Forecasting Poverty and Shared Prosperity Using Cell Phone Data

Cellphones generate large datasets of "digital footprints" from a population, which can be analyzed using data mining and computer-learning techniques to reveal behavioral patterns that can then be used to estimate and forecast poverty and shared prosperity. This proposal presents an affordable, practical, and scalable solution for mapping poverty based on the aggregate behavioral patterns of cellphone users, which will be piloted by the Government of Guatemala. The World Bank has initiated an innovative partnership with Telefonica Research and the iSchool of the University of Maryland to develop computer algorithms based on anonymized cellphone call records gathered by Movistar Guatemala, the country's largest cellphone provider. This initiative aims to produce detailed and reliable information at a lower cost.