The Internal Socio-Economic Polarization of Urban Neighborhoods: The Case of the Municipality of Nice

Abstract

In continuity with the research on social segregation and the phenomenon of urban gentrification, this article examines the cohabitation patterns of populations with diametrically opposed incomes within the same neighborhood, typically observed in the city center. This phenomenon is defined here as internal socio-economic polarization. It is measured through the combination of two original indexes (poverty and wealth indexes) constructed based on income deciles per consumption unit for the year of 2017. The analysis focuses on the municipality of Nice, characterized by a low demographic dynamic, a relative concentration of seniors, and a strong tourist attractiveness, particularly in the highly polarized neighborhoods that occupy almost the entire city center. This study is complemented by a principal component analysis summarizing the characteristics of the population and housing stock in the neighborhoods of Nice. The main objective of this research is to identify and locate polarized neighborhoods within the urban context of Nice, to analyze the distinctive traits of their population and housing stock, and, finally, to highlight potential trends in the population’s socio-economic status. Moreover, the economic trajectories of polarized neighborhoods, in connection with their population and housing characteristics (such as the secondary use of a portion of the housing stock, often low-quality old buildings, social housing, and the overrepresentation of retirees), help explain the forms of socio-economic polarization observed in these neighborhoods (such as the indications of gentrification, unfinished gentrification, and sustainable cohabitation).

1. Introduction

1.1. Urban Social Segregation: An Extensive Term to Characterize the Urban Space

Since the 1980s, the term socio-spatial segregation has been frequently used in France in public debates and scientific works, notably in geography and urban sociology. The concept was first proposed by the Chicago school in the 1920s to understand the recomposition of urban space, particularly areas grouping populations of the same social, ethnic, and racial origins. The Chicago school interprets this phenomenon through the processes of domination–invasion–succession, reflecting individual behaviors and the elements of urban territorial structuring (Decamps 2009). The use of the notion of segregation to characterize today’s French urban space no longer corresponds exactly to the notion mobilized by the Chicago school in 1920. Its current more polysemous meaning (Grafmeyer 1994) concerns not only the physical separation that may be observed but also the processes that have generated this separation and those that are its consequence (Brun 1994; Grafmeyer and Authier 2019; Frantz 2010). To characterize this broader set of processes, most current definitions describe urban social segregation as the spatial expression of social relations. This is referred to as the “social division of space” (Roncayolo 1994) or the “social differentiation of urban territories” (Oberti and Préteceille 2016; Grafmeyer and Authier 2019). The term was initially used to designate neighborhoods accumulating social problems, often corresponding to neighborhoods targeted by urban policy generally located on the outskirts of large French cities. Some studies have also analyzed affluent neighborhoods as another facet of social segregation (Pinçon and Pinçon-Charlot 1989; Zalio 1999; Grafmeyer 1992). Urban social segregation is most often interpreted according to two theoretical approaches. The Marxist approach interprets social segregation as the spatial expression of the reproduction of the socio-economic structure. The marginalization of disadvantaged populations is indicated by the economic development and public policies in urban planning (Préteceille 1995; Lehman-Frisch 2009; Frantz 2010). The other approach is structured around households’ preference for a homogeneous neighborhood in cultural and socio-economic terms, a preference that arises not only from the privileged classes (Pinçon and Pinçon-Charlot 2014) but also from the working classes (Lehman-Frisch 2009). Often, the intertwining of different aspects of socio-spatial segregation necessitates the superimposition of the two theoretical interpretations (Madoré 2015; Lehman-Frisch 2009). These two theoretical interpretations find their extension in the phenomenon of “touristification” (Gotham 2005) more recently observed in large metropolises where the housing stock intended for the temporary accommodation of visitors tends to develop and add to an already growing set of secondary and occasional residences.

1.2. Internal Social Polarization of French Urban Spaces

While the neighborhoods targeted by urban policy and affluent neighborhoods are considered the most striking and frequent configurations of segregative phenomena in France, there are, however, other forms of social inequalities within a city that do not correspond to this dualization of space (Oberti and Préteceille 2016; Brun 1994). The deployment of various processes within urban space shapes the geographical diversity of the city’s social composition, producing spaces where socially different groups coexist, particularly in the city center.

The aim of this article is to analyze, within the city of Nice, these situations here referred to as the internal socio-economic polarization of neighborhoods, as well as to identify their population characteristics and real estate stock. The emphasis on the opposition between “rich” and “poor” neighborhoods can distort the understanding of the overall socio-urban fabric. Therefore, this study of these socio-economically polarized neighborhoods aims to shed light on these under-researched cases within the field of urban geography. To sum up, this article seeks not only to identify these polarized neighborhoods and their characteristics but also to analyze the evolution of household income distributions in these areas, in order to position the neighborhoods within broader socio-spatial processes.

The Different Manifestations of the Geographical Coexistence of Social Groups in City Centers: Between Life Course Transitions and Spatial Transformations

The local configurations and specificities of each French urban context create a diversity of situations that do not correspond to a global and uniform trend driving the organization and functioning of cities (Maloutas and Spyrellis 2019; Oberti and Préteceille 2016; Bourdin 2008). The unique political, economic, social, and urban history of each city shapes the distribution of socio-economic groups within its space, resulting in both segregative forms and forms of coexistence of different populations (Bouzouina 2007). Thus, the city center presents a great geographical diversity of social combinations, the result of various urban, sociodemographic, and economic processes with rich neighborhoods and poor neighborhoods, as well as neighborhoods with a heterogeneous internal socio-economic composition (Oberti and Préteceille 2016; Préteceille 2006a; Bouzouina 2007; Van Criekingen and Decroly 2003). We can observe not only rich neighborhoods and poor neighborhoods but also neighborhoods with an internally heterogeneous socio-economic composition (Bouzouina 2007). These heterogeneous neighborhoods concentrate both low-income and affluent populations (Préteceille 2006a; Oberti and Préteceille 2016; Authier 2003; Bouzouina 2007). This coexistence of socio-economically distant households does not always imply great social mixing. This co-presence may not indeed lead to truly profound social relations but rather to a more passive coexistence. Social hierarchies continue to exist, reminding us of the absence of equivalence between social interaction and spatial proximity (Tissot 2011; Lenel 2011; Authier and Lehman-Frisch 2013; Bourdin 2008; Giroud 2015; Oberti and Préteceille 2016; Préteceille 2006b; Bouzouina 2007; Roncayolo 2001; Audren et al. 2016; Collet 2015; Park 2004).

