44th Congress of the European Regional Science Association: "Regions and Fiscal Federalism", 25th - 29th August 2004, Porto, Portugal
This paper describes the research approach and first empirical results of the estimation of discrete choice models that describe the location decision of moving firms. The model is based on random utility theory and features systematic choice sets to account for the choice context at the highest level of spatial detail (address-level). Firms are analysed categorised to their mobility profile. These mobility profiles are homogenous groups of firms with similar mobility characteristics that are a priori assumed. The models are tested on an extensive revealed preference dataset with firm migration observations in South Holland. To avoid correlations between variables a variety of composed accessibility variables have been constructed that describe the distances to the physical infrastructure or that are an aggregated form of potential accessibility. The location attributes of alternatives have been completed with the business environment type and the rental level. The results are first of all valuable for the development of a simulation model for firm location but the empirical results also yields insight into the spatial behaviour and location preference of firms. Although further research is necessary, the presented addresses some challenges in modelling the spatial behaviours of firms in an urban environment. Therefore the presented approach holds seems valuable for the development of a simulation model for location decisions of moving firms and offers good possibilities for future research.