Post-disaster ruins

Introduction

Housing policies in cities with high-pressure real estate markets often struggle to meet demands for affordability, accessibility, and autonomy. When urban social and public housing still exists, it is often inaccessible due to long waiting lists (Kadi & Musterd, 2015). Homeownership, for its part, is out of reach for many households (Ronald et al., 2023) and longterm debt in private home ownership reduces households' abilities to develop autonomous life choices. Private rentals suffer from insecure short-different tenurial forms but always involves direct collaboration between residents and other stakeholders in the production of the estate. It generally foregrounds values beyond profit (e.g., community empowerment, sustainability, and care (Tummers, 2016)). Moreover, it attempts to combine affordability with easy accessibility and resident autonomy to make decisions in estate management (Czischke et al., 2020). There are multiple models of collaborative housing, depending on the institutional context of the city or country. Therefore, the term collaborative housing alone cannot determine whether housing estates will be affordable, accessible, or autonomous in the long term.

Affordability is often addressed by collectivising property (e.g., by creating cooperatives) (Balmer & Gerber, 2018;Gerber & Gerber, 2017). However, collectivising property does not inherently reduce profit-seeking strategies (Hinton, 2021;Vestbro, 2000). Similarly, while cohousing may have the potential to foster solidarity among people of different nationalities, ages or socio-economic backgrounds, it can also turn into identity and/or income-based enclaves (Chiodelli, 2015;Czischke & Huisman, 2018;Droste, 2015;Sørvoll & Bengtsson, 2020). Finally, collaborative housing that is managed in the same way as state-led social housing, may have little room for tenants' self-organisation. Put simply, questions remain about collaborative housing's ability to offer affordable homes in an overheated market, specifically when it comes at the expense of other social targets like accessibility or autonomy.

Existing work usually approaches these problems through single in-depth case studies or larger sample comparisons. The former has demonstrated the importance of everyday practices of collaboration in developing a culture of autonomy, affordability and accessibility in collectives (Chatterton, 2016;Droste, 2015;Jarvis, 2019). However, these studies do not indicate how their models might be replicated in other contexts. The latter goes beyond contextual differences to define the field's terminology-particularly focusing on tenure forms-but cannot explain why some cases perform better than others (Czischke et al., 2020;Lang et al., 2020;Tummers, 2017). With one exception (Ferreri & Vidal, 2022), scholars have not translated the common features of collaborative housing into context-specific explanations of performance to identify key factors of success in affordability, autonomy and accessibility.

To fill this gap, this paper develops a medium-sized comparison of eight cases from two cities in two countries. It investigates the interactions between affordability, accessibility and autonomy and the factors helping and hindering single projects balance these goals. Thereby, this paper compares different approaches to organise affordability, autonomy, and accessibility in collaborative housing. It studies how affordability is organised through collaboration. It then analyses how accessibility unfolds through the selection criteria to accommodate new residents. Secondly, it explores autonomy looking at how groups self-govern, including processes of self-design of living space. Each of these dimensions is operationalised in relation to collaborative housing. We distinguish four types of collaborative housing based on the initiator subject (intentional communities vs. third parties) and tenure form (distributed vs. concentrated property rights). We selected eight cases (four in Amsterdam and four in Milan) to illustrate each type and compare the cases' histories and presents through an in-depth qualitative analysis.

As the existing literature notes, collaborative housing involves trade-offs between affordability, accessibility, and autonomy. However, we add that the community-identified values of the housing estate are the most important factor in navigating these trade-offs, followed by the urban regulatory context. We also call on scholars and policymakers to distinguish between collaborative housing projects using their socio-political identity, not only their tenure form or management mode. This paper first outlines the key terminology, including 'collaborative housing' and affordability, accessibility, and autonomy within this model. We then review the theoretical approaches used to select the cases. After presenting the cases, we develop a comparative analysis of the trade-offs and conclude explaining why the values identified by the communities themselves are the most important factor in managing them.

Navigating the collaborative housing field: mapping terminologies

The term 'collaborative housing' is generally used in housing research and policy as an umbrella term to encompass co-living, cooperative housing, limited equity trusts, community land trusts, developer-led cohousing and more (Czischke et al., 2020). Collaborative housing is defined by 'the presence of a significant level of collaboration amongst (future) residents, and between them and external actors and/or stakeholders [where] the term collaboration stands for coordinated action towards a common purpose' (Czischke et al., 2020, p.6). understanding how these projects organise for different common purposes (i.e., affordability, accessibility, and autonomy) is crucial since some methods achieve lasting impacts, while others do not.

using collaborative housing as a catch-all term allows the field to remain open to emerging forms of collaboration. However, it can also cause confusion and the perception that all collaborative forms are generally better at dealing with high-pressure and unaffordable housing markets than non-collaborative forms. Cohousing and cooperative housing are two frequently used categories within the collaborative housing umbrella but can mean different things depending on the historical, geographical and institutional context (Jakobsen & Larsen, 2019;Tummers, 2015). Cohousing in Italy and the Netherlands is often privately owned and can be freely traded on the market (Chiodelli & Baglione, 2014), but include some shared spaces and (often) a common lifestyle identity (Lang & Stoeger, 2018).

