Lessons learned & the way forward

‍

Vanishing Trees

The trees are falling, one by one
Centuries-old giants now come undone
Nature's balance thrown into disarray
As forest homes are cleared away
Valuable wood now turned to cash
Short-sighted gains that cannot last
If we don't act now to conserve
What remains of forests to preserve

– Open AI’s GPT-3

‍

Recent headlines and debates surrounding carbon credits from avoided deforestation motivated this review of a project type that has been subject to critical discourse since its inception. This publication aims to provide background information on REDD+ projects, summarize the current discussion, and outline possible directions for this type of project.

Our forests are under threat

Forests store approximately 400 gigatons of carbon, accounting for 92% of terrestrial biomass worldwide (Kayler, 2017). Recent research shows that these ecosystems have an even greater cooling effect than previously thought, highlighting their crucial role in keeping temperatures down (Lawrence et al., 2022). 

However, trees are disappearing rapidly across the globe. Land use change (principally deforestation) accounted for approximately 5 gigatons of CO₂ annually between 2007–2016. This is equivalent to as much as 10% of the total global greenhouse gas emissions (IPCC, 2019). In fact, some tropical forests now emit more carbon than they absorb, as deforestation and degradation have turned them from carbon sinks into sources (Baccini et al., 2017). 

While the specific agents and causes of deforestation vary by region and over time, economic forces fundamentally drive deforestation. The rising global demand for commodity crops, timber, and minerals incentivizes hundreds of millions of economically disadvantaged people who are seeking to fulfill their basic needs in and near tropical forests to convert land for more profitable uses. Simply put, as long as deforestation delivers more economic value than conservation for the deforestation agents, forest loss will not be avoided at the rate and magnitude required to limit global warming to 1.5 °C by the end of this century.

Without rapidly halting deforestation this decade, humanity has little to no chance of avoiding climate change's most catastrophic impacts. The landmark Stern Review concluded that reducing deforestation is the “single largest opportunity for cost-effective and immediate reductions of carbon emissions” (Stern, 2007). 

Emergence of REDD+ 

In response to this pressing need, policymakers have designed a series of policies – collectively known as Reducing Emissions from Deforestation and Degradation (REDD) – to provide a financial incentive to governments, businesses, and communities to conserve forests. 

After Costa Rica and Papua New Guinea placed this topic on the 2005 COP11 agenda, the idea generated widespread enthusiasm for a new approach to forest conservation through promoting performance-based incentives at scale. By placing an economic value on the carbon stored in forests, REDD can encourage sustainable forest management and make forests more valuable standing than cut down. Through the sale of carbon credits, REDD developers aim to provide local communities with revenue for maintaining the forest in which they live. This drives them away from other potential sources of income that would involve the destruction of the forest, such as palm oil cultivation, cattle ranching, or illegal logging.

At the COP13 in 2007, the term was expanded to “REDD+” to reflect wider activities and co-benefits. As of August 2022, 624 individual REDD+ projects and programs have been initiated — with about two-thirds still active (Simonet et al., 2020). The majority are paid for by multilateral and bilateral donors, including the World Bank and the UN-REDD initiative. 

Only a handful of these projects have received funds through the voluntary carbon market (VCM). In 2011, the Kasigau Corridor project in Kenya became the first project to issue carbon credits for REDD+ under the Verified Carbon Standard (VCS). 

As of March 2023, the global voluntary carbon market (VCM) project data available to CEEZER indicates that there are nearly 200 REDD+ projects recorded at either Verra or Plan Vivo. 91 of these projects are currently active, i.e. they have completed the registration process and have either issued or will soon issue credits. The majority of these projects are located in South America (54), followed by Africa (21), Asia (8), North America (7), and Oceania (1). Over 60% of all registered REDD+ projects have associated co-benefits, which is significantly higher than the 23% seen in the general VCM landscape.

