the pool of similar landowners, those that would not have harvested in the credited years are most likely to participate, because they can be paid for what they would have done anyway. To provide a quantitative example of adverse selection, let’s say that ten small landowners each have a modeled 50% chance of harvesting this year. We don’t know which ones would have harvested and which would not have; we only know that they all have a 50% chance of harvesting under current conditions. When we offer these ten landowners the chance to sell carbon credits for not harvesting this year, those who actually would not have harvested are likely to be the first to respond. Ideally the funds would be sufficient to convince some of the other landowners, who would have harvested, to decide to postpone harvesting by one year. But if participation is less than 100%, there is a good chance that more than half of the participants are from the set that would not have harvested anyway.
Proposed Change: Verra might consider selling something other than offset credits, perhaps “tonne-year carbon credits.” This would involve estimating effects programmatically, based on discernible changes in land management over the pool of participating lands and adjusting discount rates as needed to accurately reflect overall program impact.