Analyzing the Impact of Public Housing Privatization on Immigrant Micro-Segregation in Milan

Abstract

In several Western European countries, a significant share of social rental housing stock has been sold since the 1980s as part of government policies aimed at promoting homeownership societies. Research has shown that tenure conversion has contributed to increasing socio-spatial segregation of lower-income groups, with diverging spatial patterns of homeownership among immigrants. This paper examines the impact of recent public housing privatization schemes in Milan in relation to micro-segregation and peripheralization processes of foreign populations, which represent distinctive features of immigrant residential distribution in this city. By employing name analysis, an unconventional approach in segregation studies, I inferred the geographical origins of homebuyers and mapped their distribution across the city. The findings reveal divergent purchasing behaviors, whereby Italians predominantly acquire properties in semi-central areas currently undergoing urban regeneration. In contrast, immigrants tend to concentrate their acquisitions in peripheral post-war public housing neighborhoods or in areas predominantly inhabited by residents with similar geographical origins. This paper contributes to the existing literature on ethnic residential segregation in Southern European cities by shedding light on the underexplored role of public housing privatization policies in shaping specific residential patterns and housing outcomes among different groups.

1. Introduction

In several Western European countries, a significant share of social housing stock has been privatized since the 1980s as part of government policies aimed at promoting homeownership societies (Gibb 2003; Gimat et al. 2023; Pattison and Cole 2020). Among the different pathways of privatization, one of the most popular forms has been tenure conversion, whereby social or public dwellings have been converted into private, owner-occupied houses, often by granting occupants the right to buy at discounted rates (Forrest and Murie 1988; Elsinga et al. 2014)1. In the UK, more than 1,500,000 dwellings were sold off to sitting tenants under the well-known Right to Buy program (Kleinhans and Van Ham 2013), representing one of the most remarkable examples of the neoliberal turn in housing policy in Western societies (Whitehead 2012).

The privatization of social dwellings has been supported by several assumptions asserting the “superiority” of homeownership over rental housing. According to policymakers, homeowners exhibit greater self-sufficiency and take better care of their dwellings compared to tenants. Homeownership is associated with positive socio-spatial outcomes as it is linked to lower residential turnover, which, in turn, is believed to positively impact the quality of social interactions, leading to increased social capital and trust among residents (Kleinhans and Elsinga 2010). In addition, the sale policies of social housing were motivated by nonprofit housing organizations’ need to compensate for the reduction in financial resources caused by government cutbacks in the social housing sector, which limited their ability to effectively manage the housing stock (Gibb 2003).

Regarding the effects of social housing privatization, studies have emphasized the widening socio-spatial divides between neighborhoods (Kleinhans and Van Ham 2013; Murie 2014). Sales policies have contributed to the increased residential segregation of socially disadvantaged groups, including immigrants, who have often remained confined to lower-quality housing concentrated in less attractive neighborhoods. Conversely, thanks to favorable market conditions created by these policies, more affluent tenants have gained access to better quality homes in the most desirable areas, where owner-occupancy rates were already high (Forrest and Murie 1988; Kleinhans and Van Ham 2013). One long-term consequence of sales policies is gentrification. When privatized dwellings are later resold or rented on the free market, they are rarely accessible to first-time buyers and low-income households due to the higher prices set. This phenomenon has been observed in certain areas of London, where privatized properties were increasingly occupied by younger, more affluent households with shorter-term residential mobility patterns (Gimat et al. 2023).

An underexplored aspect in this debate pertains to the outcomes of privatization policies on the residential distribution of immigrant populations. In their study conducted in Amsterdam, Boterman and van Gent (2014) found that tenure conversion has contributed to increasing ethnic polarization between the historical center and postwar peripheries of the Dutch capital. In addition, the authors have highlighted that each migrant group exhibits divergent patterns of purchase, suggesting the relevance of a more thorough examination of the factors that shape tenure conversion outcomes, such as the structure of the neighborhood housing market and the purchasing behaviors of homebuyers.

A deeper understanding of the outcomes of social housing sales for the immigrant population can help clarify the role of this policy in addressing a crucial component of social integration within the host society, namely housing, with homeownership being often considered a strong indicator of such integration (Constant et al. 2009; Kauppinen et al. 2015; Vono-de-Vilhena and Bayona-Carrasco 2012). In addition, examining this process through a spatial lens can shed light on the locational advantages and disadvantages associated with migrants’ purchases, which extend beyond the tenure dimension. This is closely related to debates on neighborhood effects, which consider the extent to which neighborhood socio-spatial characteristics shape individual life chances and outcomes (Bolt et al. 2010; Fajth and Bilgili 2020; Gijsberts and Dagevos 2007).