It is interesting to point out that the "carbon efficiency" (i.e. the ability to avoid emissions per area) of these REDD+ projects varies greatly. Projects with higher carbon efficiency will avoid more emissions (and thus generate more credits) per hectare of land. According to our data, the average estimated annual emission reductions per hectare is around 4.4 tons of CO₂ – but many projects have a factor between 0.1 and 10 times that. The extent of this variance depends greatly on the local context of the project, including whether the project area is facing unplanned or planned deforestation, or parameters such as biomass density. While carbon efficiency seems to be an accessible indicator for project quality, the high variance illustrates one of the main challenges of this project type: the identification of a suitable baseline that fundamentally influences the carbon efficiency value. 

Criticism and accusations - the baseline dilemma

Carbon accounting associated with REDD+ projects is complex, partly because the hypothetical risk of forest loss without the carbon project must be quantified. One important parameter to assess a project’s effectiveness is the “baseline” – i.e., what would happen to the forest without the existence of the project. Baseline scenarios are calculated by using similar areas as a reference or counterfactual. Deforestation activity observed in those reference areas will then be translated to the actual project area. The identification of adequate reference areas is a critical step for the emissions reduction equation as it determines a (ideally very accurate) proxy for the degree of deforestation that will have ultimately been avoided by the project activity. 

Ever since the first REDD+ projects were implemented, discussions about effectiveness, accuracy, and social impacts have played an increasingly central role in defining the discourse across multiple registries (Badgley, 2021; Gifford 2020; Global Forest Coalition 2020; Duchelle et al. 2018; Pan et al., 2022).

A recent article published in The Guardian presented academic research that questioned the accuracy of baselines used to calculate project emission reductions in several Verra-certified projects (West et al. 2020; West et al. 2023; Guizar Coutiño et al. 2022). The underlying papers used an alternative way of identifying reference areas (synthetic controls) to create proxies for individual REDD+ projects as opposed to identifying individual reference areas as required for developing a baseline scenario under the VCS standard. The article claims that only a few Verra REDD+ projects showed evidence of deforestation reductions and indicated that 94% of the credits had no climate benefit. According to underlying research, the threat to forests had been overstated by about 400% on average.

While critical voices and open discourse are valuable means to motivate the evolution toward more granular baseline assessments and specific calculation methodologies, critical studies should be subject to the same rigor and scrutiny that is being demanded from carbon credit methodologies. In the case of The Guardian article, for example, it is worth noting that the broad datasets used in the cited studies are not very accurate at the individual project level. Verra’s VCS methodology typically requires locally calibrated data at a granular level to achieve greater accuracy. 

In response to the article, many actors within the VCM commented on the referenced publications and the general importance of REDD+ projects, among them project developers and resellers like Everland, South Pole, and Respira, but also the International Emissions Trading Association (IETA), rating provider Sylvera, and also Verra published its responses to the article.

Summary of responses from key stakeholders 

• IETA responded to these criticisms, urging more balanced coverage to include a broader range of expert views. They stated that methodologies are built on a combination of fact and expert judgment. “There might be different expert views, sometimes based on new information, but often they simply have a contrarian view that needs to be scientifically tested.”
• Sylvera, a carbon rating provider reacted with reference to their own in-depth research showing that the number of seemingly worthless credits has been dramatically overstated by the cited publications. They dissected the synthetic baseline technique used in which control areas were selected solely based on physical characteristics (i.e. distances from roads, rivers, settlements, slopes, etc.). “Those physical characteristics, relating to accessibility and proximity to markets are all relevant factors. However, the single and obviously most important factor has been completely omitted—the proximity to the active front of deforestation”. Furthermore, Sylvera points out that forests need to be protected to deliver our climate goals and that we “must not sabotage the financing of projects that are delivering sound climate, social, and biodiversity benefits”. 
• Verra itself formulated a response to the press coverage criticizing the applied methodology of synthetic controls and emphasizing the continuous improvement of its methodologies based on the best available science and technology through rigorous consultations with academics and experts. In fact, Verra will release a new consolidated REDD methodology in Q3 of 2023. The updated methodology aims to “ensure that forest conservation credits maintain the highest level of integrity”. It is an evolution of “existing procedures for robust and transparent quantification of greenhouse gas emission reductions from high-quality forest conservation activities”, including allocated activity data using a single deforestation dataset for the entire jurisdiction and crowdsourced data for deforestation risk. 
• Pachama presented their solution to compute rigorous baselines in their latest research brief as a reaction to the article. The algorithm they use ensures that carbon credits equal real emission reductions by matching the carbon project to a control area within a search region. The matching features consider key deforestation risk factors and enable baseline validation against independent observations using placebo projects. Over the coming months, Pachama announced plans to incorporate feedback, extend the analysis to more projects, and finalize detailed technical documentation for peer review.
• Carbon Direct’s response mainly supports The Guardian’s findings referring to own, in-house research (unfortunately not presented or linked). Carbon Direct argues that many segments of the economy still struggle to decarbonize at scale while the IPCC stresses the clear need for large-scale carbon removal.

