Shared Housing as Public Space? The Ambiguous Borders of Social Infrastructure

The Folkhem era in Sweden set high architectural standards for social infrastructures dispersedly located in cities. Over the past two decades, however, Swedish planning, when it comes to the localization of social infrastructure, has been increasingly characterized by privatized social infrastructures added to housing. Methodologically, this article draws on a compilation of architectural designs of shared housing that includes social infrastructure, 12 interviews with developers, and 22 interviews with residents. The article argues, first, that two historical approaches can be identified: one in which porous borders support urban social life in and around the housing complex and another where distinct boundaries form an edge where things end. Secondly, the article argues that in recent shared housing complexes, the infrastructures of fitness, health care, and privatized services—previously available solely in the public realm—have moved physically and mentally closer to the individual, largely replacing residents’ everyday use of public space. The article concludes that in recent shared housing complexes, ambiguous borders are formed. Ambiguous borders allow a flow of goods and people, but the flow is based on the needs and preferences of residents only. Overall, such privatization counteracts the devel‐ opment of urban social life while adding to housing inequality, as this form of housing is primarily accessible only to the relatively wealthy. Furthermore, there is a risk that urban planning may favour such privatization to avoid maintenance costs, even though the aim of planning for general public accessibility to social infrastructure is thereby shifted towards planning primarily for specific groups.


Introduction
A substantial amount of research has theorized social infrastructures and how shared public spaces can counteract "inequality, polarization and the decline of civic life" (Klinenberg, 2018), how "eyes on the street" support safety in cities (Jacobs, 1961), and how the design of public places supports a shared "life between buildings" (Gehl, 1971).Social infrastructure comprises publicly accessible places and includes a varied typology of public institutions, commercial centres, places of worship, transport infrastructure, and public places such as squares, parks, and pavements (Klinenberg, 2018;Latham & Layton, 2019).Housing, however, has been almost entirely overlooked in social infrastructure studies.This lack of attention may seem self-evident since social infrastructure forms the "background structures and systems that allow social, economic, cultural and political life to happen" (Latham & Layton, 2019).Nevertheless, housing complexes can include several of the architectural typologies, organizations, and institutions identified as social infrastructure.A place of worship or a corner store can be located in multistorey housing.Larger housing complexes may include a community centre and shared outdoor spaces for recreation and urban gardening.
In Sweden, planning for social infrastructureschools, health centres, and nurseries, as well as parks and playgrounds-has been central to urban planning since the Folkhem (people's home) era.Between the 1940s and the 1970s, and especially up until the 1960s, the Folkhem design approach of simplicity, honesty and purposefulness, coupled with an assertion by developers that "only the best is good enough for the people," resulted in high-quality residential architecture and urban design (Nylander, 2013).Urban planning and design became a tool for providing social infrastructure to the Swedish population during a period of increasing socio-economic equality.High architectural standards were set for libraries, community centres, sports arenas, schools, nurseries, and health centres, all of which were located so as to be accessible for all residents (Eriksson, 2001).Over the past three decades, however, Swedish cities have been increasingly characterized by privatized social infrastructures.Since the 1990s, Sweden has seen the deregulation and privatisation of such social infrastructures as schools and care services, as well as sports arenas and pharmacies, all of which were previously provided and organized by national and local governments (Hedin et al., 2012).
The relationship between urban planning and social infrastructure in Sweden must be understood in relation to the Swedish welfare regime.During the Folkhem era, the ruling Social Democratic party developed a welfare regime that benefitted both the working class and the white-collar middle class (Esping-Andersen, 1990).According to Esping-Andersen (1990), this explains the extraordinarily high costs of Swedish welfare, which provides population-wide free or heavily subsidized social infrastructure in the areas of education (skola), health care (vård), and the care of children and the elderly (omsorg).The provision of welfare and welfare institutions in the areas of education and health care, which was organized by national and local governments, coupled with the responsibility of municipalities to provide affordable housing and recreational areas, all meant that urban planning had to cover a broad range of social infrastructure.Additionally, Swedish municipalities have a planning monopoly, which during the Folkhem era put the municipalities in a position to plan the entirety of social infrastructure-everything from playgrounds and pavements to schools and hospitals-within a planning paradigm of "normative rationality" (DeVerteuil, 2000).As mentioned, however, recent decades have seen a privatisation and marketisation of education, health services, and care service.This development has occurred in parallel with an increasingly social and geographic polarisation of Swedish cities (Grundström & Molina, 2016;Hedin et al., 2012).The privatised forms of social infrastructure compete in their respective markets.From a perspective of localising social infrastructure, the result is that private schools and private health care centres are located in areas that are deemed attractive.Pharmacies are often located in well-to-do downtown areas while vulnerable areas experience a retraction of welfare institutions (Urban, 2016).In addition, housing has entered onto the scene as a new actor providing social infrastructure.