Investigating the socio-spatial outcomes of sale policies remains a pertinent issue in light of ongoing debates surrounding the shortage of affordable housing (Wetzstein 2017) and concerns about the rising levels of socio-economic and ethnic segregation in cities (Tammaru et al. 2016). While some countries, such as the UK and Germany, are attempting to repurchase previously sold stock to address the current housing crisis, other countries, such as France and Italy, are accelerating the sale of social housing (Gimat et al. 2023). Taking Italy into consideration, this policy occurs in a context where the availability of social housing is minimal (only 5% of the stock) and where residential ethnic segregation assumes unique forms. Like in several Southern European contexts, Italian cities traditionally feature high levels of social segregation and low levels of spatial segregation (Arbaci 2019; Mugnano and Costarelli 2018; Zajczyk 2003).

This paper delves into the privatization policies of publicly owned dwellings in Milan. Since 2014, the local provider has put more than 2000 vacant dwellings up for auction at discounted rates, adding to long-standing policies that have reduced the availability of social units by about 30,000 since the 1990s (Report Milano 2021). Building on previous studies that have shown increasing trends of micro-segregation and peripheralization of immigrants in this city (Costarelli and Mugnano 2017; Mugnano and Costarelli 2018), this paper aims to examine the effects of public housing sales on the spatial distribution of the immigrant population. The research question guiding this paper is as follows: what are the socio-spatial outcomes of public housing privatization in the urban context of Milan?

To address this question, I identify and compare the purchasing behaviors and strategies of both native and immigrant groups, with a specific focus on two key migrant communities, namely Africans and Asians. These communities represent the largest segments of foreign citizens in Milan and have also accounted for the most significant number of public housing purchases among immigrant populations. Using a name-based classification method (Lan et al. 2021; Mateos 2007), I inferred the geographical origins of buyers based on their names and geolocalized the properties associated with these names to analyze and compare the spatial distribution of property purchases made by both native and immigrant populations, as well as Asian and African buyers.

The findings reveal a notable prevalence of property acquisitions by Italian citizens in residential estates located within central or semi-central areas that are currently undergoing urban regeneration projects bringing improvements to the built environment. Conversely, immigrant property acquisitions display a pronounced concentration in peripheral neighborhoods. Significantly, purchasers originating from African countries manifest a concentration of acquisitions within post-war public housing districts, whereas Asian buyers tend to converge in areas predominantly inhabited by residents with similar geographical origins. This underscores the existence of a discernible link between the outcomes of this policy and the distinctive residential patterns exhibited by Asian buyers, in particular the Chinese community.

This paper is divided into five sections. The following section contextualizes the present study within the framework of housing policies in Italy and their influence on the residential patterns of immigrant communities, with a specific focus on the city of Milan. Section 3 outlines the research design, encompassing data collection, methods, and analysis. In the fourth section, I present the findings, which are subsequently discussed in the concluding section of this paper.

2. Italian Housing Policies and Their Influence on Immigrant Residential Patterns in Milan

Since its inception, Italian public housing (edilizia residenziale pubblica) has offered tenants various advantageous opportunities to access homeownership, such as rent-to-buy (affitto a riscatto) and right-to-buy schemes (piani di vendita e alienazione). According to estimates, the rent-to-buy schemes facilitated the conversion of at least 50,000 units, while approximately 100,000 units have been sold as part of the right-to-buy schemes introduced in 1993. The latter encouraged public landlords to sell up to 50% of their stock to sitting tenants with at least five years of occupancy, resulting in a 22% reduction in the availability of public housing stock and generating approximately EUR 2.4 billion in earnings (for a stock valued at EUR 9 billion).

Despite the increasing tenure mix, privatization has brought about additional managerial challenges for public landlords2. In post-privatization multi-tenure buildings, roughly 70,000 (Federcasa 2015), public landlords have found it increasingly difficult to align their managerial strategies with the requests and needs of new homeowners. In the majority of these estates, these landlords retained only a minority share of dwellings. Consequently, public landlords were compelled to accept homeowners’ decisions, such as initiating renovation works, which posed further challenges to the financial stability of these organizations, already affected by budget reductions. Therefore, sales policies have primarily targeted this segment of the public housing stock.

In Italian cities, right-to-buy schemes have also contributed to the exacerbation of socio-spatial segregation through filtering processes that increasingly marginalized lower-income households, including immigrants who were unable to buy, in unattractive and low-quality estates. Meanwhile, better-off tenants were able to purchase higher-quality housing stock (Bergamaschi and Maggio 2020). Rent-to-buy and right-to-buy schemes continue to shape current housing policies in Italy, which have been reconfigured on a regional basis since 2001, granting local authorities more autonomy to define their agendas (Costarelli and Mugnano 2017).

