With the increasing need for urban green space in urban area to improve the climate adaptation of cities, the spatial planning of residential land-use faces new challenges. Therefore, land-use allocation models offer a useful tool to shed light on the trade-offs and generate suitable solutions for housing allocation problems. A critical prerequisite for housing allocation models is, however, that the value-function is specified such that it accurately represents buyers’ willingness-to-pay for dwelling and location characteristics in the housing market. Hedonic price analysis is the predominant method to estimate willingness-to-pay values based on housing transaction data. Due to high correlations between spatial factors, however, the ability to identify the parameters of spatial factors involved in such value functions is limited. The objective of this study is to apply an alternative method to accurately measure households’ preferences of housing location and its neighborhood characteristics that is based on a stated choice experiment.

In this paper, we present the results of a stated choice experiment that we developed for this purpose. The population consists of homeowners (households) in middle-sized to large-sized cities in the Netherlands. The experiment consists of two parts to measure preferences for neighborhood characteristics and for accessibility of urban amenities respectively. The price of the dwelling is an attribute in both experiments so that a single discrete choice model can be estimated based on the pooled data from the two experiments. The experiments are implemented in an on-line survey and data is collected for a large national sample of homeowners in the Netherlands. This study will provide quantitative insight into homeowners’ preferences (willingness to pay) regarding spatial characteristics of a dwelling. By doing this, we obtain an empirically estimated housing land-use allocation model. This model offers a tool to municipalities and housing developers to optimize urban housing development taking into account financial, climate, and social objectives (match demand and supply of housing). In the paper, we describe the design of the experiments, the data collection, the results of a loglikelihood estimation and we show how the estimation results can be used to specify a state-of-the-art housing allocation model.