Reproducible Research Repository
Reproducible Research Repository
  • Home
  • Repository
  • Collections
  • About
    Home / Repository / PRWP / RR_CAF_2025_520
PRWP

Reproducibility package for Unveiling Hidden Hardships: Leveraging Alternative Data To Map Multidimensional Vulnerability In The Central African Republic

2026
Get Reproducibility Package
Reference ID
RR_CAF_2025_520
DOI
https://doi.org/10.60572/163h-6v45
Author(s)
Pierre Jean-Claude Mandon, Vincent Nossek, Walker Kosmidou-Bradley, Frederic Mortier, Baptiste Cheville
Collections
World Bank Policy Research Working Papers
Metadata
JSON
Created on
Jan 09, 2026
Last modified
Jan 16, 2026
  • Project Description
  • Downloads
  • Overview
  • Reproducibility Package
  • Description
  • Scope and coverage
  • Disclaimer
  • Access and rights
  • Contacts
  • Information on metadata
  • Citation
  • Overview

    Abstract

    In fragile states such as the Central African Republic, where conflict and institutional fragility severely constrain traditional data collection, mapping multidimensional vulnerability and potential deprivation poses a significant challenge for designing targeted interventions. This paper presents an innovative geospatial dashboard that harnesses alternative data sources—including nighttime light intensity, other relevant satellite imagery, geocoded infrastructure inventories, and critical event records—to develop high-resolution indices (at a 5×5 km scale) of economic capacity, access to essential services (education, health, and water), flood exposure, and lethal conflict risks.
    By employing a Bayesian state-space model to disaggregate sectoral GDP and friction-based accessibility metrics, our analysis uncovers pronounced spatial disparities: economic activity remains concentrated in urban hubs such as Bangui, while rural areas suffer from compounded vulnerabilities, including limited economic opportunities and poor service access. Cross-validation with the 2021 Harmonized Household Living Conditions Survey confirms the predictive validity of these indices for household wealth, with economic and service indicators positively correlated with welfare outcomes. Conversely, exposure to lethal conflict appears paradoxically associated with higher-value targets, potentially reflecting rent-seeking dynamics.
    These tools enhance the precision of policy targeting in data-scarce environments, providing scalable and actionable insights for poverty alleviation in conflict-affected, low-income countries.

    Reproducibility Package

    Scripts
    Readme Get Reproducibility Package
    Link: https://reproducibility.worldbank.org/catalog/438/download/1261/README.pdf
    Reproducibility package for Unveiling Hidden Hardships: Leveraging Alternative Data To Map Multidimensional Vulnerability In The Central African Republic
    File name
    RR_CAF_2025_520
    Zip package
    RR_CAF_2025_520.zip
    Title
    Reproducibility package for Unveiling Hidden Hardships: Leveraging Alternative Data To Map Multidimensional Vulnerability In The Central African Republic
    Date
    2026-01
    Dependencies
    R dependencies are listed in the file renv.lock.
    Instructions
    See README in reproducibility package.
    Notes
    Computational reproducibility verified by Development Impact (DECDI) Analytics team, World Bank.
    Source code repository
    Repository name URI
    Reproducible Research Repository (World Bank) https://reproducibility.worldbank.org
    Software
    R
    Name
    R
    Version
    4.5.2

    Reproducibility

    Technology environment

    Paper exhibits were reproduced on a computer with the following specifications:
    • OS: macOS Sequoia
    • Processor: Apple M4 Pro
    • Memory available: 24 GB

    Technology requirements

    Runtime: ~10 minutes

    Reproduction instructions

    To reproduce the findings in this paper, a new user should follow the steps below:

    1. Obtain access to the restricted datasets and place them in the appropriate folders, as described in the Data Availability section.
    2. Open the R Project located at Scripts/Scripts.Rproj.
    3. Restore the R environment using renv::restore() and follow the prompts. Alternatively, install the required packages manually using the information provided in the renv.lock file.
    4. Open the main script and run the code to execute the full workflow.
      Because some datasets are restricted and therefore not included in the reproducibility package, the results obtained by the replicators are provided in the Outputs folder. These outputs allow users to verify the findings against the results reported in the paper.

