ABSTRACT
A facet of retail geography that has latterly attracted widespread attention in Western countries is the assessment of future shopping needs and the development of retail-location models. This important trend derives primarily from the concern of planners with the provision of shopping facilities and in particular the location of new centres. But the growth of model building also stems from the concern of social scientists to evolve more rigorous theory relating to the fundamental relationships of urban structure. Underlying both approaches is the assumption that consumer behaviour in the aggregate can be predicted on the basis of mathematical probability, even though the behaviour of the individual consumer or of distinctive socioeconomic groups is not normally predictable. Almost all models disregard the structure of retailing.
