Abstract

1. The challenge
In West Africa, traditional censuses and surveys (general population and housing censuses, household surveys, etc.) are extremely costly, take 3 to 5—or even 10—years to complete, and quickly become obsolete. Yet, the state requires timely data to develop policies and make decisions.
2. The new alternative
Utilizing two existing types of big data:
a) Mobile telephony data (CDRs – Call Detail Records)

• Every call or SMS leaves a trace via mobile network towers.
• This makes it possible to determine where people sleep, work, and travel, without needing to survey them directly.
• Anonymized, yet highly powerful.

b) Geospatial technologies

• Satellite imagery (housing, nighttime lighting, crops).
• GPS, drones, and OpenStreetMap.

3. Concrete applications for official statistics

• Population: Estimating population between censuses by counting rooftops via satellite and analyzing mobile phone density.
• Mobility: Measuring rural-to-urban migration, movement between geographic areas, and the impact of a crisis.
• Poverty: Identifying areas with no electricity at night and low mobile activity (indicative of potential poverty).
• Agriculture & isolation: Mapping rural roads, cultivated land, and access to markets.
• Disasters: Tracking population movement in real time following events like floods.

4. Challenges in West Africa

• Legal: Data protection laws (involving operators like Togocom and Moov).
• Technical: A shortage of data scientists at national statistical institutes (such as INSEED).
• Trust: Public concern that this data might be used for surveillance.
• Methodology: The need to demonstrate that these new data sources are as reliable as traditional surveys.

5. Conclusion
In short, the use of mobile telephony data and geospatial technologies does not mean the end of traditional surveys; rather, it serves to complement them. Traditional surveys will become more streamlined and targeted, supported by mobile data and geospatial technologies that provide a real-time snapshot of the phenomenon under study. This represents the future of national statistical institutes.