Spatial dependence in commercial property prices: micro evidence from the Netherlands

Following a hedonic framework, this paper constructs various transaction-based commercial property price indicators for the Netherlands. Using quarterly data from the Investment Property Databank (IPD), the analysis covers a total of 10,000 listed properties over the period 2001-2011. The study contributes to the empirical literature by introducing a spatial econometric methodology into a hedonic framework, via a spatially lagged explanatory variable (spatially lagged valuations per square metre). The results provide significant evidence of the presence of spatial dependence in unit valuations in all sub-sectors of the commercial property market, namely retail, office, industrial and residential. Accordingly, high (low) priced commercial properties tend to be geographically clustered rather than randomly distributed over space. The comparison of the alternative transaction-based indices shows a systematic upward bias in the baseline transaction-based indicator that relies solely on prior appraisals. In addition, compared to the baseline indicator, the spatially augmented transaction-based price indicator appears to fluctuate less and is more robust to small sample sizes. These results are robust for alternative spatial weights matrix specifications.