Derivation of predictor variables for spatial explicit modelling of ‘urban shrinkage’ at different scales

Demographic change has become a major topic for Central, Eastern and Western European cities and urban regions which will be on a "non-growth or decline path" for at least the coming 25 years. Growing and shrinking communities will exist simultaneously side by side within densely urbanised regions. The trend towards further suburbanisation and dispersion observed in the 1980s in Western Europe and the 1990s also in Central (Eastern) Europe accompanying the transition process will decrease without being balanced by a sustained re-urbanisation. This urgently requires rethinking on the growth-oriented approaches of urban development and alike modelling. They must be enhanced by a "decline or shrinkage paradigm." The paper focuses on the discussion of predictor variables that explain urban shrinkage and residential decline in form of residential vacancy. In the first part the phenomenon of urban shrinkage will be embedded into the historic and current European urban development, most of all looking at examples of the demographic situation in Eastern Germany. In a second part, the paper expands briefly on discussing obvious spatial effects of ‘non-growth’ and ‘shrinkage’ in the urban landscape. The major part of the paper concentrates on the identification and testing of predictor variables that explain the quantity and spatial distribution of residential vacancy. The statistical analysis proves that outmigration, share of aged people, unemployment rate and share of social welfare recipients as social variables as well as spatial pattern of urban structural type, distance to the urban (sub-)centers and to urban green explain the spatial distribution of residential vacancy in the city of Leipzig. The paper ends with some initial ideas on the integration them into a GIS-based spatial explicit rule-based model.