Evaluating a Poverty Reduction Program in Cameroon with Satellite Data

Study Context 

Understanding the impact of poverty reduction programs is crucial, but measuring their success can be difficult, especially in fragile and conflict-affected areas where reliable data is limited.  The Grassfield Participatory and Decentralized Rural Development Project (GP-DERUDEP) in Cameroon, implemented in two phases from 2005 to 2020, aimed to improve households’ agricultural productivity, income, and social infrastructure through community-driven development. While preliminary reports suggest positive outcomes, no rigorous study has measured the program’s causal impact on reducing poverty. Researchers will use satellite images, machine learning, and on-the-ground checks to measure the effects of GP-DERUDEP on poverty levels in remote and conflict-affected villages.

Study Design 

Researchers will analyze satellite images of the region from 2000 to 2024, covering a period before, during, and after the project. These images will be combined with household survey data from the Cameroon Demographic and Health Surveys (DHS) and Multiple Indicator Cluster Surveys (MICS), along with on-the-ground observations, to estimate changes in poverty levels. By comparing areas where the program was implemented earlier to those where the program was started later, researchers can assess its impact. A machine learning model will be trained to identify and analyze changes to types of buildings, roads, and farmland in satellite images.

Results and Policy Lessons 

Results forthcoming.