Dynamic adjustment process of retail store density in cointegrated panels: evidence from Japan

Using recently developed econometric techniques for testing the existence of unit roots and cointegration in panel datasets, this article investigates the dynamic adjustment process of the number of retail stores per person by business type. Specifically, the major objective of this article is to examine whether there is a long-term relationship between retail store density and consumers' ability to transport and store goods; that is, the relative costs of performing distributive tasks, as suggested in previous studies relating to store density. The first conclusion from this article is that a cointegration relationship exists between store density for most types of retail businesses and proxies for consumers' ability to transport and store goods. Second, the estimation results of an error correction model reveal that the density of stores dealing in convenience goods exhibits a slow adjustment speed to recover the long-run equilibrium density level, while the density of stores selling shopping goods shows relatively fast adjustment.