Housing segregation in Sweden's three largest metropolitan areas, Stockholm, Gothenburg, and Malmö, is considered a concern with a large societal impact and a driver of inequalities at large (Hedin et al., 2012).On the one hand, research points towards increased precarity: a decrease in housing standards and in the size of newly constructed dwellings (Grundström, 2021b), a displacement of vulnerable groups due to renoviction (Baeten et al., 2016), a lack of affordable housing and an emerging housing precariat (Listerborn, 2018).On the other hand, research shows an increase in housing wealth among privileged groups (Christophers & O'Sullivan, 2018) and a housing sector engaged in constructing dwellings for middle-income and wealthier groups.In this context of housing segregation and inequality (Dorling, 2014), shared forms of housing are marketed and sold based on their incorporation of residents-only infrastructures, such as private lounges, spas, gyms, cinemas, billiard rooms, restaurants, winter gardens, or gardens with places to play tennis or boules.Health care is provided in the home, residents may avail themselves of cleaning, and catering and childcare services are provided by hired staff.Outside the realm of shared housing, private housing associations (bostadsrättsföreningar) have also begun adding these more exclusive spaces.As the cost of housing has increased, dwelling size has decreased, which has led to a demand for social infrastructure that adds value to housing, as explained by a realtor in Stockholm (Nordlander, 2019).Housing is thus helping to reformulate which spaces form part of the public realm and which ones are private-which spaces are part of urban planning and which spaces are part of housing design in the private market.
The incorporation of social infrastructure into housing has its longest historical trajectory in shared forms of housing, such as co-housing.Co-housing has primarily been conceived of as a form of housing based on togetherness and sharing reproductive work (Hayden, 1981;Lang et al., 2020).Furthermore, co-housing has been considered a form of housing that supports de-growth and social sustainability (Kries et al., 2017;Vestbro, 2010).In recent years, however, new forms of shared housing have also come to include both very exclusive dwellings (Grundström, 2021a;Westholm, 2019) as well as precarious housing (Bergan et al., 2020).Recent forms of shared housing include social infrastructures such as a cinema, wine cellars, meeting rooms, a restaurant, a spa, swimming pools, a gym, tennis courts, and boules and barbecue areas.Residents can work remotely and receive medical and health care in their homes, while goods and other services, such as cleaning services, can be provided on-site.The incorporation of social infrastructure into housing raises questions about how "boundaries and borders" (Sennett, 2018) in cities are reformulated and shaped, and the potential consequences this has for urban planning and inequalities in cities.How has the design and planning of shared housing evolved historically?To what extent has shared housing been built and lived behind "distinct boundaries" (Sennett, 2018) or through open, "porous borders" (Fainstein, 2010;Sennett, 2018)?How does the incorporation of social infrastructure influence the daily practices of residents and accessibility of the general public?Who has access to which places, to which services-to which social infrastructure?Despite increasing socio-economic differences and the inclusion of social infrastructure in housing, few studies have analysed precisely which types of previously public spaces are being moved from the public realm into housing.And few studies, if any, have analysed the potential consequences of urban planning and housing inequality.The first aim of this article is to analyse which types of social infrastructure have historically been included in shared housing.The second aim is to analyse how social infrastructures in housing influence residents' daily practices and what the consequences may be for future urban planning.The article argues, first, that two historical approaches in incorporating social infrastructure into shared housing can be identified: one in which "porous borders" (Fainstein, 2010) support urban social life and another where "distinct boundaries" (Sennett, 2018) form around the housing complex.Secondly, based on two current cases, the article argues that social infrastructure for everyday social interaction, fitness, health care, and privatised services-previously available solely in the public realm-has moved physically and mentally closer to the individual.Residents lounge with the like-minded, they tend to prefer spontaneity over commitment, and their daily interactions take place primarily within the housing complex.Even though the inclusion of social infrastructure in housing supports co-presence, friendship, care, and collective experiences, this introverted way of life means that civic engagement in the city outside is reduced.The article concludes that social infrastructure in shared housing to a large extent replaces residents' everyday use of public space.Ambiguous borders are formed that seemingly allow a flow of goods and people, but the flow is based on the needs and preferences of residents only.
Thus, the boundaries, borders, and scalar implications of investments in social infrastructure need to be analysed as welfare regimes change and the relationship between the private and public realms is reformulated.The inclusion of social infrastructure in recent forms of shared housing risks contributing to housing inequality, as it is primarily the relatively wealthy who can access this form of housing, even as social infrastructure in vulnerable areas is decreasing due to austerity policies.There is also a risk of urban planning tending to favour privatisation and incorporation of social infrastructure into housing in order to avoid costs for maintenance, even though doing so shifts the aim of planning for a general public towards planning primarily for specific groups.

