STUCK IN THE MIDDLE: THE TRANSITION FROM SHELTER TO HOUSING FOR REFUGEES IN BELGIUM

Given the increasing number of people seeking refuge in Belgium and across Europe more broadly, it is essential that governments assume responsibility for the integration of refugees into society. An essential aspect of this process is the provision of a durable path from shelter to housing. Choosing a place of residence, and the success in obtaining suitable housing conditions, are factors that signi!cantly shape the integration process. In the context of refugee settlement in Belgium, however, asylum and migration policies to date have been predominantly focused on tackling temporary crises with little consideration paid to long-term integration and housing strategies. Due to separated policy competences (reception at the federal policy level, and housing at the regional level), and the absence of a sense of responsibility from both Flemish (regional) and federal government, voluntary organisations have developed signi!cant roles at the local level in the transition from shelter to housing, and in further housing needs of recognised refugees. Achieving this transition is impeded by capacity problems, discrimination, lack of local social networks, and limited timeframes. Using the concept of path dependency as a theoretical starting point, this article employs qualitative methods to highlight the impact of both asylum policies and the spatial characteristics of reception centres on transitions towards more permanent housing. Recommendations for refugee accommodation are made to enhance the transition from shelter to housing. Finally, it addresses alternative housing projects that are conducive to social integration as well as to the transition of refugees to the regular housing market.


Introduction 1.From an Asylum to a Housing Crisis for Refugees
Following the sharp increase in refugee ows in 2015, the issue of refugees and migrants has taken a prominent place on the political agenda.In 2015 and 2016, almost 60 percent of the people in shelters in Belgium were formally recognised as refugees or subsidiary protected people (CGVS, 2017).Most of them were young single men or families with children.The so-called asylum crisis was proclaimed to be over at the end of 2016.In practice, however, it soon became clear that the most signi cant challenges still lay ahead as the asylum crisis gradually evolved into a more long-term housing crisis for refugees.To date, refugees have been stranded in a transition from shelter to regular housing.In other words, refugees are 'stuck in the middle'.The lack of access to long-term housing acts as a hindrance towards refugee integration.Indeed, the home and living environment is an essential indicator and the rst step towards the integration of newcomers (Murdie, 2002;Francis and Hiebert, 2014;Firang, 2018;Teixeira and Drolet, 2018;Balampanidis, 2020).In addition, spatial characteristics of housing and reception centres (including available meeting spaces, mobility infrastructure, etc.) generate opportunies and resources that potentially promote upward mobility.These opportunities and recources include social systems; markets; institutions; human, social and ethnic networks, etc. (Balampanidis, 2020).Taken together these characteristics can form part of what Balampanidis (2020) terms a dynamic 'opportunity framework'.
The basis of this evolution from an asylum crisis towards a housing crisis for refugees has various causes.First, in Belgium, there is hardly any relationship between the strictly de ned policy domains of asylum and reception which are organised at the federal level, and domains such as integration and housing which are implemented at the regional level (Wyckaert, 2017).Due to this policy gap, no state institution is responsible for administering the transition of recognised refugees into the housing market (Vandevoordt, 2019).Unlike neighbouring countries such as Germany and the Netherlands, Belgium does not provide initial housing and newcomers must, therefore, nd accommodation by themselves.For many, this quest is extremely challenging due to numerous societal barriers.Speci cally, they have limited social networks, no steady income, often face ethnical discrimination by landlords, and a common language between landlord and refugee is absent.Moreover, the Belgian housing market is characterised by both a shortage of a ordable private rental housing and a scarcity of social housing (Saeys et al., 2018).
While the causes of the refugee housing crisis in Belgium can in part be attributed to the above mentioned policy gap and to shortfalls in the housing market, a less-apparent explanation can further be found in the recent reforms to Belgium's integration policies (Debruyne, 2019).Via a state-led transition, several competences were shifted towards the regional (Flemish) level and then gradually restructured.As did the task of ensuring integration which was, as a consequence of vote-oriented reasons and political convictions, gradually reformed into a policy grounded by individual responsibility, language acquisition, and citizenship (Debruyne, 2019).A substantial body of integration support, including securing access to housing as part of the nexus between reception and housing, was dropped from the regional political agenda.Numerous responsibilities ended up with local authorities, causing a lot of pressure within the local contexts of individual cities and smaller municipalities that was not at all compensated for by the regional level.It is at this local level that informal players such as voluntary and citizens' organisations arose as important actors developing strategies to ll these gaps (Mayblin and Poppy, 2019;Schrooten et al., 2019).This contribution identi es transit-housing projects as part of these strategies and proves that these projects not only represent a crucial extension of the transition period to regular housing but can also form an essential spatial and social environment for (further) integration.
Secondly, this contribution assumes the presence of path dependency in the housing trajectory of newcomers.In this context, reception situations, as well as potential transit accommodation, followed by primary (regular) housing in the host country, constitute a housing trajectory.'Path-dependency' is an aspect that is actively present in economic and social sciences, but is also valuable within housing research (De Decker et al., 2011).It states that 'housing history matters' in terms of the existing built-up environment and the neighbourhood as well as with regards to institutional arrangements and policy (De Decker et al., 2011).'Path-dependency' also exists, as Heringa et al. (2018) describe, because future behaviour, in this case in relation to both residential choices and increased opportunities in obtaining primary (regular) housing, is embedded in past actions and current situations whilst also depending on the 'knowledgeability' (language, spatial environment, social systems) of individuals and the information and values that they have obtained through interactions with others.Institutional arrangements in terms of, for example, organising reception, and the implementation of policy by means of whether or not to activate or mobilise asylum seekers, have a direct e ect on how such interactions take place and what social possibilities are available to asylum seekers.In addition, the built environment serves as an opportunity framework (Balampanidis, 2020) for interaction, as well as maintaining and expanding social ties and networks in the context of reception structures and transit-housing projects.

