Hierarchical spatial interactions in the Aveiro-Ílhavo housing market Retrieving W from data under an asymmetrical assumption

Space is an important element for understanding housing values: it assimilates the complexity of locational preferences and translates them through market price dynamics. In housing price modeling, space tends to be incorporated through the lens of spatial heterogeneity and spatial dependence, which need to be explicitly modeled. The scientific debate on the field has been encouraging more relaxed assumptions regarding the specification of socio-economic-spatial relations, with a data-driven (ex-post) W specification gaining importance over the years.

This work draws from theoretical and empirical insights on spatial diffusion, particularly the so-called ripple effects in housing market prices, combining them with an approach to estimate W from data. It applies a strategy to retrieve a spatial dependence structure under the minimal assumption that they are hierarchical, using a Monte Carlo simulation and an empirical application to a local housing market in Portugal (Aveiro-Ílhavo). It shows this technique's feasibility in producing estimates of spatial dependence at the local scale to identify complex but meaningful territorial interaction structures. The revealed spatial interactions are compatible with the notion of ripple effects and also surpass geographical proximity.