The Boundaries and Borders of Social Infrastructures in Shared Housing
Access to social and technical infrastructure has been an issue of global concern for decades.Organizations such as UN-Habitat and the World Bank have raised the question of access to infrastructure for the poor.Although infrastructure may be physically located close to where urban poor groups live, it is rarely accessible to the poor, as access is not free of charge (Grundström, 2009;UN-Habitat, 2020).The meaning of infrastructure as a key issue in theoretical work was pinpointed by Graham and Marvin (2001) in their work on "splintering urbanism.''Their analysis of infrastructure, and in particular private systems for infrastructure provisionincluding electricity, water, and telecommunication systems, but also streets and highways as well as skywalks and plazas-shows that infrastructure segregates as much as it connects.Graham and Marvin identify infrastructure primarily as digital and material connections, but also raise social implications of use and access.The field of infrastructure studies has since come to include a broad range of issues.In particular, Klinenberg (2018) has focused on social infrastructure, broadening its meaning to include not solely the provision of welfare services like education and health care, but also public spaces and institutions.
Social infrastructure is capaciously defined by Klinenberg (2018) as public institutions, public places, community organizations, and commercial establishments.This definition includes both physical and social spaces and both public and private institutions.Libraries, sidewalks, nurseries, and corner stores are all included if they support "urban social life" and "socializing between different socio-economic groups" (Klinenberg, 2018).According to Layton and Latham (2021), social infrastructures are "the in-between structures and systems that afford and support action" and collective experience.Theorising infrastructure draws attention to the facilitation of activity: how facilities are provided and how spaces are practised (Layton & Latham, 2021).Investigations of social infrastructure provide an argument for the importance and affordances of public space.Social infrastructure affords social connections, supports urban social life, and contributes to people's well-being as well as their mental and physical health (Klinenberg, 2018).Social infrastructure in the form of neighbourhood houses supports the social capacity of newcomers to Canada (Lauer & Chung Yan, 2022), community organizations support connections that combat loneliness among groups of ageing adults in the US (Fried, 2020), and a park in London can support several modalities of sociality, from co-presence to civic engagement (Layton & Latham, 2021).The positive connotations and concerns related to social infrastructure have resonated with scholars in architecture, urban planning, and urban design.Concepts such as "functions," "affordances," and the "facilitation of activity" that have been used to analyse social infrastructure (Latham & Layton, 2019) are also key in the work of architects and planners who argue for the importance of "cities for people" and "liveliness" in cities (Carmona et al., 2010;Gehl, 2010;Jacobs, 1961).
Despite such arguments for the positive connotations of social infrastructure, several authors have pointed to the challenges of social infrastructure, both in its "built" and in its "lived" form (Lefebvre, 1974(Lefebvre, /2007;;Sennett, 2018).Concerns have been raised about treating the forms of sociality valued by some as "universally valued" and have suggested that empowering forms of infrastructure at the local scale may "simultaneously constrain its transformative potential" in other scales (Middleton & Samanani, 2022, p. 4).Civic stewardship groups in New York City differed in their ability to create a more sustainable city, depending upon the degree of group connectivity and the scale at which the groups worked (Campbell et al., 2022).A study of a Dutch library showed that co-presence and co-mingling were the dominant types of socializing even in an exemplary public space such as a library (van Melik & Merry, 2021).The authors conclude that "the actual-taking place of meaningful encounters is difficult to arrange (or to 'infrastructure')" (van Melik & Merry, 2021, p. 17).In urban planning, the importance of supporting urban social life and allowing people to meet with differences has also been raised by scholars who have lamented inequality, segregation, zoning, and division in cities. Fainstein (2010, p. 174) suggests that public space should be "widely accessible and varied" and "borders between districts should be porous."In a similar vein, Sennett (2011), identifies a distinction between boundaries and borders.Sennett argues that distinct boundaries "establish closure through inactivity, while the-edge-as-border is a more open condition" which leads to more events and more liveliness (Sennett, 2011, p. 265).
Boundaries versus borders, or porosity versus closure, are key in defining accessibility to social infrastructure.While the boundary is "an edge where things end," the border is "an edge where different groups interact" (Sennett, 2018, p. 220).Nevertheless, a black-and-white distinction between the boundary and the border is too crude.Rather, the function of a border can be understood as similar to that of a cell membrane, which selectively lets matter flow in and out.Such urban membranes can be made of bricks and mortar, of solid walls, and of the social spaces that form around them (Sennett, 2018).Boundaries, borders, and porosity can be analysed on an urban or a neighbourhood level, but also at the level of the relationship between buildings and the streetscape, according to Sennett, who argues that "making buildings more porous…could make buildings more truly urban" (Sennett, 2011, p. 266).This would include housing, which, through the inclusion of social infrastructures, can be made truly urban and can be part of the urban fabric and urban social life.The social infrastructure in the form of stores, pizzerias, nurseries, and small offices has been shown to support a social street life (Gehl, 1971;Jacobs, 1961).Furthermore, feminist critique has shown that for many women, housing is a place of work embedded in and dependent on an open relation to urban social life (Hayden, 1981;Sangregorio, 1994).Borders as membranes include both the built and the lived space, or as Sennett (2018) defines it, the "ville and the cité."Thus, the relationship between housing and social infrastructure is an issue not solely of built form and typology, but also of sociality, including how residents live their daily lives, the routines that shape their every day and their social interactions, and their spatial practices.According to Lefebvre (1974Lefebvre ( /2007)), spatial practice is part of constructing hierarchies in society, and captures the relation between physical and social space-the built and the lived-in the daily lives of residents.In a similar vein, Sennett (2018) argues that both the ville (the built) and the cité (the lived) form the ethics of building and dwelling in cities.
