Why Biofuels Policy Must Focus on Real-World Land Use Risk

By Veronica Bradley, Director of Environmental Science

The recent collapse of Brazil's Amazon Soy Moratorium (ASM) demonstrates a critical weakness in how U.S. biofuels policy addresses induced land use change (ILUC). As major traders prepare to abandon a pact that researchers estimate saved rainforest area equivalent to the size of Ireland, the episode underscores why feedstock-level ILUC penalties and exclusions fail to capture the dynamic political and economic forces that determine actual deforestation outcomes.

The Fragility of Supply Chain Commitments

Brazil's soy moratorium, established in 2006, represented one of the world's most successful voluntary deforestation controls. It prohibited signatories from sourcing soybeans grown on land deforested after July 2008 in Brazil's Amazon biome. For nearly two decades, this agreement functioned as a practical bulwark against soy-driven forest clearing, with companies sharing monitoring costs and enforcement mechanisms.

Yet this landmark agreement now teeters on collapse, not because of environmental failure, but due to a 2023 Mato Grosso state law that removes tax incentives worth billions of reais from companies that participate in conservation programs deemed more restrictive than federal Brazilian law. The incentives at stake are substantial: between 2019 and 2024, grain traders received approximately 4.7 billion reais ($840 million) in tax benefits.

Mato Grosso's law took effect January 1, and we have already seen Brazilian Association of Vegetable Oil Industries (ABIOVE)-member companies choosing to preserve their tax credit eligibility over their zero-deforestation commitments. This decision effectively ends the pact despite its documented success.

Why Current ILUC Approaches Fall Short

Current biofuels policy in many jurisdictions attempts to address ILUC through carbon intensity scoring systems that assign fixed values to different feedstocks based on modeling estimates. These approaches attempt to reduce land conversion risk by reducing demand for crop-based feedstocks via higher carbon intensity scores; however, they typically calculate a single ILUC penalty for commodities like soy, regardless of sourcing region or changing market conditions. Similarly, other regimes, like the European Union's Renewable Energy Directive (EU RED), eliminate entire feedstocks like palm oil or, most recently, soybean oil because high-value ecosystems (e.g., tropical rainforests in South America, peatlands in Southeast Asia) have been converted to produce these commodities regardless of the drivers of deforestation.

The European Union's recent proposal to phase out soy-based biofuels from its renewable energy targets exemplifies this blunt approach. While the EU cites high ILUC risk and deforestation concerns, this blanket exclusion fails to distinguish between sourcing regions with vastly different environmental profiles and economic drivers behind land use change.

But the dynamics within sourcing regions are incredibly important to determining land use change risk. A 2022 study in Nature Communications found that domestic leakage undercut 43-50% of avoided deforestation from zero-deforestation supply chain policies in Brazil's soy sector. At the same time cross-border leakage remained below 3% due to market segmentation between Brazilian and competing exporters. Different jurisdictions have variable enforcement capacity, institutional credibility, and sustained implementation over time. This paper shows that where these factors are depleted, intraregional deforestation can remain rampant. Where these factors are robust, farmers and their supply chain partners find other means to satisfy demand.

Regions like the United States, on the other hand, are achieving increased soy production on less land. U.S. soybean yields have increased significantly through improved agricultural practices and technology, allowing farmers to meet growing demand without expanding cultivation area. In fact, a 45% increase in soybean yields since the 1990s generated an additional 15 billion pounds of soybean oil in 2025. That growth exceeded total U.S. biomass-based diesel demand in 2024, without requiring a single acre of new cropland. Yet under the EU's approach, this efficient, lower LUC risk production is treated identically to soy sourced from regions with active deforestation pressures.

The ASM demonstrated that transparent engagement through targeted protocols creates stronger incentives for environmental protection. Demand signals that clearly communicate expectations and consequences influence outcomes inside their value chain. By excluding all soy rather than incentivizing responsible sourcing, on the other hand, the EU foregoes opportunities to leverage its market influence toward better environmental outcomes. The former scenario tells us that sustained engagement decreases deforestation risk, while the latter leaves that less scrupulous buyers to continue purchasing from recently deforested areas without resolve.

Climate change, however, requires resolutions. We need to decarbonize now when it matters most. Waiting even just 5 years to act will require 13 times more emission reductions to have the same climate impact.

