This research will integrate and generalize conflicting views of the role of inter-facility flow in location problems. An over-riding question is 'why is it so difficult to construct a model to represent a phenomenon that occurs so frequently in reality?' Much research effort has been devoted to developing models for the location of interdependent hub facilities. In these problems, for reasons that are not understood, non-integer solutions are sometimes encountered. Some procedures have produced highly fractional solutions, although others have fared better. It is important to answer this question because, for theoretical and practical purposes, interdependence between facilities is a crucial source of clustering/agglomeration in the space economy. The research will refine mathematical models of hub systems, thereby integrating a wide range of recent literature in an effort to enhance scientific understanding of a fundamental problem in economic geography, namely the optimal location of interactive facilities.
Methods to be employed include optimization, mathematical programming, and network analysis. The potential impact on advancing knowledge is as follows. The failure of location models for the case of interdependent activities is a major puzzle for theorists: this may in fact be repaired as a result of a careful, creative examination of the way in which these issues play out in the case of hub and spoke networks. Hubs serve as conduits for interactions between geographically dispersed nodes. Within the capacity limits of aircraft, trucks, pipelines, or network links, the marginal cost of an extra unit of flow is small (perhaps close to zero). Furthermore, there are lower variable costs where greater flows are handled by higher capacity infrastructure. Thus, providers of transportation services reduce their average unit costs by bundling flows.