Domopolitics and its Implications
Political and institutional arrangements concerning asylum and organising reception are grounded in what Walters (2004) describes as a 'domopolitical' logic -a logic as to how to govern the state as a home.Domopolitics is a governmental approach in which the security and management of the mobility of people seeking protection are central points of attention.Speci cally, it produces a series of modes (and moments) of governance concerning the lives of, in this case, asylum seekers (Darling, 2011).Modes of governance appear to have a signi cant impact on the future housing trajectories of newcomers in their host countries.Darling (2011) discusses three ways, from the national to the more local level, in which domopolitics are re ected in asylum policy in the UK which is very similar to the Belgian approach: the ltering of refugees and the management of their mobility; the regulation of dispersal; and the disciplining e ect of accommodation.Within the dispersal system, and especially within the disciplining through refugee accommodation, it is not only the local implementation of asylum policy, such as whether or not to authorise and support local activity and mobility, which is considered necessary.There is also the in uence of the spatial context (the built environment) of the areas in which refugees are dispersed, and particularly the spatial characteristics of both the refugee accommodation and its immediate surroundings.Starting from reception locations as the rst residential environments in the housing trajectory of refugees, both aspects (local implementation of asylum policy and spatial context) are signi cant for the development or non-development of social networks.In a highly competitive situation on the housing market, social capital, in terms of social networks and the social recourses they comprise, has proven to be crucial for newcomers' ability to access housing in the transition from shelter to housing (Murdie, 2002;Firang, 2018;Adam et al., 2019).
Belgium's domopolitics consider large-scale and isolated collective reception initiatives as the standard.These are unfavourable conditions in accordance with the nexus between reception and housing from a path dependency point of view.This topic is discussed later on, theoretically and by assessing some collective centres as case studies.The contextualisation and problem de nition above formed the basis for a one-year master dissertation research in urbanism and planning.This research tried to answer the question 'how important the (social, spatial and political) nexus is between the stages of reception and housing for recognised refugees' access to the regular housing market?'