The Case for A More Nuanced Risk-Based Approach

Instead of assigning fixed ILUC values to feedstocks regardless of origin or excluding entire feedstock classes, biofuels policy should adopt frameworks that:

• Assess institutional robustness on the ground

  Policymakers should evaluate the actual strength and stability of deforestation controls in sourcing regions. Do protections rely on voluntary corporate commitments vulnerable to economic pressure, or are they embedded in enforceable law with demonstrated political support? The recent breakdown of Brazil's ASM reveals how even long-standing, successful voluntary measures can collapse when subnational governments create financial disincentives.

  The Nature Communications research reinforces this point: the study found that leakage rates and net deforestation outcomes depended heavily on market share thresholds and the extent of trader participation. Success hinges on whether companies actually transform entire supply chains rather than segregate compliant and non-compliant streams.

• Capture dynamic changes in governance regimes
  

  
  A robust system would monitor for warning signs of weakening protections. Legislative proposals threatening environmental pacts, antitrust investigations of conservation agreements, shifting political coalitions favoring extractive industries, or declining enforcement capacity all increase risk. The deterioration of the ASM was visible months before the actual collapse; yet, static ILUC models would continue treating U.S. and Brazilian soy identically until well after the damage in the Brazilian Amazon biome occurred.

  As Rhett Ayers Butler, writing in Mongabay, put it, at the end of the day,

• Recognize that high carbon stock forests face different risks

  Not all agricultural expansion carries equal deforestation risk. Biofuels sourcing from regions with extensive remaining forest cover and weak protection mechanisms poses fundamentally different concerns than sourcing from long-converted agricultural landscapes. The critical variable is not the intrinsic carbon intensity of soybean cultivation. The question is whether soy demand will drive conversion of high-carbon ecosystems both within and across jurisdictions. As the Nature Communications research suggests, risks within a weakly protected region are much different than risks across borders.

  Blanket penalties and exclusions for crop-based feedstocks miss the opportunity to drive better outcomes on the ground. Preventing sourcing from recently deforested land actively reduces deforestation risk within supply chains. Applying a risk-based approach encourages producers and local stakeholders to strengthen their institutions and improve governance in pursuit of market access.

Practical Implications for Policy Design

What would a risk-based, qualitative approach mean in practice?

Feedstock certification based on landscape-level protections: Rather than assigning carbon intensity scores by crop type, certification should evaluate whether sourcing regions maintain effective barriers against conversion of high-value ecosystems. This shifts emphasis from "what is grown" to "where it's grown and under what governance structures."

Dynamic reassessment requirements: Certification status should require regular verification that protective frameworks remain intact and effective. The Mato Grosso situation demonstrates how a compliant sourcing region can become high-risk within a single legislative session. Policies should include regular reassessment and allow for rapid decertification when protective regimes weaken.

Integration with broader governance indicators: Biofuels sourcing decisions should consider countries' broader environmental governance trajectory. Are protections strengthening or weakening? Is enforcement capacity growing or declining? The distance between policy commitments and conditions on the ground can be substantial, so it is important to weigh implementation capacity over formal commitments.

Support for jurisdictional approaches: Rather than relying solely on individual project-level verification, support should flow to subnational jurisdictions demonstrating comprehensive land-use planning and enforcement. The soy moratorium functioned through shared monitoring costs and aligned trader incentives, which work at scale rather than farm-by-farm. Policies that engender this system create market-backed accountability that can scale.

Moving Forward

The evidence points toward a fundamental reorientation of how biofuels policy addresses land use change risk. There is nothing in current policy that will prevent the end of Brazil's soy moratorium from impacting deforestation in the Amazon. There is also nothing about our policies that motivate the traders from reneging on their voluntary commitments outside the U.S. Rather than attempting to model ILUC through complex yet ultimately static economic simulations, policy should:

1. Prioritize risk-based certification that can respond quickly to deteriorating protections
2. Support jurisdictional approaches that align economic incentives with conservation at regional scales
3. Recognize that institutional stability matters more than snapshot assessments of carbon intensity

The lesson for biofuels policy is clear: focusing on penalties at the feedstock level while ignoring the political and economic forces that determine actual land use outcomes at the subregional level leads to systematic and asymmetric application of deforestation risk. A U.S. soybean is treated the same as an Amazonian one. As climate goals become more urgent, the role clean fuels play in reducing emissions becomes more important, but low-risk clean fuels supply chains are currently treated the same as high-risk ones. Current ILUC penalties effectively de-risk high-risk feedstocks be assigning blame to the entire feedstock class. At the same time, blunt instruments like the EU's Renewable Energy Directive prevent low-risk clean fuels from helping to further decarbonize the economic zone. Policy frameworks must evolve to capture these dynamics. We cannot wait to decarbonize or end deforestation, and we can use our policy levers now to do both.