How Satellite Data Reveals Poverty and Risk in the Central African Republic

In the Central African Republic, collecting reliable data is often as difficult as delivering aid. The country has not held a full census since 2003. Years of conflict have weakened institutions, displaced hundreds of thousands of people and made large parts of the territory hard to reach. Yet poverty is widespread, especially in rural areas, where most households live below the international poverty line.

A new World Bank Policy Research Working Paper, Unveiling Hidden Hardships: Leveraging Alternative Data to Map Multidimensional Vulnerability in the Central African Republic, offers a new way to understand who is vulnerable and where they live. Researchers from the World Bank Group, CERDI, the University of Clermont Auvergne, the University of Twente, CIRAD in Montpellier, and Phoenix Consulting International combined satellite images, infrastructure maps, flood models, and conflict data to build a detailed vulnerability map of the country.

Instead of relying only on traditional surveys, the team used data visible from space and other digital sources to fill critical gaps.

Mapping the Economy, Grid by Grid

The researchers began by estimating economic activity. They used satellite images of nighttime lights, which are often linked to income and production. Brighter areas usually signal more economic activity.

Using a statistical model, they broke down national GDP into three sectors: primary activities like agriculture and forestry, secondary activities such as processing and manufacturing, and tertiary services. The model estimates that about 44 percent of GDP comes from the primary sector, 23 percent from the secondary sector and 33 percent from services.

They then spread this information across the country using a 5-by-5-kilometer grid. Cropland data helped estimate farming activity. Forest and deforestation data captured logging. Built-up areas, population density and mine locations helped identify urban and industrial activity.

The result is clear: economic activity is heavily concentrated in Bangui, the capital. Outside a few towns such as Bambari and Berbérati, most rural areas show very low economic density.

Measuring Access to Basic Services

Income is only one part of vulnerability. The study also looks at access to schools, health facilities and water points.

Using detailed location data from the national statistical office, the team calculated how long it would take people to walk from each grid cell to the nearest school, clinic or water source. They considered roads, terrain and off-road travel conditions to estimate realistic travel times.

The findings reveal uneven access. Primary education is relatively widespread compared to other services, especially along main transport routes. Health access is far more limited, with many remote areas facing long travel times. Water access sits in between, with better conditions near towns but large gaps in rural zones.

Each of these measures was converted into a simple 0 to 100 scale so they could be compared directly.

Living with Floods and Conflict

The study also accounts for environmental and security risks. Flood exposure was measured using high-resolution flood models. Areas along major rivers, including the Ubangi River near Bangui, face higher flood risks.

Conflict exposure was measured using data on lethal violent events. The researchers created a rolling index to show which areas have experienced repeated violence. Unlike flood risk, which is concentrated along rivers, conflict risk is more spread out and often overlaps with economically important towns and mining areas.

When these risks are combined with income and service access, a clearer picture emerges. Bangui appears less vulnerable than most of the country by national standards, though it still faces serious challenges. Western and southwestern regions show some of the highest vulnerability levels, driven by weak economic activity, poor service access and exposure to shocks.

Do the Maps Match Reality?

To check whether the satellite-based measures reflect real living conditions, the researchers compared them with data from more than 6,000 households surveyed in 2021.

The results are strong. Areas with higher economic activity scores tend to have wealthier households. Better access to schools, health facilities and water is also linked to higher household wealth.

One surprising finding is that areas with higher conflict exposure also tend to show higher wealth. This does not mean conflict improves lives. Instead, it suggests that armed groups often target economically valuable areas. Prosperous locations may become more attractive targets in fragile settings.

Overall, the study shows that combining satellite data with local infrastructure and conflict records can help governments and aid agencies target support more effectively. In a country where traditional data systems are weak, this approach offers a practical way to see where hardship is deepest and where help is most urgently needed.