Methods and Overview
The paper is based on two types of data.First, the research draws on a study of secondary literature on refugee integration and an analysis of recent policy documents on reception, housing and integration.Second, several case studies are used comprising in-depth interviews with residents, neighbours, sta members of reception and housing initiatives, voluntary networks' leading volunteers and policy makers, as well as spatial analyses through 3D views, schemes and mappings.Case studies include both large-scale collective centres, which serve as a standard for the Belgian reception structure, and transit housing projects which may operate as critical social and spatial environments.The latter are aimed at assisting the transition into the regular housing market.Cases are situated in both the Netherlands and Belgium.Although in the Netherlands a similar domopolitical logic equally leads to large-scale collective centres as the norm, the government ensures a nexus with regular housing by organising rst (transit) housing.In Belgium, this transit housing is mostly organised by informal actors.
The rst part explores, through speci c research on the transition from shelter to housing, how governance in the reception stage is, or is not, in line with housing policy.It shows how NGOs and civil society organisations are an essential link in this transition process.This Belgian context is mirrored with the approach in the Netherlands.
The second part elaborates on the determinants of path dependency between reception and primary (regular) housing as part of the housing trajectory of newcomers.A rst section gives insight into the present 'domopolitical' logic and its e ect on the governance of refugees' mobility in Belgian asylum policy.The framework employed by Darling (2011) for his assessment of UK's asylum policy was used, as well as a comparison to the Dutch asylum policy.Social networks prove to have, in addition to supporting NGOs and civil society, an important role in providing access to housing.Therefore, as an introduction to the case studies, a second section elaborates on the interrelationship between social networks, spatial contexts, local institutional arrangements, and policy implementations (domopolitics) within reception and transit housing projects that shape the conditions for the development of these social networks.This is followed by an analysis of four cases, including two large-scale collective reception centres, and two transit housing projects.These case studies are used to assess the role of local spatial contexts and local policy implementations in determining residents' engagement with the local community and neighbourhood.It is argued that this engagement is necessary for developing and maintaining social ties and laying down roots in the community -essential for achieving path dependency between reception and regular housing.

The Absent Nexus Between Reception and Regular Housing from a Governance Perspective
Belgium has a complex state structure characterised by strictly de ned federal and regional policy areas regarding asylum and migration, integration, and housing (Martiniello, 2013).As a consequence of the gap between federal shelter and regional housing policies in the Belgian federal state context, the responsibility to assist refugees in their search for housing is continuously passed from one authority to another.Therefore, unlike other European countries, such as the Netherlands, Germany, Sweden, and Denmark, Belgium's governments take little or no responsibility for providing accommodation for refugees (D'Eer et al., 2019).After recognition, a transition period of two to four months starts in which refugees are expected to nd housing.As refugees are mainly self-reliant in this search, they can search anywhere in the country.For many newcomers, however, this term of two months appears to be unfeasible (VVSG, 2017) as, on average, the search for housing takes ve to six months (Vluchtelingenwerk, 2017).
Refugees face many barriers in their highly pressured search for housing.First, accessing social housing is challenging due to the criteria of local anchoring and language, as well as long waiting lists.As a result, refugees mainly depend on the regular housing market, but similar de ciencies exist in the low-cost housing segment of the market.In addition, many landlords refuse to rent to ethnic-cultural minorities or people with migrant backgrounds.'Taste-based discrimination' (Loopmans et al., 2014) or 'static discrimination' -the distrusting of ethnic minorities to ful l tenant obligations, is mainly caused by anxiety, language barriers, and an insu cient guarantee of adequate income (Vanderslycke, 2016).The stigmatisation of refugees hampers their access to the regular housing market.In the end, this leads to the segregation of refugees in speci c city neighbourhoods and abominable living conditions that prevent refugees from participating fully in society.
Housing is considered to be an integral part of the integration of refugees into society: accommodation, as well as living environments, are the contexts in which integration in other domains takes form (Ager and Strang, 2008).This insight does not appear to be re ected in Belgian policies.On the contrary, as a result of the evolution of integration policies, securing access to housing for refugees has disappeared completely from the regional policy agenda.The remainder of this section brie y sets out how the nexus between reception and housing was organised within regional integration policy, and the e ects of this organisation on refugee housing today.
It has been more than 50 years since, in the absence of a coherent political vision on integration, civil society organisations set up primary integration practices.Between 1980 and 1990, after shifting integration to the regional (Flemish) level, both civil and state-actors collaborated closely with one another to establish an integration policy (Vandevoordt, 2019).However, since then, a politically-driven evolution, based on a control and instrumentalise approach, has reformed the integration sector (Debruyne, 2019).In addition, nancial cutbacks in recent years have ensured a decrease in support for newcomers.These cutbacks have made support more conditional (based on language acquisition and citizenship) and, as a result, integration initiatives, including support in the search for work or housing, have been particularly a ected (Debruyne, 2019).Because of this downsizing, a lot of responsibilities for integration, including support to access housing, shifted to the local level (CPASs, CAWs, SVKs, VDAB).Due to a lack of volunteers and resources, or lack of political will, local authorities are not always able to take up these housing responsibilities (Vandevoordt, 2019).Sublimating the support to access housing as a part of regional integration policy, there is now no public authority responsible for the organisation of individual's transition from reception to housing.The gap that exists between these di erent policies remains uncovered.
History has repeated itself.Initiatives and projects by non-pro t NGOs and civil society organisations have arisen (Schrooten et al., 2019).These initiatives involve both support in the search for housing, as well as transithome projects and living-together types that can help support integration (Wyckaert, 2017).These residential initiatives often provide temporary housing, as an extension (in months or years) of the intermediate stage between shelter and regular housing.As of 2015, a substantial body of integration support is once again the focus of various citizens' initiatives and associations in Flanders.Today, there are about 60 organisations, initiatives and associations (D'Eer et al., 2019).
In the Netherlands, asylum seekers also stay in centralised (but often remote) reception centres, with all the typical inconveniences and consequences that such accommodation envelops.However, there is a connection between reception and settlement as there is one state-led response to overcome barriers.It is a model that starts from the hypothesis that 'one of the basic prerequisites for social inclusion is having adequate housing from which to live one's life in the community' (Quilgars and Pleace, 2016, p.5 in Allsopp et al., 2018).Part of the dispersal plan is that municipalities in the surroundings of reception centres are obliged to provide housing for refugees after recognition (Vluchtelingenwerk Nederland, 2017).Nonetheless, there are some negative implications to this dispersal system because the mobility of newcomers is again limited or determined by linking the individual refugee to housing in a speci c municipality.In addition, due to shortages in housing supply, these dwellings are often temporary housing or units, and refugees are assigned to houses scattered all over the country ( Van der Horst, 2004).As a result, the (limited) ties and preferences that people had or have developed remain unaccounted.Spatial and social links, demanded further in this article in relation to path dependency, are absent because of this often-arbitrary distribution system and how reception is organised.