    Data

    Datasets
    Grid-Level Indices of Economic Activity, Services Accessibility, and Risk Exposure for the Central African Republic
    Name
    Grid-Level Indices of Economic Activity, Services Accessibility, and Risk Exposure for the Central African Republic
    Note
    Source: Authors’ construction using QGIS; underlying sources as documented in Appendix Table A.1 (EOG, Colorado School of Mines; LandScan/ORNL; MODIS; University of Maryland; Minex; U.S. Geological Survey; IPIS). File location: Inputs/Base_CAF_v1.csv. Access date: 20/11/2024.
    Access policy
    Data is included in the reproducibility package
    Citation
    Authors' construction. (2025). Grid-Level Indices of Economic Activity, Services Accessibility, and Risk Exposure for the Central African Republic [dataset].
    Population Point Exposure to Lethal Conflict for the Central African Republic
    Name
    Population Point Exposure to Lethal Conflict for the Central African Republic
    Note
    Authors’ construction using QGIS based on Armed Conflict Location & Event Data Project (ACLED). Since redistribution of this dataset is not allowed by the licensing terms, a list of variables is included in the README file. File location: Inputs/CAF_Pop_pts_MTI_SCIR.csv. Access date: 15/11/2024.
    Access policy
    Data is publicly available but does not allow redistribution, and it is not included in the reproducibility package.
    License
    ACLED End User License Agreement
    License URL
    https://acleddata.com/eula
    Data URL
    https://acleddata.com/conflict-data
    Citation
    Armed Conflict Location & Event Data Project (ACLED). (2025). Population Point Exposure to Lethal Conflict for the Central African Republic [dataset]. https://acleddata.com/conflict-data
    Fathom Global 3.0 Flood Risk
    Name
    Fathom Global 3.0 Flood Risk
    Note
    Population point–level exposure to flood risk, extracted at each population point. Flood hazard exposure is based on Fathom Global 3.0 flood risk datasets (30 m; 2020–2030). Since the data is limited-access and not included in the reproducibility package, the list of variables used is included in the README file. For more information regarding this dataset, users can contact the authors at pmandon@worldbank.org. File location: Inputs/CAF_Pop_pts_MTI_Flood2.csv. Access date: 15/11/2024.
    Access policy
    Data is limited-access and is not included in the reproducibility package.
    Data URL
    https://www.fathom.global/newsroom/world-bank-collaboration/
    Citation
    Fathom, World Bank. (2025). Fathom Global 3.0 Flood Risk [dataset]. https://www.fathom.global/newsroom/world-bank-collaboration/
    Grid-Level Indices for the Central African Republic
    Name
    Grid-Level Indices for the Central African Republic
    Note
    Data was created manually using QGIS with information from ICASEES Surveys. This dataset provides population point–level information on access to health facilities, primary schools, and water points. Data was extracted at each population point available. A list of the variables is included in the README file. For more information regarding this dataset, users can contact the authors at pmandon@worldbank.org. File location: Inputs/CAF_Pop_pts_MTI_stats_wWater.csv.
    Access policy
    Data is restricted and is not included in the reproducibility package
    Citation
    Authors' construction with information from Institut Centrafricain des Statistiques et des Etudes Economiques et Sociales (ICASEES). (2025). Grid-Level Indices for the Central African Republic [dataset].
    Sectoral Gross Domestic Product and Nighttime Lights Time Series for the Central African Republic
    Name
    Sectoral Gross Domestic Product and Nighttime Lights Time Series for the Central African Republic
    Note
    This dataset was manually compiled by the authors in QGIS. It includes historical GDP data disaggregated by sector, together with nighttime lights data. Sectoral GDP data are drawn from the World Bank World Development Indicators, National Accounts data from the national statistical institute ICASEES, and internal computations for older series. Nighttime lights are computed using the DMSP series from the Colorado School of Mines Earth Observation Group (EOG) (https://eogdata.mines.edu/products/dmsp/). As the data is restricted, the list of variables is included in the README file. For more information regarding this dataset, users can contact the authors at pmandon@worldbank.org. File location: gdpSectLight_1960To2022.RDS. Years included: 1960-2023. Access date: 05/06/2024.
    Access policy
    Data is restricted and not included in the reproducibility package
    Citation
    Authors' construction. (2025). Sectoral Gross Domestic Product and Nighttime Lights Time Series for the Central African Republic [dataset].
    Household-Level Regression Dataset Matched to Grid-Level Indices
    Name
    Household-Level Regression Dataset Matched to Grid-Level Indices
    Note
    Dataset resulting from a spatial join between household-level data from the 2021 Harmonized Household Living Conditions Survey and grid-level indices produced by script '2_Vulnerability_indices.R'. The survey is collected by the national statistical institute (ICASEES) with the support of partners, including the World Bank Group, and is available internally only. Files location: reg_data_light.dta. Since the data is restricted and not included in the reproducibility package, the list of variables used is included in the README file. For more information regarding this dataset, users can contact the authors at pmandon@worldbank.org.
    Access policy
    Data is restricted and not included in the reproducibility package.
    Citation
    Authors' construction. (2025). Household-Level Regression Dataset Matched to Grid-Level Indices. [dataset]
    Global Administrative Areas Database
    Name
    Global Administrative Areas Database
    Note
    The data were transformed into a regular 5×5 km grid covering the full territory of the Central African Republic by applying a grid function to the national boundary shapefile from the Global Administrative Areas database (GADM 4.1). For mapping purposes, the authors also created a simplified version of this grid containing fewer administrative variables. Files location: Input/GIS/gadm41_CAF_1_2020.shp (original national boundary); Input/GIS/Grid_5km_CFA_v4.* (5×5 km grid); Input/GIS/BaseGrid_NewPrefs.shp (simplified grid for mapping).
    Access policy
    Data is publicly available and included in the reproducibility package
    License
    GADM license
    License URL
    https://gadm.org/license.html
    Data URL
    https://geodata.ucdavis.edu/gadm/gadm4.1/shp/gadm41_CAF_shp.zip
    Citation
    University of California, Davis. (n.d). Global Administrative Areas database [dataset]. https://geodata.ucdavis.edu/gadm/gadm4.1/shp/gadm41_CAF_shp.zip
    Central African Republic - Subnational Administrative Boundaries
    Name
    Central African Republic - Subnational Administrative Boundaries
    Note
    Files location: input/GIS/caf_admbnda_adm3_200k_sigcaf_reach_itos_v2.*. Access date: 22/13/2023.
    Access policy
    Data is publicly available and included in the reproducibility package
    License
    Creative Commons Attribution for Intergovernmental Organisations (CC BY-IGO)
    License URL
    https://data.humdata.org/faqs/licenses
    Data URL
    https://data.humdata.org/dataset/cod-ab-caf
    Citation
    United Nations Office for the Coordination of Humanitarian Affairs (OCHA), Agency for Technical Cooperation and Development (ACTED). (2024). Central African Republic - Subnational Administrative Boundaries [dataset]. https://data.humdata.org/dataset/cod-ab-caf
    Main Cities of the Central African Republic
    Name
    Main Cities of the Central African Republic
    Note
    Spatial layer of main cities in the Central African Republic, manually reproduced as a shapefile in QGIS based on the publicly available map of main cities published on Wikipedia (https://en.wikipedia.org/wiki/List_of_cities_in_the_Central_African_Republic#/media/File:Central_African_Republic-CIA_WFB_Map.png). Files location: input/GIS/CAR_main_cities.*.
    Access policy
    Data is included in the reproducibility package
    Citation
    Authors' construction. (2025). Main Cities of the Central African Republic [dataset].
    Data statement

