Zur Modellierung von räumlichen Kaufkraftströmen in unvollkommenen Märkten

Abstract

One of the main theorems of the New Economic Geography is concerned with incomplete markets caused by increasing returns. The commonly used model to predict spatial interaction probabilities, i.e. the Huff-model, is based on equilibrium assumptions. Thus there is a gap between new theory and old model which may be closed by introducing e.g. a new interaction model, if it would be possible to show an empirical relevance of increasing returns. This latter is worked out for 171 locations in Hesse (Germany). But instead of developing a new interaction model a very simple modelling trick is used to optimise a priori the measures of attractiveness used with the Huff-model. The solution presented here as algorithm I gives a very close approximation of commercial turnover in incomplete markets. The second influence on the probabilities of interaction in the Huff-model is the distance between source and destination. The mostly used euclidian distance is based on the unrealistic assumption of an isotropic surface. Therefore it is proposed to use a digital network of roads channelling the traffic flows on their shortest paths. If such a network is not available, it may be derived from a TIN which must be slightly modified in respect to show more realistic features of the traffic routes. Solutions are given by algorithms II and III. It is concluded that Economic Geographers should concentrate just in one part of their work on the classical theories of Lösch and Christaller, particularly on his Dynamics of the Central Place System. Then they may give New Economic Geographers a lead, because they have not yet comprehended these parts of the classical theories.