The presence of high-income populations in the city center is explained by the concentration of urban amenities and its symbolic image (Bridge 2006; Zukin 1998; Brueckner et al. 1999), but the proximity to the job market and the support networks offered by centrality also attract low-income populations. These heterogeneous neighborhoods within the same district reflect differentiated preferences and stages of life (Tissot 2011; Lenel 2011; Bouzouina 2008). This is particularly the case for young adults in professional and family transitions who temporarily occupy central neighborhoods in a rental market of highly variable quality while waiting for professional and family stability, leading a significant portion of them to settle in a family housing neighborhood, often in the suburbs (Van Criekingen and Decroly 2003; Bourdin 2008; Lenel 2011). There is therefore a constant renewal of this type of population, largely linked to the presence in the city center of a housing stock mostly composed of small apartments for private rent (Van Criekingen and Decroly 2003; Authier 2003). These young adults in professional transition do not constitute a uniform group because, on the one hand, they do not all have the same qualifications, which strongly differentiates their salaries, and, on the other hand, they are not all at the same stage of this professional transition, which again determines their financial resources. The relative specialization of the housing stock in the central districts of large cities thus makes it a space where young adults in transition concentrate, who, due to the diversity of their previous paths and their progress in this transition, generate a certain socio-economic heterogeneity among the inhabitants of the same neighborhood.

At the same time, most large cities began urban renewal policies as early as the 1970s, but much more clearly during the 1990s and 2000s, particularly in initially poor neighborhoods where the housing was often degraded. Similarly, following in the wake of “pioneer” social groups, such as artists or gays (Giraud 2010), some initially poor neighborhoods in large cities were invested in by new populations with higher financial means. These socio-spatial transformation processes, which can overlap, are generally considered through the notion of gentrification. Ruth Glass defined this notion of gentrification in the 1960s, and this term now encompasses a wide range of urban changes (Bourne 1993; Van Criekingen and Decroly 2003). Always involving the settlement of materially wealthier populations in initially less socio-economically advantaged urban environments, gentrification results, at least temporarily, in the cohabitation of households with very unequal resources. In this context, the internal socio-economic polarization of urban neighborhoods would just be a stage of gentrification leading to the eviction of low-income populations. While many neighborhoods in large French cities can illustrate this process, it is extremely simplistic to consider all situations of the cohabitation of different socio-economic groups within the same neighborhood as the intermediate phase of an inexorable gentrification (Giroud and Ter Minassian 2016; Oberti and Préteceille 2016; Authier 2003; Bouzouina 2007; Grafmeyer and Authier 2019). Indeed, comparing the gentrification processes initiated in the neighborhoods of different cities and/or within the same city reveals quite diverse evolutions (Van Criekingen and Decroly 2003). These gentrification processes are not always linear and can lead to different situations. In some cases, the replacement of low-income populations with higher income populations appears almost complete, while, in other cases, social changes can lead to more or less lasting internal social polarization (Giroud and Ter Minassian 2016; Oberti and Préteceille 2016; Authier 2003; Bouzouina 2007; Jeanmougin and Bouillon 2016). Thus, the massive but dispersed presence of degraded buildings difficult to transform, the presence of social housing, and the hyperspecialization of housing into very small apartments as a result of a large student population, can limit urban transformation and favor the retention in the neighborhood of economically disadvantaged households who then cohabit with newly settled wealthier households. The internal socio-economic polarization of these neighborhoods then constitutes a lasting state, even though the changes to part of the housing stock and the redevelopment of public spaces have led to the arrival of wealthier classes. The terms sporadic, incomplete, or even marginal gentrification are used to describe these neighborhoods where high-income households live alongside pockets of poverty (Jourdan 2008; Bacqué and Fijalkow 2006; Authier 1995, 2003).

Finally, the example of the countries severely affected by the 2008 economic crisis also shows (Maloutas and Spyrellis 2019) that, in some neighborhoods, internal socio-economic polarization can also be a transitional stage of pauperization. Although the intensity of the economic and social crisis experienced by France is far from comparable to what Greece or other countries in Europe, particularly in the South, faced, these processes of potential impoverishment in urban neighborhoods undergoing gentrification have also been observed (Jeanmougin and Bouillon 2016).

1.3. The Context of Nice

Nice is a metropolitanized tourist area (Christofle and Hélion 2018) characterized by particularities related to its topography. On one hand, the physical constraint of the terrain influences its urban development. On the other hand, it does not follow the general socio-spatial trends of major French cities (Schwabe 2007). This singularity, partly due to its Mediterranean character, makes it particularly interesting to study from the perspective of socio-economic polarization within urban neighborhoods.

Thus, the municipality experiences a strong tourist attraction, resulting in a significant proportion of secondary residences within its housing stock. This leads to competition in the housing market between temporary and permanent populations. At the same time, moderate demographic growth is accompanied by a significant proportion of elderly people compared to what is observed in the central cities of most major French urban areas.

1.3.1. Demographic Diversity Within the Nice Côte d’Azur Metropolitan Area: Relative Population Stagnation in Nice and Demographic Dynamism in the Suburbs and Hinterland

Nice is located at the foot of the Alps and on the Mediterranean coast, near the French–Italian border and the Principality of Monaco. Constrained by the terrain, the development of the Nice area follows the valleys (Decoupigny et al. 2007; Fusco and Scarella 2008b). It is the fifth most populous municipality in France, behind Toulouse, and the second in the Provence–Alpes–Côte d’Azur region, after Marseille. In 2017, the Nice Côte d’Azur metropolitan area (NCA) comprised 49 communes, with a population of 537,999 and a surface area of 1400 km². Its average density, which is relatively moderate, is 367 inhabitants/km² (INSEE 2020), combining the very dense coastal municipalities with the very low-density mountain municipalities that include the Mercantour National Park. Unlike other French urban areas (Bordeaux, Lille, or Paris), the municipality of Nice has a predominant weight within the metropolis. It hosts more than half of the metropolis’s population, nearly 63% or 340,017 inhabitants (Figure 1).

Over the decades within the Nice Côte d’Azur metropolitan area, different areas have experienced varied demographic trajectories. Since the late 1960s, the population growth of the urban area has been less sustained than the suburban and hinterland areas, which have proven to be more dynamic (Figure 2). In the 1990s, the metropolitan population continued to grow at a slower pace, mainly due to the other municipalities, as the municipality of Nice experienced relative stagnation. In the early 2000s, the metropolitan population saw a slight increase marked by less contrasting demographic dynamism between its different parts. Between 2007 and 2017, the population of Nice decreased very slightly, following the overall trend of the Nice Côte d’Azur metropolitan area, while the population of the other municipalities in the metropolitan area varied little.

1.3.2. Relative Concentration of Seniors and a Slight Deficit of Young Adults

In 2017, the population of the municipality of Nice maintained a singular demographic structure, characterized by a relatively high concentration of people aged 65 and over (Conseil Communautaire 2010; Schwabe 2007), while remaining well below the levels observed in certain small French seaside towns such as Saint-Raphaël, Arcachon, Royan, or La Baule. This trend has resulted in a generational imbalance, with nearly 24% of the population aged 65 and over, while those under 18 represent about 19%. Compared to other large French cities, Nice shows a notable deficit in the 18 to 25 age group and, to a lesser extent, the 25 to 30 age group (Figure 3). This deficit of young adult populations can be explained by the lower concentration of higher education institutions compared to other large French cities, such as Lille and Toulouse.