In Milan, public housing is jointly owned by the City Council (46%) and the regional public housing company ALER (Azienda Lombarda per l’Edilizia Residenziale) (54%)3 (Petsimeris 2018; Report Milano 2021). The current privatization policy primarily targets vacant dwellings owned by the latter, constituting approximately 10% of the entire stock (7200 units in 2020). These sales policies aim to bolster the financial stability of the organization, which currently struggles to cover the expenses of necessary renovation work required for the allocation of these dwellings to eligible households (Costarelli 2022). However, this policy has further diminished the availability of public rental solutions that were already challenging to access for immigrant groups. This difficulty arises from the requirement of at least five years of regular registration and a stable presence in the region (anzianità di residenza). In 2017, immigrants accounted for 16.8% of total tenants living in the City Council’s public housing stock, with the most prevalent nationalities being Moroccan (2.9%), Filipino (2.3%), Sri Lankan (1.6%), Peruvian, and Ecuadorian (1.1%) (Petsimeris 2018). The combination of a limited supply of public housing, stringent allocation rules, and immigrants’ vulnerable position within the labor market are crucial factors in explaining both residential patterns and housing conditions of foreign populations (Benassi et al. 2019; Mugnano and Costarelli 2018).

While rent remains the most prevalent tenure choice among immigrant households, accounting for 63.5% at the national level, recent years have witnessed increased attention on the homeownership trends within this population, which rate stood at 20% in 2021. The typical social profile of immigrant homeowners corresponds to long-term residents (with at least 10 years in the country) holding stable job positions (Scenari Immobiliari 2021). The highest rates of homeownership are recorded among Albanian immigrants, who were among the earliest significant immigrant groups in Italy, dating back to the early 1990s. They are followed by Chinese, Moldovan, and Moroccan immigrants (Ministero dell’Economia e delle Finanze and Agenzia delle Entrate 2019).

However, the mere access to homeownership by immigrants does not necessarily indicate full social integration. Variables such as the conditions of their dwellings, mortgage burdens, and property locations should also be taken into account. In fact, limited economic resources and a lack of credit market guarantees have directed immigrants toward smaller and lower-quality housing in the homeownership market, compared to native households (Alietti and Agustoni 2013).

In addition, immigrants’ home purchases have been strongly influenced by economic downturns resulting from recent crises and subsequent increases in immigrant employment rates. The incidence of home purchases by immigrants declined from 17.3% in 2017 to 8.9% in 2019 and further dropped to 5% in 2021 following the onset of the COVID-19 pandemic. Similarly, during the period 2004–2007, immigrants from non-European countries accounted for 8.2% of the total mortgages disbursed. However, this percentage decreased to 4.5% between 2008–2011 (Banca d’Italia 2012).

When looking at the spatial implications of these trends, recent data show that immigrants’ home purchases are more frequent in small-sized towns and peripheral areas of the largest cities. The incidence of foreign purchases in peripheral areas increased from 31.7% in 2017 to 35% in 2021, and from 53.1% to 57% in greater peripheral areas (provinces). Meanwhile, in central and semi-central areas, it dropped, respectively, from 8% to 3.7% and from 7.2% to 4.3%. These trends underscore significant changes compared to past situations in the largest cities, where access to homeownership had enabled immigrants to move from high-immigrant concentration neighborhoods into more socially diverse areas (Scenari Immobiliari 2021).

In Milan, recent research has pointed out a growing trend of peripheralization among ethnic minorities in the city (Bergamaschi et al. 2021; Costarelli and Mugnano 2017). This socio-spatial process, common to other Southern European metropolises, has been described as an “urban diaspora” by Arbaci (2019). While immigrants are increasingly choosing to reside in peripheral areas, they also tend to cluster together, leading to a process of micro-segregation. Unlike many American and Continental European cities, where ethnic concentration often results in large homogeneous neighborhoods in the outskirts (Maloutas and Karadimitriou 2022), Milan’s micro-segregation occurs at smaller scales. It is observed at the level of streets, individual buildings, or blocks of buildings. Several contextual factors contribute to the formation of micro-segregation, including the relatively low proportion of public housing and its scattered distribution. Research indicates that these micro-segregation areas in Milan are expanding, giving rise to larger clusters where the divisions along income or ethnic lines are becoming more pronounced (Mugnano and Costarelli 2018).

