Program Targeting with Machine Learning and Mobile Phone Data: Evidence from an Anti-Poverty Intervention in Afghanistan

TL;DR

This study demonstrates that machine learning models using mobile phone data can effectively identify ultra-poor households in Afghanistan, matching survey-based accuracy and improving classification when combined.

Contribution

It introduces a novel approach combining mobile phone logs with survey data to enhance targeting accuracy in anti-poverty programs.

Findings

Mobile phone data can identify ultra-poor households nearly as accurately as surveys.

Combining survey data with mobile phone logs improves classification accuracy.

Machine learning methods effectively leverage mobile data for program targeting.

Abstract

Can mobile phone data improve program targeting? By combining rich survey data from a "big push" anti-poverty program in Afghanistan with detailed mobile phone logs from program beneficiaries, we study the extent to which machine learning methods can accurately differentiate ultra-poor households eligible for program benefits from ineligible households. We show that machine learning methods leveraging mobile phone data can identify ultra-poor households nearly as accurately as survey-based measures of consumption and wealth; and that combining survey-based measures with mobile phone data produces classifications more accurate than those based on a single data source.