Applying spatial autocorrelation concept for pawnshop business insight

Abstract

Background: Pawnshop businesses play a strategic role in supporting financial inclusion and regional liquidity; however, nationally aggregated indicators may obscure substantial variations in business activity at finer geographical scales. Methods: This study examines the spatial patterns and geographic dependence of pawnshop business activity in the Jabodetabek metropolitan region from May 2024 to May 2025 using a combination of Local Moran’s I (LISA), Local Getis–Ord Gi*, and Local Geary’s C. Due to the absence of district transaction data, regency/municipality-level customer and loan totals from the Financial Services Authority (OJK) were disaggregated using population and land-area proxies from Statistics Indonesia (BPS). These proxy-based estimates were then analyzed within a GIS framework using a row-standardized spatial weight matrix to capture local spatial dependence. Findings: The results reveal a segmented spatial structure, characterized by strong customer and total-loan concentrations in Central Jakarta and Tangerang City, transitional and outlier patterns in surrounding areas, and a high-intensity corridor of average loan value per account (AL) in Bogor City and adjacent districts. Conclusion: The findings demonstrate that pawnshop business development should not rely solely on national or metropolitan averages but should adopt cluster-based and context-sensitive strategies for branch expansion, service optimization, and regional resource allocation. Novelty/Originality of this article: The originality of this study lies in integrating three complementary local spatial statistics at the district level to transform pawnshop indicators into geoeconomic business insights, while explicitly acknowledging that the proxy-derived values represent exploratory estimates that require further validation using branch- or transaction-level data.