Chapter 9 addresses what many would see as the key question of economic geography: when there is agglomeration of economic activities because of locational externalities, will that agglomeration take place in locations which are recognisably central in a spatial sense? It is a curious feature of some of the recent literature on ‘centre–periphery’ development that it lacks the structure to make sense of its own terminology and to answer that key question.

The model presented here assumes that there is no mobility of factors between locations. Factor mobility is a powerful force in spatial agglomeration (Ottaviano and Puga, 1997), and by abstracting from this force, Venables aims to show us how the externalities associated with backward and forward linkages between imperfectly competitive industries are sufficient to give rise to rich and systematic spatial patterns of economic activity.

The results are derived by numerical simulation and are intuitively appealing. In the kind of model used in the chapter, agglomeration is associated with scale economies and imperfect competition, so with more than one industry it is natural for the less competitive industry to tend to be the one which occupies the central region. There is also a well established phenomenon of centripetal forces being strongest at intermediate levels of trade costs. Here, as trade costs change, the changing balance of centrifugal and centripetal forces as trade costs change produces switches in the spatial pattern of production.