Regional retail landscapes emerging from spatial network analysis

Abstract

ABSTRACT In the tradition of Christaller’s central place theory, policy and scientific studies delineate retail markets around predefined central places. Evolutions on the demand and supply side have changed where consumers shop and where stores are located. We use spatial network analysis techniques (Leiden detection algorithm) that allow for a bottom-up approach. The data were sourced from a questionnaire on shopping in Flanders (Belgium). The results show that community detection is able to deal with the geographical complexity of retail. Communities for daily goods shopping remain small, for recurring goods have become very large, even bigger than those for exceptional goods.