In contrast, cooperative housing involves collective property and is praised as the most decommodified and autonomous housing strategy (Balmer & Bernet, 2015;Balmer & Gerber, 2018). It is widely believed that cooperatives organise affordable and sustainable housing on a people over profit basis (Forrest & Williams, 1984;Gerber & Gerber, 2017). yet, in a strict sense, cooperatives only promote common property; they do not necessarily enable the sharing of spaces nor develop care or empowerment (Vestbro, 2000). Cooperatives are not always even affordable. The Italian cooperativa was an early 20th-century large-scale organisation rooted in workers' struggles and mutual aid practices that aimed to provide affordable housing units for its members. These often involve private property without sharing (Baiges et al., 2020). Alternatively, Dutch wooncoöperaties are smaller-scale initiatives that promote shared spaces, communal living, and common property (Baiges et al., 2020). These cooperatives created an association that aims to counter the individual appropriation of housing values, and both residents and non-residents can become members (Gerber & Gerber, 2017). In sum, collaborative housing includes many sub-categories that differ internally. This diversity of collaborative housing is reflected in the proliferation of single case studies (Cabré & Andrés, 2018;Lang et al., 2020). Such studies outline how ideals are translated into practice on an interpersonal scale (Tummers & MacGregor, 2019). They also explore how collaborative housing facilitates social mixing between different nationalities (Czischke & Huisman, 2018) or ages (Droste, 2015). yet, these studies' in-depth and situated knowledge is often difficult to translate to other contexts. Studies that go beyond the single case study tend to lump similar collaborative housing practices together at an international scale (Czischke, 2018). These studies compare urban policies across countries (Tummers, 2015) and create international typologies (Crabtree-Hayes, 2023; Griffith et al., 2022). However, they cannot explain case-specific dynamics. Few studies have approached collaborative housing through a mid-scale sample, that is, considering multiple cases within a single country (Balmer & Gerber, 2018) or within multiple countries but focusing on a single collaborative housing model (Ferreri & Vidal, 2022). The few studies that do exist suggest that the availability of specific incentives, land policies, and sectoral regulations are crucial for facilitating the emergence of collaboration. However, the field of collaborative housing studies has yet to determine which factors matter most for simultaneously ensuring affordability, accessibility, and autonomy in the long run and across countries. Collaborative housing projects, as we shall show below, navigate the everyday tradeoffs between affordability, accessibility and autonomy by self-identifying their values. It is the dynamic and historically changing relation between these values and the regulatory context that influences the long-term affordability, accessibility, and autonomy of the projects. The next section outlines why affordability, accessibility, and autonomy need to be studied simultaneously within specific types of collaborative housing.

Combining affordability, accessibility, and autonomy in collaborative housing studies

Affordability has been generally used to describe a number of issues related to access to qualitatively good housing for what can be generally understood as a decent price. The term is therefore widely applicable, yet context-dependent (Galster & Lee, 2021). Affordability is, therefore, best understood in relation to the economic principles and social norms about living standards in a specific area, including both ratio-based economic standards and normative-based life-quality indications (Haffner & Hulse, 2021). Ratio-based economic standards often refer to quantifiable principles such as the rent-to-income ratio that should not exceed 30%, even though this 'objective' number has been criticised for its arbitrariness (Haffner & Boumeester, 2014;Haffner & Heylen, 2011). A normative-based measure instead can place other non-economic values of housing at the centre, such as togetherness, sustainability or sharing. These values are particularly important for collaborative housing projects (Mulliner et al., 2013).

In collaborative housing, affordability can be defined as the set of social norms and formal rules around costs and quality of living. In other words, it is defined by the strategies and values that each project follows to manage, balance, and cover the costs for its residents over time (Brysch & Czischke, 2022). This could include common budgets, transfers among households, internal subsidies, or a deliberate reduction of expenses depending on specific households' incomes. In this paper, we scrutinise affordability looking at the collective strategies (if any) that each case adopts to deal with the cost of living.

