Selected Country Experiences​
Burkina Faso faced challenges during its 2019 census due to security issues in the north and east, which left some areas unenumerated. To address this, the National Institute for Statistics and Demography (INSD), in collaboration with the United Nations Population Fund (UNFPA) and WorldPop, developed predictive methods to estimate populations in inaccessible regions. These methods produced gridded population estimates for the entire country, helping to fill critical data gaps.
Papua New Guinea had to cancel its 2021 census due to COVID-19 disruptions. In response, the NSO partnered with WorldPop to use geospatial modelling for population estimation. This approach integrated satellite imagery, census preparation data, and malaria survey data with over 50 geospatial datasets. A two-stage modelling process was used to account for settlements hidden under forest canopies. The final model provided population estimates with 95% confidence intervals across all administrative levels. Despite technical challenges like data interoperability and boundary mismatches, the results were openly shared through global data platforms to support future census planning.
Cameroon has not conducted a national census since 2005 and faces significant demographic data gaps, especially in conflict-affected regions. To address this, the National Institute of Statistics collaborated with WorldPop to produce high-resolution 2022 population estimates using geospatial modelling. This method combined satellite imagery, survey data, and advanced Bayesian techniques. The resulting gridded data has been used for survey planning, vaccination campaigns, and election logistics. It also supported the creation of a digitised national sampling frame. While not a substitute for a full census, these estimates provide the only granular population data co-developed with the government and are vital for planning and service delivery.
Zambia postponed its 2020 census to 2022 due to the pandemic and funding issues. In the interim, the Zambia Statistics Agency (ZamStats) worked with WorldPop under the GRID3 project to develop bottom-up population estimates using household survey data and geospatial information. These estimates were used for malaria control and COVID-19 vaccination planning. Ahead of the 2022 census, ZamStats also explored household count estimates to refine Enumeration Areas (EAs). The 2022 census was Zambia’s first fully georeferenced census. Afterward, WorldPop helped produce top-down gridded estimates aligned with preliminary census results, offering detailed population distribution insights. These estimates will be updated with final census data in 2025, and comparisons with earlier modelled estimates are ongoing.
Ethiopia has not conducted an official census since 1974, resulting in major data gaps. In 2023, the Ethiopian Statistical Service (ESS) and UNFPA partnered with WorldPop to model national population estimates. Given the availability of comprehensive 2015–2017 cartographic data, a top-down modelling approach was chosen. Household counts were combined with geospatial covariates like land cover, climate, infrastructure, and night-time lights. Residential characteristics were derived from global settlement data to exclude non-residential structures. The resulting national population estimates are currently undergoing validation and refinement before publication by ESS.