In France, since many years, urban disparity based on income is increasing in most metropolitan areas (INSEE 2023). This urban disparity refers to the unequal distribution of social groups in the urban space (Rhein 1994). This phenomenon is explained by 2 main factors: first, since the financialization of the French real estate market in the 90’s (Nappi-Choulet 2013), many households and investment funds are turning to real estate for investment, as it is presented as a safe haven, particularly luxury real estate. This has led to a significant concentration of property wealth by a small number of households, particularly in bourgeois districts (INSEE, 2021). Second, we observe an “entre-soi” phenomenon, particularly within the upper social classes (Préteceille 2006). The location of the housing became central for the upper social classes, to the point of becoming the most important characteristic in the housing choice process (Bouzols 2019). These social characteristics of luxury housing are not well documented in the real estate literature, and not quantified.

The marketing literature brings elements to determine whether a product is considered as luxury or not. Based on the conceptual framework of luxury of Bachmann, Walsh, and Hammes (2019), which details different constructs of the luxury product , we will pose the following research question: is the social dimension of luxury the most important dimension in luxury real estate perception? Is location a luxury?

To answer this question, we will develop a method in 3 steps. First, based on advertisement data, we will evaluate the proportion of properties that are posted on luxury ad sites in each municipality. This will determine if the municipality is perceived as luxury location by the real estate agent or not. Second, we will implement a mass appraisal model following XGBoost algorithm from an exhaustive transaction database (DV3F) to estimate the value of the whole stock of dwellings in Île-de-France (from land declaration database). Finally, we will develop spatial econometrics models (GWR, SDM) based on the estimated database from land declaration, to determine which factors determines what is a luxury housing. We run these models including each municipality defined as luxury by real estate agents.

The implications of this work are two-fold: first, it will bring the conceptual model of Bachmann, Walsh, and Hammes (2019) in the poor luxury real estate literature: is luxury real estate only a question of social construct? This conceptual model will be modified according to the results. Second, it raises questions about growing spatial segmentation and “entre-soi” dynamics.