In collaborative housing, the accessibility of a project can the defined as both social norms embedded in a specific community and formal rules formalised through statutes and contracts. Since collaborative groups often follow a self-selection procedure, we must seriously examine both which groups consider collaborative housing to be an actual option (Lang et al., 2020) and what barriers exist to accessing collaborative housing for specific social groups. The level of accessibility is determined by the collective values reflected in selection criteria, monetary barriers, the cultural profile of the group as well as the degree of physical access to the estate's public common spaces. Some collaborative projects aim to fulfil a neighbourhood function, while others are focused on self-provision and architecturally resemble gated communities (Dang & Seemann, 2021). Therefore, this research assesses the accessibility of each project based on the requirements to join and physical barriers for non-residents 1 .

The third crucial dimension of collaborative housing is autonomy. This study understands autonomy as strategies and values the project holds with regards to how each project (i) determines the rules that govern their co-living and (ii) how the values that drive this self-regulation are negotiated within the evolving context of market and state relationships. This definition includes housing collectives' relative freedom from market dynamics (e.g., increasing housing prices) and political influence (e.g., changing orientations of state bureaucracies). The degree of autonomy defines which values and strategies a group holds to determine their own rules for organisation, including self-design of the dwellings and common spaces (Sørvoll & Bengtsson, 2020). Autonomy within collaborative housing is always precarious (Vidal, 2019). Collaborative housing projects risk being co-opted by the market when individuals decide to exploit their housing for profit or the state if grassroots initiatives are incorporated into policies that limit their self-determination (uitermark, 2009). Within capitalist and state structures, autonomy can be achieved by pursuing autonomous decision-making while relying on the state or certain market mechanisms for support (Bianchi, 2022;Ganapati, 2010). Ferreri & Vidal (2022) demonstrated that housing cooperatives can only move beyond their niche status with state support. Each collaborative housing project's relationship with the state and market is always in motion, and autonomy happens during this process of defining one's position vis-à-vis the market and governments (Savini, 2022). Thereby, collective strategies and sets of rules to avoid some market mechanisms such as speculation or privatisation are an important aspect of cultivating and maintaining autonomy.

Affordability, accessibility and autonomy always involve trade-offs. In conventional state-led social housing, high affordability usually comes at the cost of autonomy. In small-scale communities, affordability and autonomy may come at the cost of accessibility if members select new residents based on common self-determined rules (Bianchi, 2022;Sørvoll & Bengtsson, 2020). Self-design grants residents autonomy to create housing better aligned with their desires and future needs; yet, it does not necessarily guarantee affordability since the margins created by cutting out a third party like a developer can be used to pursue luxury instead. Therefore, it is important to consider how different collaborative housing types respond to these trade-offs.

Types of collaborative housing

Collaborative housing can be subdivided into groups based on tenure form(s) and project initiator(s). Since collaborative housing can encompass multiple forms of tenure (Lang et al., 2020), it is useful to organise the field along a continuum, ranging from concentrated to distributed rights. Tenure identifies the multiple relations between individuals and their housing. These relations can reflect different (legal) rights, such as use or property rights. The nexus of relations ultimately defines a property regime (see also Bossuyt, 2021). To understand the multiplicity of collaborative housing, it is useful therefore to consider whether and to what degree rights related to ownership, income and management are distributed (or dispersed) across different subjects (Bossuyt & Savini, 2018).

Concentrated and distributed rights are a typology that allows one to distinguish between different organisational formats, yet each case shows specific articulations. Concentrated rights are held by one subject, for example, a collaborative housing form in which members own private apartments while co-owning shared spaces. Rights are distributed when the property and its manifold rights of use belong legally to different subjects that use, manage, run, or earn income from the estate (see also Griffith et al., 2022). For example, the municipality might own the estate but grant use-rights and management-rights to a collective, including the right to self-select new members. In a case of even more distributed rights-what we call common housing-the rights to the estate may be held by a wider network of associations and decision-making power is distributed between different legal subjects (e.g., a cooperative made of residents' sub-coop and a main coop, see also Card, 2020).

Initiator(s) refers to the (group of ) subject(s) that initiated the projecteither an intentional community or a third party, i.e., an organisation that is not made of or by the (future) residents. Initiators establish the social and material foundations of the project and influence its trajectories. Collaborative housing includes collaboration between multiple stakeholders at different levels (Blandy et al., 2006). Therefore, it is not always possible to categorise it according to familiar top-down or bottom-up distinctions (Droste, 2015;Ferreri & Vidal, 2022). However, the subjects that initiated the project determined its original motivation and mission. For example, intentional communities often invest voluntary and unpaid hours to create the project. This contribution could make communities more committed to selecting other committed tenants and avoiding commodification in the long run. Initiators are not always clearly defined: projects could emerge from a housing association's intentional community that functions as a third party for other residents. This third party then buys property for other intentional communities.

Dividing the field along these two axes-tenure form(s) and initiator(s)-creates four ideal-types: self-promoted, common, co-managed and developer-led housing. We understand these as categories that in practice may partially overlap in reference to real, existing cases. We matched two cases to each type to better cover the full spectrum of collaborative housing.

Self-promoted housing is initiated by an intentional community wanting to live collectively and collaboratively. They represent a desire to live in close vicinity and collaboratively but maintain private ownership (Blandy et al., 2006). Each member has rights in the form of individual property, which they may freely sell or lease on the market. Members often co-own spaces such as the garden, a guest room, and sometimes a common (extra) kitchen.

Common housing represents projects initiated by an intentional community with distributed rights. Common housing can take multiple organisational shapes. often it is a cooperative consisting of residents and/or non-residents. The building is owned by the cooperative as a single subject, effectively making the residents both landlords and renters: two different subjects, one with income rights and one with management rights (Aernouts & Ryckewaert, 2018;Chatterton, 2016). In practice distributed rights can overlap: similar rights can be exercised by different groups; a non-resident may still influence day-to-day management through collective decision making even though management is intended to be the exclusive right of residents. yet, this separation is a crucial strategy towards decommodification as residents no longer directly and individually benefit from the sale or lease of their estate (see also Peredo & McLean, 2020).

Co-managed housing projects are also characterised by distributed rights but are formed by a third party, either publicly or privately initiated. This third party holds the right to extract rent over the estate and makes decisions about sales and leasing. However, the inhabitants hold an autonomous position and decide how to design/use the building and on new members (Bossuyt, 2022). In other words, the third party sets the terms and conditions of the rent, while the inhabitants can decide which extra selection criteria they uphold for new members. The inhabitants may also have the freedom to create solidarity systems to pay the rent in accordance with one's income, yet, the level of the overall rent is decided upon by the third party. The degree of autonomy among residents can differ substantially. Some projects are run in a top-down manner by a third party, giving the residents little freedom beyond determining the use of some shared spaces.

Developer-led housing projects are initiated by a single third party that holds the rights to sale and lease. Residents have no say over who can become a member. This category generally provides little autonomy for residents to determine spatial characteristics or organisational matters (Anund Vogel et al., 2016). In these cases, collaboration takes place around shared use and general management decisions over selected common spaces.

Contexts, cases, and methodology

Sharply rising living costs have made Amsterdam and Milan unaffordable for lower-to-middle-income groups (Belotti & Arbaci, 2021;Hochstenbach, 2022b). In Amsterdam, the average home sale price surged by 82% from 2013 to 2020 (Hochstenbach & Aalbers, 2023, p.20), while Milan saw a 47% rise in mean rental prices from 2004 to 2020 (Coppola et al., 2023).

While both cities are encountering comparable trends, their housing provisions vary considerably. Amsterdam features a relatively large social rental system (46%) and actively encourages the collaborative housing sector (Van Duijne & Ronald, 2018). However, the city has experienced a decrease in social renting (61% in 2012 to 47% in 2021 (ANP, 2022)) and an increase in home ownership, further widening the welfare gap (Boelhouwer & Priemus, 2014;Hochstenbach, 2022a). In Milan, over 70% of households are homeowners; the rental supply consists of residual public housing (i.e., 30% of rentals and less than 8% of the entire housing stock) and a private, unregulated supply mostly in the hands of small landlords (Belotti & Arbaci, 2021;Coppola et al., 2023;Poggio & Whitehead, 2017).

Shared living groups, known as woongroepen, have long existed in the Netherlands but remain a niche phenomenon. Collectief Particulier Opdrachtgeverschap (CPos)-the blending of individual home ownership with shared spaces like gardens-gained popularity in the late twentieth century after the government began to promote home ownership. While CPos were initially intended to be accessible to lower-to-middle-income groups, the cost of home ownership has become a barrier to entry 2 . Currently, the municipality encourages housing cooperatives and hopes they will account for 10% of the total housing provision within 25 years, equating to approximately 40,000 units by 2045 3 . To facilitate this objective, the municipality designated 13 plots of land and established a 50,000 euro fund per cooperative 4 . However, at the time of writing, the cooperative sector faces challenges due to changing market conditions and higher building costs and rents.