In 2022, foreign citizens constituted 20.61% of Milan’s total population, representing a diverse array of over 140 nationalities. The largest groups among these foreign citizens were Egyptians, Filipinos, and Chinese, accounting for 14.91%, 13.77%, and 12.11% of the immigrant population, respectively. These were followed by Peruvians, Sri Lankans, and Romanians, each making up approximately 5–6% of the immigrant population (City Council Register). After a substantial increase during the 2003–2013 decade, the growth rate of the migrant population in Milan has gradually declined, indicating that many immigrants have become integrated into the city’s social fabric.

Immigrant groups tend to occupy older and lower-quality public housing units, particularly in historic public housing neighborhoods like San Siro, Mazzini, Calvairate, and Stadera, which were built before World War II. In these areas, immigrants share spaces with other vulnerable populations such as elderly individuals living alone, households with invalidating health problems, and squatters. A minority of immigrants (less than 15%) reside in post-war public housing neighborhoods like Gratosoglio, Gallaratese, and S. Ambrogio, located in more peripheral areas but offering better housing standards (Alietti and Agustoni 2013). Micro-segregation is also noticeable in areas like Bovisa and Via Padova, where the housing structure is predominantly private, but the low quality of dwellings makes it unattractive to middle-income households.

Regarding the residential patterns of some of the most prominent immigrant groups, Egyptians have tended to settle in the southern and western public housing peripheries (San Siro, Giambellino, Lorenteggio, and Mazzini), as well as in areas dominated by private housing. Chinese citizens have exhibited a more uniform distribution in the northern periphery, while clusters of Sri Lankans and Filipinos have been observed in both multiethnic neighborhoods in the northeastern private housing periphery of Via Padova and affluent condominiums in the city center, where they often work as caretakers (Mugnano and Costarelli 2018).

While some immigrants have reached similar conditions to the native population, others live in precarious arrangements that expose them to greater risks. This polarization in the housing outcomes of the immigrant population is a result of market-led housing policies that have been able to integrate the majority, albeit at higher costs, while systematically excluding another part (Tosi 2017). Given the significant role of public housing in shaping immigrant residential patterns and the critical interaction with the private housing market in drawing socio-spatial boundaries along ethnic lines, in this paper, I am interested in investigating the impact of tenure conversion policies of public housing on the ongoing dynamics of micro-segregation among immigrant populations in Milan.

3. Data and Approach

Between 2014 and 2016, the Lombardy regional government initiated two plans aimed at auctioning off 2955 vacant public housing units owned by the local public housing company ALER, within the Milan metropolitan area. These plans focused on units in multi-tenure buildings, where the public landlord maintained a minority share of dwellings, including units in prestigious or highly lucrative buildings, as indicated in official documents. The dwellings were made available through monthly public calls, which provided information regarding size, location, opening bids, and various dwelling characteristics.

These calls also featured images of the dwellings and outlined eligibility criteria for prospective homebuyers. Both EU and non-EU citizens were allowed to participate. For non-EU participants, possessing a valid residence permit and having worked continuously in Italy for at least two years were prerequisites. Additionally, applicants must not have any criminal convictions or have been involved in the illegal occupation of public or private dwellings within the last five years.

The starting bid was set at 20 to 30% below the market price, depending on the condition of the dwellings. Aspiring homebuyers were obligated to provide a 10% down payment of their bid as a guarantee, and in the event of a successful bid, they were prohibited from reselling the property for the following ten years. After this period, the public landlord retained the right of first refusal if an auctioned dwelling was to be resold on the open market.

The following analysis is based on the outcomes of the auctions, specifically the information extracted from official documents that report auction prices, the location of sold units, and the names of buyers4. Comprehensive data for the first sale plan (2014–2019) were already accessible in official records when data collection commenced in January 2022. Data pertaining to the outcomes of the second plan (2016–2022) were supplemented through direct observation of public auctions for units sold between February and May 2022.

Observation occurred during the monthly public web-streamed openings of bid envelopes. Two public officers were responsible for unveiling these envelopes, disclosing the number of bids received, revealing bidders’ names, and verifying the completeness of documents before assigning units to the highest offer. In instances where two or more applicants offered the same price, priority was given to first-time homebuyers, defined in policy documents as those who bid for a house to reside in and do not own any other property in the country. As a result of these plans, 2255 dwellings were sold, with 85% (1928) located in the city of Milan. This generated earnings of EUR 257,598,182.7, approximately EUR 80 million more than the total opening bid.

From both documents and direct observation, the following data were collected: (a) the precise location (address) of 1928 sold dwellings in the city of Milan; (b) the surnames and forenames of 901 buyers; (c) the initials of 1027 buyers’ surnames and forenames5. To determine the geographical origins of buyers and infer the socio-spatial outcomes of sale policies, I relied on name–origin classification drawing from previous studies in the field of residential segregation (Mateos et al. 2006). This approach aimed to associate buyers with the most likely geographical origin of their names. To enhance the reliability of the information for inferring such origin, only buyers with full surnames available were considered, necessitating the removal of the set of data involving only initials, namely 1027 cases.