Domopolitics in Belgium's Asylum Policy
Drawing on Darling's (2011) analysis of asylum governance in the UK, the analysis of the domopolitical logic in relation to asylum seekers in Belgium is divided into three parts: (i) the ltering and management of mobility, (ii) the regulation of dispersal and (iii) the disciplining e ect of accommodation.This analysis is conducted from a national to a local level.
4.1.1.Assuring Security Through the Filtering snd Management of Mobility Darling (2011) characterises domopolitics in the UK as being primarily based on the guarantee of safety and the limitation of mobility.The Belgian situation is not di erent.Security comes in the form of economic security allowing refugees who can boost and secure the local economy on the one hand, and in the form of the protection of internal security and personal safety of citizens on the other.In short, both the UK and Belgium are looking for highly educated migrants and 'real' refugees (Darling, 2011).That is why the system of domopolitics is not based on completely closed borders, but on the ltering of refugees through categorisation and selection as well as restrictions of mobility (Darling, 2011;Fisher et al. 2019).The categorisation is also re ected in the assignment of a particular integration path or route to be followed by the asylum seeker in which his/her mobility is controlled to varying degrees.

Regulating and Ordering Mobility Through Asylum Dispersal
In Belgium, as in the Netherlands and the UK, accommodation for asylum seekers is dispersed across the country.An asylum seeker is, depending on their individual characteristics (e.g.nationality, risk degree, or family situation), assigned to a speci c location.This limitation of mobility has grown and been strengthened over the years.Asylum policies have evolved from laissez-faire self-governance (Darling, 2011), which allowed refugees to settle anywhere and to bene t from nancial support, to an increasingly regulated system.Since 2000, this has resulted in a compulsory connection of material support (bed/bath/bread) to an appointed location (Pleysier, 2011).This dispersal system is a product of a domopolitical logic that states that the mobility of refugees should be limited and managed to avoid new migrants being attracted to the state because of its possession of an overly free social migration policy (Vanderbruggen et al., 2014).This logic and its related outcomes appear to be based on an incorrect hypothesis of pull factors (La eur and Marfouk, 2019).

