<![CDATA[Zakat, as an instrument of financial redistribution in the Islamic economy, has significant potential for poverty alleviation; however, the spatial distribution patterns of its impact in Indonesia have not yet been extensively analysed using spatial modelling. In this paper, we measure the spatial pattern of the impact of zakat on poverty data in Indonesia and form a zakat cluster in Indonesia using the Bayesian geoadditive model approach. In this study, we collected data from the Indonesian Central Statistics Agency and the National Amil Zakat Agency (BAZNAS) in 2024, respectively, to map and model socio-economic and spatial determinants, namely the relationship between zakat and poverty in Indonesia. We found strong support for our approach to mapping and modeling flexibly to include spatial analysis. The results of the analysis indicate that zakat has a significant effect on poverty, with a posterior mean of 21,042.1 (SD = 28,142.4; 95% PCI: 12,665.3–30,185.0), where the entirely positive credible interval indicates a consistent spatial association between the distribution of zakat and the concentration of the poor. Furthermore, the structured spatial effects reveal regional heterogeneity with a significant East–West divide: the Papua region exhibits the highest positive spatial effect (up to +317.67), indicating a poverty burden far exceeding the model’s predictions, whilst Java and parts of Kalimantan show a negative effect (up to −317.67), reflecting the greater effectiveness of zakat distribution. The model identifies distinct regional cluster structures, demonstrating significant spatial heterogeneity across regions in Indonesia. The Bayesian geoadditive approach proved superior to conventional models in capturing these non-linear patterns and spatial inequalities. The resulting maps could serve as a new analytical method for policymakers to design more targeted zakat collection and distribution strategies in support of national poverty alleviation.]]>