Building the different pieces of the puzzle to address these challenges

In the early days of REDD+, technology such as satellite data, artificial intelligence algorithms, or cloud computing was not available. In other words, tools to establish baselines with high certainty were lacking. Indeed, lengthy PDFs, spreadsheets, and manual measurements are still the core tools for measuring and verifying the bulk of the forest carbon credits today. But like all sectors of the global economy, the VCM is busy digitizing itself. 

For instance, Verra created the Digitalizing Monitoring, Reporting, and Verification (DMRV) Working Group, and is running a pilot with Pachama that will harness remote-sensing data to measure forest carbon. Moreover, the use of remote sensing, deep learning, and other innovative technologies allows for monitoring forests and the carbon stored in them with unprecedented accuracy. Carbon quality rating agencies that specialize in nature-based projects – such as Sylvera and Renoster – are providing an additional layer of quality assurance to the REDD+ market by utilizing such technologies. 

New ways of assessing the business-as-usual estimate of expected carbon emissions in the absence of a carbon project are being proposed, too. For example, dynamic baselines use an approach that is somewhat similar to the studies that were featured in The Guardian. Mimicking randomized control trials used in drug testing, a control area is algorithmically selected as a suitable comparative reference for the carbon project. By comparing the project's effectiveness to that of the control group over time, the project’s emissions reduction can be determined. 

While more research is needed in the field, blockchain also has the potential to improve verifiability, reduce transaction costs, and assist in addressing additionality and permanence concerns for REDD+ projects (Kotsialou et al., 2022). The Open Forest Protocol is an example of this. It is a blockchain-based MRV solution for forestation data that provides clearly attributable measurement and verification data. However, this effect only becomes significant if MRV is actually more accurate, as otherwise it would replicate the current problems of the central ledger system.

Another trend in the REDD+ landscape is the shift from project-level to jurisdictional approaches. Jurisdictional REDD+ covers an entire local government area or jurisdiction, such as the Amazonas state in Brazil, with a single reference baseline or reference level of deforestation. By taking a more holistic approach, these programs address deforestation drivers more effectively, potentially reducing exposure to non-permanence risks and reversals (Schwartzman et al., 2021). Previously, these approaches were mainly used for results-based finance agreements between countries or multilateral organizations such as the World Bank (e.g., through their Forest Carbon Partnership Facility). 

While facilitating the transition to jurisdictional approaches, so-called “nesting” provides a practical solution to the criticisms of project-level REDD+. Such nested projects are aligned with jurisdictional baselines, which help to manage leakage and enforce environmental safeguards across programs and projects. However, the transition to a jurisdictional approach does not come without risk either: while political leaders easily sign-off environmental pledges, their enforcement is not necessarily ensured (Duchelle, 2018; Nepstad et al., 2021). Effective jurisdictional strategies are needed that translate finance – or the prospect of finance – into forest-friendly changes in the land-use system that are embedded in public policies and programs aligned to achieve.

The requirement to comply with state-of-the-art accounting and crediting standards will drive up costs to develop and maintain projects significantly. According to a recent report by the UNEP, a clear signal in the form of a carbon price of USD 30–50 per ton of CO₂e for forest emissions reductions is necessary to incentivize countries to invest in the protection of their forests.