Regulation Through Accommodation
It has to be noted, however, that the dispersal of asylum seekers all over Belgium is not really managed in a structured way; it is often determined by crisis management (Pleysier, 2011) and by the availability of vacant infrastructure at the time of crises.Because of the organisation of reception infrastructure as a crisis measure, these infrastructures are mainly large-scale vacant infrastructures such as holiday parks, nursing homes, hospitals and military barracks; sometimes in remote locations.Fedasil, the Belgian federal reception agency, signs contracts with local authorities and private partners.Additionally, there are also smaller local reception infrastructures (LRIs).As a result of ltering, categorisation, and the restriction of mobility, asylum seekers are assigned to a particular type of infrastructure.Highly-protected Syrians, as well as people in short-term resettlement programmes, are housed in small-scale local reception initiatives; Afghan refugees are allocated to large-scale, geographically more isolated centres for an extended procedure of about one year.This is evident from the following statement by a policy o cer: 'The reception model must ensure that the local reception infrastructures are able to proceed more quickly with the integration of asylum seekers who are in any case allowed to stay in Belgium, and, conversely, those who have to stay in the asylum centre know what awaits them' (San, 2016, p.1).As a result, there is little or no focus on the integration of asylum seekers residing in collective reception centres, even though there are also refugees who start in this type of reception centre.In summary, Belgian asylum policy allocates social security to speci c locations and uses allocation as a way of regulating, limiting, or determining mobility.
In addition, the accommodation itself can also be considered as an instrument for regulation within domopolitics.In this respect, a 'home' discourse or the perception of asylum seekers as 'residents' are almost entirely absent in the interpretation of policymakers and o cials ( Van der Horst, 2004).This results in geographically isolated locations or speci c types of infrastructure, as well as speci c local implementation of asylum policies.Both signi cantly in uence the daily lives of asylum seekers (Platteau et al., 2016) and the extension and maintenance of social networks.The latter is inherent to the transition from reception to housing based on the concept of path dependency.In addition to the spatial context in and around the centres, such as the presence of social, economic or transport facilities or places for encounters, the implementation of policies by local sta members of the reception centres is very important (discussed in more detail below).Another less obvious aspect is the stereotyping of asylum seekers and refugees in politics and policy which can have local impact.This discourse and a lack of proper communication has led to massive protests against new asylum centres or the creation of hostility between locals and residents of the centres.Allport (1954), Blommaert et al. (2004) and Soenen (2006) have already proven that contacts in terms of small encounters are crucial for the development of mutual trust and the further integration of newcomers.More attention is paid to this issue in the analysis of the case studies.
Despite the need to invest in integration during the pre-recognition phase, strongly regulated large-scale collective centres are politically considered as the ultimate standard.As a result, a strange paradox occurs.During their stay in these asylum centres, refugees are not supposed to integrate into society; and yet, immediately after their formal recognition, politicians expect refugees to integrate as quickly as possible.

Social Networks in the Transition from Reception to Housing
Aigner (2018) describes four di erent gateways for refugees to the housing market: migrant-assisted entry, non-assisted entry, welfare-assisted entry, and local-assisted entry.In Belgium, refugees receive support from social assistants of both the regional Integration Agency and local public social welfare centres.As already described, due to nancial cutbacks and the increased pressure on local authorities, the amount of time that these people can or want to invest in the search process for housing is limited (VVSG, 2017;Saeys et al., 2018).The 'non-assisted entry' has proven to be unsuccessful because of limited language skills, scarce nancial resources (CPAS support), discrimination, and a lack of both social housing and a ordable (private) housing.The increasing importance of actions by civil society has already been mentioned and can be referred to as 'local assisted entry'.
In addition to the assistance by NGOs, social networks often prove to be crucial in terms of nding adequate housing in highly competitive situations (Hanley et al., 2018;Adam et al., 2019) through the so-called 'migrantassisted entry'.According to Moris and Loopmans (2015) nancially weaker tenants facing prejudice solely nd housing within speci c niches.These properties are rare in the regular rental housing market and are often accessed through personal contacts within private networks.Such transactions can also be observed within the migrant-assisted pathway where refugees, via other migrants or their limited social networks, nd a place to live at the moment of transition to the regular housing market.As a result, refugees coming from remote large-reception centres, often relocate to speci c parts of cities, also known as 'arrival neighbourhoods'.These areas are often the rst points of entry for refugees to these cities, and give them access to cheaper housing, enable them to earn an income through informal forms of employment, and to participate in already existing ethnic institutions and so on (Park, 1915;Burgess, 1928).A high level of social cohesion often characterises these arrival neighbourhoods.They are transition zones from which migrants leave as soon as they acquire more advantageous socio-economic positions (Massey, 1985;Saunders, 2010).At the same time, it appears to be di cult for newcomers to extend their social networks beyond these arrival areas (Beeckmans, 2017).Despite the great solidarity between migrants and the opportunities o ered by these neighbourhoods, this entrance may also bring potential harm, such as the danger of ending up in a network where refugees nd, via mediators and in exchange for money, a dwelling that is too small and of low quality (Saeys et al., 2018).In addition, these neighbourhoods are often stigmatised urban areas, which can result in the stigmatisation of refugees through their association with these areas (Darling, 2016).

The Interrelation Between Spatial Contexts, Social Networks and Domopolitics Defining Path Dependency
The connection between structural integration, in terms of access to the housing market (actual dwelling), and social integration (social networks) has already been made explicit (Esser, 1980).Additionally, the living environment of the (broad) surroundings of a dwelling, and in particular, its social-spatial dimension, also provide necessary conditions for integration through interaction (Ager and Strang, 2008;Francis and Hiebert, 2014;Balampanidis, 2020).
Life paths, daily or long-term, are de ned by socialisations which take place in speci c spatial environments.These socialisations in uence subsequent choices and possibilities in residential location, activity patterns and social networks (Heringa et al., 2018).Despite acknowledgement that domopolitics has an impact on the daily lives of asylum seekers, there is a total absence in asylum policy of vision concerning asylum seekers as 'residents' and reception contexts as 'living environments'.The speci c socio-spatial characteristics of a given living environment, such as the presence of (public) meeting places, de ne opportunity frameworks (Balampanidis, 2020) that allow or limit interactions with others in terms of socialisation and the extension or maintenance of social networks.Otherwise, social ties create, as Heringa et al. (2018) state, familiarity with certain spaces, activities, and types of people.
In other words, personal capital, such as language, money and social networks (Putnam, 1993, in Ager andStrang, 2008;Aigner 2018), political constraints and spatial opportunity frameworks determine the conditions for integration in terms of daily spatial and social interaction.Existing social capital is part of these social ties.However, the employees of centres can also act as social ties within the reception stage.The same applies to initiators of transit housing projects in the intermediate stage.These people implement or cope with certain policies and, as a result, impose certain (political) constraints.However, they also organise activities in the area and provide links to both people and places in the given neighbourhood.A lot depends on how centres, and volunteer organisations, choose to deal with top-down policy, or the lack of it.
During the reception phase, Fedasil manages reception centres through an approach which 'integrat [es] the centre in the neighbourhood' through neighbourhood initiatives.The main aim of these initiatives is to integrate the centre as an institution and not the asylum seeker as an individual (interview Fedasil Brussels, personal communication, April 2017).This approach is, with regard to prominent political discourses, in favour of the development of mutual trust among asylum seekers and locals as generalised groups, but does not contribute to the strengthening of individual social and spatial interactions.However, research on collective reception centres in Belgium by Platteau et al. (2016) indicates that there are varying local implementations of top-down imposed policies.This diversity is related to the interpretation of the 'concept of integration' by community workers and by the management of the centres.Their vision has a signi cant impact on the organisation of activities that can interconnect centres with neighbourhoods (Platteau et al., 2016).The nature of these activities di ers when centres wish to focus solely on the integration of the reception centre with the neighbourhood or, the integration of individual residents in the community.Platteau et al. (2016, p. 54) reveal this with a statement by a centre director: 'Our task is not the individual integration of residents; they are here only temporarily.The centre, on the other hand, is structural, which is why the centre needs to be integrated into the institutional landscape'.Much also depends, of course, on the asylum seekers and local residents that the centre wants to reach with these activities and on the location where they take place (Platteau et al., 2016).

Case Studies
The spatial context of reception centres in Belgium (Poelkapelle) and the Netherlands (Houthaven, Amsterdam) were analysed with regard to their embeddedness in the neighbourhoods as opportunity frameworks for individual (social) integration of the refugees.
Similarly, housing types and environments in the intermediate stage can be considered as a framework for social integration through contacts in day-to-day situations.Housing types that facilitate refugees and locals living together can result in an extension of social networks.This can be of great value in the search for (permanent) housing.In order to demonstrate their value in terms of extending refugees' social networks, two cohabitation projects were analysed.

