Site icon Conservation news

REDD+ projects falling far short of claimed carbon cuts, study finds

Rainforest in Ecuador

Rainforest in Ecuador.

  • New research reveals that forest carbon credits are not offsetting the vast majority of emissions that providers claim.
  • A team of scientists looked at 26 REDD+ deforestation-prevention project sites on three continents, leading to questions about how the developers calculate the impact of their projects.
  • The researchers found that about 94% of the credits from these projects don’t represent real reductions in carbon emissions.
  • Verra, the world’s largest carbon credit certifier, said the methods the team used to arrive at that conclusion were flawed, but also added it was in the process of overhauling its own REDD+ standards.

A study has found that the vast majority of the carbon credits from a set of forest conservation projects on three continents aren’t “offsetting” the emissions of the companies and individuals who purchase them.

The researchers found that about 94% of the credits from the projects they looked at don’t represent real reductions in carbon emissions.

“They’re not reducing as much deforestation as they claim to be,” study lead author Thales West, an assistant professor of environmental geography at Vrije Universiteit Amsterdam in the Netherlands, told Mongabay. Some projects don’t appear to be diminishing forest loss at all, he added.

The study was published Aug. 24 in the journal Science. An earlier version was a basis for reporting by The Guardian and several other news outlets beginning in January 2023 concluding that many carbon offsets were “worthless.”

Offsets that don’t represent reductions in deforestation and emissions threaten the flow of funding for forest conservation, Julia P.G. Jones and Simon Lewis, two scientists based in the U.K. who were not involved in the study, wrote in a related commentary for Science, also published Aug. 24. Once-interested companies may shy away from investing in carbon credits if they think they might be called out for trying to “greenwash” their image, Jones and Lewis added.

West and his colleagues analyzed the performance of 26 REDD+ project sites in Southeast Asia, Africa and South America, focusing on 18 that had enough publicly available information on the deforestation baselines used for their calculations.

Illegally harvested logs confiscated by authorities at Keo Seima Wildlife Sanctuary in Cambodia.
Illegally harvested logs confiscated by authorities at Keo Seima Wildlife Sanctuary in Cambodia. Part of the sanctuary overlaps with a REDD+ project that was part of the study by Thales West and his colleagues. Image by John Cannon/Mongabay.

REDD+ is short for reducing emissions from deforestation and forest degradation in developing countries. The idea is that monetizing and protecting threatened forests that would otherwise be logged or degraded, by issuing carbon credits, helps to ward off a further increase in the global temperature. In turn, selling those credits provides an income stream that can be invested back into forest conservation, which proponents say is vital to protecting not only the carbon that forests hold, but other ecosystem services, biodiversity, and vital resources for many forest-dependent societies.

Jones and Lewis pointed out that deforestation in the tropics accounts for the lion’s share of the 5 billion metric tons of carbon dioxide released as a result of land-use change every year.

Ensuring that each credit sold actually represents the equivalent of a metric ton of carbon dioxide that would have been released into the atmosphere were it not for the project — what’s known as “additionality” — is vital to ensuring their integrity. Satellite monitoring and other tools make it “relatively simple” to calculate deforestation rates and account for the carbon forests hold, West said. But working out how much additional carbon is either retained by the forest or removed as its trees grow — the sources of carbon credits — is trickier.

The process involves projecting a baseline, typically based on extrapolating the deforestation data from the previous decade for a given area, which is then compared to the actual deforestation observed. Carbon credits are then issued based on the difference.

The projects that West and his colleagues looked at used four different methods for calculating how much additional carbon stemmed from the project’s presence.

“They create baselines in different ways,” West said. And different algorithms could lead to very different conclusions about the amount of deforestation that would have happened without the project. That can lead to projecting higher hypothetical deforestation rates than would have happened. Such flexibility is a problem, he added, because it’s in the interest of developers to claim their project had the largest possible impact, and in turn sell a greater number of credits.

Tropical timber awaiting export in New Ireland province, Papua New Guinea.
Tropical timber awaiting export in New Ireland province, Papua New Guinea. Proponents of REDD+ say such projects and financing through the sale of carbon credits provide an alternative to logging for landowners and governments. Image by John Cannon/Mongabay.