Applying Name–Origin Analysis

Conventional methods for measuring residential segregation rely on statistical indices designed to capture specific dimensions that characterize the proximity and separation of groups. These indices include measures of evenness, exposure, concentration, clustering, and centralization (Massey and Denton 1988). The application of these indices necessitates the availability of reliable data on variables of interest, such as citizenship, ethnicity, or other relevant ones, at appropriate temporal, spatial, or nominal resolutions. In cases where such data are unavailable, proxies are employed. For instance, in the UK, country of birth has been utilized as a proxy to determine a person’s ethnicity. However, this classification overlooks the increasing numbers of second-generation ethnic minorities born in the “destination” or “host” countries, as well as migrants born in “intermediate” countries. In Spain and France, an alternative proxy used is nationality, although it is important to note that the latter can change over time, and individuals may possess more than one, including the host country’s nationality, especially for second- or third-generation migrants (Mateos 2007).

Name–origin represents an alternative approach to studying ethnic residential segregation. This approach is grounded on the idea that classifying surnames and forenames into ancestral groups of origin (referred to as ethnic classification) offers a feasible alternative to subdividing populations or classifying neighborhoods based on common origin. It introduces novel ways to measure residential segregation in cases where official information and statistics are unavailable. Instead of relying on statistical data and census information, these approaches rely on the likely ethnic origin of names. It is important to note that while this methodology cannot completely replace self-assigned ethnicity information, scholars have acknowledged that it provides a sufficient level of confidence in classifications to be utilized in measuring inequalities (Lan et al. 2021; Mateos 2007).

The current limitations of name–origin classification methods stem from several factors. Firstly, these methods are constrained by the availability of a limited number of names in their reference datasets. This limitation means that less common or newer-generation names may not be adequately represented, potentially affecting the accuracy of the analysis. Secondly, processes like transliteration, where names are converted from one script to another, or the “acculturation of a name” in a host country, can alter the way names are written or perceived. Additionally, intermarriage between different ethnic groups can result in names that do not clearly indicate a single ethnic origin, making it challenging to classify individuals with “mixed ethnicity” (Mateos et al. 2007; Mateos 2007).

In this paper, the verification of the origin of the buyers’ surnames was conducted using the Worldnames tool6. This analysis attributed 551 units to Italian buyers and 350 to immigrants. Among the immigrant buyers, it was possible to determine the geographic origin of 266 surnames, corresponding to 97 buyers from Africa, 108 from Asia, 41 from Europe, and 20 from America. However, there were 84 buyers whose surnames, in principle, were attributed to immigrants, but accurate verification was not possible. As a result, these cases were excluded from the analysis. Data reported in

Table 1. Categories of buyers according to attributed geographical origin.

Attributed Origin	Number of Buyers	Share of Total Buyers (%)	Average Purchase Price (EUR)
Italy	551	67	112,283
Africa	97	12	99,062
Asia	108	13	110,897
Europe (excluding Italy)	41	5	80,764
America	20	2	90,753
Total	817	100	108,421

synthesize the results of name–origin analysis.

The subsequent steps involved calculating (a) the incidence of immigrant purchases at the neighborhood level and comparing it to the overall distribution of auctioned dwellings; (b) the incidence of purchases by two specific immigrant communities, namely Africans and Asians—who were the most frequent buyers within the immigrant population—and comparing it with their overall residential distribution at a neighborhood scale. In the first case, the incidence was calculated by determining the proportion of dwellings purchased by immigrants in a specific neighborhood in relation to the total number of dwellings sold citywide, multiplied by 100. In the second case, for both Africans and Asians, the incidence was calculated by assessing the proportion of dwellings purchased by each group within the neighborhood relative to the total number of dwellings bought by immigrants across the entire city, also multiplied by 100. The residential distribution of these communities was determined by comparing the number of residents from each group to the total population of the neighborhood, multiplied by 1007. The definition of neighborhood aligns with the administrative unit, known as the Nucleo d’Identità Locale (NIL), of the city. The findings from these calculations will be presented in the following section.