Uncertainty will remain inherent to REDD+ in the future 

While the approaches described above are valuable improvements to the verification process and address some systemic risks, carbon accounting for REDD+ projects will continue to rely on probabilities and estimates in the future. Fundamental challenges facing carbon credits, such as defining additionality, cannot be solved by digital technologies alone. In other words, it will never be possible to have a certain understanding of what could happen to a specific area in the absence of the project. 

As practices improve, we will become better at calculating probabilities. This will increase the accuracy of our estimations for carbon credits even further. New technologies and approaches will allow the market to course correct itself. Yet, continuous improvement also means that it is somewhat inevitable that what is viewed as good practice in the carbon market today will change next year. 

What is the risk associated with purchasing REDD+ credits today?

After the broad press coverage that REDD+ received in early 2023, there has been a growing misconception that all credits from these projects available in the VCM bear a greater quality risk than credits from other project types. Given the uncertainty described previously, some buyers now question whether REDD+ credits should be excluded from their portfolios altogether. 

Needless to say, poor-quality projects should be called out and not used for compensation purposes. There is no place for them in a robust carbon market. However, many REDD+ projects have had significant positive impacts on climate, biodiversity, and local communities. The factors that determine the quality of these projects are diverse and complex. To ensure the social and environmental integrity of their purchases, buyers need the knowledge and tools to conduct additional due diligence beyond exclusive reliance on standards. We see a wide range of expertise within buyer organizations and – related to that – different levels of guidance needed to master the VCM jungle. Before being accepted for trading on the platform, every project undergoes a rigorous individual quality screening by CEEZER. 

For REDD+ projects, the following quality parameters are especially important to assess during the screening:

• Location: Questions such as “where is the project based (country, jurisdiction)?”, “what is the current legislation regarding deforestation activities?” or “how is legislation being enforced?” help determine factors like baseline and additionality. The Human Development Index (HDI) and Corruption Perceptions Index (CPI) also provide useful information.
• Methodology: Whilst assessing the baseline deforestation risk, it is crucial to determine whether the project concerns avoided planned or unplanned deforestation, i.e. a more conservative (unplanned) or less conservative (planned) scenario.
• Baseline scenario: Defined by the corresponding methodology, baseline scenarios need to be calculated. It is critical to calculate the baseline as accurately as possible to avoid over- or under-crediting. The guiding question is: What would have happened to the land without the project activity?
• Leakage: Assessed to determine the risk of unintended carbon emissions in other areas as a result of the project.
• Permanence: Derived from registry crediting period, scientific publications, and proprietary research, permanence indicates the expected longevity of carbon storage.
• Additionality: The carbon project exists because of the carbon credit revenues and would not have occurred in a business-as-usual scenario. A risk of additionality for REDD+ projects may come from existing regulation implemented in the project area, such as through governmental conservation programs. 
• Co-benefits: Co-benefits indicate environmental, social, and cultural benefits beyond climate effects. Additional co-benefit certification (e.g. through VERRA-CCB, SD Vista) ensures that the additional benefits have been validated by an external body. 
• NDC relevance: This assesses whether the project is affected by the host country’s NDCs under the Paris Agreement and, thus, if the credits may be claimed to meet national climate targets. 

CEEZER works with over 2 million data points to assess the individual performance of projects and enable data-driven comparability across and within project types, regions, and specific quality indicators. On top of that, CEEZER offers access to independent third-party project quality ratings from leading providers such as BeZero.

REDD+ remains a crucial tool

There is clearly not one single silver bullet for combating global deforestation, but the VCM plays an important role in directing investments toward forest conservation, biodiversity, and local communities. REDD+ projects are among the most prominent types of projects in the VCM, and they are likely to continue to play a significant role in the future. However, to produce effective and impactful carbon credits from REDD+ projects, it is necessary to develop robust methodologies that accurately reflect the complexity of the ecosystem as well as create MRV systems that integrate the breadth of knowledge and technology available to verify calculated emission reductions. 

CEEZER provides buyers with the tools necessary to differentiate between different projects and to find the best match for their individual compensation strategies. Get in touch or book a demo today to learn more about how CEEZER can help you navigate the VCM and identify high-quality REDD+ projects as part of your portfolio.