    Some data is restricted and not included in the reproducibility package. For more information, see the README file.

    Description

    Output
    Unveiling Hidden Hardships: Leveraging Alternative Data To Map Multidimensional Vulnerability In The Central African Republic
    Type
    Working Paper
    Title
    Unveiling Hidden Hardships: Leveraging Alternative Data To Map Multidimensional Vulnerability In The Central African Republic
    Description
    Policy Research Working Papers (PRWP)
    Authors
    Author Affiliation Email
    Pierre Jean-Claude Mandon World Bank pmandon@worldbank.org
    Vincent Nossek World Bank vnossek@worldbank.org
    Walker Kosmidou-Bradley World Bank wkosmidoubradley@worldbank.org
    Frederic Mortier CIRAD, Montpellier, France frederic.mortier2@gmail.com
    Baptiste Cheville Phoenix Consulting International, Paris, France bastien.cheville@gmail.com
    Date of production

    2026-01-08

    Scope and coverage

    Geographic locations
    Location Code
    Central African Republic CAF
    Keywords
    Alternative Data Development Economics Fragile States Multidimensional Poverty Satellite Imagery Conflict
    Topics
    ID Topic Parent topic ID Vocabulary Vocabulary URI
    C55 Large Data Sets: Modeling and Analysis C5 Journal of Economic Literature (JEL)
    I32 Measurement and Analysis of Poverty I3 Journal of Economic Literature (JEL)
    O12 Microeconomic Analyses of Economic Development O1 Journal of Economic Literature (JEL)
    O55 Africa O5 Journal of Economic Literature (JEL)
    R12 Size and Spatial Distributions of Regional Economic Activity R1 Journal of Economic Literature (JEL)

    Disclaimer

    Disclaimer

    The materials in the reproducibility packages are distributed as they were prepared by the staff of the International Bank for Reconstruction and Development/The World Bank. The findings, interpretations, and conclusions expressed in this event do not necessarily reflect the views of the World Bank, the Executive Directors of the World Bank, or the governments they represent. The World Bank does not guarantee the accuracy of the materials included in the reproducibility package.

    Access and rights

    License
    Name URI
    Modified BSD3 https://opensource.org/license/bsd-3-clause/

    Contacts

    Contacts
    Name Affiliation Email
    Pierre Jean-Claude Mandon World Bank pmandon@worldbank.org
    Reproducibility WBG World Bank reproducibility@worldbank.org

    Information on metadata

    Producers
    Name Abbreviation Affiliation Role
    Reproducibility WBG DECDI World Bank - Development Impact Department Verification and preparation of metadata
    Date of Production

    2026-01-08

    Document version

    1

    Citation

    Citation
    loading, please wait...
    Citation format
    Export citation: RIS | BibTeX | Plain text
    Back to Catalog
    The World Bank Working for a World Free of Poverty
    • IBRD IDA IFC MIGA ICSID

    © The World Bank Group, All Rights Reserved.