The conditions and circumstances of people around the world cannot be reduced to national averages alone. Within the same country, economic development can be unbalanced, with some regions benefiting from growth while others lag behind. Understanding these differences at a detailed level is an essential part of monitoring well-being and knowing where to focus development efforts where they are needed most.

The World Bank’s new Geospatial Poverty Portal is an important tool in these efforts, providing indicators of poverty at the local level, inequality, and multidimensional poverty indicators over time and at the global level. This new website is part of the World Bank’s global poverty monitoring, which appears in the Poverty and Inequality Platform (PIP).

The portal features an interactive map based on data from the newly released Subnational Poverty and Inequality Database (SPID), covering more than 1,600 subnational regions from 141 economies. The statistics in SPID are direct calculations from available survey data at their levels of representation, meaning that the frequency of data varies by country. Some countries have subnational time series of up to 10 years, allowing users to compare changes temporally and spatially over a long period.

The SPID shows that countries can have large differences in prosperity across subnational regions. Differences can be surprisingly wide in countries with large development gaps between rural and urban centres. Remote regions still often rely on low-productivity agriculture, but many of these countries also have dynamic and growing urban centers. For example, in the case of Thailand, the World Bank upper-middle-income country poverty rate ($6.85 per day in 2017 PPP) in Bangkok in 2020 is actually 0 percent, while the poverty rate in Mae Hong Son Province, the poorest province, has a score of 47.3. percent.

Figure 1. Interactive map based on the Subnational Poverty and Inequality Database

Source: Speed

The availability of survey data varies between countries. Some countries publish data annually, while others have large gaps between surveys. To compare subnational regions across countries, it is important to be able to compare them at the same time. This is done by “classifying” poverty rates at the national level using the same methods used to monitor global poverty and then focusing on poverty estimates at the subnational level.

The Global Subnational Atlas of Poverty (GSAP) is a map of subnational poverty rates in more than 1,700 regions across more than 160 economies projected in a common year. The latest edition of the Global Poverty Plan shows poverty rates at the subnational level through 2019. While only 12 countries had extreme poverty rates above 50 percent in 2019, the Global Poverty Plan shows that there are 182 sub-regions of 26 countries that are rising It has poverty rates of more than 50 percent. percent. Among them are 162 in sub-Saharan Africa, 19 in the Middle East and North Africa, and one in East Asia and the Pacific.

One way the Global Anti-Poverty Program has been used is to identify the number of poor people who face the additional challenge of living in locations with high climate risks or exposure to natural hazards. A 2022 article in Nature used the 2020 edition of the Global Action Plan and found that up to 1.8 billion people are directly exposed to floods once every 100 years, and 780 million of them live below the poverty line in the upper middle-income countries identified by the Bank. International. .

Poor households have fewer coping mechanisms and safety nets, making them more vulnerable to climate risks. Exposure to climate shocks can lead to long-term losses in the well-being of their families. Figure 2 shows that there is variation in patterns of poverty rates at the subnational level (measured at the international poverty line set by the World Bank) due to different levels of flood risk (low, medium or high) in selected countries. Subnational areas with high poverty rates and numbers of poor are sometimes, but not always, areas exposed to high natural hazards. Therefore, it is important to understand the relationship between poverty and climate risks in different country contexts.

Figure 2. Exposure to flood risk in five sub-Saharan countries by sub-national region

A common challenge is the desire to obtain detailed information for climate vulnerability analysis, but statistics on poverty and other social statistics are not available at this level of detail. Even when census or household survey data are available, information may be very limited.

Small area estimation techniques provide some options to address this issue. The first technical paper for small area estimation was in 1994 (Ghosh and Rao). Since then, the World Bank has helped popularize and operationalize the use of small area estimation (SAE) techniques by conducting further research (see our previous blog) and developing tools such as PovMap 15 years ago, and more recently open source Stata. Sai code. Today, the Stata sae code has become the World Bank’s gold standard for accurately estimating poverty in small areas.

These resources appear on the Geospatial Poverty Portal alongside a growing archive of historical poverty mapping work by the World Bank and national statistical offices. We hope that this archive of small area estimation data will enable further spatial analysis and research on granular poverty and other spatial data. The Stories page also showcases innovative examples of work that tell a more detailed story of poverty and vulnerability in a country. An example from Vietnam below demonstrates the power of combining small-area poverty estimates with climate and environmental data (report).

Figure 3. District-level small-area poverty estimates in Vietnam, 2019

District-level small-area poverty estimates in Vietnam, 2019

Source: Figure 6.7,

We are excited about the ability of geospatial data to tell a more accurate and compelling story about how the world’s most vulnerable groups are affected by changing climate patterns. Further improvements, features and additional data are planned for the geospatial poverty portal.

If you would like to have your work recommended for our Stories page, or to learn more about this initiative, please contact or

Leave a Reply

Your email address will not be published. Required fields are marked *

%d bloggers like this: