Spatial dynamics of logistic allocation in large urban centers has gained relevance in urban economics, public policy formulation and transport engineering. A crucial aspect of these spatial, agglomerative, and dispersive phenomena are the causal links that such processes generate or by which they are generated. This study investigates the causal relationships linked to the allocation of logistics warehouses in the context of the sub-regions of the São Paulo Metropolitan Region (SPMR) based on the theory of agglomeration economies. Two variables related to the urban distribution of goods were selected: (a) Gross Domestic Product (GDP) and (b) Cargo Theft (CT) and contrasted with the number of logistics warehouses (NW). Five regions of the SPMR were considered both for spatial descriptive analysis and for the study of causality based on the Granger methodology, adapted by Hurlin, and on the cointegration methodology and Panel Vector Error Correction Model. For SPMR, the relationships obtained demonstrate, in the short term, unidirectional causalities between GDP/NW. In the short term, the causal relationship between CT/NW for the metropolitan region was unidirectional from the amount of cargo theft to the number of warehouses. The subregions showed strong heterogeneity in causal relationships, both in the short and long term, demonstrating the complexity of the logistical allocation phenomena in the SPMR. • Cargo theft is a crucial decision factor when installing a logistic warehouse. • The demand in high-GDP regions increases the need for nearby logistics services. • The number of warehouses affects long-term economic growth. [ABSTRACT FROM AUTHOR]