jdnmiguel/abbds_ajae_2025

This repository contains data and replication materials for the American Journal of Agricultural Economics article entitled, “Buyers’ Response to Third Party Quality Certification: Theory and Evidence from Ethiopian Wheat Traders” by Gashaw Tadesse Abate, Tanguy Bernard, Erwin Bulte, Jérémy Do Nascimento Miguel, and Elisabeth Sadoulet. # abbds_ajae_2025 This repository contains data and replication materials for the American Journal of Agricultural Economics article entitled, “Buyers’ Response to Third Party Quality Certification: Theory and Evidence from Ethiopian Wheat Traders” by Gashaw Tadesse Abate, Tanguy Bernard, Erwin Bulte, Jérémy Do Nascimento Miguel, and Elisabeth Sadoulet. Contact Jérémy Do Nascimento Miguel at jeremy.dnmiguel@gmail.com if you have any questions. The folder titled *“Analysis”* contains Stata do-file, R Script, and dta files to replicate all the Figures and Tables in the article. First, open the *“Project.stpr”* file to open a do-file editor window with the four do-files. The first step is to **update row 20 of the “0_Path”** do-file before running it to create all the paths. Run the do-files *“1_cleaning”* and *“2_montecarlo”* if you want to **start from the raw datasets** and obtain the cleaned dataset used in the analysis (Nb: computing time preferences take some times to run >25min). You can directly **run the do-file *“3_paper”* to get all the tables and figures** in the manuscript and appendix, except for Figure 8. Each output will be labeled with the number it has in the final version of the paper. Open the *“Machine Learning.R”* R-script to run the machine learning analysis and obtain Figure 8. You should **update the raw 27 with your path**. There are notes at the top of the do-files and R scripts describing what that specific segment of code codes. Within the “Analysis” folder, there are two folders: *“Data” and “Results”*. “Data” contains all the necessary datasets to replicate the result*, organized into two subfolders: “raw”, which includes the anonymized raw datasets, and “clean”, which contains the cleaned datasets used for analysis. The second folder “Results” contains two subfolders, “graphic” and “table,” where figures and tables are stored.