The Making of the Systematic Assessment of the Achieved Emission Reductions of Carbon Crediting Projects

In April 2021, as the voluntary carbon market was experiencing unprecedented growth, a small group of researchers from the University of Cambridge and ETH Zurich began asking a crucial question: what does the academic evidence actually tell us about carbon credit quality?

The timing wasn't coincidental. Companies were making ambitious net-zero pledges, and carbon credit prices were surging. Yet, while individual studies had identified quality issues in specific project types, no one had systematically synthesized the evidence across the entire field. Professor Laura Díaz Anadón and Prof. Andreas Kontoleon from Cambridge's Department of Land Economy, along with Dr. Benedict Probst, Professor Volker Hoffmann and Dr. Malte Toetzke from ETH Zurich's Group for Sustainability and Technology, saw an opportunity to fill this critical knowledge gap.

What started as a focused collaboration soon expanded. As we dove deeper into the literature, we realized we needed additional expertise. We reached out to leading researchers who had conducted groundbreaking studies on specific project types - from cookstoves to forestry projects. Dr. Barbara Haya and Annelise Gill-Wiehl from UC Berkeley brought crucial insights on cookstove and improved forest management projects, Dr. Lambert Schneider from Öko-Institut contributed deep expertise on industrial gas projects, Prof. Philipp Trotter on renewable energy projects, and Prof. Thales West on forestry projects. Prof. Jan Minx supported the methodological implementation of the systematic assessment.

The proliferation of Machine Learning over the last years transformed our research process. What initially seemed like an overwhelming task - screening over 64,000 potentially relevant papers - became manageable with AI-powered tools. The ASReview platform, which uses machine learning to prioritize relevant papers, helped us identify key studies we might have missed through traditional methods.

But finding the studies was just the beginning. We needed a framework that could integrate diverse types of evidence, from randomized controlled trials to rigorous observational studies.

Through numerous iterations and feedback from our expanding team, we developed a quantitative framework centered on the "offset achievement ratio" - a measure comparing actual emission reductions to issued credits.

The results were sobering. Less than 16% of analyzed credits represented real emission reductions. This finding emerged consistently across different analytical approaches and robustness checks. While we had expected to find quality issues, the magnitude surprised even our experienced team.

Throughout the process, we were guided by a commitment to academic rigor and practical relevance. We wanted our findings to be not just publishable in a top journal, but actually useful for improving carbon markets. This dual objective shaped everything from our methodology to our presentation of results.

The paper's publication in Nature Communications marks a beginning rather than an end. With Carboncredits.fyi, we're transforming our systematic review framework into a living platform. Using the same AI tools that aided our initial research, we're building a system to continuously incorporate new evidence as it emerges.

Looking back at those first meetings in 2021, we couldn't have predicted how our project would evolve. What began as an academic collaboration between Cambridge and ETH Zurich grew into a global effort involving multiple institutions. As carbon markets continue to expand, we hope our work provides both a wake-up call about current quality issues and a framework for ongoing assessment.

The challenges in carbon markets are significant, but so is the opportunity for improvement. With continued collaboration between researchers, better data, and new technological tools, we can work toward carbon credits that truly deliver their promised climate benefits.