Several analyses and studies converge to describe a significant concentration of elderly individuals along the coastal front of the metropolis, particularly in the city center of Nice (Conseil Communautaire 2010; Chauvot and Pougnard 2014). Due to the small size of their households, they predominantly reside in small-sized accommodations within the collective rental housing stock in the urban core of Nice (Conseil Communautaire 2010). The proximity of the services offered by the urban centrality are also likely to contribute to this presence of elderly people in the city center, who often coexist with low-income populations, including isolated unemployed individuals (Conseil Communautaire 2010; INSEE 2014; Chauvot and Pougnard 2014). Families more frequently reside in the periphery in larger accommodations that better meet their needs. The INSEE’s infra-communal mapping analysis of the Nice Côte d’Azur metropolis in 2011 reveals an overrepresentation of individuals aged 60 and over in the coastal part of one of the zones where the unemployed are overrepresented (Figure 4) (INSEE 2014). In this central coastal zone, the average fiscal income varies considerably from one micro-area to another.

1.3.3. Touristification of the Residential Space in Nice: A Trend Towards Secondary Residences with High Housing Vacancy

The urban landscape of the Nice Côte d’Azur metropolitan area, especially its coastal part, is influenced, like many other territories in the Mediterranean basin, by the tourist economy dating back to the 18th century, based on heliotropism (Christofle and Hélion 2018; Centi 1993; Goujard 2002). This is evidenced by a pronounced overrepresentation of secondary residences in the housing stock. In 2017, the municipality of Nice counted 31,344 secondary residences, representing 51% of the 61,949 secondary residences in the Nice Côte d’Azur metropolis. These secondary residences are concentrated along the coast, particularly in the collective housing stock of the municipality of Nice, where their condition is better than in the rest of the territory (Christofle and Hélion 2018; Conseil Communautaire 2010; Métropole Nice Côte d’Azur 2016a, 2016b; Métropole Nice Côte d’Azur 2023). They constitute 14% of the housing stock in the city center, compared to 9.5% nationally and between 3% and 5% in other comparable territories1. The competition between the primary and secondary uses of housing in downtown Nice can create tensions in the real estate market, potentially reducing or blocking the availability of primary residences for the local population and likely increasing property prices (Fusco and Scarella 2008a; Scarella 2014; Conseil Communautaire 2010; Métropole Nice Côte d’Azur 2016a). Therefore, it is conceivable that socially disadvantaged populations face difficulties in finding housing in a tight real estate market.

In 2017, the vacancy rate of housing in the municipality of Nice reached 14%, exceeding the national average (8.3%) and that of other central cities in the major provincial metropolises (ranging from 7% to 9%). This abundance suggests a mismatch between residential and tourist demand and supply (Arrighi 2018). In 2013, 24% of the vacant dwellings in Nice were considered to be in poor condition, largely due to their age (Métropole Nice Côte d’Azur 2016a, 2016b). Despite the General Interest Programs (PIG) from 2008 to 2023 aimed at combating substandard housing, vacancies persisted in 2017 (Figure 5). The high presence of small apartments in downtown Nice can lead to high turnover rates and frictional vacancies (Métropole Nice Côte d’Azur 2016b; Conseil Communautaire 2010). The significant aging of the population in Nice also contributes to the high vacancy rate; in 2013, 56% of the owners of unoccupied homes were over 60 years old (Métropole Nice Côte d’Azur 2016a, 2016b). Thus, in 2017, a vacancy could result from the owner moving to a retirement home. Vacancies may also occur due to the death of the owner, complicating the property’s return to the market by heirs. The complexity and duration of settling estates can lead to varying periods of vacancy before properties are put up for sale, rented out, or occupied by heirs (Conseil Communautaire 2010). Additionally, some of these vacant properties may be used as second homes (Métropole Nice Côte d’Azur 2016a, 2016b). The particularly high vacancy rate also suggests the potential for gentrification, but the deterioration of some of these properties, often associated with their original popular character in certain neighborhoods (Ter Minassian 2009), makes their renovation costly and complex. Therefore, it is possible that some of the vacant properties in poor condition will eventually be offered for rent, contributing to the maintenance of populations with modest resources. This could further foster socio-economic polarization within certain neighborhoods of Nice, potentially leading to their impoverishment.

2. Materials and Methods

The methodology used to classify neighborhoods into 4 categories (“poor”, “rich”, “polarized”, or “other”) is based on the wealth and poverty indexes calculated for each IRIS2 INSEE of the city. This calculation is carried out for the years 2012 and 2017 using the income deciles derived from the FiLoSoFi file (a system on localized fiscal and social incomes), which provides information on the distribution of populations in all the IRIS of France based on their disposable income per consumption unit. The construction method for these poverty and wealth indexes, their theoretical and empirical justification, and their application are presented in a text systematically addressing all French provincial big metropolitan areas (Gripsiou and Bergouignan 2022). The poverty index is obtained through a linear combination of the relative differences using the lower values between the first three deciles of disposable income per consumption unit for each IRIS and the first three deciles of disposable income per consumption unit for all of metropolitan France (Gripsiou and Bergouignan 2022). The wealth index is obtained through a linear combination of the relative differences using the higher values between the last three deciles of disposable income per consumption unit for each IRIS and the last three deciles of disposable income per consumption unit for all of metropolitan France (Gripsiou and Bergouignan 2022). The classification of the IRIS into “poor”, “rich”, or “polarized” is determined based on whether their “poverty” and “wealth” indexes exceed certain thresholds (Gripsiou and Bergouignan 2022). An IRIS is considered “poor” if its poverty index is greater than 4% and its wealth index is less than 4%. This 4% threshold for the poverty index statistically corresponds to an IRIS whose poverty rate is equal to the poverty rate of all of metropolitan France. In other words, an IRIS with a poverty index exceeding 4% has a probability greater than 95% of having a poverty rate higher than that of all of metropolitan France. An IRIS is considered “rich” if its poverty index is less than 4% and its wealth index is greater than 4%. An IRIS is considered socio-economically “polarized” if its poverty index is greater than 4% and its wealth index is also greater than 4%. An IRIS with both poverty and wealth indexes below 4% forms a residual category where the socio-economic situation of its inhabitants closely resembles that observed for all of metropolitan France.

The use of census data (a synthesis of the 2017 population census) allows for the more detailed characterization of a socio-economically “polarized” IRIS in Nice compared to other categories of IRIS in the municipality. The sociodemographic variables utilized primarily characterize residents in terms of their age, number of children, educational attainment, employment status, residential and daily mobility. They also describe housing in terms of type (apartment or house), size, age of the building they belong to, and the status of their occupants (owner, social housing tenant, or private tenant). These characterizations will be considered from a bivariate perspective and then through a multivariate approach using principal component analysis to better reveal the potential factors of internal social polarization within the IRIS areas.

