Spatiotemporal dynamics of housing development: the contribution of spatial analysis methods to the understanding of new housing prices in the Metropolises of Lyon and Toulouse (2005-2019) - Archive ouverte HAL Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Spatiotemporal dynamics of housing development: the contribution of spatial analysis methods to the understanding of new housing prices in the Metropolises of Lyon and Toulouse (2005-2019)

Résumé

Property markets can be seen as social constructs. In this area, the logic of property pricing is particularly resistant to classical microeconomic analysis (volume of demand does not always reduce price levels, price levels are relatively insensitive to production costs, etc.). In the case of the new housing market, the selling price is determined by the choice of developers. Part of their activity is to set the selling price of their dwellings in order to maximise their margin (Bardet, Coulondre and Shimbo 2020), a calculation that can be compared to a compromise between selling high and selling fast. The hypothesis underlying this work, inspired by previous qualitative work (Ausello, 2022; Le Brun, 2023), is that the way developers adjust the selling prices of their dwellings to their characteristics is sensitive to the housing market situation and dynamics on the one hand, and to their location on the other. The aim of this paper is to test this hypothesis using spatial analysis methods. It implies being able to objectify, according to the phases of the economic situation, the role played by the geographical context in the price setting rules of housing developers. This hypothesis is tested on two sites, the metropolises of Toulouse and Lyon, for the period 2005-2019. It is based on private databases produced and made available by the company Adéquation. These databases, which are georeferenced (XY coordinates) and fully disaggregated (1 line = 1 dwelling), provide information on the housing production of private property developers and are considered almost exhaustive (Juillard, 2019). On the basis of these data, we have constructed databases of buildings (1 line = 1 building) marketed by private developers between 2005 and 2019 in the Toulouse and Lyon Metropolises. An initial analysis of these databases makes it possible to divide the period under study into several phases for each location. The periodisation adopted has attempted to associate each phase with coherent trends in terms of sales volumes and price variations. On this ground, our protocol combines three spatial analysis methodological tools: multiscale geographically weighted regression (MGWR), regionalisation and multilevel modelling. Our method follows the work of Marco Helbich et al. (2014), who proposed an empirical segmentation of real estate markets for the Austrian case, and that of Thierry Feuillet (2021) on the hybridisation of local and contextual spatial models in health geography. The protocol is divided into four steps: [1] We first apply a hedonic price model for each phase of each plot, where the price per square metre of each building is explained by four variables: distance from the centre, whether or not the building is located in a ZAC (urban development zone), the proportion of assisted home sales, and the proportion of studios and 1-bedroom units in each building. We justify the use of a geographically weighted regression method by measuring the spatial autocorrelation of the residuals of this first (non-spatial) model. [2] We use an MGWR to provide a continuous exploration of the spatial heterogeneity of the statistical relationship between price and its determinants. [3] We carry out a regionalisation (SKATER method) on the basis of the coefficients assigned to each building by the MGWR. The relevance of this regionalisation is verified by comparing it with a declarative sub-market segmentation (produced by Adéquation on the basis of discussions with housing developers). [4] We use multi-level modelling to calculate the proportion of total price variance explained by each submarket segmentation (ICC). The result objectifies the degree of spatial mimicry by developers for each location and phase. This complex methodological protocol makes it possible to draw theoretically significant conclusions on the spatial and temporal sensitivity of new housing. In the case of Lyon, the weight of the spatial context follows 'pro-cyclical' variations: the variance explained by sub-market segmentation is higher in the 'high' phases of the cycles (high volumes, rising prices) than in the low phases. This suggests that during low phases, developers lose confidence in the importance of different local contexts. In the case of Toulouse, the result is the opposite: the variance associated with segmentation is "counter-cyclical". In this configuration, we therefore hypothesise that, when faced with a fall in sales volumes, developers opt for cautious strategies and are more likely than usual to follow the prices offered locally by competing developers. Conversely, periods of recovery or expansion seem to encourage more adventurous pricing strategies. This opportunistic approach may be due to the centrality of tax exemption to property development in Toulouse. References Ausello, N., 2022, « Technologie et usage de la carte foncière : vers une normalisation de la production urbaine ? », Mappemonde, n°133. Bardet, F., Coulondre, A., Shimbo, L., 2020, “Financial natives: Real estate developers at work”, Competition & Change, vol. 24, no 3 4, p. 203 224. Feuillet T., 2021, Les effets de contexte en géographie. Des fondements théoriques à la modélisation statistique, HDR, Université Lyon 3 Jean Moulin. Juillard, C., 2019, Le tournant numérique des données immobilières, research report iread, LIFTI, PUCA, Urbanics. Helbich, M., Brunauer, W., Vaz, E., & Nijkamp, P., 2014, “Spatial Heterogeneity in Hedonic House Price Models: The Case of Austria”, Urban Studies, vol. 51, n°2, p. 390-411. Le Brun, P., 2023, « L’expert·e et l’opérateur·ice : produire et vendre des études de marché à des promoteurs immobiliers en France », Métropoles, n°32.
Fichier non déposé

Dates et versions

hal-04648907 , version 1 (15-07-2024)

Identifiants

  • HAL Id : hal-04648907 , version 1

Citer

Nicolas Ausello, Pierre Le Brun. Spatiotemporal dynamics of housing development: the contribution of spatial analysis methods to the understanding of new housing prices in the Metropolises of Lyon and Toulouse (2005-2019). 23rd European Colloquium on Theoretical and Quantitative Geography, Research Centre for Territory, Transport and Environment of the Universities of Coimbra and Porto, Sep 2023, Braga, Portugal. ⟨hal-04648907⟩
0 Consultations
0 Téléchargements

Partager

Gmail Mastodon Facebook X LinkedIn More