Adding a basis for sustainable poverty monitoring: The indicator systems and multi-source data of multi-dimensional poverty measurement

Poverty is a pressing social and economic issue that demands attention. Reducing and eliminating poverty are shared objectives globally. Current studies often rely on existing theoretical models or survey questions when selecting dimensions and indicators, which may have limitations. Few researchers have offered a comprehensive framework for indicator selection based on data sources. The measurement of multi-dimensional poverty primarily relies on the use of the Multi-dimensional Poverty Index (MPI). However, there is a lack of systematic research on the selection of indicator systems and their respective weights. This study used in-depth literature analysis and systematic literature review (SLR) methods to sort out and integrate indicator systems and evaluation methods of multi-dimensional poverty evaluation (MPE). The results indicate that: (1) The indicator systems of MPE should include the following dimensions: economic, health, education, living standard, social relationship and natural environment for both household and regional level; (2) MPE based on multi-source data could reduce the difficulty and costs of field survey, increase comparability and convenience, and be more objective and reliable, which would be an important direction for future related research; And (3) Hybrid method would be more reasonable than the single weighting method, which could minimize the loss of information and make the weighting result as close as possible to the actual result. We propose to establish a dynamic monitoring system based on multi-source data, which could offer new insights for sustainable poverty monitoring.