Data for armed conflict accelerates cropland loss and food insecurity worldwide

This repository contains the underlying data and code used for the spatial analyses, difference-in-differences (DID) estimation, and the propensity score matching (PSM) analysis in the study. Global distribution of armed conflict events (Conflict_event_point_density.zip) 1.1 A raster dataset in GeoTIFF (.tif) format representing the global spatial distribution of armed conflict events used in Fig. 1. 1.2 Administrative-level summary of armed conflict and cropland loss A vector dataset at the global fifth-level administrative-unit scale containing aggregated information on armed conflict events and associated cropland loss. The attribute table includes the summarized number of conflict events and cropland loss for each administrative unit. The raster and vector spatial datasets can be opened, visualized and analysed using ArcGIS Pro or other compatible GIS software. DID analysis dataset (did_stata.dta). This Stata dataset contains the core panel data required to reproduce the baseline difference-in-differences analysis reported as Model 4 in Supplementary Table 3. The file can be opened and analysed using Stata. PSM analysis dataset (PSM_data.csv). PSM_data.csv contains the core dataset used for the propensity score matching (PSM) analysis in this study. The file can be opened and analysed using R or other software capable of reading CSV files. The corresponding analysis code is provided in the Supplementary Information, allowing users to reproduce the PSM analysis reported in the study. The code provides the main procedures used to reproduce the DID and PSM analyses. The DID analyses and placebo tests were conducted in Stata using the did_stata.dta dataset, whereas the PSM analysis was conducted in R using the PSM.csv dataset. Required packages should be installed before running the code.