Location and Co-location in Retail - A Probabilistic Approach Using Geo-coded Data for Swedish Metropolitan Retail Markets

This paper offers a unique empirical approach to detect co-location patterns in the retail sector. We analyse the co-location of retail service establishments in Sweden by using geo-coded data. We pinpoint each establishment in Sweden down to a 250 by 250 metre square. Our analysis captures a general pattern for the co-location of different types of retailing activities, also taking into account several spatial attributes of location. We study the three major retail markets in Sweden (Stockholm, Malmö, and Gothenburg). We position the paper within a framework in which the presence of an economic activity in space is explained by the spatial attributes of the place as well as the nature of the economic activity. Our empirical design follows a probabilistic approach, whereby the probability of finding a particular type of retailing activity in a square is explained by the presence of similar and different kinds of retailing activities in the respective square, as well as by the characteristics of their location. We select three major and distinct types of retailing stores: clothing, household appliances, and specialized stores. Our findings are well in line with the propositions of bid rent theory and central place theory for retail markets. We further document negative location tendencies between shops that sell frequently purchased products and shops that sell durables. Moreover, our results highlight the importance of demand in the close surroundings, which is particularly strong for small-scale establishments.