<![CDATA[This study examines the spatial spillover dynamics of the Human Development Index (HDI) in West Java Province, Indonesia, using the Enclave Panel Threshold Spatial Autoregressive (E-PT-SAR) model and panel data from 27 districts during 2021–2025. The model extends the conventional PT-SAR framework by incorporating enclave identification, distinguishing core prosperous regions, prosperous enclaves, and poor regions. The estimated poverty threshold (γ = 8.5124%) separates West Java into prosperous and 13 poor districts, while the enclavity index and a stationarity-constrained enclave threshold (τ = 0.7143) identify a single prosperous enclave, Ciamis. The results reveal that spatial spillovers differ markedly across regimes. Crucially, the spillover among prosperous districts is insignificant in the conventional two-regime model (ρ = 0.4656, p = 0.240) but becomes significant once the isolated enclave is separated (core prosperous ρ = 0.6344, p = 0.025), revealing a spillover that the standard threshold model masks. A strong spillover is also found in the poor regime (ρ = 0.9740), whereas the enclave shows no statistically detectable spillover, consistent with the hypothesis that spatially isolated prosperity does not diffuse to surrounding poorer areas. The effects decomposition confirms that indirect (spillover) effects dominate in the poor regime. Because the enclave regime comprises only one district, its results are interpreted as exploratory. These findings imply that regional development policies should be tailored to different spatial regimes. Keywords: Spatial Econometrics; E-PT-SAR; Enclave; Spatial Spillover; HDI; West Java]]>