4. Socio-Spatial Impacts of Public Housing Privatization in Milan

In this section, I examine the socio-spatial outcomes of public housing privatization in Milan by analyzing the distribution of purchased dwellings among buyers of different geographical origins and comparing the average purchase prices. As indicated in Table 1, the name-based classification approach reveals that Italians acquired the majority of auctioned dwellings, accounting for 551 units, in contrast to the 350 units purchased by immigrants. Among the immigrant population, the highest number of units was purchased by Asians and Africans, totaling 205 units, while Europeans and Americans contributed only minimally with 61 units in total. In terms of the average auction price, Table 1 highlights more similar values for Italians (EUR 112,283) and Asians (EUR 110,897) compared to the other groups: Africans (EUR 99,062), Americans (EUR 90,753), and Europeans (EUR 80,764). These findings suggest the existence of distinct participation patterns within this specific housing market, as well as disparities in both the quantity and value of purchases.

These patterns are shaped by various factors, including the distribution of economic resources, which can vary significantly both between and within each group. The results indicate that, among all immigrant communities, individuals of Asian origin were able to mobilize substantial economic capital, enabling them to actively participate in auctions and seize the residential opportunities they aspired to. In this regard, their performance closely resembles that of Italians. These observations support previous findings in the Italian context (Alietti and Agustoni 2013; Tosi 2017), which indicate that certain immigrant groups have achieved comparable socio-economic conditions to native residents.

The spatial distribution of privatized dwellings (see Figure 1) highlights the highest concentrations in peripheral neighborhoods of the city, specifically in the northwest (Gallaratese, Quarto Oggiaro, Comasina, Villapizzone), west (Baggio), south (Barona and Gratosoglio), and east (Taliedo). In contrast, a smaller share is located in central and semi-central areas. Many of the estates in these neighborhoods were once entirely publicly owned housing. They have been turned into multi-tenure buildings as a result of tenure mixing processes initiated a long time ago by sale policies. The examined sale plans represent a significant step in the ongoing process of public housing dismissal in these areas, accelerating the transformation of these neighborhoods into fully privately owned condominium blocks.

A pivotal question to investigate in the context of current tenure conversion operations is whether these sale plans will result in increased social diversity, particularly in terms of changes in the socio-economic, demographic, and ethnic characteristics of residents. Studies conducted in the UK context have indicated that in the short term, privatization policies primarily altered the tenure balance, as targeting sitting tenants did not necessarily lead to increased residential mobility (Kleinhans and Van Ham 2013). However, our study deals with the sale of vacant public dwellings, which represents a distinct form of privatization and implies a twofold change in the short term. On the one hand, this change pertains to the tenure dimension, transitioning from public to private ownership. On the other hand, it has social implications, involving new residents occupying previously empty units.

The name analysis approach discussed in the previous section allowed us to infer the geographical origin of 901 buyers, constituting 47% of the total sales, providing insights into both social and tenure change dynamics triggered by this policy. I will begin by presenting the incidence of purchases by immigrants compared to the overall distribution of privatized units.

Figure 1 illustrates that the highest incidence of immigrant buyers is observed in peripheral neighborhoods (highlighted in red), some of which feature a significant presence of public housing (such as Barona, Comasina, and Gallaratese). In contrast, central and semi-central areas have the lowest incidence (close to 0). Immigrants tend to gain access to homeownership in the outermost areas, while Italians are more successful in securing dwellings in central locations. For instance, in the case of Porta Ticinese, the incidence of Italian buyers is 2.46, compared to 0 for immigrants. Similar disparities, but with a reverse situation, can be observed in other areas, such as Barona and Gratosoglio in the south, where the incidence of immigrant purchases is higher than that of Italians (3.01 against 0.68 and 2.33 against 0.68, respectively). These data indicate a pattern of differentiation, between natives and immigrants, regarding the outcomes of sale plans.

In particular, purchases made by Italians are concentrated in areas targeted for urban regeneration, which are receiving significant investments in infrastructure and housing renovation, making them attractive and accessible hubs for various urban functions. For instance, apartments in the semi-central blocks of public housing estates, known as Quartiere Filzi near Città Studi (east), have predominantly been bought by Italians, with an incidence of 8.07 compared to 1.09 for immigrants. Urban renovation efforts in these areas are in full swing, including the introduction of new metro line stops, which enhances the accessibility of both existing and future households. Similarly, in Taliedo, the incidence of immigrants is relatively high (3.01), while that of Italian buyers is even more notable (7.39). Here, Italian buyers appear to have strategically targeted apartments in public estates to benefit from the spillover effects of urban transformations in neighboring areas like Rogoredo and Santa Giulia. In these adjacent areas, new residential units and improved connectivity to railways and motorways are rapidly increasing the residential property values.