This framework will assist the following analysis of which architectural typologies of social infrastructure are included in shared housing, how daily life is practised, and how boundaries and borders form in shared housing.

Methods
Geographically, the research is focused on the metropolitan regions of Stockholm and Malmö.The selection strategy was information-oriented (Flyvbjerg, 2006).Stockholm and Malmö represent the wealthiest and poorest of the metropolitan regions and both cities have a substantial housing deficit.Furthermore, they have a high percentage of one-person households.This is important, since the rise in one-person households, i.e., "singletons," is changing how people are housed in cities and which spaces and resources they share.Singleton populations have soared in all the metropolitan regions of Europe, North America, and Australia (Klinenberg, 2012).Sweden has one of the largest singleton populations: 54% of households in Malmö and 55% in Stockholm are singletons (Statistics Sweden, 2021).Shared housing was selected for analysis since it is the most information-rich housing form, as well as the one with the longest history of including a variety of social infrastructures.The three existing types of shared forms of housing (Grundström, 2021b)-co-housing, co-living, and residential hotelsare built in both Malmö and Stockholm.
In the first phase of the investigation, information about the three architectural types of shared housing was compiled.Statistics that show exactly how many individuals live in shared housing are limited.However, using government-gathered statistics from Statistics Sweden on "household type," along with the Income and Wealth Register, the Population Register, and the Property Register, made it possible to isolate the number of households made up of individuals living with individuals other than their spouse/registered partner and/or children (Statistics Sweden, 2021).In total, 1,003,563 individuals in Sweden currently share housing with people other than their family members, which amounts to almost 10% of the total population.The compilation of the three types of shared housing was based on previous research, presentations of shared housing from the national co-housing network, web pages about shared housing, and information from the Swedish National Board of Housing and Planning (Boverket).In addition, 12 interviews with 15 operators and developers of shared housing were conducted.Interviewees were project leaders and founders and co-founders of shared housing.They were selected based on their engagement and digital visibility as actors involved in shared housing.The interviews lasted between 40 and 80 minutes and were structured around themes that included the interviewee's model of shared housing, the social aspects and physical design of shared housing, the target groups the interviewee had identified, and what these groups shared.The interviews were conducted during the spring of 2021, and due to restrictions imposed by the Covid-19 pandemic, all interviews were conducted via Zoom or Microsoft Teams.
During the second phase of the investigation, 22 residents of shared housing were interviewed.Since age and housing type were strongly related, one housing case developed for seniors and one developed for younger residents were selected.For the senior case, four women and six men living at the Bovieran complex in the Malmö region were interviewed.Bovieran, which translates to "riviera living," aims to mimic Southern Europe in a Swedish climate.It includes fully equipped two-and three-room apartments plus a winter garden, boules court, and community space, and it is aimed at residents aged 55 and up.All of the interviews at Bovieran were carried out on-site.The second case was co-living housing, i.e., shared housing developed for younger residents.Co-living complexes include bedrooms for one or two residents plus shared spaces for eating and relaxing, gym and yoga, and co-working and socializing.In all, 12 interviews with co-livers in Stockholm and the Malmö region were conducted.Nine of the interviews were conducted with current residents and three with residents who had moved out.Among the co-livers interviewed, six were women and six were men.In each case, the selection of interviewees was based on an initial interview with a contact person, followed by snowballing and balancing numbers of women and men.The interviews lasted between 30 and 70 minutes and were structured around themes that included how residents socialized inside and outside their shared housing, what made people fit in or not, which types of social infrastructure residents shared and what they did not want to share, regulations residents had to follow, and whether residents experienced changes in their daily habits after moving in.The interviews were conducted during 2021, and due to restrictions imposed by the Covid-19 pandemic, the majority of the co-living interviews were conducted via Zoom or Microsoft Teams.For a limited period, it was possible to conduct interviews on-site; therefore, a total of five co-living interviews were carried out on-site.When interviewed on-site, residents offered to show their shared spaces as well as their private rooms or apartments.All interviews were anonymised, and the names of the two co-living complexes investigated have not been revealed in order to preserve anonymity.
All interviews with developers and residents were transcribed in full and then both deductively and inductively coded.In addition, to reflect on the interviews, notes from interviews and discussions with residents as they showed us around were compiled, and policy documents and web pages about shared housing were reviewed.

