Data-Driven Designs for Poverty Targeting in Social Protection: A Systematic Review and Meta-Analysis of Evidence from 2020–2025
DOI:
https://doi.org/10.70882/noun-ijcea.2026.1117Keywords:
Algorithmic fairness, Big data, Machine learning, Poverty targeting, Proxy Means Test, Social protectionAbstract
The effective targeting of social protection programmes is critical for poverty reduction, yet traditional methods often suffer from high inclusion and exclusion errors. Data-driven approaches, including machine learning and big data analytics, are increasingly proposed as solutions, but their real-world effectiveness and associated challenges require systematic evaluation. This study synthesizes evidence from recent literature (2020–2025) to examine the performance, trends, and limitations of data-driven designs for poverty targeting. A systematic review was conducted following PRISMA guidelines. Searches across Scopus, ScienceDirect, and Google Scholar yielded 47 peer-reviewed studies and high-quality working papers that met the inclusion criteria. Findings were synthesized narratively and presented in a meta-analysis table. Data-driven methods—such as Proxy Means Tests, machine learning algorithms, and novel data sources (mobile phone, satellite)—generally improve targeting accuracy over traditional community-based or categorical methods, particularly in economically diverse contexts. Innovations like Togo's mobile data-driven cash transfer programme demonstrate the potential for rapid, scalable crisis response. However, persistent challenges include data quality decay, algorithmic bias, the systematic exclusion of "invisible" populations (e.g., the phoneless or homeless), and threats to transparency and public trust. While data-driven targeting is a powerful tool for enhancing social protection efficiency, it is not a panacea. Effectiveness is highly context-dependent, and a purely algorithmic approach risks undermining equity and rights. The evidence strongly supports hybrid models that combine data-driven precision with community validation and robust grievance mechanisms. Future research should prioritize long-term impact evaluations, fairness audits, and strategies to reach populations systematically missed by digital data.
References
Aiken, E., & Blumenstock, J. (2024). How AI helped deliver cash aid to many of the poorest people in Togo. Yahoo! Finance.
Aiken, E., Ashraf, A., Blumenstock, J. E., Guiteras, R. P., & Mobarak, A. M. (2025). Poverty targeting at scale: Algorithmic vs. traditional approaches.
Aiken, E., Bellue, S., Karlan, D., Udry, C., & Blumenstock, J. E. (2022). Machine learning and phone data can improve targeting of humanitarian aid. Nature, 603(7903), 864–870.
Aiken, E., Ohlenburg, T., & Blumenstock, J. (2023). Moving targets: When does a poverty prediction model need to be updated? In ACM SIGCAS/SIGCHI Conference on Computing and Sustainable Societies (COMPASS '23).
Altındağ, O., O'Connell, S. D., Sasmaz, A., Balcoglu, Z., Cadoni, P., Jerneck, M., & Foong, A. K. (2021). Targeting humanitarian aid using administrative data: Model design and validation. Journal of Development Economics, 148, 102564.
Barca, V., et al. (2021). Inclusive social protection info systems [Rapid review, interviews].
Barca, V., et al. (2023). Promote use of admin data in social protection [Review and case studies].
Corral, P., Henderson, H., & Segovia, S. (2023). ML poverty mapping vs. traditional methods in Mexico [XGBoost vs. parametric models].
Follett, L., & Henderson, H. (2023). Improve poverty targeting via hybrid PMT-CBT [Bayesian modelling].
Hall, O., Dompae, F., Wahab, I., & Dzanku, F. M. (2023). A review of machine learning and satellite imagery for poverty prediction. Journal of International Development, 35(7), 1753–1768.
Hall, O., Ohlsson, M., & Rognvaldsson, T. (2022). A review of explainable AI in the satellite data, deep machine learning, and human poverty domain. Patterns, 3(10), 100600.
Han, P., Zhang, Q., Zhao, Y., & Li, F. Y. (2021). High-resolution remote sensing data can predict household poverty in pastoral areas, Inner Mongolia, China. Geography and Sustainability, 2(4), 254–263.
Hillebrecht, M., Klonner, S., & Pacere, N. A. (2023). The dynamics of poverty targeting. Journal of Development Economics, 161, 103033.
Human Rights Watch. (2023). Automated neglect: How the World Bank's push to allocate cash assistance using algorithms threatens rights.
Human Rights Watch. (2025). World report events of 2024.
Kidd, S., Athias, D., Kidd, S., Fitzpatrick, C., Catling, L., Baker, L., & Treacy, K. (2020). Hit and miss: An assessment of targeting effectiveness in social protection.
Martinez, R. G., & Cooray, M. (2025). Enhancing poverty targeting with spatial machine learning: An application to Indonesia. arXiv preprint, arXiv:2503.04300.
Nkurunziza, F., Kabanda, R., & McSharry, P. (2025). Enhancing poverty classification in developing countries through machine learning: A case study of household consumption prediction in Rwanda. Cogent Economics and Finance, 13(1), 2444374.
Poulin, C., Trimmer, J., Press-Williams, J., Yachori, B., Khush, R., Peletz, R., & Delaire, C. (2022). Performance of a novel machine learning-based proxy means test in comparison to other methods for targeting pro-poor water subsidies in Ghana. Development Engineering, 7, 100098.
Schnitzer, P., & Stoeffler, Q. (2021). Targeting for social safety nets: Evidence from nine programmes in the Sahel (World Bank Policy Research Working Paper).
Shi, B., Huang, W., Dang, Y., & Zhou, W. (2024). Leveraging social media data for pandemic detection and prediction. Humanities and Social Sciences Communications, 11(1), 1–10.
Smythe, I. S., & Blumenstock, J. E. (2022). Geographic microtargeting of social assistance with high-resolution poverty maps. Proceedings of the National Academy of Sciences, 119(32), e2120025119.
World Bank. (2024). Scaling up social assistance where data is scarce: Opportunities and limits of novel data and AI. www.worldbank.org
Xu, J., Song, J., Li, B., Liu, D., & Cao, X. (2021). Combining night time lights in prediction of poverty incidence at the county level. Applied Geography, 135, 102552.
Zhang, W., Lei, T., Gong, Y., Zhang, J., & Wu, Y. (2022). Using explainable artificial intelligence to identify key characteristics of deep poverty for each household. Sustainability, 14(16), 9872.
Downloads
Published
Issue
Section
License
Copyright (c) 2026 Moses K. Orfega, Francis B. Osang, Godwin U. Inyang (Author)
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