To identify trends in the income distribution within the “polarized” IRIS of Nice, we examine the evolution of wealth and poverty indexes between 2012 and 2017 for the IRIS identified as “polarized” in either of these two years. An increase in the wealth indexes accompanied by a decrease in the poverty index could indicate potential gentrification, with the growth of high-income populations in a neighborhood originally occupied by low-income populations, some of whom may still reside there. Conversely, an increase in the poverty index alongside a decrease in the wealth index could signify a trend towards impoverishment, with a reduction in high-income populations. We may also observe varying levels of internal socio-economic polarization, potentially increasing with the coexistence of populations with diametrically opposite economic resources within the same IRIS, often associated with socio-urban characteristics and heterogeneous buildings.

Regarding the data used in this study, derived from the FiLoSoFi system and the population census, they present certain limitations that should be noted before analyzing the results obtained. One of the main limitations of the fiscal data is that students who are fiscally attached to their parental household, as well as the majority of secondary residents3 and temporary tenants, do not appear in the files for the distribution of disposable income per consumption unit. This may potentially introduce a bias in the interpretation of the internal socio-economic polarization of the neighborhoods. Another limitation of the data mobilized in this study is the absence of micro-data linking income and residential mobility, which would allow for a more in-depth analysis of gentrification. Indeed, for confidentiality reasons, the FiLoSoFi system does not provide this type of data. Furthermore, the proportions of households recently settled in their homes derived from the population census pose interpretative challenges, since these data do not reveal whether the new residents already lived in the same neighborhood or whether they came from another neighborhood within the same municipality. The lack of information regarding the duration of housing vacancies presents another challenge for interpreting the evolution of internal polarization in the studied urban neighborhoods. After this brief overview of the limitations of the data mobilized here, our attention will turn to the constraints associated with the variables and the characteristics linked to the internal socio-economic polarization of the neighborhoods in Nice, which we will address through the discussion in this article.

3. Results

3.1. A City with Pronounced Polarization at Its Core

In 2017, the majority (75%) of the IRIS in the metropolitan area of Nice Côte d’Azur are classified into the “rich” and “poor” categories. Nearly 17% of the IRIS areas exhibit a socio-economic profile similar to that of metropolitan France, with an income distribution slightly more concentrated around the central values. Internal polarization affects only 8% of the IRIS areas, primarily located in the city of Nice, especially in the downtown area and along the coast, forming a continuous area encompassing almost the entire city center (Figure 6).

The “poor” IRIS, representing 46%, are mainly concentrated within the municipality of Nice, forming pockets of poverty. These areas of poverty are particularly pronounced in the IRIS areas bordering the west and east of the municipality of Nice, as well as around the central sector of the municipality, where poverty indexes reach very high levels (Figure 7). A very small portion of the “poor” IRIS (5% of the “poor” IRIS areas) are located in the peripheral municipalities, such as Cagnes-sur-Mer, Carros, Saint-Laurent-du-Var, or Vence.

However, the IRIS considered “rich”, representing 28% of all IRIS, show a more scattered distribution within the metropolitan area of Nice. They dominate in the peripheral municipalities characterized by a low density, such as Carros, Cagnes-sur-Mer, Saint-Laurent-du-Var, or Vence. Within the municipality of Nice, they cluster into four main sectors alternating mainly with the “poor” IRIS from the west to the southeast of the municipality, where the wealth index reaches its maximum level (Figure 8).

The IRIS classified as “other” are fewer in number compared to the “poor” and “rich” IRIS, and they are scattered throughout the metropolitan area, with a particularly marked concentration in the east and west. The municipality of La Trinité is exclusively composed of “other” IRIS, making it the most homogeneous municipality in terms of income distribution, both between neighborhoods and within neighborhoods.

Due to the relative uniformity of the peripheral space, mainly occupied by “rich” and “other” IRIS, and the absence of “polarized” IRIS within the peri-urban municipalities, our study focuses on the IRIS of the municipality of Nice. The municipality of Nice encompasses all “polarized” IRIS, totaling 14 out of 139 IRIS in 2017.

3.2. Neighborhoods of the Municipality of Nice with Pronounced Internal Socio-Economic Polarization

Although “polarized” IRIS are not very frequent, their geographical distribution shows a marked concentration around Old Nice, near the coastline, with the exception of one IRIS on the outskirts of the city forming the border between Nice and La Trinité. This concentration of “polarized” IRIS fits into the pattern of the other two categories, “poor” and “rich” IRIS, which themselves also form “blocks” within the urban space of the city of Nice. It can be observed that the distribution of wealthy and poor households follows the topography of the area, with the affluent households located on the hills where the density is lower and the low-income households situated along the valleys with a higher density. Thus, the socio-economic geography of Nice is characterized by large areas of socio-economic coherence that repeat regularly, except in the city center, due to significant income disparities within its neighborhoods. The existence of this pattern confirms the hypothesis that the neighborhoods with strong internal socio-economic polarization are located in the urban core of cities and often raises questions about possible ongoing gentrification.

Some characteristics of the housing stock and its residents are closely linked to the internal socio-economic polarization observed in the 14 “polarized” IRIS of Nice in 2017. This correlation is measured using the Cramér’s V coefficient (the V of Cramér4). In the city of Nice, four proportions expressing binary qualitative variables show a correlation with the four categories of IRIS, defined based on the original indexes of wealth and poverty, with a Cramér’s V of 0.20 or higher (Figure 9). These are as follows:

• The proportion of flats belonging to buildings completed before 1945;
• The proportion of secondary and occasional residences;
• The proportion of households renting in social housing;
• The proportion of employed people walking to work.

Among these four variables, three are closely related to the “polarized” IRIS, highlighting specific characteristics of these IRIS with high internal socio-economic polarization. On average, the “polarized” IRIS show higher proportions of flats belonging to buildings completed before 1945, exceeding by 29 percentage points the average measured in “poor” IRIS and nearly 40 percentage points higher than the average in the other two categories of IRIS. Similarly, the “polarized” IRIS exhibit higher proportions of secondary and occasional residences on average, emphasizing their tourist character compared to the other three categories of IRIS. Furthermore, an IRIS with strong internal socio-economic polarization also include higher proportions of employed people walking to work on average, indicating their central location within the municipality. This proximity to the city’s central functions sets them apart from the other three categories of IRIS.

The proportions of households renting in social housing contrast the “poor” IRIS, where this proportion is relatively high on average, with other categories of IRIS. In fact, the proportions of households renting in social housing within “polarized” and “rich” IRIS are equal and significantly lower than the average observed in a “poor” IRIS. The very limited presence of social housing in a “rich” IRIS can be easily explained by the criteria for allocating such housing, which primarily depend on low-income levels. Regarding “polarized” IRIS, the scarcity of social housing reflects more their central geographic position within the municipality, where there are no major social housing complexes. In other words, the coexistence of low-income and high-income populations in a “polarized” IRIS is not due to isolated social housing buildings forming “poor” enclaves amidst relative material affluence. More generally, “polarized” IRIS distinguish themselves from both “rich” and “poor” IRIS without necessarily constituting an intermediate category, but rather delineate their own specificity as urban central areas with older buildings primarily oriented towards tourism.

