Skip to content

Urban Poverty Mapping in Mozambique: Is census or geo-spatial data better?

This paper contributes to an expanding pool of knowledge by estimating household per capita consumption at a very detailed level within all urban areas in Mozambique.

We build a prediction model based on characteristics from the census, another model based on geo-spatial data and finally, a model that combines the two data sources, all with per capita consumption from the Inquèrito sobre Orçamento Familiar (IOF) 2019/20.

This study is the first to generate high-resolution poverty maps of all urban areas in Mozambique, an output that may in itself be valuable for the Government of Mozambique and others.

Moreover, the comparison of the predictive capacities of census variables vis-a-vis geo-spatial data at the lowest official unit of aggregation provides important insights for the continuous development of poverty mapping methods in the academic literature.


Available at: