Integration through Collaborative Housing ? Dutch Starters and Refugees Forming Self-Managing Communities in Amsterdam

Since 2015, Europe has experienced an unprecedented influx of people fleeing countries facing political turmoil. Upon receiving asylum status, refugees in the Netherlands are currently regionally dispersed and individually housed in public housing. Themunicipality of Amsterdam has recently adopted an alternative approach, whereby young adult refugees and Dutch young adults are brought together in collaborative housing (Czischke, 2018). This article presents findings from a case study of the pilot project, launched in 2016,which houses over 500 young adults, half refugees and half Dutch together in temporary dwellings. The goal is to provide refugees with social and cultural tools to integrate in the host society by interacting with their peers through collective self-organisation. Compared with more traditional forms of housing refugees, integration through collaborative housing is expected to deliver results. Our study aims to examine this assumption by looking at the daily reality of collaboration and self-organisation amongst tenants in this pilot project, and interrogates how this approach may help the integration process. The analytical framework draws on Ager and Strang’s (2008) core domains of integration, which emphasises the role of social connections in the integration process. An ethnographic research design was adopted, including interviews and participant observation as data collection techniques. Preliminary findings indicate the gradual formation of social connections such as social bonds, social bridges and social links. Ultimately, we expect findings to inform better policies and practices in the field of housing and urban planning that help the integration of young refugees in European societies.


Introduction
The number of forcibly displaced persons worldwide has increased from approximately 45 million in 2012 to 65 million by the end of 2016 (United Nations High Commissioner for Refugees [UNHCR], 2017).Most people fleeing violent political conflict find shelter elsewhere in their own country or in neighbouring countries, but a small minority applies for asylum in Europe.As one of the European Union member states receiving a comparatively large number of asylum seekers (Eurostat, 2018), the Netherlands struggles to find adequate affordable housing for those refugees who successfully acquire a residence permit.The recent rise in the influx of refugees further increases the pressure on affordable housing in popular parts of the country.Upon receiving asylum status, refugees in the Netherlands are regionally dispersed (as in other European countries) and individually accommodated in social rental housing.However, the effectiveness of this approach vis-à-vis integration goals is questioned (Bakker, Cheung, & Phillimore, 2016).More generally, the integration of refugees into Dutch society in terms of education, employment and psychological wellbeing is often considered suboptimal (Bakker, Dagevos, & Engbersen, 2017;Korac, 2003;SER, 2018).
The municipality of Amsterdam has recently adopted an alternative approach to housing refugees, whereby young adult refugees and Dutch young adults are brought together in collaborative housing, keeping an even mix of each group (50% refugees, 50% Dutch).The ambition is to provide refugees with social and cultural tools to integrate in the host society by interacting with their peers through daily practices of collective self-organisation.This article presents initial findings from a study of the Startblok, the pilot project of this approach, launched in 2016.The project attempts to tackle several of the abovedescribed challenges at once: the lack of affordable housing for young adults and for recent refugees and the integration of refugees into the host society.
The overall research question guiding this article is: how could collaborative housing help the integration process of refugees?To that end, we examine the case of the Startblok project through the lens of the following questions: what shape does the self-organisation and selfmanagement of the tenants take?How does integration of refugees via social mix and self-organisation in a housing project work out in practice?
In what follows, we first summarize how refugees are currently received in the Netherlands and touch upon the recent emergence of collaborative housing in the Netherlands in the context of new roles for traditional housing providers.We then define and discuss different elements of integration as a two-way process.Our choice of research design-a case study with ethnographic research-is explained in the next section.Subsequently we address the research questions, starting with a full outline of the Startblok project and then analysing the self-organisation of the tenants, the role of the housing corporation and the integration of the refugees.Our tentative conclusion is that, compared with current alternatives, integration through collaborative housing appears to be an innovative and effective approach.

Refugee Integration in the Netherlands
Between 2012 and 2017 the number of refugees applying for asylum in Europe per year rose from over 300.000 to just over 700.000, with peaks due to the intensification of the Syrian war of respectively 1.3 million in 2015 and 1.2 million in 2016 (Eurostat, 2018).The Netherlands experienced a similar increase in asylum requests, and the number of requests that was granted rose accordingly, from 6.000 in 2012 to 34.000 in 2016 (Statistics Netherlands, 2018).In the Netherlands, asylum seekers that are successful in their applications receive a fiveyear residence permit, after which they can apply for permanent residency.This group is the focus of this article.To distinguish them from asylum seekers who are still awaiting a decision on their requests, and from those who have received a negative decision, in the remainder of this article we will refer to asylum seekers that have been granted a residence permit as refugees.
Upon obtaining a residence permit, refugees acquire the right to work and become entitled to most of the welfare arrangements available to Dutch citizens, such as the right to social housing, social services allowances, health care and loans for pursuing further education (up to 30 years of age).They receive coaching from municipal social services who attempt to place them into suitable trajectories towards education, employment and/ or volunteering.
As all immigrants from non-EU countries to the Netherlands, refugees have to pass the Dutch exam in 'inburgering'.This concept is often translated as integration, but Besselink (2006, p. 14) points out that it is "very much like the term 'enculturation' but having a root [in the Dutch term] 'burger', which means 'bourgeois' or 'citizen'".The exam consists of two or three parts, namely literacy training if applicable, basic proficiency in the Dutch language and knowledge about Dutch society.The exam has to be successfully passed, or at the very least demonstrable attempts to pass have had to be made within three years of obtaining a residence permit.Sanctions include fines and (theoretically) non-renewal of the residence permit.Refugees can borrow 10.000 euros from the government to finance the courses, and the loan will become a gift upon successfully passing the exam.In 2013 the Dutch government liberalised the market for 'inburgering' courses, resulting in a proliferation of new and sometimes less qualified companies offering such training.The pass rate dropped from 78% in the previous years, to 39% (Netherlands Court of Audit, 2017, p. 40).Several Dutch municipalities thereupon decided to become more actively involved in the integration processes of refugees in their area.
Similar to Denmark, Germany, Sweden and the United Kingdom, the Netherlands utilises a policy of dispersal to distribute the perceived economic 'burden' of asylum seekers and refugees evenly over the country (Darling, 2017).Every six months the government allots a quota of refugees that recently obtained a residence permit to every Dutch municipality.The quota is proportional to the number of inhabitants of the municipality.Refugees are furthermore spread out randomly over neighbourhoods with social housing, depending on where homes are available upon their arrival.

Housing Refugees in the Netherlands
In recent decades successive Dutch governments have introduced reforms to make the housing sector more market-conform, by encouraging tenure conversion, (i.e., transforming rental housing into owner-occupancy); allowing sharp rent increases via the deregulation of part of the rental market; and the introduction of temporary renting contracts (Huisman, 2016).Housing corpo-rations, not-for-profit foundations who have a long tradition of close cooperation with government, own the majority of Dutch rental housing.In recent years their core task has been redefined to focus on housing those who cannot support themselves on the open market, such as low-income households and disadvantaged groups (Hoekstra, 2017;Mullins, Milligan, & Nieboer, 2018)including recent refugees.These changes have resulted in an ongoing residualisation of the once large regulated housing stock.Regulated rent as a proportion of the total housing stock declined from 58% in 1985 to 34% in 2015 (Blijie, Gopal, Steijvers, & Faessen, 2016).
As a consequence, waiting times for social housing have lengthened, especially in regions with employment opportunities.In popular cities such as Amsterdam they have risen to more than ten years.Housing corporations allocate their homes partly through waiting lists and partly through giving priority to people with urgent needs (e.g., homeless people).The priority housing arrangements further reduce the proportion of houses available to those on the regular waiting list.Aware of the potential tensions of this situation, in 2015 the Dutch housing corporations made an appeal to the government for assistance, arguing that given the sharp increase in asylum seekers, the existing housing allocation model for refugees was unsustainable (Gualthérie Van Weezel, 2015).This model requires that each refugee household should be accommodated in their own autonomous, affordable home with a permanent rental contract.In response, the Dutch government introduced a number of financial measures and relaxed the law to allow refugees to be housed in shared accommodation and with temporary rental contracts-as long as, after some years, the refugees would then be allocated housing under the preexisting model (Blok, 2015).
These developments take place against the political discourse in the Netherlands that currently emphasises the need to move away from a welfare society towards a participation society ('participatiesamenleving'), with a broader societal focus on opportunities for selfdetermination (Uitermark, 2015).In the field of housing, this translates into a gradual retreat from large-scale housing developments (Nieboer & Gruis, 2016) and an interest in self-provision, tenant empowerment and collaboration (Czischke, Zijlstra, & Carriou, 2016).To this end there is growing interest amongst some Dutch housing corporations for innovative rental models such as different forms of collaborative housing (Bokhorst & Edelenbos, 2015;Elliott, 2018;Platform31, 2017).

'Top-Down' Collaborative Housing
Collaborative housing (Czischke, 2018;Fromm, 2012) is an umbrella term that comprises a wide range of collectively self-organised and self-managed housing forms.These include, for example, resident-led housing cooperatives, cohousing and Community Land Trusts (CLTs).These different housing forms are characterised by high degrees of residents' participation spanning the conception, development and management of the housing project, and the establishment of reciprocal relationships, mutual help and solidarity.Common motivations behind these projects include high levels of environmental sustainability, mutual provision of care for children, senior citizens, and other people with special needs and, in some cases, a redefinition of gender roles in the household (Lang, Carriou, & Czischke, 2018).In addition, in the aftermath of the 2008 global financial and economic crisis, affordability and social inclusion of disadvantaged groups have emerged as new drivers of many collaborative housing projects.
While the original models of collaborative housing emerged as bottom-up initiatives, i.e., people joining forces to jointly provide housing for themselves and by themselves, in recent years we have seen the emergence of more 'top-down' approaches.These correspond to housing projects initiated by a professional housing provider, be it a social housing organisation, a private developer, or a foundation or similar organisation.A top-down initiated collaborative housing project would typically involve a professional entity either owning a building or a plot of land, or being in a position to acquire either of these, for the future (re)development into a collectively self-managed housing project.
In this type of projects, initiators usually act as developers and managers, and convene a group of residents under a shared vision of a collectively self-organised and self-managed project to be sustained in the long term (Czischke, 2018).The opportunity is given to residents to propose their own common activities.Physical spaces for these collective activities and uses are usually co-designed with the residents and financed by the providers.Thus, a landlord or professional housing provider/developer is in a strong position to enable the development of a collaborative housing project and support the group of residents throughout the initial stages of the collective living arrangements.However, given the relative newness of these initiatives, there is no conclusive evidence yet on the longer-term outcomes of 'top-down' versus more typical 'bottom-up' approaches in terms of, e.g., community cohesion, effective self-maintenance and self-management, or resident satisfaction.

The Role of Social Connections in Refugee Integration
When considering the reception and establishment of migrants into their host societies, scholars distinguish between integration and assimilation.Both concepts can be understood as specific forms of social inclusion, a more general term referring to the "ability of individuals to participate in the community" (Dukic, McDonald, & Spaaij, 2017), and the process whereby minority or disadvantaged groups overcome their previous exclusion from society.Integration can be viewed as an interactive process, whereby the receiving society and the migrant mu-tually adapt to each other.This implies that both parties have to be prepared to accommodate each other.Bakker, Dagevos and Engbersen (2014, p. 432) for instance define integration as: A multidimensional two-way process that starts upon arrival in the host state.This process requires from immigrants a willingness to adapt to the lifestyle of the host community, and from the host country a willingness to facilitate integration (i.e., access to jobs and services) and an acceptance of the immigrants in social interaction.
In contrast, assimilation can be regarded as a onedirectional effort, solely by the migrant, to become completely incorporated into the host society (Strang, Baillot, & Mignard, 2018).Both concepts have normative implications, and in how far migrants should integrate and what constitutes successful integration is a recurring topic in contemporary political debates.
In this article, we focus specifically on the integration of refugees.Compared with other migrant groups, such as family or labour migrants, refugees start at a disadvantage.They had to flee their country of origin, and often suffer from traumatic experiences.In the Netherlands, the long stay in asylum seeker reception centres and the insecurity experienced during the often-lengthy wait for a decision on their asylum request compound this negative starting point (Bakker et al., 2014).Furthermore, like other non-Western migrants, they lack culture-specific skills and knowledge, and it is difficult to have their educational and professional credentials from their country of origin recognised.In the Netherlands, only a small proportion of refugees find employment.After two years of stay, 25% is employed for eight or more hours per week, and this rises to 50% after eight years of residency (Bakker et al., 2017).Although more than 50% of refugees in the Netherlands has an average to high educational background, only 10% finishes a language course suitable to their level, and this underachievement is structural (Netherlands Court of Audit, 2017).
Overall, Dutch refugee integration policies focus mainly on tangible outcomes such as housing, work and education.These outcomes are often considered insufficient (SER, 2018), for instance by the current government (VVD, CDA, D66, & ChristenUnie, 2017), and by refugees themselves (cf.Korac, 2003).We argue that one of the missing links in attaining more successful refugee integration might be found in a less tangible aspect of integration, namely social connections.This notion originates from Ager and Strang's ( 2008) conceptual framework concerning the core domains of refugee integration (see reproduction in Figure 1).They distinguish between three forms of social connections (second row from above in Figure 1): "social bonds (with family and co-ethnic, conational, co-religious or other forms of group), social bridges (with other communities) and social links (with the structures of the state)" (Ager & Strang, 2008, p. 70).Korac (2003) found that refugees value education and employment, but also stress being connected with the host community through social contacts.She therefore emphasizes the need for policies that focus on the building of social connections: This research strongly indicated that [refugees'] personal satisfaction and assessment of integration success goes beyond simple, measurable indicators, such as individual occupational mobility or economic status.It importantly includes indicators such as the quality and strength of social links with the established community....Policies and interventions facilitating settlement and full participation in the receiving society should address the issues of integration in community by promoting strategies for building 'bridging social capital', that is, links between the established community and the newcomers.(Korac, 2003, p. 63) Following from the above, we posit that collaborative housing forms might be more helpful than others for the process of mutual adaptation, as they would facilitate interaction between inhabitants more than traditional forms of housing.The underlying assumption is that more frequent and closer contact between residents will tend to foster the formation of social bonds and social bridges, which in turn might help refugees (and although perhaps to a lesser extent, those already established in the country) to engage with and navigate more tangible elements of integration, such as education and employment.

Employment
Taking the above concepts as a basis, we have developed a simple analytical framework to help us systematize the Startblok's approach as well as to identify its preliminary outcomes.As illustrated in Figure 2, this framework rests on the principle of structured self-organisation amongst tenants to carry out a number of tasks related to the management and maintenance of the housing.The underlying assumption is that structured self-organisation will to lead to regular social interactions between refugees and Dutch tenants, thereby providing both groups with opportunities for a wide range of social connections, which ultimately helps refugees to integrate in the host society.The selforganising principle is built on two main pillars or 'necessary conditions', each based on a specific assumption related to a specific desired integration outcome, namely: 1. Demographic homogeneity and social bonds: Having something in common promotes bonding between people.All tenants are singles without chil-dren in the age range 18-28.People in this category and age bracket tend to be at the same stage in their life courses, and thus have similar life styles, compared to other age groups.This holds for both Dutch and refugee tenants.The assumption underlying this condition is that demographic homogeneity (in this case, age and household composition) is a necessary condition to facilitate social bonding across diverse cultural and/or ethnic backgrounds; 2. 50/50 mix and social bridges: Through an even allocation of half of the flats to refugees and the other half to Dutch tenants, the assumption is that both groups will have the opportunity to meet and interact on a regular basis and on an even footing, leading to the formation of social bridges between them.

Method
This case study is part of a larger research project which focuses on the role of Dutch housing corporations in supporting residents' groups in the context of selforganisation and self-management and governance practices.The project aims to shed light on the extent to which this collaborative housing approach to refugee integration can help integration.We opted for a case-study in order to capture the specificities of the approach; the combination of the different elements that define the Startblok model is fairly unique, in that it brings together housing for refugees, collective self-organisation, and social housing allocation policies at the municipal level.Our chosen methodology has some constraints.The case is still in statu nascendi, which enables us to investigate only the first year and half of it.Further, it is a unique case, which prevents comparison and generalisation at this stage.In addition, the limited conceptual knowledge about this type of approach has prompted us to adopt a more explorative approach to this single experiment.

SELF-ORGANISATION:
The field work started in February 2018 and will continue until December 2018.In line with the case study research design, we apply ethnographic data collection techniques including semi-structured interviews with residents and with representatives from the housing corporation.In addition, we have conducted participant observation on site, e.g., during residents' meetings.This allows us to supplement the interviews with observations on how the different actors interact with each other.We have interviewed both refugees and Dutch tenants.In parallel we have reviewed secondary data, including the project website, policy documents, newspaper articles, and audiovisual material.This review also includes the regular monitoring of developments in Dutch integration policy.Additionally, the housing corporation made available recent survey data on tenant satisfaction in this project.

Refugee Integration through Collaborative Housing: Preliminary Findings
In this section we present our preliminary findings, according to the analytical framework presented in Section 2.4.We start with a description of the organisational structure of the Startblok project, followed by our initial results on 'outcomes' related to each of the 'necessary conditions' outlined in our analytical framework.

The Startblok Project
When in 2015 the Amsterdam municipality had to deal with an unexpected redoubled influx of refugees to house, they looked for innovative approaches.A local councillor came up with the idea of mixing young adults with young refugees (Van Veen, 2016).The city council supplied the grounds and the infrastructure: roads had to be laid and electricity and sewage installed.Amsterdam housing corporation De Key was responsible for moving and installing the housing units.De Key has recently, as one of the first Dutch housing corporations, changed its official status from a general social housing provider, to one that only caters for young adults, defined as those in the age category 18 to 27.Other partners involved in the Startblok project are Vluchtelingenwerk Nederland, an NGO that receives government funding for helping refugees with their integration into Dutch society, and the municipal social services, that are also involved in the integration trajectories of refugees.
The Startblok is situated in the South of Amsterdam, on the site of a former sports accommodation consisting of grass fields.Although in the outskirts, the site enjoys good public transport connections.The housing consists of retrofitted container units, which have been used for another project before.By July 2016 the first tenants moved in-half of them are young refugees recently given a residence permit and the other half Dutch young adults.The aim is to give all these young people a springboard into a successful adult life, hence the name 'Starting block'.The goal is to create a community by letting residents organise and manage the project themselves.The Startblok consists of 463 bedsits, 48 shared apartments, a small office and a clubhouse.
Eligible for living in the project are lower-income singles without children from 18-27 years of age, who are in education, looking for employment or already employed.The majority of the refugees are from Syria and Eritrea.In line with the demographic composition of the recent influx of refugees in the Netherlands, they are mostly male.The Eritrean refugees usually have a low degree of literacy and low educational levels.The Syrian refugees, in contrast, tend to have middle to higher educational levels.The gender distribution among the Dutch tenants is more even, with a slight overrepresentation of female tenants.They mostly have the Dutch nationality, but there is a small minority of tenants with a Moroccan or Turkish background.The Dutch tenants reflect the Amsterdam population in that they are often highly educated.All tenants obtain a five-year lease.
The project's organizational structure is illustrated in Figure 3.The buildings are divided into 19 corridors, each encompassing between 16 and 32 bedsits.Each corridor has a shared communal space and each bedsit contains a separate bathroom and a kitchen unit in the room.Refugees and Dutch tenants are mixed throughout the corridors, ideally alternating every bedsit, so one Dutch, one refugee, one Dutch and so on.Two of the tenants on each corridor, one Dutch, one refugee, are the group managers, responsible for managing the corridor.They are the first port of call if problems arise on the corridor.They receive a small discount on their rent as compensation for this.The tenants on a corridor are expected to meet each other weekly, for instance while sharing a meal.Tenants who do not comply with the house rules, such as no littering of the corridors, can be given a fine by the group managers.The practical management of the grounds and the housing is performed on a daily basis by the grounds team and the 'klusteam' or maintenance team.Each of these teams consists of five members, and they receive a discount on their rents similar to the group managers.
At the next level of the hierarchy is the project team, consisting of five tenants employed on a part-time basis; the social managers.The team manages the project on a day-to-day basis, interacting with the group managers and the grounds team in case issues are not resolved at the corridor level.Together with the onsite manager, the project team selects the new Dutch tenants.The selection process involves registration followed by obligatory information meetings and written applications.Refugee tenants in the project are matched by central government bodies.Furthermore, there are two PR managers; an administration manager who handles the paperwork with the leases; a maintenance manager who coordinates the maintenance team; and a community development officer.All of these people are employed part- time by the housing corporation.Finally, there is Actief (not included in the organogram), a more autonomous group of five tenants whose goal is to stimulate tenants to become active in organizing social events and so forth.
A central tenet of all these various roles is that they are all appointed from within the tenant population; one has to live in the Startblok to be involved at the organizational level.At the site only one professional is present who is not a tenant, the onsite manager (project coordinator) employed by the housing corporation who liaises with the tenants.The monthly meeting of all these groups constitutes the highest instance in the project.The principle of community formation through regular interaction between tenants is built into the DNA of the project.Upon arrival, tenants are asked to sign a manifesto (Figure 4) endorsing these ideas, and they are continually reinforced by the 50/50 principle, the day-to-day visibility of the group managers and the attempts by the project team and other active groups to directly engage with tenants and to organize social events in the club house and the shared outdoor space.The physical organisation of the housing further promotes this.On the corridors each room has its own kitchen and bathroom, and is thus in principle independent, but due to the fairly small size of the rooms the tenants also make use of the shared common room available to each corridor.In this way the project strikes a seemingly effective balance between tenant autonomy and community formation.

Preliminary Outcomes
In this section we present initial findings on integration outcomes of the Startblok model, focusing on the presence of different types of social connections: social bonds, social bridges and social links.We then reflect on the attainment of social connections so far, and reflect critically on the assumptions underlying the model.

Social Bonds
Our initial findings show that the first necessary condition of the Startblok model, namely 'demographics', translates in fact into age-related bonding.Most tenants indicate that they feel connected to the other residents in Figure 4.The Startblok Manifesto.Source: startblokriekerhaven.nl the project.One refugee stated: "We make friends here, and for me, I feel like having family here.We are more than just neighbours or friends.We respect each other" (open answers to the survey, KWH, 2018).Tenants for instance eat together with their corridor neighbours or have drinks.Friendships develop and some tenants organise social activities together such as barbecues or soccer games.That similar age plays a large role in this, is illustrated by how one of the refugees put it, when reflecting on why the Startblok succeeds at being a community: "All the people have the same age.Same mind same thinking; way of thinking is the same.Here a lot of people have the same interest" (open answers to the survey, KWH, 2018).
In addition to age-related bonding, our findings suggest the formation of other types of social bonds; while the 50/50 principle ensures some level of continuous interaction between refugees and Dutch tenants from the same age groups, there is also opportunity for the refugees to interact with people from the same cultural background (and with those with other backgrounds).For example, the Syrians in the project often interact with other Syrians.This allows refugees to leverage their own cultural support network (Van Kempen & Şule Özüekren, 1998), while at the same time being part of the wider, mixed Startblok community.The risk that refugees become isolated, or (at the other extreme) become completely segregated from the rest of Dutch society, is therefore mitigated.This is in sharp contrast to the traditional form of refugee housing, where people are often dispersed even within neighbourhoods.Ultimately, the Startblok manages to achieve a relatively high concentration of refugees in a small geographical area, without this being viewed as problematic by those in the surrounding neighbourhoods.

Social Bridges
The 50/50 principle of the project, which permeates all levels, has been actively maintained from its start.This holds not just for the housing, but also for the formal organisational roles that tenants undertake.For example, on corridors care is taken to ensure that at all times half the tenants are refugees and half are Dutch.One of the two group managers assigned to each corridor is always a refugee, and the other is Dutch.Likewise, the composition of the project team also reflects this principle, including three Dutch and two refugee members (or vice versa).This means that interaction between refugees and Dutch people is built into the model.Furthermore, the active observance of the 50/50 principle prevents the drift over time towards homogenisation sometimes observed in other forms of shared housing.The required involvement of refugees in the organisational roles prevents that all or most expert and organisational roles are undertaken by Dutch tenants.This ensures that both refugees and Dutch tenants share a sense of direct ownership of the project.Their active inclusion in organisational roles also contributes to the refugees' integration pro-cess.Through their collaboration with the Dutch young adults, they are helped in learning the Dutch language and understanding the local culture.For instance, the importance of the Dutch norm of being on time for appointments is passed on to the refugees in organisational roles.More generally, for those with part-time jobs in the project this work is usually their first experience of employment in the Dutch context, in terms of how payment, contracts and responsibilities are organised.

Social Links
The attainment of social links is described by Ager and Strang (2008, p. 181) as "the connection between individuals and structures of the state, such as government services".Access to such services was found in Startblok in the form of the on-site presence of Vluchtelingenwerk, the government agency that provides support for refugees.While refugees in conventional housing have to go to their offices elsewhere in the city, the immediate presence of Vluchtelingenwerk significantly lowers the barriers to seek this type of assistance.In this way, the advice on integration courses, opportunities for education and work that the agency offers, becomes more accessible.
While not explicitly considered as part of the 'Startblok model', we found the location of this housing project playing a significant role in the acquisition of social links by refugee tenants.Although in the suburbs of Amsterdam, the Startblok is well-connected to other parts of the city via public transport.A good location is key to facilitate access to employment and education opportunities, and to social networks that are expected to help refugees to form social links with the wider Dutch society and to prevent (economic) isolation.Furthermore, Ager and Strang (2008, p. 181) highlight: The benefits of living in areas where refugee settlement [is] more established, in that local services [are] seen as more capable of dealing with refugee's specific needs, thereby ensuring levels of access more in line with those of other residents.The Startblok's accessibility to the large and sociallydiverse city of Amsterdam provides opportunities for refugee tenants to form not only social links, but also to extend their social bonds and social bridges.

Conclusions
In this article we have presented initial findings from an ongoing study of an innovative approach to refugee integration through collaborative housing.This approach, launched by the Municipality of Amsterdam and housing corporation De Key in 2016, brings together young refugees and Dutch young adults in an even mix, following a 50/50 principle.The underlying assumption is that, through structured self-organisation, the daily interactions between people from each group will progressively lead to the formation of social bonds and social bridges, social connections that are key to the integration process.Our preliminary findings suggest that social connections are indeed being formed between both groups: as expected, social bonding happens across ethnic and cultural backgrounds by virtue of belonging to the same age group and household type (i.e., young single people).In addition, we found evidence of social bonding on the basis of common cultural and/or ethnic backgrounds, i.e., refugees bond with each other.This, however, doesn't stand in the way of the creation of social bridges between refugees and Dutch tenants, which can be explained to a large extent by the inbuilt social mix of the 50/50 principle.In addition to social bonds and social bridges, we established the formation of social links due to the accessible location of the project-an aspect that was not explicitly considered part of the model when first conceived.Despite not being in a central location per se, the accessibility to public transport connections to the city of Amsterdam provides refugees with opportunities to access not only education and employment opportunities, but also wider social networks, including with ethnic communities akin to theirs that can provide different types of support.These findings resonate, with literature that emphasises the importance of proximity to their own cultural and ethnic communities as part of the integration process of new arrivals, be it refugees or other types of migrants (Andersen, 2017;Van Kempen & Şule Özüekren, 1998).Given the importance of integration outcomes for contemporary European societies, for refugees themselves as well as their receiving countries, and the promising preliminary results, we argue that further research into refugee integration through collaborative housing is timely and urgent.Further analysis into the mechanisms of self-organisation and social mix in shared housing could provide part of the missing link of social connections in current integration theory, and could inform better policies and practices in the field of housing and urban planning to help the integration of young refugees in European societies.