The Architecture and Planning of Social Infrastructure in Shared Housing
Adding social infrastructure to housing has a long history in architecture, urban planning, and design.Two historical approaches can be identified.One approach involves combining social infrastructure and housing with the aim of bringing urban social life and private dwellings closer together through porous borders.The other involves privatizing social infrastructure for the sole use of residents by drawing clear boundaries between the housing complex and the city outside.Both approaches are clearly discernible in Swedish shared housing.
Co-housing sprang out of ideas developed by Charles Fourier in the early 19th century.Fourier's phalanstère, or "social palace," elaborated in 1808 (Helm, 1983), was the first architectural housing design that included what Klinenberg (2018) terms social infrastructure.Fourier proposed large building complexes where communities consisting of a mix of professions and social classes would live and work together according to the principles of collective property, social interaction, and sexual freedom.The phalanstère was a utopian architectural design for an entire community that included all forms of social and technical infrastructure, from libraries to nurseries, schools, walking galleries, and governmental offices, all under one roof.Fourier's concept was never built in its entirety, but evolutions of the concept formed the basis for various later combinations of housing and social infrastructure.These include housing in support of workers, housing in the housekeeping and cooperative movements (Kries et al., 2017), and feminist home design and community planning (Hayden, 1982).One of the internationally best-known examples of a combination of social infrastructure and housing is Unité d'Habitation by Le Corbusier.Built in Marseille, France, between 1947 and 1952, it included apartments, shops, restaurants, a nursery, and a health centre combined with indoor streets.The phalanstère as well as the unité d'habitation are examples of setting a distinct boundary (Sennett, 2018) outside the housing complex while porous borders were thought to be formed within.
One of the earliest shared housing complexes in Sweden was the so-called "Markeliushus," a kollektivhus (collective house) that was designed and built in Stockholm in 1932-1935.The Swedish kollektivhus of the early 20th century was developed by leading pioneers of the modern movement who argued that housing would contribute to a new, modern, rational, and democratic citizenry.The overarching principle for the new society was the collective-hence the name kollektivhus.Women and men would work outside of the home and participate in political meetings and debates, while sports and leisure facilities in dispersed locations would contribute to a healthy population (Vestbro, 2010).Housing needed to be organized and designed to support this new collective organization of society (Hirdman, 2000).In addition to the 50 apartments of the Markeliushus, the entire ground floor of the six-storey building comprised primarily "social infrastructure," including a nursery, a grocery shop, and a restaurant.Even though the Markeliushus was run by a housing association and served its residents, the ground floor was publicly accessible and would have supported "social activity generated by street-level commerce" (Klinenberg, 2018, p. 76).People going to the restaurant and picking up their children from the nursery would contribute to the social "life between buildings" (Gehl, 1971).The addition of social infrastructure to housing complexes is also an approach that supports a "porous" border (Fainstein, 2010) between urban public life and private homes.The ground floor of the Markeliushus constituted a border as a membrane (Sennett, 2018); residents, employees, and citizens could all flow in and out of these spaces.
During the 1960s and 1970s, the focus of shared housing shifted towards co-housing and the sharing of reproductive and maintenance work.A system in which housework was done collaboratively and by both women and men could help reduce time spent on housework and shape a society based on gender equality.According to Sangregorio (1994, p. 73) the fundamental ideas were to "save material resources and liberate human resources."The overall design of co-housing included fully equipped apartments and was built on the idea of "more for less" (Kärnekull, 1991).If 40 households gave up 10% of their square footage, residents could instead have a library, TV room, dining room, sauna, laundry room, table tennis room, and workshops.The kitchens, designed to cater to communal cooking, were the hearts of these housing complexes, and residents jointly prepared and shared meals during the week.Shared housing became less publicly accessible during this period.Although the struggle for gender equality was a political goal of broad societal concern, the design of co-housing focused on residents rather than communities and thus contributed less to "urban social life" (Klinenberg, 2018).Libraries, saunas and urban gardens were accessible solely by residents, thus constructing more distinct "boundaries" (Sennett, 2018) towards the urban environment.Even though some co-housing complexes may invite non-residents to take part in activities such as cooking or sharing meals (Westholm, 2019), co-housing is generally focused on residents only.
The newest shared housing complexes seek to fulfil a rising demand for comfortable housing with services included for professionals who travel extensively or for retirees who enjoy leisure and lounging.Several initiatives have been launched, including exclusive residential hotels (Grundström, 2021a), complexes such as Bovieran for people aged 55 and up, and co-living hubs for international knowledge economy workers "who can work anywhere as long as they have a laptop" (Müller, 2016, as cited in Bergan et al., 2020Bergan et al., , p. 1208)).Here, the focus is no longer on the "collective" or the "collaborating community," but rather on the "creative class" (Florida, 2001).Sweden's first residential hotel, Victoria Park, was inaugurated in 2009 and included a staffed reception area, a lounge, a billiard room, a cinema, wine cellars, meeting rooms, a restaurant, a spa, swimming pools, a gym, tennis courts, and boules and barbecue areas (Grundström, 2021a;Victoria Park, 2007).Based on this first example, two forms of social infrastructure have also spread into less exclusive housing types, including the Bovieran complex and co-living housing.The first are places to socialise.At Bovieran, social spaces include a winter garden with seating and a place to play boules; co-living spaces may include a shared living room for hanging out or watching movies.The second is fitness facilities such as gyms and yoga studios, accompanied by services to support well-being and health.In addition, digitalisation has blurred the notion of which activities are part of the use of a dwelling and which belong to the public realm.Goods or services, such as cleaning or health care services, are ordered online and brought to the home or carried out within the private sphere of the housing complex.Yet another consequence of digitalization is that increasing numbers of people work from home.Overall, these new forms of shared housing are primarily geared towards residents, supporting a form of "club goods" (Manzi & Smith-Bowers, 2004).While there may be a need for social infrastructure for vulnerable groups, as Klinenberg (2018) exemplifies using the case of U.S. barbershops, the majority of residents in the Swedish examples discussed here are not socio-economically vulnerable.Rather, the main consequence of this housing is the privatization of social infrastructure-access to health care services and to public places such as gyms, parks, and places to socialise-that previously contributed to urban social life.
In sum, the historical trajectories show that despite the existence of co-housing in support of "urban social life" (Klinenberg, 2018;Westholm, 2019), both "distinct boundaries" and open, "porous borders" (Fainstein, 2010;Sennett, 2018) have existed as social infrastructures have been added to shared housing complexes.Importantly, the historical examples identified above had little or no bearing on the planning ideologies of their respective eras.Shared housing has been built as part of urban blocks in historical inner cities, as freestanding slabs in modernist areas, and as complexes in the urban periphery or in the countryside (Vestbro, 2010;Westholm, 2019).In the most recent forms of shared housing, the "built" and the "lived" are practised in complex and contradictory ways.

Dwelling and Practicing Social Infrastructure in Shared Housing
The tendency towards more enclosed shared housing complexes, such as Bovieran and the co-living complexes, poses challenges and represents a counter-development to the role of social infrastructure in fighting "inequality, polarization and the decline of civic life" (Klinenberg, 2018).Although differences exist between Bovieran and co-living residents, the two groups also hold several practices in common.
First, there is a tendency towards socializing or lounging with the like-minded, partly due to the fact that these are rather specific forms of housing that most residents actively choose to live in.Residents' perceptions of how similar or different they are in relation to other residents vary, but they tend to mention certain similarities.Anne-Marie, a Bovieran resident in her mid-70s, acknowledged that such similarities exist, saying: "People are fairly similar here and it is possible that it is a certain type of people who are attracted to this.It is people who are very active."Another Bovieran resident, Nils, who was in his mid-80s, "And, you are the same age or the same generation, and you have a social exchange of things."Bovieran residents are elderly and some receive medical and health care in their private homes.Pernilla, a resident in her mid-80s, explained that many of the residents would "live [t]here until the very end."One consequence of this awareness of illness and old age is that residents check in on each other and make sure that everyone is fine.Anna, a co-liver in her 30s, thought that co-living was a "very active choice and it [was] definitely people that value[d] social interaction a lot."This goes for both Swedish as well as international co-livers.Anders, a co-liver in his mid-20s, thought that there were differences between people in his co-living complex, but there were also "groups of almost modern hippies.A lot of raves, a bit spiritual…when it started it was sort of for people in the tech industry and for start-ups."Bovieran and the co-living complexes thus appear to be forms of housing primarily for social, extroverted people who also are very active in organizing activities.Bovieran has groups for cooking, gardening, exercise, and boules as well as a group that organizes parties and activities for well-being.Co-living residents organise social activities such as watching TV together (sports, series, or shows), having coffee in the shared living room, cooking together, going out to dinner or a museum, or going for a walk or a run.The similarities that residents experience and the activities organised among them sup-port several modalities of sociality found in urban places: "co-presence, sociability and friendship, and, care and friendship" (Layton & Latham, 2021).But simultaneously, and in contrast to the city and its value as a place where to meet difference (Fainstein, 2010), these communities lead to less opportunity on a neighbourhood or urban level to "foster contacts, mutual support and collaborations among friends and neighbours" (Klinenberg, 2018, p. 18).Certainly, cohesion, care and friendship develop between residents, but these do not support the development of urban social life to any great extent.
Secondly, residents' social life is centred around their daily interactions within the housing complex itself.Residents spend a substantial amount of time in their homes at Bovieran because most residents are retired, and in co-living complexes, because many residents work remotely, either full-time or for long periods.Furthermore, they live in housing designed for people to meet in as part of their daily lives.This apparently leads to more social interaction among residents and less interaction with non-residents.Anna explained that she mainly socialised at home, with other co-livers, saying: "I mean, I feel like I'm very bad in keeping my relationships outside the house…it is very easy because, like, you have a lot of people that you like, you get along with and you live with them."Margareta, who was in her late 60s and lived in Bovieran, explained that she had no lack of friends outside Bovieran, but said that at Bovieran, "there are so many people to talk to and it's really very nice."Anders explained: "There have been occasions when I have been invited to visit someone [outside of the co-living complex].And then it turns out that there is something going on here, and then I feel more like being here with the people I live with."He went on to say that this sounded very harsh, and he did invite his outside friends to visit, but even so, one became very close to the other residents.Ingrid, a Bovieran resident in her mid-70s, explained that the architectural design made it easy to meet other residents and start a conversation if one wished to do so, saying: "In this place, you meet people to talk to just by going to the post box to pick up your mail."The design of these housing complexes affords socialising in the lounge, in the winter garden with its clusters of seating, at the gym, or in the shared kitchen where people can choose to cook together.Both forms of housing also support residents' health and well-being.In addition to its winter garden, Bovieran also includes spaces for boules, a gym and a sauna, and the co-living complexes include a gym and yoga studio.These are all "places to gather" (Klinenberg, 2018), but they are solely for residents, who actively arrange activities and spend most of their time in their home environment.Klinenberg (2018) gives the example of the library and the urban park as examples of social infrastructure that offers a wide range of activities for diverse groups of people.Places such as winter gardens, gyms, and yoga studios support the health and well-being of residents, but they could play a more inclusive role as places where non-residents could also meet and interact if they were made publicly accessible rather than privatised.These housing complexes do not have a design based on porosity (Fainstein, 2010) which would "make buildings more urban" (Sennett, 2011, p. 266), and supportive of events and liveliness.
Third, residents tend to prefer spontaneity over commitment.Social infrastructure in both physical and digital forms is close at hand.Most residents mention the importance of not needing to sign up or plan for activities in advance.In contrast to the mandatory tasks in kollektivhus housing, the Bovieran and co-living housing forms are based on choice, voluntary participation, and casual day-to-day socialisation.Groups that cater to residents' interests and well-being organise social activities that residents can join without signing up in advance or on just a few minutes' notice.Ingrid, a Bovieran resident in her mid-70s, stressed the importance of not being forced to participate in any of the activities organized by residents, such as going on a group walk, cooking dinner together, playing boules, or watching a sports game on the big-screen TV in the winter garden.She said: "It's important that you don't feel forced to do things, everything is voluntary.All the time.There will be more of the fun stuff when you are not compelled to do things.It is important that it is voluntary."Residents of Bovieran felt relieved that they no longer had a large house and garden to care for and argued that they wanted to use their newfound freedom to choose which activities to join.Co-livers argued that they were busy and needed to be able to make decisions quickly both at work and during their leisure time, since "something [might] come up."One aspect of social life in Sweden is that plans tend to be made well in advance.Anna complained that she might be invited to visit friends "a month in advance" and that it made her feel bad to have to commit to something instead of being able to join spontaneous activities in the co-living complex.She said: "And the thing is, there are a lot of things that are happening spontaneously.And you can do all of this with no effort.Basically, you will get out of your room and you join whatever is happening."Digital platforms are used to share information about activities that come up.Anna explained: "You check on Slack if there are events, if people are going somewhere, and then you decide what to do." Klinenberg (2018) critiques the community-building vision of social media, arguing that social media platforms cannot substitute for social infrastructure, nor can social media provide a safety net or a gathering place.In co-living and Bovieran, however, social media and gathering places overlap and form a strong connection to the housing complex as well as to social interaction without much effort.This contrasts with previous notions of shared housing, such as the Swedish kollektivhus, which were built on porous borders that made buildings urban.The combination of shared housing and remote work, or retirement, adds another dimension to daily life.But in these cases, social media and the design of the housing complexes overlap and add to an introverted and rather enclosed community.
In sum, the daily practices (Lefebvre, 1974(Lefebvre, /2007) of these two resident groups form in close proximity to their dwelling complex.Residents interact with others who are similar to themselves, and their daily life plays out in the home or in the social infrastructures for lounging, fitness, socialising, and co-working that are physical and digitally close at hand.In these recent, introverted forms of shared housing, encounters with other people from different socio-economic circumstances are reduced.The city is still out there, but socializing happens close to home.

Shared Housing as Public Space? The Ambiguous Borders of Social Infrastructure
This article traces the historical trajectory of shared housing and identifies how shared housing in Sweden has increasingly become more introverted as the originally porous borders between private dwellings and social infrastructure (Helm, 1983;Kries et al., 2017) have shifted towards a more introverted sociality.The shared housing complexes analysed in this investigation comprise a complex network of social infrastructure, the materiality of which includes, e.g., a winter garden, a lounge, a billiard room, a cinema, wine cellars, meeting rooms, a restaurant, a spa, swimming pools, co-working spaces, a gym, tennis courts, boules, and barbecue areas and a staffed reception area.The institutions and organisations offering and managing social infrastructure include private housing associations, restaurant owners, and a wide range of businesses offering services-ranging from cleaning and dog-walking to childcare and in-home eldercare-of which some are privately operated and others are tax-funded and operated by local authorities.What is evident within this complexity is the recent inclusion, within shared housing, of varied forms of social infrastructure previously found solely in the public realm, where they were accessible to all.
