SUR-SAR Modelling of the Human Development Index (HDI) of Regencies and Municipalities in West Java Province, Indonesia

Abstract

This study examines regional disparities in the Human Development Index (HDI) across 27 regencies and municipalities in West Java Province from 2022 to 2024 using a Spatial Seemingly Unrelated Regression–Spatial Autoregressive (SUR–SAR) model. HDI in West Java exhibits both spatial clustering and temporal persistence, indicating that development outcomes are shaped by interregional interactions and structural continuity across years. To capture these dependencies, the analysis incorporates eight socio-economic variables such as poverty, availability of health facilities, population density, inequality, unemployment, consumption level, GRDP per capita, and regional minimum wage into a multi-equation spatial system estimated using Maximum Likelihood. Four model classes are evaluated sequentially that is OLS, SUR, SAR, and SUR–SAR. The results demonstrate that the SUR–SAR specification provides the best overall fit and effectively accounts for both spatial dependence and inter-year error correlations. Population density consistently emerges as the strongest positive and highly significant determinant of HDI, while poverty shows a stable negative effect across multiple models. Minimum wage becomes significant only within the SUR–SAR framework, indicating that the influence of labour income on human development is more apparent once spatial and temporal linkages are jointly modelled. The spatial lag parameter is positive and significant across all equations, confirming the presence of HDI spillover effects among neighboring regions.