To evaluate these projects, the team compared measured deforestation within the bounds of the project to a set of control areas, instead of the baselines that the developers had originally used. These “synthetic” controls comprised a combination of actual places selected because they were similar in “the most important characteristics,” West said, such as size and a threat of deforestation that are similar to the project area.

The researchers found that the REDD+ projects tend to overestimate the amount of deforestation that would have occurred without their presence. By the team’s calculations, the project developers’ estimates would have allowed them to claim some 89 million credits.

The methods in practice and that flexibility are allowed under the standard-setting Verified Carbon Standard (VCS) program by Verra, the world’s largest carbon credit certifier.

But because the credits were based on what West and his colleagues say was overestimated hypothetical deforestation, only 6.1% of credits in their analysis could be tied to additional reductions in carbon emissions. About a third would come from projects that had led to some deforestation reductions, though not as much claimed, and 68% — more than 60 million credits — came from projects that didn’t significantly reduce deforestation at all, according to the team’s analysis.

West, a former REDD+ project developer and auditor himself, said that after a project has been in place for a while, developers should use methods “like the ones that we used” to provide a clearer picture of how a project’s interventions have affected the forest. Currently, that’s not required by Verra, he added.

In response to The Guardian’s reporting in January 2023, Washington, D.C.-based Verra raised questions about the methods that West and his team used in the study, which at that time hadn’t been peer-reviewed or published. In a technical review, the organization said the study used only a handful of characteristics to create the synthetic controls and that it used too few geographic areas that also didn’t face “any serious threat of deforestation.”

“Our initial analysis of this version [published in Science] indicates that, despite some minor changes, the overall methodology, results and conclusions are the same — and, therefore, the significant concerns we flagged earlier this year still hold,” Verra told Mongabay in an emailed statement.

A river cuts through lowland rainforest in western Papua New Guinea.
A river cuts through lowland rainforest in western Papua New Guinea. Image by John Cannon/Mongabay.

Robert Nasi, the chief operating officer of CIFOR-ICRAF, a global research and development organization, noted the difficulty of forecasting what might have happened in these scenarios.

“Perfection doesn’t exist, especially when it comes of this sort of prediction,” added Nasi, who was not involved in the West-led study or Verra’s technical review.

West agreed: “How can you use simulation models to create better baselines? I don’t have an answer for that,” he said. “It’s not something that can be easily done.”

Still, the synthetic control method used by West and his colleagues revealed “with some robustness” that the methods currently in place for calculating baselines are “even less correct,” Nasi told Mongabay.

Along with the organization’s initial response to The Guardian’s reporting and West’s work, Verra has also taken public steps to change its REDD methodology as part of an ongoing process that the organization says has been in the works since 2020. The group plans to release the update this year.

“While REDD+ projects have indeed achieved enormous impact to date, we recognize the areas for improvement in the current system and are committed to fostering that ongoing evolution,” the organization said in the statement to Mongabay.

Nasi said offsets would be needed “one way or the other” to address climate change, for example, to balance out the emissions of hard-to-decarbonize industries like shipping and aviation.

“Does it mean that the idea of carbon offset is bad? No,” he added. “It’s the way people implement it.”

The focus of future research should be on finding a solution that makes offset calculations more accurate, Nasi said.

Banner image of a rainforest in Ecuador by Rhett A. Butler/Mongabay.

John Cannon is a staff features writer with Mongabay. Find him on Twitter: @johnccannon

Forest carbon offsets are a tool, not a silver bullet (commentary)

Correction: Robert Nasi is the chief operating officer of CIFOR-ICRAF. He is no longer the interim CEO.


Jones, J. P. G., & Lewis, S. L. (2023). Forest carbon offsets are failing. Science, 381(6660), 830–831. doi:10.1126/science.adj6951

West, T. A. P., Wunder, S., Sills, E. O., Börner, J., Rifai, S. W., Neidermeier, A. N., Frey, G. P., & Kontoleon, A. (2023). Action needed to make carbon offsets from forest conservation work for climate change mitigation. Science, 381(6660), 873–877. doi:10.1126/science.ade3535

FEEDBACK: Use this form to send a message to the author of this post. If you want to post a public comment, you can do that at the bottom of the page.

Exit mobile version