Shared housing, such as Bovieran, co-living complexes, and other similar forms, have a material demarcation, a distinct wall, against the city outside.Characteristically, a single entrance leads into the housing complex where all the social infrastructures that the residents share are located.The city's public space is thus in a sense incorporated into the housing complex, in proximity to the dwellings.The shared space does not remain public, as public accessibility does not exist.But for the residents, the social infrastructure functions to a certain extent as a replacement for public space in their everyday life.This incorporation of public space in the housing complexes, and the demarcation against the city outside, may seem like a shift towards a "distinct boundary" from a "porous border" (Sennett, 2018), as seen in previous forms of co-housing.However, with respect to the flows of goods, people, and services that move in and out of today's shared housing complexes, what is evolving is an ambiguous border.
Ambiguous borders function primarily in a one-way direction, and for the benefit of a specific group of residents.According to Sennett (2018), the "porous border" can be understood as a "membrane" that allows the flow of people and goods in both directions.The ambiguous border seemingly allows a flow of goods and people, but the flow is based on the needs and preferences of residents only.The ambiguous border supports activities that were previously carried out in other parts of the city or in the public realm.Instead of leaving the housing complex, residents can work remotely, receive in-home health and medical care, chat with neighbours, provide a meeting place for customers or for work, give large dinner parties or play boule with relatives and friends, and receive delivery of services and goods at home.People and goods from the outside enter through the ambiguous border to share in the community within or to offer their services.The inclusion of social infrastructure through ambiguous borders supports a variety of modes of sociality, from "co-presence, sociability and friendship, care and kinship to kinesthetic practices and collective experiences" as identified in public places by Layton and Latham (2021, p. 12).Missing are the "carnivalesque" and the "civic engagement" modes of sociality which are key to the public, urban social life.
The tendency to live and socialise others like oneself is a clear sign of segregation.This tendency is apparent not only in shared housing but also at the urban level, as polarisation and socio-economic differences have increased in Sweden since the 1990s (Christophers & O'Sullivan, 2018;Grundström & Molina, 2016;Hedin, et al., 2012).At present, the amount of shared housing with social infrastructure is still limited and future development uncertain, but there is undoubtedly growing interest from developers and the private sector in marketing and selling housing with social infrastructure included (Nordlander, 2019;Westholm, 2019).It should be noted that rooftop terraces, saunas, and gyms, along with other services, may be included not only in shared housing but also in other types of privately owned housing complexes.This form of investment in and provision of social infrastructure in housing suggests yet another dimension to issues of housing inequality.Even as a concentration of assets (Dorling, 2014) is taking place for residents in shared housing, a simultaneous polarisation process risks deepening socio-economic vulnerability by decreasing housing standards (Grundström, 2021b) and access to social infrastructure (Urban, 2016) in poorer neighbourhoods.
From a planning perspective, the ambiguous borders of shared housing show how important it is for urban planning and design to identify boundaries and borders and at what scales they exist.Localising social infrastructure has been, and still is, central to urban planning and design.When shared housing and privately owned housing associations add social infrastructure to their dwellings, those decisions are made by private entities, meaning that public urban planning has little or no influence on such localisation of social infrastructure.This raises issues of public accessibility.How will future urban planning take the localisation of social infrastructure into account when certain groups-the middle and upper-middle classes-can provide such infrastructure for themselves through housing?Furthermore, digitalisation and the experiences of the Covid-19 pandemic have led to increased numbers of people working from home.This rise in remote working has gone hand in hand with soaring numbers of singletons in the metropolitan regions of advanced economies (Klinenberg, 2012).What kind of demand will there be for co-working spaces, leisure and fitness centres, daycare, and cleaning services in future housing?And what will be the response from developers?Politicians in local government may favour privatising certain social infrastructures in order to reduce maintenance costs.There is a risk that the combination of housing and social infrastructure, although it could contribute to making buildings more urban through porous borders, will instead have a stratifying effect on segregation and add to housing inequality.There is also a risk that it will shift the aim of planning for general public accessibility towards planning primarily for specific groups.
How and where we live in cities matters, and that includes which social infrastructure we share.Our cities have room for shared housing and counter-communities that focus on residents' well-being and on support for small groups of residents.It is correct that shared housing has been, and still is, based on de-growth and sharing reproductive work, which is much needed in many communities.But at the same time, the number of enclosed housing complexes is increasing, and ambiguous borders continue to evolve.In this context, the role of urban planning to support public accessibility and the fair distribution of social infrastructure is crucial for urban social life and for our cities to become more equal and just.