World Bank develops method to stitch together poverty data from different surveys

Poverty researchers have a problem: they measure income in one survey and health or education in another, leaving gaps in what they actually know about who is poor. This paper proposes filling those gaps by making a statistical assumption rather than waiting for data that perfectly matches. The method works better than existing approaches and lets governments compare poverty across countries using the data they already have.

For decades, poverty measurement has been split between surveys that ask different questions at different times, making it impossible to know which people are poor in multiple ways at once. This method lets researchers use existing data to build a complete picture without waiting years for new surveys designed from scratch. That matters because development banks actually allocate money based on poverty counts. More accurate counts mean money flows differently.