Reception Stage
Large-scale reception centres can improve social and spatial interaction that increase the likelihood of successful social integration and structural integration in terms of access to housing.From an urban planning point of view, these conditions for improvement mainly refer to the geographic location and the spatial characteristics of the direct and wider surroundings (as opposed to the architecture of the centres themselves).That is why this contribution focuses on the di culties and opportunities of spatial location and organisation (scale, geographical context, and so on) as an opportunity framework (Balampanidis, 2020) as well as on the programmatic organisation of centres in terms of initiatives to connect them to their respective neighbourhoods.

Local Functioning in Line or Against Federal Policy
The rst case is a Belgian state-driven initiative in the remote rural municipality of Poelkapelle.Neighbourhood activities, such as guided tours and movies, mostly inform neighbours of the centres' functioning and the situation of its residents in an attempt to change attitudes and create a sense of security.As one community worker stated: 'Until the establishment of the centre, no stranger had been seen walking the street in Poelkapelle' (Interview Fedasil Poelkapelle, personal communication, April 2017).However, current community activities do not lend themselves to actual encounters between residents and centre residents.Despite this lack of encounters, interviews with centre residents at one of these neighbourhood initiatives (a movie night) indicated this to be their main reason for participating: 'I participate because I want to talk to people a lot outside the centre and this is the only way, but there are not a lot of days like this and not a lot of people are participating' (interview at resident centre Poelkapelle, personal communication, April 2017).Platteau et al. (2016) stress the importance of activities that are also useful for centre residents.By focusing on activities that respond to the interests of both centre and neighbourhood residents, both parties will be more inclined to participate.It results in more of what Valentine (2008) has described as 'meaningful contacts': contacts that change values and attitudes in the form of positive respect, more than mere tolerance (Valentine, 2008, in Van Kapel andNuis, 2014).Recently, investments are being made to engage asylum seekers in local activities through, for instance, voluntary work.However, this remains only a small part of the mission of the Belgian government since asylum seekers will, according to their own perceptions, stay only temporarily in the reception centre as there is no certainty concerning their likelihood of asylum recognition.In addition, the rural municipality of Langemark-Poelkappelle has little means of support (facilities, leisure and sympathisers) to integrate 300 asylum seekers through participation in local associations and sports clubs or through personal ties with residents (interview Fedasil Poelkapelle, personal communication, April 2017).
In the second case, a project of the Dutch NGO 'Ondertussen' in the former Houthaven in the city of Amsterdam is trying to connect a (future) reception centre to the adjacent neighbourhood.Contrary to the governmental approach in the rst case, the civil society organisation is trying to create a win-win situation for both the existing neighbourhood residents and the future asylum-seekers.That is why the mission of 'Ondertussen' is, at rst, to invest in the reinforcement of the surrounding neighbourhood, starting from the needs and interests of current residents, such as the construction of a new community centre.The network of activities is de ned by the organisers as an intermediate environment in which future asylum seekers will be more engaged in relation to the neighbourhood.Residents of the asylum centre will be linked to this community network through personal interests, making contacts more meaningful (Valentine, 2008).The engagement of asylum-seekers starts from their talents and their ability to actively take charge of their own lives (interview with initiator of Ondertussen, personal communication, April 2017).The vision of 'Ondertussen' is that these e orts may also be bene cial if asylum seekers return to their country of origin because they are more likely to be mentally prepared (interview with initiator of Ondertussen, personal communication, April 2017).This perspective is in sharp contrast with the Belgian government's approach that invests little or nothing in integration based on personal interests because of the uncertainty of nal recognition.

Difficulties and Opportunities in the Spatial Context
Each geographical context has its pros and cons.Though centres in remote rural contexts face less opposition from nearby residents, they face more challenges to get these people involved in their daily operations and also face problems when trying to stimulate contacts between local residents and centre residents (Platteau et al., 2016).There are no public meeting spaces within a radius of one kilometre from the centre in Poelkapelle.The social and economic facilities in the area are limited, which results in insu cient services for the capacity of the reception centre.Poor access to public transport adds to this problem: 'We can't leave here and during the weekend we do nothing … there are only three buses the whole day and there are no activities in or out the centre… we do nothing' (interview at resident centre Poelkapelle, personal communication, April 2017).A new approach is needed, but the scale and location of centres such as Poelkapelle make this challenge less achievable.The urban context of the (future) centre of Houthaven, on the other hand, o ers more potential for integration through interaction at, for example, public meeting spaces.Furthermore, good access to public transport is bene cial for the development of local networks in other parts of the city.

