<![CDATA[Capturing multidimensional poverty through conventional poverty statistics is challenging in view of their limited spatial resolution and focus on monetary indicators. In Indonesia, poverty measurement remains largely expenditure-based, potentially obscuring localized deprivations in education, health, and living standards. The objective of this present study is to address this limitation by developing a granular spatial mapping framework for the Multidimensional Poverty Index (MPI) in East Java Province. Employing the Alkire–Foster approach and Susenas 2023 data, the provincial MPI is estimated at 0.0479, and MPI values are spatially predicted at a 3 × 3 km grid resolution by integrating geospatial indicators of infrastructure accessibility, education and healthcare facilities, nighttime light intensity, and population density. The spatial models demonstrate strong predictive performance (R2 ≈ 0.97; AUC ≈ 0.98), revealing pronounced fine-scale variation in multidimensional poverty and identifying deprivation clusters that are not observable in administrative-level statistics. Areas characterized by geographic isolation and limited-service accessibility consistently exhibit elevated predicted MPI values. The findings of this study highlight the significance of high-resolution multidimensional poverty mapping in facilitating the development of more spatially targeted and evidence-based poverty reduction policies at the local level.]]>
Copyrights © 2025
