Mapping the Spatial Distribution of Poverty Using Satellite Imagery in Thailand
This report presents the results of a feasibility study on high-quality poverty statistics in Thailand using satellite imagery, geospatial data, and advanced algorithmic techniques to complement conventional survey methods.
The "leave no one behind" principle of the 2030 Agenda for Sustainable Development requires appropriate indicators for different segments of a country’s population. This entails detailed, granular data on population groups that extend beyond national trends and averages. ADB collaborated with the National Statistical Office of Thailand and the Word Data Lab for the feasibility study, which aimed to enhance the granularity, cost-effectiveness, and compilation of high-quality poverty statistics in Thailand.
- How Are Poverty Statistics Estimated?
- Tapping Computer Vision Algorithms for Predicting Poverty Rates
- Using Random Forest Estimation to Compile Grid-Level Estimates of Poverty Head Counts
- Key Findings
- Estimating Structural Models as an Alternative Method
- Summary and Conclusion