The principal component analysis (PCA) allows for a multivariate perspective, not just a bivariate one (as previously discussed), to capture the specificities of the population and housing characteristics across different categories of IRIS. This PCA is applied to the wealth and poverty indexes as well as to the full set of proportions of population and housing characteristics in 2017 across the 139 IRIS of the municipality of Nice5.

The histogram of the eigenvalues from the PCA reveals that the first two axes summarize 56% of the statistical information (Figure 10). The first axis, encompassing around 33% of the information, contrasts an IRIS with the characteristics of poor populations, with overrepresentations of inhabitants without diplomas, inactive individuals, workers, immigrants, overcrowded housing, and high poverty index values, against an IRIS with the characteristics of rather affluent populations, with overrepresentations of inhabitants with higher education degrees, managers, intermediate professions, homeowners, and high wealth index values. In other words, it highlights the socio-economic contrasts between urban areas.

The second axis, summarizing 23% of the initial information, is mainly correlated with characteristics related to housing, the daily home-to-work mobility of employed individuals, and residential mobility. It contrasts a central IRIS, characterized by an overrepresentation of small, older dwellings occupied by small households, high residential mobility, a higher frequency of soft daily mobility, and a greater presence of secondary residences, with IRIS areas further from the city center characterized by family-oriented populations and an overrepresentation of large dwellings. This confirms the singularity of downtown Nice and the overrepresentation of secondary and occasional residences within it, a consequence of two centuries of tourist fame. Other variables have relatively high coordinates in the first factorial plane without being specifically associated with either axis, occupying an intermediate position between the two dimensions. In one quadrant of this first factorial plane, the characteristics often relate to precariousness, particularly among young adults (the frequency of unemployment or part-time work and the proportion of 18–24-year-olds) or adults of parental age (proportion of single parent families). In another quadrant of this first factorial plane, the characteristics are more associated with large families occupying social housing (the proportion of families with four children, proportion of social housing, and proportion of those under 18).

When examining the projection of the 139 IRIS on the first factorial plane, it is observed that the “poor” IRIS form a vast cloud extending from the central part of the first factorial plane to the entire area corresponding to the positive coordinates on the first axis (Figure 11). Conversely, the “rich” IRIS are concentrated in the area corresponding to the negative coordinates on the first axis, forming a more restricted cloud. The “polarized” IRIS and the “other” IRIS are arranged along a “band” positioned between the “poor” and “rich” IRIS. The “polarized” IRIS are located exclusively at the level of the positive coordinates on the second axis. The “other” IRIS primarily occupy the part of the cloud corresponding to the negative coordinates on the second axis. Thus, the correlation of sociodemographic characteristics and the indexes of wealth and poverty used to form the four categories of IRIS is confirmed by the relatively distinct positioning of these categories on the first factorial plane.

Analyzing the distribution of the “polarized” IRIS on the first factorial plane (Figure 11), we observe a relatively compact and less extensive cloud compared to the other IRIS categories, indicating a significant correlation with the characteristics related to centrality and tourist function (the overrepresentation6 of small households, small and old apartments, secondary residences, soft daily mobility, and high residential mobility). The “polarized” IRIS exhibit seemingly contradictory traits, a logical consequence of their internal polarization, such as social status (the overrepresentation of higher education graduates) and precarious employment (the overrepresentation of part-time contracts). While the position of the “polarized” IRIS on the first factorial plane shows some homogeneity, one IRIS clearly deviates from the cluster of “polarized” IRIS (Figure 12). This is the “Vinaigrier” IRIS, which stands out due to a very low proportion of secondary residences and, conversely, a significant proportion of social housing. This is particularly notable, because, in the majority of “polarized” IRIS, social housing is under-represented while secondary residences are highly overrepresented. This is evidenced by its position on the first factorial plane but also by its geographical position on the outskirts of the city of Nice, far from the “cluster” of “polarized” IRIS.

Because of its singular character, which may seem contradictory to the trends prevailing during the 2010s in terms of housing costs, socio-economic polarization could be considered a priori as a transitional stage in gentrification, which in the case of Nice could also come under the heading of touristification. With regard to gentrification, we can attempt to verify whether Nice’s “polarized” IRIS are moving in this direction by examining the levels and trends between 2012 and 2017 in the poverty and wealth indexes of “polarized” IRIS in 2017.

3.3. Internal Socio-Economic Polarization and the Gentrification Process

Examining the changes in the poverty and wealth indexes between 2012 and 2017 for the “polarized” IRIS in the municipality of Nice in 2017, we find that none of them experienced a simultaneous decrease in both the wealth and poverty indexes. Although the poverty and wealth indexes of the “polarized” IRIS do not necessarily follow identical trends over time and do not start from the same levels of poverty and wealth, this observation suggests that, from 2012 to 2017, the inhabitants of the “polarized” IRISs did not tend to economically homogenize.

It is notable that only two out of the fourteen “polarized” IRIS in 2017 saw their wealth index decrease between these two years. This observation might suggest that the internal socio-economic polarizations observed in most of the affected IRISs in Nice actually constitute a stage in the gentrification process of the city center. While an increase in the frequency of high-income populations in many of these IRIS is an undeniable reality, it is not certain that these developments will lead to a massive replacement of low-income populations by socio-economically advantaged populations.

In certain IRIS, the increase in the wealth index, even reaching very high levels, does not accompany a decrease in the poverty index or the poverty rate. For example, the “Saleya-Château” IRIS saw its wealth index rise from 0.26 to 0.37 between 2012 and 2017, which in 2017 corresponded to the income per consumption unit of the top 10% of wealthiest individuals, being over 50% higher than the income per consumption unit of the top 10% of the wealthiest individuals in the entirety of metropolitan France. However, alongside this significant growth in the upper end of the income distribution, the poverty index in this IRIS increased from 0.07 to 0.15, which in 2017 corresponded to the income per consumption unit of the bottom 10% of poorest individuals, being more than 20% lower than the income per consumption unit of the bottom 10% of the poorest individuals in metropolitan France, also corresponding to a poverty rate of 21% (compared to 14.1% for the entirety of metropolitan France). In this “Saleya-Château” IRIS, it appears that, instead of gradually transforming into a “wealthy” IRIS, the sociodemographic changes are exacerbating the internal socio-economic polarization.

