Principal investigator: Eduarda Miller de Figueiredo
Location of the Intervention: Amazon
Sample Size:
Sector: Environment
Primary Variable of Interest: deforestation
Type of Intervention: Resolution 3.545 – Rural Credit
Methodology: Differences in Differences
Resolution 3.545 of 2008 made the granting of subsidized credit in the Amazon contingent upon proof of compliance with legal requirements regarding land titles and environmental standards. In light of this, the authors seek to evaluate the resolution's effective impact on deforestation within the Amazon biome. Using panel data composed of information from INPE (National Institute for Space Research) and the Central Bank, the Difference-in-Differences methodology was employed to produce estimates. The results are significant and suggest a 60% reduction in deforestation. Heterogeneous effects were found for different activities and regions, demonstrating that the resolution is a complementary, not a substitutive, policy.
Policy Problem
Of the latest global greenhouse gas emissions, almost 20% is attributed to tropical deforestation. Because of this, policymakers worldwide have been dedicated to implementing a set of incentive-based policies to help reduce deforestation.
This article aims to assess the impact of Resolution 3.545 on deforestation, given that this resolution represented a potential restriction on access to rural credit, which is one of the main mechanisms for supporting agricultural production.
Implementation and Evaluation Context
In 2008, the Central Bank of Brazil published Resolution 3.545, which made the granting of subsidized credit in the Amazon contingent upon proof of compliance with legal requirements for land title and environmental regulations. Furthermore, this resolution applies exclusively to properties within the administrative definition of the Amazon biome, meaning that properties outside the biome are not subject to this policy.
Resolution 3.545 determined that eligibility for access to rural credit must be conditional upon the recipient of the credit presenting: (i) the Rural Establishment Registration Certificate, which proves that the property hosting the project to be financed is duly registered in federal records; and (ii) a document issued by the State attesting to the regularity of the establishment hosting the project to be financed, as well as a declaration that the property is not under any encumbrance.[1] This stemmed from deforestation. Producers who feared the resolution could affect their future access to credit might signal an intention to change their deforestation behavior in the future, thus being considered compliant with current environmental regulations.
It is important to point out that soon after the mandatory adoption of the resolution, new measures loosened the requirements for granting rural credit to small producers, mainly through the inclusion of new groups of small producers in the list of credit recipients who were exempt from complying with the original requirements of Resolution 3.545.
Policy/Program Details
For the analysis, a panel dataset from 2003 to 2011, broken down by municipality each year, was used. A georeferenced map containing the location of the municipalities and the boundaries of the Amazon biome was used, which allowed the creation of subsamples of municipalities, both inside and outside the Amazon biome, located at specific distances from the biome's boundary, as shown in the figure below:
Figure 1: Brazilian Amazon Biome and Sample of Benchmark
The deforestation data were constructed from satellite images from the National Institute for Space Research (INPE) within the scope of the Legal Amazon Deforestation Monitoring Project (PRODES). The authors define deforestation as the area of forest in square kilometers cleared in the twelve months preceding August of a given year. For this reason, credit loans and all other variables were recoded for this period, in which the year t will refer to the twelve months preceding August of t.
Data from administrative contract levels compiled by the Central Bank were also used to construct rural credit variables at the municipal level from data in the Common Registry of Rural Operations. This set of administrative microdata encompasses all records of rural contracts negotiated by official banks and credit cooperatives.
Method
The fact that Resolution 3.545 only applies to properties located within the Amazon biome has created an explicit geographical division between two groups of municipalities: (i) those entirely within the Amazon biome and therefore subject to the resolution; and (ii) those entirely outside the Amazon biome, which are not subject to the resolution. Therefore, the treatment group will consist of the municipalities in the first group, while the control group will consist of the municipalities in the second group.
Thus, the authors use Differences-in-Differences to assess the impact of Resolution 3.545 on deforestation. The variable of interest in the study is the interaction of a binary variable indicating whether the municipality is located within the Amazon biome with a variable marking the period after the implementation of the resolution. Fixed effects of municipality and year were added to the model to control for persistent characteristics of the municipalities and common temporal trends, respectively. Furthermore, variables were added to control for other environmental policies.
Main results
The main results demonstrated that, in the absence of Resolution 3.545, total deforestation would have been 2.000 square kilometers greater than that actually observed from 2009 to 2011 in the 100 km sample of municipalities, representing a 60% reduction compared to the 2008 baseline deforestation. In other words, the resolution has played an important role in containing deforestation in the Amazon biome.
Furthermore, the results suggest that the inclusion of environmental monitoring and law enforcement controls does not affect the impacts of the resolution; that is, the resolution did not affect credit granting via a reduction in demand from credit recipients. Therefore, in general, the results indicate that the effects of Resolution 3.545 on deforestation directly reflect the reduction in deforestation as a response to reduced access to rural credit.
However, the authors emphasize that the policy may have had different effects in different regions. Thus, regional heterogeneity is explored, observing how the relationship between credit and deforestation may differ between municipalities with different activities. In livestock-raising municipalities, the point estimate is quite similar to the other specifications. In contrast, the estimates suggest that Resolution 3.545 had no impact on deforestation where agriculture is the main activity. According to the authors, this is consistent with reports documenting that agricultural production in Brazil has become less dependent on credit and, after various technological improvements, has increased its production on an intensive margin. Regarding credit lending, the results indicate that credit concessions for use in livestock farming were the most affected by Resolution 3.545 compared to other activities, suggesting that access to credit and deforestation are particularly correlated in livestock farming.
Therefore, the authors conclude that Resolution 3.545 helped to contain deforestation in the Amazon biome. However, the resolution had negative effects on both deforestation and credit availability, and the actual implementation of the new policy supports the assumption that the resolution only affected deforestation through the credit channel.
Lessons in Public Policy
The study suggests that conditioning rural credit is an effective policy instrument in combating illegal deforestation. However, the heterogeneous effects across sectors and regions suggest that it is a complementary, rather than a substitute, policy for other conservation efforts.
[1] In the Amazon, embargoes are an administrative sanction that can be applied to landowners as punishment for illegal deforestation on private property.
