Régulation et Environnement

Du 10/05/2023 de 12:00 à 13:15

Salle R1-14, Campus Jourdan, 75014 Paris

LORIS André(PSE)
YANGSIYU Lu(PSE)

"Green finance and deforestation reduction in Brazil: a PVAR analysis of the Amazon Fund"

Yangsiyu Abstract: Whether regulation is good or bad for growth is a long-standing question in economics. In this paper, we study the impact of an environmental regulation on industrial firms’ productivity in China, examining not only firms in the highly-polluting industries but also those that are less polluting but still regulated “cleaner” firms. We show that doing so allows us to capture the net impact of environmental regulation on economic activity. Using a heterogeneous difference-in-differences approach, we find that the productivity of "cleaner" firms increases by 4%, while there is no effect on productivity for highly polluting firms. Firm exit, output effect, and process and practice innovations are the primary drivers of these productivity impacts, and the mechanisms differ between private firms and state-owned enterprises. ******* Loris Abstract: The Amazon rainforest is emitting more carbon dioxide than it can absorb due to deforestation since 2021, leading to significant impacts on global warming. The loss of biodiversity due to land use change in the Amazon biome is also a major issue. Legal Amazon is an administrative area in Brazil that encompasses 64% of the Amazon biome and nine federal states. The Amazon Fund is the main international climate finance vehicle that operates in Legal Amazon, but its disbursements have decreased significantly in recent years due to disagreements between donors and the Brazilian government. This paper aims to assess the impact of the Amazon Fund's projects on reducing deforestation and other factors such as national environmental agency sanctions and agricultural production. By using satellite observations and microeconomic data, a panel dataset has been constructed to analyze the evolution of various environmental features, climate finance, regulation, and production from 2002 to 2020 across 760 municipalities in Legal Amazonia. A Panel Vector AutoRegression (PVAR) is used to model a stylized economic system in which variables can affect each other at different lags. The study's main findings suggest that the Amazon Fund's disbursements significantly reduce deforestation rates, with state-managed projects proving more effective than those managed by municipalities or universities. The most efficient projects involve land use planning, which includes the development and the protection of local autochthonous communities. Overall, the study highlights that the Amazon Fund's operates with a very low abatement cost (under $1 per ton of CO2).