Assessment of Spatial Dependence Using Spatial Autoregression Models: Empirical Analysis of Shopping Center Space Supply in Ohio

Abstract

Trade area (TA) analyses are useful for selecting the highest and best-use sites for shopping center investments and identification of the level of shopping center impact on existing retail, transportation, housing, and environment systems. This study offers a unique approach to assess the relation between shopping center attributes and TA market segmentation using zip code units (ZCU) as the level of analysis. The research methodology integrates gravity-based retail allocation models with spatial statistics, in particular, spatial autoregression models. The research findings reveal that accounting for spatial dependence in regression models offers a reliable assessment of retail supply–demand relations and the level of retail saturation in ZCUs. This contributes to site selection of shopping centers because it provides information on the existing nature of retail markets and generates comprehensible visual results to both public and private sector decision makers. Evaluating the potential success of shopping center investments is important in business strategy and urban policy-making.