Milan lacks a unified policy to promote collaborative housing. In the 2000s, new types of social housing supply began emphasising collaborative housing and social mixing (Alexandri et al., 2023;Bricocoli & Coppola, 2013). These initiatives were limited in number, highly diversified, and led by different actors. Sharing spaces have become common in such social housing projects however, there is no statutory mandate that these services be accessible to external users 5 . Furthermore, social housing projects including forms of collaboration still tend to exclude the lowest-income groups, that are instead eligible for public housing (Bricocoli & Cucca, 2016). Private and other kinds of subsidised housing projects can also create shared spaces, and can lead to reductions in the payment of development fees if they are made available to external users and perform a public service. This arrangement leaves the terms of public interest relatively unclear, so negotiations are made on a case-by-case basis, and historical performance is highly varied 6 .

Within these municipal and national contexts, this study selected the following eight cases (see Table 1). We conducted 32 in-depth semi-structured interviews between September 2022 and June 2023 with residents, architects, and collaborative housing experts in both cities. Interviews were conducted in English, Italian, or Dutch, and all quotes were translated into English. The participants have been anonymised and numbered, though the project names are real to allow for follow-up research and cross-referencing. Names of architects and experts in the housing field were shared with approval.

Affordability: from radical collectivisation to informal solidarity funds

The eight cases evidence different strategies to address affordability. Projects with distributed rights might best meet affordability objectives, with the most collaborative projects creating property collectivisation and solidarity funds. However, concentrated rights cases can also manage affordability if there is a strong, explicitly stated value in the housing project's public image and initial statutes.

The Nieuwe Meent cooperative locates property rights at the level of the cooperative to create a project with affordable homes, according to the Dutch social housing standard. only residents can join the cooperative, and they collectively make decisions about what happens with the property and its investments. Residents rent their apartments from the cooperative, effectively becoming both renter and landlord-a dual position that guarantees the legal barrier against individual capitalisation of the asset and thus prevents the residents from treating their house as a commodity.

However, whether a project moves towards housing commodification in the long run depends on its values and institutional context. Values and policies might change over time, altering the desires and possibilities of collective housing projects. Base Gaia in Milan began with a collective property model but is now transitioning to individual property. This is meant to provide residents with a stronger social safety net and long-term welfare provision in an increasingly expensive housing market. The move from collective to individual property tenure forms is often driven by the need for economic security as the project matures. As a resident of the estate explains: at the beginning, we decided for undivided property because this was a guarantee against speculation (…) but in Italy, the real God is an apartment and people will be more relaxed about their children if they have an apartment. Soon, we will have new rules about the sale of the house 7 under such circumstances, the project's affordability and accessibility targets are likely to change. This also occurred in Amsterdam, despite its different welfare context. The self-promoted project Vrijburcht was designed as a CPo (i.e., individual private apartments with some common shared areas) to provide affordable housing with conviviality. Its low-tech design and a partnership with a social housing provider to fund the facility encouraged affordability. yet, according to residents, the project did not reach the lower range income groups, for example, those in social renting. That is why, currently, Vrijburcht is a very white people collective. Preferably, we would have created a more diverse group (…) but homeownership has created a burden 8

Another resident explained that the initial affordability targets of the projects turned out to be a leverage to capitalise and profit from a rapidly growing housing market in the city. As units are privately owned and sold on the market, those interested in living there nowadays are facing different prices per unit:

[the architect] has built it relatively cheap and the residents who are now moving out are profiting from this as they can sell it for a high price on the free market… I do not think we could have currently bought our home 9

As a self-promoted project, Vrijburcht's tenure rights were (unquestioningly) concentrated with resident-owners. Private property now presents a barrier for low-income households. Homeownership continues to create a burden for new members as prices per unit have increased significantly.

The other self-promoted housing project, Base Gaia, featured more diverse income levels. It included a solidarity loan system between the starter families, where families with higher incomes or capital carried more of the financial burden. The more radical Villapizzone, with its distributed tenure rights, also encouraged diverse income levels with an anonymous system where families contribute to a common pot; each family can also anonymously take what they need to cover living costs. The residents of NieuwLand established similar income-based rent levels but maintained a base rent.

Affordability goals must be translated into organisational values and property management, and this seems easier with distributed rights. For instance, Villapizzone has provided an affordable housing project since 1978, with income sharing to achieve that affordability. yet, its existence is now facilitated by the use rights gained over the estate, granted by the municipality. yet, its future will depend on the winning of a tender to be opened in 2028 10 . The use rights allowed residents to freely use the building and surrounding park, while only bearing the maintenance costs, not the lease on the land. This was in fact one of the reasons they could ensure affordability.