A more balanced allocation of purchased dwellings was observed in the northern neighborhoods of Comasina (4.1 Italians and 3.83 immigrants), Quarto Oggiaro (3.28 Italians and 3.15 immigrants), and Niguarda (1.92 Italians and 2.19 immigrants), as well as in the western district of Baggio (2.6 Italians and 2.19 immigrants). In these working-class areas, differences between the two categories are rather small, possibly indicating that apartments were sold to households with similar socio-economic backgrounds, despite their different origins. Unlike the dynamics observed in the eastern neighborhoods mentioned earlier, the outcomes of public housing sales in these areas serve to maintain the existing social fabric rather than facilitating the potential influx of gentrifiers into highly sought-after areas.

A Focus on African and Asian Communities

When examining the outcomes for the immigrant population, it becomes evident that Africans and Asians emerge as the two groups with the highest share of dwelling purchases, accounting for 36% and 41% of attributed sales, respectively. Figure 2 and Figure 3 show the incidence of purchases by each category and the spatial distribution of residents from the same continent. The statistics presented in Section 2 indicated that Egyptians, Filipinos, and Chinese constitute the three largest components of the total immigrant population. Therefore, it is reasonable to associate these countries of origin with African and Asian buyers identified in this study.

Prior research (Mugnano and Costarelli 2018) has highlighted that Egyptian residents are primarily concentrated outside the city center, particularly in the proximity of pre-World War II public housing estates known as quartieri storici (historical neighborhoods) such as Stadera, Mazzini, and San Siro. Additionally, there is a notable Egyptian presence in the multicultural neighborhood around Via Padova in the north of the city. The Chinese community is particularly concentrated in the northern periphery, characterized by a working-class background and predominance of owner-occupied housing. This periphery has seen an increase in Chinese residents in the first two decades of the 2000s, coinciding with commercial gentrification in the nearby Via Paolo Sarpi, traditionally known as the Milanese Chinatown. Concerning the third largest immigrant community in Milan, Filipinos, together with Sri Lankans, exhibit a distinct residential pattern. Due to their employment in domestic service and caretaking, often for affluent households in prestigious residential areas, they tend to concentrate in the city center. However, many of them are gradually transitioning from situations of dependence, such as co-residence with their employers, to independence. This shift is reflected in their increasing presence in post-war housing blocks in peripheral districts like Quarto Oggiaro and Gratosoglio.

Taking the localization of purchases by African buyers, Figure 2 highlights that the highest incidences are observed in Barona, Gratosoglio, and Taliedo (highlighted in red). This suggests that this group has adopted purchasing strategies aimed at acquiring apartments located in areas already inhabited by a significant proportion of residents from the same group (or close to), particularly evident in Barona (incidence of 6.45) and Gratosoglio (incidence of 4.03). In addition, considering the notably low incidence of purchases by Italians in these neighborhoods (0.68), the outcomes of privatization are likely to further accentuate the ongoing dynamics of socio-ethnic differentiation between these peripheral neighborhoods and the rest of the city.

In the other districts, the outcomes of privatization exhibit notable variations. It is interesting to observe that, despite a relatively low presence of African residents in Gallaratese (northwest), the incidence of African buyers is higher (2.42) than in the well-known public housing neighborhood San Siro (2.02) where the incidence of this population is among the highest in Milanese neighborhoods (19.63). While in certain neighborhoods privatization appears likely to exacerbate socio-ethnic polarization, in others, the outcomes can contribute to greater social diversity. However, it is crucial to emphasize the importance of micro-level analysis to comprehensively understand the dynamics of socio-spatial inequality in urban contexts like Milan. A closer examination of the spatial distribution of sold apartments at the street or estate level is necessary to draw more robust conclusions. Indeed, while a mixed allocation of privatized units may indicate increased social diversity at the neighborhood level, it can also result in greater separation at a more granular scale of analysis.

Turning our attention to the second immigrant group considered in this paper, the Asian community, Figure 3 reveals that the highest incidences of house purchases are concentrated in the northern periphery, particularly in the neighborhoods of Comasina (6.85), Gallaratese (7.26), Niguarda (4.84), and Quarto Oggiaro (4.44) (highlighted in red). These areas were found to have a balanced allocation between Italians and the whole migrant population. However, a closer examination of the geographical origin of immigrant buyers suggests a strong homogeneity in terms of background. Asian individuals tended to purchase properties in areas with the highest concentration of residents from the same group. An illustrative case is Comasina, where both the incidence of Asian-origin residents and their property purchases are notably high. Here, the outcomes of housing privatization policies reflect, and reinforce, pre-existing residential patterns among Asians, particularly the Chinese community.

