SERIE RESEARCH mEmORMIDR A SYNTHESIS BETWEEN MACRO AND MICRO MODELS IN SPATIAL INTERACTION ANALYSIS, WITH SPECIAL REFERENCE TO DYNAMICS

This paper deals with the correspondence between macro and micro spatial interaction and choice models, in both a static and a dynamic context. After a review of recent literature, special attention is devoted to the interpretation of aggregate spatial interaction models (both the conventional gravity or entropy models and the more general theory of movement devëloped by Alonso) in the framework of stochastic discrete choice analysis, in particular multinomial logit models. It is demonstrated that both types of spatial interaction models are in agreement with the multinomial logit model originating from random utility theory. In addition, it is shown that the Alonso model has also a correspondence with a nested logit model. In the paper special emphasis is placed on dynamic spatial interaction analysis. First, a dynamic entropy model formulated as an optimal control model is devëloped, in which also feedback effects from the supply side are incorporated. This leads to a generalized dynamic spatial interaction model, which is again in agreement with discrete choice theory. It is also shown that a dynamic version of the Alonso model can be interpreted in an analogous way.