The likelihood that groups will organise for long-lasting affordability thus depends on whether and how rights are distributed and, if they are not, whether the mission of affordability is translated into formal mechanisms of cost-sharing and solidarity. Distributed rights shared across a network of cooperatives (or other agencies) is one organisational tactic to assure affordability against speculation (Nieuwland or Nieuwe Meent); others created internal loan systems that flourish from strong internal solidarity (Villapizzone). However, these are precarious balances that can be distorted without supportive property relations that safeguard the project's affordability. Without the free use of public land, Villapizzone would not have had the means to create affordable homes even if have a strong community centred around low-cost living in solidarity.

Accessibility: from organising one's own housing to establishing a housing movement

Each project's accessibility was determined using two elements: the criteria for new residents to join the estate and non-residents' access to common spaces. Projects initiated by its future residents-as opposed to projects initiated by a broader intentional community or third party-can foreclose the capacity to reach out and be accessible unless the community explicitly identifies as part of a housing movement. Cases involved in housing justice networks tend to value the accessibility to collaborative housing for diverse social groups. They translate these values into specific strategies to achieve this goal. This was more common in distributed third party-initiated projects and in projects that maintained clear values over time. In other words, accessibility was not a consequence of collaboration but was itself a value to be nurtured in the everyday practices of running the projects.

Base Gaia, urban Village Bovisa, and Vrijburcht did not see their projects as a political statement nor an attempt to provide for a group beyond their own. Future residents are selected based on their ability to pay the housing market price. These estates featured individual ownership with shared gardens only accessible to the residents. This decision was often motivated by a desire to form a close-knit community and create a safe space where doors can be left unlocked and children can freely play 11 . However, a Vrijburcht resident expressed that this desire has turned such establishments into enclaves that are inaccessible to the neighbourhood 12 . Groups often express a desire to serve the neighbourhood while also seeking a close-knit community, which can lead to tensions 13 .

Conflicts between the collective and the neighbourhood emerged over the closure of Vrijburcht's private dock, which was initially open to the public. According to a resident, that dock is ours; it is on our property and we paid for it but when we closed it neighbours responded as if we took something away from them, yet, they had never realised we had given them something in the first place 14 Private ownership influences the relationship between projects, their shared spaces, and accessibility. While Vrijburcht may not publicly share the dock, they do run a semi-professional theatre-one of the only cultural spaces in the neighbourhood. In contrast, urban Village Bovisa keeps all of its common spaces closed. Base Gaia's common spaces (a spare room, kitchen, and gated courtyard) are officially recognised as public services, allowing the property to claim discounted development fees for these square metres. However, the municipality does not detail how such public interest spaces should be used and does not supervise their use, so their accessibility largely depends on the collective's values. The residents determine how often the space is accessible to the wider public (and for whom). Base Gaia has opened their space for some public events and to facilitate neighbourhood gatherings 15 .

De Nieuwe Meent, Villapizzone, and Cenni di Cambiamento all moved beyond a single enterprise to help establish a network of projects. All these projects have distributed rights, which might suggest that projects for self-provision (by communities of future dwellers or by third parties like developers) are less likely to strive for accessibility unless regulated as social housing with income requirements oversight (e.g., Cenni di Cambiamento). Each of these projects found its own ways to meet this challenge. De Nieuwe Meent focuses on accessibility through radical open resident selection, ensuring diversity and openness in their project and regular communication and interaction with their wide base of supporters 16 . A member describes, the people who were doing the work were not sure yet if they would live there but they wanted to make the project happen. I say that at that phase we did not have a very clear definition of who puts the work and who benefits 17 This radical open approach acknowledges the burden of self-organisation (in terms of knowledge, time, and money) and aims to create more accessibility by offering housing to new members without adopting that burden. They also prioritise a publicly accessible inner garden for educational purposes and specify in statutes that future income should be invested in similar cooperative housing projects, not the lowering of rents 18 . This approach to accessibility contrasts sharply with the Cohesion, a collaborative housing project initiated by a third party, with concentrated rights, and focused on service provision. The Cohesion offers rental units with entertainment-oriented, non-essential shared spaces like gyms and cinemas 19 . Rental contracts are short-term, resulting in high turnover rates, and units are allocated on a first-come, first-served basis. Residents appreciate the absence of waiting lists but face market-driven rents and increasing costs for new tenants 20 . The building is owned by Rockfield, a large real estate company that provides community-based living in various complexes. Residents have little say in the management and use of the shared spaces but do pay service and community costs as part of their rent 21 22 . Some residents feel exploited by this commodification of the community and less regulated service costs, which leads to higher overall living expenses 23 . ultimately, projects that were initiated by its future residents tended to be more inward-looking and less accessible, even when they had distributed rights. Therefore, it is important for accessibility to be stated as a core value of the project and that the guarantees of public access in exchange for favourable planning treatment are rigorously implemented.

