Investigating path dependence and spatial characteristics for retail success using location allocation and agent-based approaches

Abstract

Agent-based models (ABMs) have been widely used to represent and investigate complex systems and are a contemporary modelling approach used in the study of land-use change. While many ABMs have been constructed to address research associated with residential land development and human choices, agricultural land transition and farmer decision-making, and transportation networks and planning, less attention has been given to improving our understanding about the drivers and agent behaviours associated with commercial and retail competition, which subsequently affects land-use change. Among existing ABMs that represent the retail system, the focus has been on understanding consumer behaviours, but the inclusion of store competition is lacking, and most retail competition models still use a top-down modelling framework. The presented research contributes to the gap in agent-based models of commercial land use through the development and use of a retail-competition agent-based model (RC-ABM). Utilizing empirical research on consumer expenditures and retail location site selection, competition for home-improvement expenditures is simulated within the Region of Waterloo, Ontario, Canada. A model-to-model comparison of the RC-ABM against the historically used location-allocation model showed distributional equivalence. Results exhibit a high level of alignment between the RC-ABM and a traditional Location-Allocation Model (LAM) in estimating market capture and store revenue acquisition. While modelled competition cannot solely reproduce the observed spatial pattern of home-improvement stores in our study area, outcomes identified local market saturation, path-independent areas where store survivability is highest, and the relative influence of store and geographic characteristics on retail success. The presented model and case study highlight the potential to enrich future land-use models by better representing commercial development.