Two other IRIS (“Victor-Hugo-Buffa” and “Musiciens”) experience similar trends with very high wealth indexes in 2017 (0.32 and 0.47, corresponding to the income per consumption unit of the top 10% of wealthiest individuals, being, respectively, 50% and nearly 60% higher than that of the top 10% of the wealthiest individuals in metropolitan France). At the same time, the poverty indexes also increased in these two IRIS, reaching significant levels in 2017 (0.18 and 0.10, corresponding to poverty rates of 22% and 18%, respectively). Thus, like the “Saleya-Château” IRIS, the “Musiciens” and “Victor-Hugo-Buffa” IRIS are characterized both by the highest internal socio-economic polarizations and by an increase in these polarizations, primarily driven by the upper end of the income distribution but with a poverty frequency that tends to increase.

In some respects, the “Liberti-Albert 1er” IRIS shows some similarities with these IRIS (“Victor-Hugo-Buffa, “Musiciens”, and “Saleya-Château”) characterized in 2017 by the highest levels of internal socio-economic polarization. The wealth index there is very high (0.39 in 2017, indicating a median income per consumption unit of the top 10% of wealthiest individuals, being over 50% higher than that in metropolitan France). However, unlike the three other highly polarized IRIS, it is distinguished by a decrease in the poverty index between 2012 and 2017, which suggests possible gentrification, although this remains a distant prospect given the persistence of poverty in this IRIS (in 2017, the poverty index was still 0.13, with a poverty rate slightly above 20%). These four IRIS are also characterized by a high frequency of secondary or occasional residences, accounting for around 40% of the housing stock, which could increase pressures on the housing market and favor the replacement of low-income populations by affluent populations. However, the situation is more nuanced, as, even when the frequency of poor populations slightly decreases, it remains at a very high level. This apparent paradox might stem from an old, or even very old, housing stock (e.g., the “Saleya-Château” IRIS), which includes a substantial portion of mediocre-quality apartments with a much dispersed ownership structure, making major urban renewal projects difficult. The very high housing vacancy rates (between 11% and 21%) observed in these four IRIS could indirectly and partially indicate a significant portion of poor-quality7 housing. Ultimately, these four IRISs, characterized by a significant and growing presence of high, even very high, income populations, also maintain a substantial proportion of very low-income populations. While this does not constitute an insurmountable barrier to gentrification, this persistence of poverty, reflecting the peculiarities of the housing stock, delays its completion and maintains a very high internal socio-economic polarization that could endure.

Three other IRIS (“Sasserno”, “Jean Médecin”, and “Rossini”) share some similarities with the previous four, notably the persistence of significant poverty (poverty indexes above 0.10 in 2017, corresponding to poverty rates just below 20%, reflecting a slight decline between 2012 and 2017). However, they differ by having lower levels of wealth indexes (around 0.16 in 2017, which corresponds to a consumption unit income for the top 10% that is about 20% higher than that for the entirety of Metropolitan France). Although, in one of these three IRIS, the wealth index significantly increased between 2012 and 2017, the gentrification process, if confirmed, remained at a moderately advanced stage, and the lesser internal socio-economic polarization compared to the four previously analyzed IRIS mainly reflects a lesser skewing towards the upper end of the income distribution. Ultimately, for these three IRIS (“Sasserno”, “Jean Médecin”, and “Rossini”), gentrification does not appear to be an obvious process and is still far from complete. This is likely also the case for the IRIS “Promenade du Paillon”. This IRIS is somewhat of a hybrid between the IRIS “Saleya-Château”, “Musiciens”, “Victor-Hugo-Buffa”, and “Liberti-Albert 1er” and the IRISs “Sasserno”, “Jean Médecin”, and “Rossini”. The level of poverty there is quite comparable and significant (a poverty index of 0.09 in 2017 for a poverty rate of 18%) and the level of wealth is intermediate (a wealth index of 0.24 in 2017, corresponding to a consumption unit income for the top 10% that is more than 30% higher than that for the entirety of Metropolitan France). Here again, gentrification does not appear to be a very dynamic process, despite the skewing towards the upper end of the income distribution, because not only does the level of poverty remain significant but it tends to increase between 2012 and 2017 while the wealth index slightly decreases.

The IRIS “Hôpital-Saint-Roch”, “Baumettes”, and “Eglise-Russe” are those for which the internal socio-economic polarization is least marked, with, in 2017, poverty indexes all between 0.04 and 0.12 (corresponding to poverty rates between 16% and 18% and incomes per consumption unit for the richest 10% of people that are 9% to 15% higher than for the entirety of Metropolitan France). Despite significant increases in the wealth indexes between 2012 and 2017, their still moderate level after this increase seems to indicate that the gentrification process, if it continues, is still far from being complete in these areas.

For most of the IRIS in Nice characterized by strong internal socio-economic polarization, this results from a double distortion of income distribution, significant downward but even more pronounced upward. This stronger upward distortion in income distribution and its tendency to intensify between 2012 and 2017 provide an argument for assuming that gentrification processes are currently under development, while noting that they are still far from being completed due to a significant persistence of poverty, although sometimes slightly decreasing. However, there are two IRIS ("Notre-Dame" and "France-Negresco") where the strong internal socio-economic polarization primarily results from a downward distortion of income distribution, with poverty indices in 2017 around 0.30 (corresponding to poverty rates of approximately 30%) and wealth indices of 0.06 and 0.07 (which means that the income per consumption unit of the wealthiest 10% is 12% higher than the national average in mainland France). In these two IRIS, more so than in other "polarized" IRIS, any potential gentrification (the "Notre-Dame" IRIS, categorized as "poor" in 2012, is classified as "polarized" in 2017 after an increase in the wealth index) would take a long time to unfold given the persistence of significant poverty

Finally, the IRIS “Vinaigrier” stands out clearly from other “polarized” IRIS both in terms of its geographical position within the city of Nice and its sociodemographic characteristics and housing stock, as evidenced by its position on the first factorial plane. This IRIS is one of those with the most significant internal socio-economic polarization, with considerable distortions at both the lower and upper ends of the income distribution (in 2017, the poverty index was 0.19, corresponding to a poverty rate of over 26%, and the wealth index reached 0.28, indicating that the income per consumption unit of the wealthiest 10% was nearly 35% higher than the average for Metropolitan France). This strong internal socio-economic polarization, which increased between 2012 and 2017 with rising poverty and wealth indexes, is structured by the existence of two distinct housing poles within this peripheral family housing IRIS. The internal socio-economic polarization of other “polarized” IRIS located in the heart of the historic city seems to result mainly from the heterogeneous quality of a stock of very old small apartments in a complex urban fabric, where buildings in visibly varying states of repair coexist. In contrast, within the “Vinaigrier” IRIS, the buildings are much newer and include houses (nearly 50% of the housing) that are relatively large and mostly owner-occupied, as well as medium or large apartments occupied by social housing tenants (about one-third of households). These two types of housing structure the socio-economic differences among their residents and generate the internal socio-economic polarization of the IRIS “Vinaigrier”. In this IRIS, perhaps even more so than in the “polarized” IRIS in the heart of the historic city, the nature of the housing stock, with its strong social component, makes it difficult to achieve gentrification, which would lead to the replacement of low-income populations with high-income populations. On the contrary, it is very likely that this very uncommon form of internal socio-economic polarization will persist for a long time. Depending on the available land and the orientation of housing programs, this lasting polarization of income distribution could nevertheless shift upwards or downwards.