Autonomy: from common decision-making to everyday conviviality

Autonomy encompasses how the collectives self-define their rules of governance within the fragile conditions of relative freedom from market logic and state control. Cases with distributed rights, even when initiated by a third party (e.g., Cenni di Cambiamento), approached autonomy more consistently. However, autonomy (also) needed to be regulated as a core value and translated into the organisational set-up; thus, cases with concentrated rights could also include elements of autonomy (e.g., Vrijbrucht, urban Village Bovisa).

NieuwLand was created by the housing association Soweto, which aims to protect alternative spaces in Amsterdam and currently owns the NieuwLand building, a former school. According to a NieuwLand resident, right from the start Soweto was doing this as a political project. None [founders of Soweto] wanted to live here but they wanted to set up an umbrella organisation that owns different buildings while each entity is autonomous (…) Soweto has the ownership and (…) we as residents have a say in the building but we cannot sell it 24

In this case, the community intentionally maintained autonomy through an organisational set-up that mirrored its values. The organisation created autonomous layers with different decision-making powers, this organisational set-up distributed the rights of ownership over the building. These distributed rights prevented residents from becoming inward-looking and lowered the temptation to capitalise by selling the building or lowering their own rents. Moreover, NieuwLand did not accept subsidies in order to remain autonomous from political influence. They achieved this by self-renovating the building with a team of volunteers from the squatters' network. 25 De Nieuwe Meent was co-founded by some of the residents and activists involved with NieuwLand. unlike NieuwLand, de Nieuwe Meent collaborated with the municipality and found themselves in constant struggles to defend their values and maintain their autonomy. The project started with a public tender, but the conditions quickly began to feel restrictive:

[the municipality says] fix the business case like this: creating smaller apartments, less shared spaces, reduce the sustainability ambitions but really, they were reasoning like a private developer, cutting out everything that is not essential (…) it is very important to be aware of what you are doing and you always need to go back to your values in these situations of decision-making 26 De Nieuwe Meent did have the autonomy to determine how to self-govern. All decisions were made through a consensus system-people did not have to agree with decisions, but they should not be strongly opposed 27 .

Another Amsterdam location, Vrijburcht, demonstrated how organisations following the collaborative housing national blueprints (i.e., the CPo system) had less control but could still manoeuvre in their everyday management:

there are formal moments in which we discuss things as a collective but usually, it all starts with a little seed planted in the people's mind (…) we rely more on pre-discussions, ask around so that the plan is supported widely before we go to the general assembly to vote 28 Thus, residents created alternative moments of autonomy by finding common ground in line with their community values. These moments of collective decision-making paralleled the formal rules. Therefore, they were often irregular and carried out by a small sub-group of residents. According to one resident, if a specific group of about ten key individuals moved, the maintenance of Vrijburcht would become much more complicated 29 .

urban Village Bovisa also struggled for autonomy. The project renovated a pre-existing building to create a shared swimming pool, inner garden, and parking area in collaboration with a developer. The developer initially shaped the organisational set-up and limited the group's ability to affect these spaces. The collective wanted to foreground sustainability but was limited by technical difficulties. Furthermore, the collective only discovered the high additional development fees during the co-design process, which made them want to leave the developer. However, the developer's technical knowledge and (more importantly) relationship with the municipality were needed to obtain the permits. As one resident explained, 'this way, it took only two years, but it would have been more time [to do] everything without them 30 ' . once the collective left the developer and reclaimed autonomy, they 'redesigned the garden and it felt like a revenge 31 '! The group reinstated their values and redesigned some key spaces to better align with their identity. The courtyard was given more spaces for encounters and one of the storage spaces was transformed into a common laundry facility, something the developer said was impossible. This example shows the importance of value formation for autonomy, especially when dealing with concentrated rights and third-party-initiated projects. By negotiating autonomy with and against the developer, the group secured discounted development fees.

In Villapizzone, a common project with an intentional community and distributed rights, autonomy was found (not given) through an elaborate process of consensus decision-making. The group assured autonomous self-management-including managing shared income and running a neighbourhood café-through monthly gatherings. Their desire for autonomy was translated into the organisational set-up and reflected in the tenure form (i.e., the collective retains the use rights over the estate with no formal ownership). According to a resident, their autonomous self-organisation had created an alternative family that would stand the test of time 32 .

Lasting autonomy is most likely when values are translated into the project's organisational structure. Projects with intentional communities and distributed rights are more likely to include formal collective decision-making procedures that facilitate autonomy. yet, urban Village Bovisa-a project with concentrated rights, initiated by a developerproves it is possible to reclaim autonomy if the collective identifies this as a core value ( Table 2).