The purchasing strategies of Asians are not only directed to the aforementioned neighborhoods but extend to neighboring districts as well, including Stephenson, Quarto Oggiaro, Villapizzone, and Affori. Throughout the rest of the city, both the distribution of residents and the incidence of property purchases by Asians are relatively low and evenly dispersed. According to this pattern, Asians, and Chinese in particular, have specific residential preferences that can satisfy thanks to stronger financial capabilities, and lead them to cluster in certain areas of the city. This tendency, however, raises an important question pertaining to the implications for social integration opportunities. In this regard, issues to be further investigated include people’s daily life experiences in these neighborhoods, the dynamics of social interactions, and the use of public and semi-public spaces, from local parks to common areas within condominiums.

5. Discussion and Conclusions

In many Western European countries, the privatization of social rental housing has become a common housing policy aligning with neoliberal principles. Since the 1980s, this policy has been actively promoted with the aim of encouraging households to achieve self-sufficiency while simultaneously reducing public expenditures (Gibb 2003; Kleinhans and Elsinga 2010). In addition to reducing the size and scope of social housing sectors developed during the post-war era, the sale schemes have also contributed to the growing socio-spatial segregation of lower-income communities, including ethnic minorities (Kleinhans and Van Ham 2013; Boterman and van Gent 2014).

Conceptualizing segregation as a process influenced by households’ preferences, housing market structure (Musterd and Van Kempen 2009), and policies governing housing supply (Van Ham et al. 2021), this paper examined the impact of public housing sales on the socio-spatial residential dynamics of immigrant populations in Milan. In such an urban context, the public housing system plays a pivotal role in shaping the residential patterns of immigrants. Indeed, the interplay between limited housing supply, stringent allocation regulations, and the scattered distribution of public housing units contributes to the emergence of micro-segregation patterns within lower-quality housing estates, manifesting at the level of streets, individual buildings, or blocks of buildings. In addition to micro-segregation, immigrant residential patterns are also characterized by growing peripheralization, wherein foreigners are progressively settling in outlying and metropolitan areas (Arbaci 2019; Bergamaschi et al. 2021; Costarelli and Mugnano 2017), reflecting evolving housing trajectories and stabilization of socio-economic conditions, which often coincide with access to homeownership.

This paper delved into a distinctive facet of public housing privatization that characterizes current housing policy in Milan, where the public landlord has been auctioning off vacant dwellings that are unsuitable for regular allocation due to necessary renovation work. Using original data on buyers’ names and the locations of sold properties obtained from official documents, I employed a name–origin analysis to infer the geographical origins of buyers and geolocated the prevalence of purchases made by immigrants with a focus on the two groups that have bought the largest share of units among immigrant populations: Africans and Asians.

The findings highlighted a predominance of property purchases made by Italians in dwellings located in central or semi-central areas undergoing significant improvements in the built environment. Native buyers demonstrated the willingness and ability to mobilize greater financial resources to acquire properties concentrated in relatively central, well-connected areas with a growing potential for attractiveness, portraying them as early gentrifiers. Conversely, the highest concentrations of immigrant purchases are observed in peripheral areas, with some notable peaks, particularly evident among buyers from African countries, in post-war public housing areas characterized by low-quality housing conditions.

Asian buyers tend to concentrate their purchases in areas with a high proportion of residents from similar geographical origins and cultural backgrounds. This suggests a desire, as well as the capacity, to seize residential opportunities offered by this policy to reside in specific neighborhoods. These efforts are possibly linked to expectations of mutual support and resource exchange through migrant networks, which might have also played a role in disseminating crucial information about upcoming purchase opportunities within neighborhoods and estates, potentially influencing bidding strategies. This could include prior knowledge of the number of potential bidders and their offers. Therefore, in addition to financial resources, it is worth investigating how the social capital of prospective homebuyers could have played a significant role in determining bidding strategies.

These findings contribute to the growing international debate on ethnic residential segregation and immigrant homeownership in Southern European cities (Arbaci 2019; Balampanidis 2020; Benassi et al. 2020; Maloutas and Spyrellis 2020; Mazza et al. 2018) by uncovering the neglected role of public housing privatization policies, and their outcomes, in shaping the spatial distribution of immigrant populations and their housing pathways. Faced with a limited supply of social housing and increasing difficulties in accessing the private rental market due to discrimination, some immigrants actively employ alternative strategies to achieve homeownership under relatively favorable conditions compared to the “regular” housing market.

This paper has several limitations, primarily pertaining to the nature of the data and the chosen analytical approach. Despite its innovative application within the context of public housing sales, the name–origin approach lacked the capability to ascertain whether immigrants with foreign names had undergone naturalization and, if so, whether they should have been categorized as Italians. In addition, it is worth noting that the various immigrant groups may have displayed a considerable range of socio-economic conditions, which proved impossible to consider when explaining the observed spatial outcomes. Last, the analysis was based on a limited number of cases, thereby constraining the generalizability of the results, although it did serve to shed light on a relatively unexplored research subject.