Overall, it seems very clear that gentrification processes are unfolding in the historic urban core of Nice and that the internal socio-economic polarization of many IRISs within this urban core is partly caused by this unfolding. However, it seems that, while these gentrification processes may be favored by the tension in the real estate market, likely exacerbated by the secondary residence status of a significant portion of the housing stock, other aspects of the structure of this housing stock limit gentrification. In the historic urban core, the existence of old or very old buildings, whose state of maintenance or renovation can vary considerably and whose ownership structure is very dispersed, appears to result in the retention of low- or very low-income populations, even in the neighborhoods where the housing market is tight. More marginally, in a less central part of the city, the socio-economic polarization of an IRIS in Nice results from the coexistence of social housing and affluent suburban areas whose expansion depends on available land. In any case, it seems that, despite the increase in the proportion of high-income populations, low- or very low-income populations are far from disappearing, leading to unfinished gentrification and an internal socio-economic polarization that is likely to persist.

4. Discussion

The coexistence of high-income populations and low- or very low-income populations within the same neighborhood is quite common in the city center of Nice, confirming the reality of the internal socio-economic polarization of certain central urban areas. This relatively high frequency, compared to other major French cities and their socio-urban forms, partly reflects the unique context of Nice, with its historic urban core comprising buildings of highly variable quality, secondary residences, and vacant housing. Although gentrification processes are unfolding in most of these socially polarized neighborhoods, there is no indication that they will fully displace low- or even very low-income populations. On the contrary, it seems that these poorer populations remain in these neighborhoods, likely due to the persistence of mediocre-quality old housing, sometimes referred to as “de facto social housing”, whose very dispersed ownership structures limit redevelopment efforts. In some of these central areas, there is even a simultaneous increase in both poverty and wealth, or, more rarely, only an increase in poverty. In all cases, the level of poverty remains significant, with a few rare areas in this urban core of Nice still showing modest wealth levels, indicating the beginning of gentrification, whose future remains uncertain. Therefore, there are good reasons to consider that in most areas of downtown Nice, variously affected by gentrification processes, internal socio-economic polarization is likely to persist, at least in the medium term. This result, which questions the continuation of sociodemographic trends in the hearts of major cities, also raises other concerns in terms of both the extent of the observed internal socio-economic polarizations and the social consequences experienced by residents.

As the presentation of the sociodemographic context of Nice has shown, the age composition of the municipality differs significantly from that observed in other major provincial cities in France. A lower proportion of young adults primarily distinguishes it, notably students aged 18–24 and a higher proportion of people aged 60 and over. This under-representation of resident student populations is much less pronounced in Nice than in other major French cities, such as Marseille, and can be partly explained by the density of the PACA region, which encourages students to stay in their parental homes and commute daily to higher education sites. The higher proportion of people aged 60 and over largely reflects the attractiveness of the Côte d’Azur coastline and, in particular, the city of Nice for retirees. In central neighborhoods with a high internal socio-economic polarization, this high proportion of seniors is often even more pronounced than in the rest of the municipality, raising the question of the role of these retired populations in this social polarization. Are they overrepresented at one of the extreme poles of the income distribution, thereby playing a significant role in internal socio-economic polarization? Or, conversely, are these people aged 60 and over themselves subject to this social polarization within these neighborhoods?

Beyond the fact that the age of individuals contributing to the neighborhood’s internal socio-economic polarization is likely to strongly influence how residents experience their urban environment, this age also impacts the potential for population renewal in the neighborhood. Indeed, the life cycle determines residential choices and influences the social structure of the urban space through the mobilities associated with these choices. The residential strategies of elderly people, at the end of their residential trajectory, differ from those of young adults and families (Scarella 2014; Fusco and Scarella 2011). These specific residential mobilities of elderly people, particularly upon retirement or later, when possibly moving to a retirement home, combine with their mortality to affect neighborhood demographics (Scarella 2014) and thus potentially the social composition of polarized areas. The fate of homes initially occupied by elderly people who have moved or passed away depends on many factors (including the standard of living of these elderly people, property ownership status, and potential inheritance). Depending on the situation, these vacated homes may contribute to vacancy, gentrification, or pauperization. Similarly, the freeing up of homes due to the departure or death of elderly people could lead to the rejuvenation of the neighborhood or contribute to maintaining the overrepresentation of those aged 60 and over. In other major cities, such as Bordeaux, the replacement of elderly populations who have died or moved to retirement homes often with affluent families has significantly contributed to the gentrification of certain neighborhoods. However, this usually occurs in neighborhoods relatively close to the historic urban core rather than in neighborhoods that are part of this historic urban core, whose urban form and housing structure are not the most sought after by families with children.

The high vacancy rate observed in the socially polarized neighborhoods of Nice is also a factor to consider when examining the future prospects of these areas. This significant vacancy within the housing stock of neighborhoods with strong internal socio-economic polarization likely results from a combination of three factors. This is undoubtedly partly due to the frictional vacancy associated with small dwellings with a high turnover rate, since socially polarized neighborhoods are, as we have seen, characterized by an overrepresentation of households that have recently moved into their dwellings, which reflects this turnover. It is also very likely to be the case of a structural vacancy, corresponding to obsolete and depreciated housing, with some of the latter constituting the vacancies and being outside the market and the others nevertheless being occupied and constituting a ‘de facto social stock’, which seems to play an important role in the internal socio-economic polarization of the neighborhoods. The existence of such a structural vacancy in the housing stock of Nice’s urban core raises the question of how to renovate and improve the comfort of these areas for the future of the internal socio-economic polarization of Nice’s city center neighborhoods. Such transformations, which are likely to revive or accelerate the gentrification process in the city center, would require a strong commitment on the part of property owners, who are currently partly hampered by various obstacles such as financial constraints or multiple ownership. Vacancy is not always an indirect consequence of poor quality, but it can be more nominal than real, with some furnished flats being rented out on a short-term seasonal basis. For example, the increase in the proportion of vacant homes over the last ten years could be correlated with the development of this type of temporary rental.

Having data on the length of time that dwellings remain vacant, in particular via the activity of meters associated with energy or water consumption, would make it possible to answer some of the questions raised by the high proportion of vacant dwellings observed in neighborhoods with a high degree of internal socio-economic polarization. Given the current state of information, the vacancy rate remains difficult to interpret, as it could result in two opposing trends in the future. This high vacancy rate could lead to the impoverishment of “polarized” neighborhoods if vacant housing continues to deteriorate (through the letting of deteriorated vacant housing or persistent vacancy). It could also lead to an increase in the residential value of “polarized” neighborhoods, with more affluent people moving into renovated housing or turning it into second homes, which could accentuate the “touristification” of Nice’s urban landscape. These two opposing trends would not necessarily be mutually exclusive but could coexist and thus reinforce the internal polarization of these central neighborhoods.

