I wrote my master's thesis on the Noel Kempff Mercado REDD project in 2011, when I was a Fulbright student at SUNY ESF. Today Bolivia is opening to carbon projects again — and the methodological problems I flagged then could show up in the new projects being designed now. I'm publishing this reflection alongside the open-access release of the thesis, in the hope that the next generation of project designers and regulators won't have to learn the same lessons the slow way.

The baseline problem we flagged in 2011

In 2009, Bolivia's new constitution enshrined the rights of nature and rejected the commodification of ecological functions. The country withdrew from market-based carbon project pipelines. The Noel Kempff Mercado Climate Action Project — which until then had been one of the most-watched REDD pilots in the world — went quiet. International credit markets moved on. Researchers moved on. Most of the institutional memory moved on with them.

For more than a decade, Bolivia ran an experiment that few countries have run: what does forest–climate policy look like without tradable carbon credits? The answers were diverse. The non-market frameworks Bolivia developed never matched the scale of finance that the international market could mobilize, and forest loss kept accelerating in the lowlands more than ever.

Now the door is opening again. Private project developers are arriving, international buyers are pushing for Article 6 agreements, and a new generation of professionals is building project documents. Most of them weren't here in 2011.

That is the gap I want to address.

What Noel Kempff was, briefly

The Noel Kempff Mercado Climate Action Project covered roughly 1.5 million hectares of Chiquitano dry forest in eastern Bolivia. It was launched in the late 1990s as a partnership between FAN Bolivia, The Nature Conservancy, the Bolivian government, and three energy companies — American Electric Power, BP, and PacifiCorp — that funded it as a voluntary climate offset. It was the first project of its kind to attempt to generate verified emission reductions at this scale in a developing country.

For a decade, Noel Kempff was a global laboratory for REDD methodology. Every choice — how to measure forest carbon, how to project what would have happened without the project, how to count leakage, how to share benefits with neighboring communities — became a precedent that later projects would copy or contest.

What my thesis examined

The sensitivity analysis I ran in 2011 asked a simple question: if you change a small assumption inside the baseline methodology, how much does the carbon estimate move?

Three categories of assumptions turned out to matter most:

  1. The reference region. A baseline is a counterfactual — what would have happened in the project area without the project. To estimate that, you pick a “reference region” that resembles the project area but doesn't have the same protection. Which region you pick, and how you justify the resemblance, can shift the entire baseline by a non-trivial margin.
  2. The historical deforestation rate. Forest loss is not linear. Pick a five-year window and you get one rate; pick a fifteen-year window and you get another. The choice is partly technical and partly rhetorical — and it sits at the center of how many credits a project can claim.
  3. Leakage assumptions. Deforestation pressure rarely disappears when you protect a piece of land; often it moves nearby. The leakage adjustment factor is the project's admission that we didn't stop deforestation, we displaced some of it. A smaller leakage assumption means more credits issued.

The point of the analysis was not to “catch” Noel Kempff doing something wrong. It was to show how much the credit number depends on choices that have no single correct answer — choices that benefit from independent scrutiny, regardless of who is making them.

Editorial note: I deliberately keep this piece methodology-focused rather than project-specific. Readers who want the Noel Kempff case-study details can go to the open-access thesis directly. The work of this blog post is to extract general lessons that apply to whatever comes next.

When verified emission reductions are sold as carbon credits, a fragile baseline means fragile climate integrity and fragile revenue. Both problems eventually become political problems.

Five lessons we already learned the hard way