With deforestation accounting for around 12% of global greenhouse emissions, protecting forests is fundamental to success in achieving the 1.5°C objective of the Paris Agreement. Yet deforestation continues to increase and events such as the 12-year record high of deforestation in the Amazon in 2020 remind us of that. Soya production is among the main drivers of this deforestation – together with cattle breeding, palm oil growth and paper industry. An increasing number of actors, from governments to private companies, including financial institutions, have made zero-net deforestation commitments, use carbon offsetting or rely on certification schemes for agricultural products. These efforts may sound positive, however in practice their effectiveness to prevent deforestation is very limited and may even have worse impacts that the problems they are intended to solve.

Zeronet deforestation vs zerodeforestation: one small word, huge implications

Zerodeforestation means that forest areas cannot be cleared or converted, whereas zeronet deforestation accepts the clearance or conversion of forests in one area if an equal area is replanted elsewhere. While the main criticism of the zero-deforestation approach is its lack of flexibility, zeronet deforestation faces multiple criticisms relating mainly to its implementation and impacts. Some even claim that practices related to zero net deforestation commitments are fraudulent. Companies commit to protect forests that were not originally threatened or fail to provide enough guarantees that they can protect those that are threated.

However, the underlying problem with zero-net deforestation comes from the lack of definition or even the impossibility of agreeing on a common language and understanding of the stakes: how to define forest, forest loss, deforestation and even trees. As zeronet deforestation allows for new forest – natural or planted – to theoretically compensate for lost forest, it necessitates defining and measuring what would be an equivalent area of a forest to offset.

It must be considered that native forest and natural ecosystems store large amounts of “irrecoverable carbon”, carbon which, once released by forest destruction, cannot be restored rapidly enough to prevent the worst impact of climate change. Moreover, these ecosystems play a fundamental role as habitats for numerous species, provide elementary ecosystems services like water and weather regulation and have essential cultural value. All these functions cannot be substituted by plantations.

As explained by researchers Brown and Zarin “net deforestation targets, inherently and erroneously, equate the value of protecting native forest with that of planting new ones.” This does not mean that reforestation and forest restoration are not positive and should not be encouraged when relevant, but rather that they should not be done to offset the destruction of natural forests.

Certifications are not an easy way out of deforestation either

Certifications of agricultural products have gained popularity as a means of validating and homogenizing efforts towards better practices in terms of environment and social justice. These efforts are appreciated, but they do not constitute a panacea since they face numerous drawbacks.

For soya, two main certification systems exist: RTRS and Proterra. The RTRS is considered to have contributed to make the practices of some soya producers more responsible. However, RTRS soya remains a very small minority of global production since all certified soya, through different mechanisms, account for only 1.5-3% of total production. Moreover, while the eligibility criteria of these certifications may be considered ambitious, they still have major weaknesses: RTRS and Proterra allow for largescale monocultures that use huge quantities of pesticides and water resources, while creating limited employment opportunities. RTRS allows also GMOs.

The RTRS have three modalities for certification, of which only one guarantees the absence of soya linked to deforestation. These three modalities are: segregation, mass balance and country material balance. Only segregation allows for fully tracing and segregating the flux to ensure that 100% of the soya comes from non-conversion parcels. The other two modalities allow instead for a mixture of certified and non-certified soya, contribute to market segmentation and fail to fix the problem of conversion / deforestation: they are simply inefficient. However, segregation still represents a very low volume of certified soya, as a result of being more expensive than using a mass balance system.


Stopping deforestation is not as simple as committing to zero-net deforestation targets or to soy certification schemes. Financial actors must be aware of the weaknesses of these systems and act accordingly if they want to be taken seriously on their commitments around biodiversity, ecosystems, and climate change. Regarding soya production, banks, insurers and investors should require soy traders such as Cargill to revise their contracts with producers to include a commitment to no longer purchase soya from converted land after 1 January 2020.

Our demands

Sign the petition for BNP to stop bankrolling deforestation