In the municipality of Nice, with its strong tourist and hospitality economy, the significant presence of secondary residences is emerging as one of the important characteristics of neighborhoods with strong internal socio-economic polarization. The temporary use of part of the housing stock tends to reduce the availability of housing on the local property market, accentuating pressure on the land market. These potential imbalances in the housing market in high-tourism areas could lead to increasing discrimination in access to property and rental supply for the local population. This situation could lead to a decline in the permanent population and, in some neighborhoods with a strong identity, a possible loss of that identity (Lopez-Gay et al. 2021; Fusco and Scarella 2008b). In tourist areas such as the municipality of Nice, online temporary rental platforms may also have contributed, through their increased profitability, to reinforcing the secondary destination of some of the accommodation and played a decisive role in the shortage of small flats available for long-term rental (Christofle and Hélion 2018; DREAL Provence-Alpes Côte d’Azur 2018). The significant proportion of secondary residences in “polarized” neighborhoods and their possible effect on housing market tension further reinforces the idea that, in these neighborhoods, there is a segmentation of the rental stock, leading to the retention of a large proportion of low-income households. It is also conceivable that the over-frequentation of some of these neighborhoods by tourists reduces their residential appeal for affluent populations and ultimately acts as a brake on gentrification.

In any case, while secondary residences are an important component of “polarized” neighborhoods, most of the income of their owners (some of whom nevertheless declare their income at the place of their secondary residence) and the income of their temporary tenants are not included in the localized tax data for these neighborhoods. This means that the social polarization measured by the poverty and wealth indexes of the households occupying the main residences in the neighborhoods of Nice’s urban core is potentially underestimated from the point of view of the populations actually present. In fact, the high proportion of secondary residences in these neighborhoods could be reflected, in terms of the population present, by an even higher weighting of high-income households, since the vast majority of secondary residence owners are people with high economic resources and, to a lesser degree, must be temporary tenants. In other words, in terms of the populations actually present, the “polarized” neighborhoods of Nice’s urban core are probably even more so during certain seasons, with an even more marked overrepresentation of affluent populations. Similarly, certain areas of the city center or seafront, considered “poor” from the point of view of the incomes of permanent residents, are, perhaps temporarily, areas of strong internal socio-economic polarization if we consider the population actually present. In order to gain a more in-depth understanding of the role of secondary residences in the internal socio-economic polarization of these neighborhoods, it would be desirable to gain a better understanding of the nature of the dwellings in question (the size, age of construction, condition, length of occupancy during the year8, etc.) and the sociology of their owners (Scarella 2014) and temporary tenants. The geographical coexistence of “temporary” and “permanent” populations also raises questions about the dynamics of the temporal relations between neighbors, the local roots of dual residents, and the involvement of local people in the tourism sector (staff, tourist services, parallel accommodation trade, tourist agencies, etc.) (Christofle and Hélion 2018).

More generally, the socio-economic heterogeneity observed in “polarized” neighborhoods does not make it possible to fully understand neighborhood relations and the degree to which residents participate in neighborhood life. Physical proximity does not necessarily imply social proximity (Park 2004; Lehman-Frisch 2008; Chabrol and Launay 2016). It can attenuate social distances, but it can also “stage the differences” of neighboring populations (Bozon 1984; Simon 1995; Giroud 2015; Jeanmougin and Bouillon 2016). In socially polarized neighborhoods, various types of sociability emerge, ranging from deep social relationships between similar residents to mutual avoidance and even conflict. This mutual avoidance behavior between heterogeneous populations takes place by dividing the space into different uses and visibilities, as several case studies have shown (Chamboredon and Lemaire 1970; Bacqué and Fijalkow 2006; Simon 1997; Jeanmougin and Bouillon 2016; Lehman-Frisch 2008; Tissot 2010). Even if some residents appreciate social diversity in principle, their practices may not go beyond these intentions in principle and result in completely different everyday uses of space (Grafmeyer and Authier 2019; Lehman-Frisch 2003). The appropriation of neighborhood life is manifested through a multiplicity of socio-economic and cultural factors, such as participation in neighborhood associations, the occupation of public space, visibility/presence in the public sphere and residential homes, and, finally, the schooling of children and adolescents in educational establishments. It is mainly in the school sphere that several studies have noted the ambivalence of social mixing and avoidance strategies (Van Zanten 2009; Giroud 2015; Launay 2016; Maurin 2004; Audren et al. 2016; Préteceille 2006b). Despite the various forms of distancing and intermingling observed, this coexistence can nevertheless have positive effects on neighborhood social life (Oberti and Préteceille 2016; Pinçon and Pinçon-Charlot 2014).

5. Conclusions

Among large French cities, Nice is one of those where neighborhoods displaying strong internal socio-economic polarization are the least common. In Nice, these socially polarized neighborhoods are concentrated in the municipality’s dense urban core. As in other cities, the age of buildings is a characteristic that is highly correlated with the internal socio-economic polarization of neighborhoods. However, the specificity of Nice’s socially polarized neighborhoods lies in the age composition of the population in these neighborhoods, where, unlike many other large cities, very young adult students are not really overrepresented, whereas retirees and secondary residences are very strongly overrepresented. The tension and high prices prevailing on the housing market caused by this high proportion of secondary residences or temporary residences do not, however, seem to be reducing the internal socio-economic polarization of these neighborhoods, since the poverty of a significant proportion of permanent residents remains high or very high. This does not mean, however, that gentrification processes are absent from these “polarized” neighborhoods, but simply that they are hampered by the poor quality of some of the older housing, the ownership structure of which is undoubtedly very dispersed. Thus, in the urban heart of Nice, even when the proportion of high-income residents increases, this does not automatically result in the poorest residents being driven out, probably because of the highly segmented nature of the rental market. This coexistence of populations with very different incomes, undoubtedly reinforced during certain seasons by people living in secondary residences, does not necessarily imply a great diversity of experience of the city, as geographical proximity only translates into social proximity in very limited areas.

In a broader perspective, this article highlights the coexistence of populations with diametrically opposed economic resources, characterized by more or less sustainable situations. The evolution of income distribution in these neighborhoods, particularly in central areas, shows that this socio-economic polarization does not follow the uniform model of the socio-urban transformations often observed in metropolitan city centers. Contrary to a significant portion of the literature that describes gentrification as an inevitable and nearly complete process in city centers, this study nuances this view by revealing more complex realities. In particular, the case of Nice illustrates the limits of the gentrification process and sheds light on the internal socio-economic polarization of urban neighborhoods as another aspect of socio-urban dynamics that is rarely studied. Analyzing these dynamics in different urban contexts could provide a broader understanding of the factors associated with the socio-economic evolution of urban neighborhoods.