An Empirical Model of Urban Spatial Development

This paper examines land development patterns by using an integrated approach that combines Geographic Information Systems (GIS) and spatial econometric models. It presents evidence that land development sites are affected by each other, and that any attempt to analyze the land development process without consideration of spatial dependence will most likely yield biased parameter estimates, misleading significant levels, and insufficient estimates of the dependent variable. It is found that spatial lagged variables not only improve model performance substantially, but also reveal the presence of clustering land development. This paper demonstrates that incorporation of spatial statistics is made easier with the application of GIS.