Conclusions: the role of values in collaborative housing

Collaborative housing comes in many forms, and they do not all deliver affordability, accessibility, and autonomy equally. We urge studies on collaborative housing to simultaneously question formal provision systems and everyday performance. To do that, this field of study needs to appreciate the different articulations between institutional and regulative settings, the values identified by the communities themselves, and the way these two interact in the production of affordable (or not), accessible (or not), autonomous (or not) living spaces. Both tenure forms and the intentions of the initiators shape the everyday operations of the estate. Tenure forms are collectively defined and constructed within the specific institutional and regulative context, but often exist in tension with them (Singer, 2000). In our study, we discovered that it is by adopting specific identifying values that the collectives are able to navigate those tensions. In some cases, they can also challenge property relations based on conception of absolute ownership. These values, and even the regulations that shape tenure forms, can change over time, thus altering the long-term performance of collaborative housing projects. This makes it very important to develop a combined critique of the community-identified values and tenure forms in the study of collaborative housing. Collaborative housing projects must make trade-offs in affordability, accessibility, and autonomy. How these trade-offs are managed depends on both the institutional and regulative context (e.g., how housing is regulated) and the values identified by the collective. Property relations and regulations set the conditions for affordability, accessibility, and autonomy. yet, intentional communities challenge the conditions of individual property distributing their multiple rights and promoting other social values such as sharing and solidarity (see also Bailey, 2020). In practice, the values that communities self-identify help navigating the constant tension between affordability, accessibility and autonomy. It keeps the cohesion and energy of the group and allows to balance off the possible conflicts between affordability, accessibility, and autonomy. Affordability, accessibility and autonomy are goals that, as an architect interviewed explained, come with 'strings attached' and 'you need to be careful when you pull one up the other ones are not going too much down. You just need to be in balance' 33 . our results confirm that values are important in guiding voluntary and cooperative housing communities (Chatterton, 2013;Jarvis, 2015). However, our study further sophisticates this claim by showing how values affect longterm performance of affordability, accessibility and autonomy, even under concentrated third-party tenure forms. The only exception was projects with a high degree of concentration and a third party (e.g., the Cohesion), where the decision-making power rested with one subject, the developer, and the market logic dominated the affordability, accessibility and autonomy targets. We offer three specific insights for collaborative housing research.

Affordability can be instituted in different tenure forms

There are multiple ways to address affordability, including collectivising property or collectivising costs and incomes from the estate (see also Bossuyt, 2022). The state's role in discounting fees and regulating incomebased accessibility is also important. Collectivisation through distributed rights tenure can achieve long-lasting affordability when it is taken as a core value and collectives establish internal mechanisms of solidarity. Projects with concentrated rights (self-promoted or developer-led) are more likely to fall victim to market prices and reduce their affordability potential over time.

Self-promoted and developer-led projects tend to be less accessible, unless accessibility is a core value

Projects with concentrated rights tenure (i.e., self-promoted and developer-led) were less accessible to potential new residents and less accessible for nonresidents wishing to use the common spaces (e.g., urban Village Bovisa). Projects with distributed rights (common housing and co-managed housing) became accessible if they actively engaged with networks of similar projects (e.g., De Nieuwe Meent). Thus, engagements with networks explicitly oriented towards accessibility can help avoid enclaving.

Autonomy exists in different forms for collaborative housing projects

This study confirmed previous research on autonomy as a precarious space of freedom and the importance of translating autonomy into the project's organisation. As expected, the commercial developer-led cases with concentrated rights (e.g., the Cohesion) had no autonomy; but this can be challenged and changed if the community sets it as a core value, for example at the point of design (e.g., urban Village Bovisa). Projects created and maintained autonomy in various ways beyond the most radical horizontal forms of consensual decision-making. For example, NieuwLand was an autonomous unit within a multi-layer organisational set-up that prevented the building from being sold. In contrast, Vrijburcht found autonomy in everyday informal conviviality among neighbours. This may suggest that autonomy is stronger in smaller estates.

This study advanced existing research on collaborative housing by demonstrating how institutional variables (e.g., tenure form and rights distribution) interact with community values. Policy should be more context-sensitive when establishing incentives for collaborative housing since cases with concentrated rights led by single developers were not accessible, affordable, or autonomous. We did not examine why and how these community values were formed in the first place. However, the historical presence of intentional communities and the possibility of responsible and forward-looking development cultures may be extremely important in explaining differences between cities, as well as the presence of institutional and regulative settings incentivising certain values over others. This paper invites future cross-city comparisons of collaborative housing to engage with both tenure forms and the socio-political profile of communities.