Abstract :
[en] Inspired by the No Net Land Take (NNLT) or Zero Net Artificialization (ZAN) doctrine formulated by the European Commission (2011, p. 19), the Walloon Government (2019, pp. 46, 99) plans to drastically limit soil artificialization. At the same time, the Walloon Government also outlines significant ambitions regarding housing production, aiming to meet the increasing demand (Walloon Government, 2019, p. 46).
Feedback from European countries where the concept of the compact city has been implemented for many years shows a trend towards liberalization of the housing market which, coupled with land restriction measures, leads to an exacerbation of the pre-existing problems of housing affordability, such as access to property, social exclusion, spatial injustices and segregation (Bibby et al., 2020; Cavicchia, 2021). Some studies have also shown the major influence of rising land values on housing prices and the variability of this trend depending on the context (Knoll et al, 2017; Reusens & Warisse, 2018). In addition, most of the tools available to promote affordable housing, such as land value capture, require a good understanding of dynamics of the prices of the real estate market before, during, and after the implementation of a policy or development.
These literature results led us to the conclusion that it was necessary to have a clear picture of the dynamics already in place to finally answer these two main questions :
- What would be the impact of limiting land supply on housing affordability in Belgium?
- What is the impact of land on the evolution of real estate values in Belgium?
This involves, among other things, a detailed modeling of land and real estate values to establish the respective shares of land and real estate values in the housing price increase and to objectify, based on available urbanizable land, whether a decrease in land supply has led to an increase in land values and whether further supply restrictions could have the same impacts.
We therefore decided to use Multiscale Geographically Weighted Regression (MGWR) (Oshan et al., 2019) in order to model residential real estate and land prices across the entire Belgian territory, using sales data from 2008 to 2019. The results demonstrate the effectiveness of these models in addressing the problems of heterogeneity and spatial segmentation of real estate markets associated with classical models, especially in territories as vast as Belgium. By allowing the variation of regression coefficients in space, these models capture the significant spatial variability of the influence of certain variables such as land area on price. The multiscale aspect of the model also helps to nuance these variations since some variables indeed have a local effect (e.g., land area), while others have a more global effect (e.g., income level).
Inspired by the No Net Land Take (NNLT) or Zero Net Artificialization (ZAN) doctrine formulated by the European Commission (2011, p. 19), the Walloon Government (2019, pp. 46, 99) plans to drastically limit soil artificialization. At the same time, the Walloon Government also outlines significant ambitions regarding housing production, aiming to meet the increasing demand (Walloon Government, 2019, p. 46).
Feedback from European countries where the concept of the compact city has been implemented for many years shows a trend towards liberalization of the housing market which, coupled with land restriction measures, leads to an exacerbation of the pre-existing problems of housing affordability, such as access to property, social exclusion, spatial injustices and segregation (Bibby et al., 2020; Cavicchia, 2021). Some studies have also shown the major influence of rising land values on housing prices and the variability of this trend depending on the context (Knoll et al, 2017; Reusens & Warisse, 2018). In addition, most of the tools available to promote affordable housing, such as land value capture, require a good understanding of dynamics of the prices of the real estate market before, during, and after the implementation of a policy or development.
These literature results led us to the conclusion that it was necessary to have a clear picture of the dynamics already in place to finally answer these two main questions :
- What would be the impact of limiting land supply on housing affordability in Belgium?
- What is the impact of land on the evolution of real estate values in Belgium?
This involves, among other things, a detailed modeling of land and real estate values to establish the respective shares of land and real estate values in the housing price increase and to objectify, based on available urbanizable land, whether a decrease in land supply has led to an increase in land values and whether further supply restrictions could have the same impacts.
We therefore decided to use Multiscale Geographically Weighted Regression (MGWR) (Oshan et al., 2019) in order to model residential real estate and land prices across the entire Belgian territory, using sales data from 2008 to 2019. The results demonstrate the effectiveness of these models in addressing the problems of heterogeneity and spatial segmentation of real estate markets associated with classical models, especially in territories as vast as Belgium. By allowing the variation of regression coefficients in space, these models capture the significant spatial variability of the influence of certain variables such as land area on price. The multiscale aspect of the model also helps to nuance these variations since some variables indeed have a local effect (e.g., land area), while others have a more global effect (e.g., income level).
