Replication data and code for “Reconfiguring deforestation drivers in Brazil’s Legal Amazon after 2018: Climate, agricultural production, and environmental governance”
Description
This repository contains the processed data and replication code supporting the article “Reconfiguring deforestation drivers in Brazil’s Legal Amazon after 2018: Climate, agricultural production, and environmental governance,” accepted for publication in Land Use Policy. The main analytical panel covers 634 municipalities in Brazil’s Legal Amazon over 2010–2021. The replication package also includes the expanded 808-municipality robustness sample and the auxiliary data required for the supplementary analyses. Data sources include annual clear-cut deforestation from INPE/PRODES, municipal GDP and agricultural gross value added from IBGE (R$ thousands, current prices), annual precipitation and mean temperature from TerraClimate, and flora-related non-cancelled environmental infraction notices from IBAMA. Additional spatial robustness analyses use HydroRIVERS river-network density and IBGE intermediate geographic regions. The package includes Stata 17 code reproducing the main manuscript Tables 2–5 and Supplementary Tables A1–A10, including the structural-break analysis, as well as Python code reproducing the river-network-density moderation and intermediate-region-by-year fixed-effects analyses reported in Tables A11–A12. A README provides the execution order, variable definitions, software requirements, and data provenance. The supplied analytical files reproduce the results reported in the article and Supplementary Material.
Files
Steps to reproduce
1. SOFTWARE. The main econometric analyses were estimated in Stata 17. Required user-written commands are reghdfe, ftools, estout/esttab, ppmlhdfe, and xtbreak. The structural-break analysis used xtbreak version 2.2. Tables A11–A12 and the figure-reproduction scripts use Python; required packages are listed in code/requirements.txt. 2. EXTRACT THE PACKAGE. Unzip LUP_REPLICATION_FINAL_V3.zip and set the extracted LUP_REPLICATION_FINAL_V3 directory as the working directory. The package starts from the processed analytical files used for the article. Original provider data are publicly available from INPE/PRODES, IBGE, TerraClimate, IBAMA, and HydroRIVERS and are not redistributed as raw files in this archive. 3. MAIN MANUSCRIPT TABLES. From Stata, with the package root as the working directory, run: do code/01_main_tables_2_5.do This reproduces manuscript Tables 2–5. Outputs are written to output/main and the execution log to output/logs. 4. SUPPLEMENTARY TABLES. In Stata run: do code/02_supplementary_A1_A10.do followed by: do code/03_table_A7_xtbreak.do These commands reproduce Supplementary Tables A1–A10 and the structural-break analysis reported in Table A7. 5. TABLES A11–A12. From a Python environment in the package root, install the dependencies with: pip install -r code/requirements.txt and run: python code/04_tables_A11_A12.py Google Colab can be used as an alternative Python environment. Outputs are written to output/python. 6. FIGURES. To regenerate the figures supplied by the replication code, run: python code/05_figures.py Outputs are written to output/figures. 7. VALIDATION. Expected numerical results are documented in reference_outputs/EXPECTED_RESULTS.txt, and the completed replication checks are summarized in reference_outputs/VALIDATION.txt. File-integrity checksums are provided in SHA256SUMS.txt.
Institutions
- Universidade de São PauloSão Paulo, São Paulo
Categories
Funders
- Coordenação de Aperfeicoamento de Pessoal de Nível SuperiorFederal District, BrasíliaGrant ID: 88887.201003/2025-00