Conference Agenda

Tropical forests are crucial to the global environment, serving as biodiversity hubs and storing carbon, yet they face severe threats from agricultural expansion. Recognizing their role in deforestation, countries like the UK and EU have implemented trade policies to curb deforestation by requiring proof that the production of a set of agricultural commodities is deforestation-free. A key question is whether these policies will encourage farmers to adopt sustainable practices or instead push them toward less regulated buyers.

In this paper, we adopt a structural approach to evaluate the effectiveness and potential leakages of sustainable supply-chain policies in the context of Brazil’s cattle sector. First, we leverage a unique dataset on animal transport, matched with property boundaries, to document key empirical patterns of deforestation and the supply network of cattle at the farm level in the state of Pará, Brazil. Second, we build a structural model of farm-level land use decisions with an endogenous supply-chain network. Third, guided by model-derived gravity equations, we estimate the role of existing domestic supply-chain zero-deforestation commitments (ZDC) in shaping the trade of cattle. Our results indicate that a 10\% increase in deforestation by a farm is associated with a 14\% decrease in its probability of being chosen as the supplier by a ZDC slaughterhouse. Finally, we use the model to conduct a counterfactual exercise that removes zero-deforestation penalties while holding land use fixed, in order to isolate the role of supply-chain restrictions in shaping market access and equilibrium outcomes.

Agricultural expansion accounts for nearly 90% of global deforestation, motivating the adoption of Payment for Ecosystem Services (PES) programs. Existing studies establish moderate effectiveness of PES in reducing average deforestation, but evidence remains limited on how institutional design—contractual ownership and payment scheme interactions—shapes conservation performance. A PES program in Ecuador provides a unique setting by combining individual and collective contracts with a multi-threshold payment scheme and staggered adoption. Our analysis provides the first causal evidence on the effect of both contracts and pricing rules operating in a real-world PES. We construct a 24-year panel covering the universe of PES contract sites (approximately 64,800 unit-year observations) along with administrative records, to estimate treatment effects and heterogeneity of each contract type on deforestation, using Matrix-Completion with Nuclear Norm Minimization. We benchmark these estimates against staggered difference-in-differences estimators. We find that collective contracts reduce deforestation by 1.7–2.5 percentage points more than individual contracts, equivalent to roughly 129–196 fewer hectares deforested per average collective contract. This advantage persists in areas with high land-use opportunity costs, suggesting that ownership structure matters beyond baseline land characteristics. Heterogeneity analysis points to payment design as a central mechanism: low-remunerated tiers weaken conservation incentives, while a premium threshold induces strategic bunching and adverse selection by attracting higher-deforestation-risk individual contracts. Prioritizing collective contracts, smoothing nonlinear payment schemes and redesigning equity-motivated premium payments can improve the impact and cost-effectiveness of large-scale PES programs.

Conventional wisdom states that incentives targeting stocks and flows can lead to equivalent externalities correction. However, this may not hold when externalities are directly produced from stocks of capital ("stock-generated externalities") rather than flows of production. Privately-owned stocks that exhibit external effects include trees with respect to CO2 absorption, shellfish and water cleaning, and livestock and methane emissions. I present an optimal control model of externality-correcting incentives to compare policies targeting measures of stocks and flows. The setting focuses on an extractive firm which privately owns a renewable stock where joint production of a private good and social benefits occurs. I compare the theoretical performance of stock and flow policies in steady state to the social optimum and a no-policy benchmark. The model is then calibrated to industry and scientific parameters of Eastern oyster (C. virginica) production in the Chesapeake Bay, where shellfish cultivation yields water filtration benefits, to explore short run dynamics. Overall, stock-targeted incentives can achieve the socially optimal allocations of resource abundance and environmental quality in steady state. While flow-targeting may yield short run benefits, the allocations become no different than no-policy in the long run. The lack of long run efficacy from flow-targeting is likely driven by revenue constraints that disallow the firm to increase stock abundance to a substantial degree across time. This paper presents a theory of incentive design aimed at minimizing environmental damages through management of privately-owned renewable natural resources.

Shifting to healthier diets and afforestation of unforested land can reduce greenhouse gas (GHG) emissions and enhance carbon sequestration. However, when implemented unilaterally by China, the world’s largest agricultural producer and importer, and the largest annual emitter of GHGs, these measures may support domestic carbon neutrality while triggering emission leakages through trade-induced land-use changes. Using an integrated environmental-economic model, we found that the proposed dietary shift towards less meat and more dairy products in China may potentially lead to an additional 14 Mha of agricultural land use abroad and cause 364 Tg CO₂-eq of emission leakage, equivalent to almost four times the domestic GHG emission mitigated by the dietary shift. Similarly, China’s recent afforestation policy may indirectly lead to expansion of food production overseas, resulting in 16 Mha of additional agricultural land use and 424 Tg CO₂-eq of emission leakage, which is more than the domestic GHG emission mitigated by the afforestation policy. Such leakages, primarily driven by land-use change, may undermine the feasibility of the 2 °C climate target and necessitate more stringent emission mitigation in non-agricultural sectors. When combined with a stringent global climate mitigation policy consistent with the 2 °C climate target, China’s dietary shift and afforestation policies may exacerbate the global carbon tax (36%), food price (65%), the population at risk of hunger (145 million people), and gross domestic product losses ($74 billion) than the 2 °C climate target alone.