<![CDATA[Agriculture plays a strategic role in supporting regional economic growth, employment, and food security in East Java, Indonesia. However, its contribution to Agricultural Gross Regional Domestic Product (GRDP) differs markedly across districts and cities, indicating spatial and structural disparities. These variations suggest that agricultural performance is shaped by localized biophysical and socio-economic factors that global regression models often fail to capture. Previous studies have largely relied on single-commodity analyses or global approaches that assume spatial homogeneity, thereby overlooking spatial non-stationarity in the determinants of agribusiness. This study aims to analyze the spatial heterogeneity of agricultural GRDP determinants and evaluate how localized factors, including productivity, land conversion, labor, food prices, market access, agroindustry value added, and food diversification, differentially impact nine districts and cities in East Java. The novelty lies in integrating multidimensional upstream-to-downstream agribusiness variables within a Geographically Weighted Regression (GWR) framework, moving beyond traditional production-centric models to identify specific regional drivers. Using 2023 secondary data from nine districts/cities, the study employs Ordinary Least Squares (OLS) as a benchmark, followed by GWR with a Gaussian kernel and cross-validation bandwidth. The GWR model outperforms OLS, increasing R² from 0.75 to 0.89 and reducing AICc from 134.2 to 112.5, while residual Moran’s I declines from 0.32 to –0.05. Agro-industry value added shows the strongest positive effect (0.38 in Banyuwangi), whereas land conversion negatively affects urban areas (–0.41 in Malang City). These findings confirm spatial non-stationarity and support place-based agricultural policy design.]]>
