A Comparative Perspective on Urban Spatial Housing Market Structure: Some More Evidence of Local Sub-markets Based on a Neural Network Classification of Amsterdam

Abstract

When analysing the spatial housing market structure of urban areas, a frequently arising question concerns the relevant criteria for housing market segmentation: is it the transaction price or related to other, socioeconomic, demographic and physical features of the location? In this study, two neural network techniques (SOM, LVQ, Kohonen) are used for identifying sub-markets within Amsterdam, The Netherlands. Because of the inductive nature of the modelling, any theory has to be understood in an open sense: generalising the principles of classification to another context to enable elaboration of institutionally sensitive housing market theory. These findings are therefore compared with earlier findings from another urban housing market-namely, Helsinki, Finland. The comparison shows that, while the price alone is an insufficient criterion for both markets, Amsterdam is more fragmented than Helsinki.