The aim of this paper is to study the processes and factors that influence the average land price of municipalities in Austria using statistical models. For this purpose, we use a dataset of 1667 Austrian municipalities. The location is clearly one of the most important factors influencing land prices. Therefore, land price data are spatial data. When modelling spatial data, spatial effects must be taken into account. In the case of land price data this is primarily the effect of spatial dependence. Spatial dependence therefore must be incorporated in the model specification. Model specifications coming from the field of spatial econometrics, especially spatial autoregressive models, and methods from the field of geostatistics, especially kriging methods, are able to account for spatial dependence.

By comparing these spatial model specifications with classical non-spatial model specifications, one can clearly show that the model-fit can significantly be increased by spatial model specifications. This shows that the process that generates the land price is a spatial process.