Intermediate Stage
A rst project in the intermediate stage, called Startblok, is located on a brown eld in Amsterdam, and houses about 500 students and young adults including equal amounts of refugees and locals.It has been developed within the 'housing rst' principle from the Dutch dispersal system in which all municipalities have to do their part in housing refugees.The system encourages municipalities to consider alternative forms of housing and to develop methods of integration within both the immediate environment and more extensively throughout the municipality (interview Orbit NPO, personal communication, February 2017).Although Amsterdam has a signi cant problem with regard to a lack of a ordable (social) housing which a ects a lot of young people including graduates and students, it is also obliged to temporarily house asylum-seekers and provide housing for a certain number of refugees after recognition.That is why the housing association 'De Key', the municipality of Amsterdam, and 'Socius Wonen' started a container dwelling project in 2016 to tackle both issues at the same time.A similar stage of a new beginning in the life of the residents facilitates the connection to one another: 'Contacts certainly take place more often because of the equal lifestyle of residents' (interview with Startblok resident, personal communication, April 2017).However, the project also has some disadvantages.The location is remote, which allows for little or no contact with residents of other neighbourhoods.A large number of residents also results in ethnic segregation, strengthened by language barriers: 'You can see that, because of language communication problems, people of certain groups are attracted to each other.Particularly, because of the large number of residents' (interview with Startblok resident, personal communication, April 2017).To increase social integration through contact with other ethnic groups, key persons ('gangmakers' in Dutch) facilitated and stimulated contacts in this large-scale project.This type of scale and location is often seen concerning temporary housing.Container units, such as those used in this project, (architecturally) leave a temporary impression, and people prefer not to see them in their neighbourhoods.In this way, a type of ghetto is built that makes its residents feel unwanted.As a result, this type of project is often placed on the outskirts of a city or municipality (interview Floris Alkemade, Plattegrond, November 2019).Other kinds of typologies or units that are implementable in the urban fabric (e.g.inside vacant buildings) should, therefore, be explored.A second case involves a co-housing project 'Solidair wonen' in the middle-sized Belgian city of Sint-Niklaas.It accommodates vulnerable individuals, such as refugees and long-term homeless people.These people have similar urgent problems; speci cally an immediate housing need.Yet there is a di erence: homeless people need permanent housing situations, while housing projects in this intermediate stage are somewhat temporary for refugees.The intentions of the project are great; however, the (mental) vulnerability of some residents goes much further than a genuine housing need, creating a less than ideal situation for the refugees.Their integration is hampered due to the lack of a robust support group.The local Belgian residents cannot o er the stable living situation that the refugees need to participate and to be introduced to Belgian society, habits, and language.During interviews, some of the residents indicated that they had limited social networks and were also a little indisposed (isolated): 'Because of my past, I am quite reclusive and I live a bit like a hermit.Therefore, I do not receive many guests or friends' (interview with resident of Solidair wonen, personal communication, March 2017).Because of this, the expansion of (local) social networks for newcomers through other residents appears to be more di cult.
Even though shelter as well as housing in the intermediate phase are temporary situations, they should be seen as proper residential contexts.The locations, scales and typologies of these projects should be assessed in relation to potential social and spatial interactions.In need of 'home' discourses, or the perceptions of asylum seekers as 'residents', it is imperative that, through activities, an e ort is made not only to integrate centres in their neighbourhoods, as is the case in Poelkapelle, but also its residents, as is the case in Houthaven.Such an approach will contribute to improving stronger social ties and know-how of newcomers in favour of their future housing pathways.

Conclusion
This article is a plea to raise awareness for a (currently-absent) political and socio-spatial nexus between reception and regular housing.From the reception phase onwards, interventions should focus on the social integration of newcomers in terms of providing interactions with others.These interactions generate knowledgeability (language, spatial environment, social systems) that determine path dependency in terms of subsequent opportunities and choices in residential locations, activity patterns and future social networks (Heringa et al., 2018;Adam et al., 2019).The socialisation of refugees during the reception and intermediate stage is in uenced by political constraints and the spatial context as an opportunity framework for socialspatial interactions.
Reception in Belgium should be organised more structurally and not be based on crisis measures.Such an approach would take into account the aspects of spatial context and (neighbourhood) activities as the local implementation or interpretation of federal asylum policies.The reception policy (domopolitics) should be more focused on a 'home' discourse in which, instead of the current emphasis on stagnation and isolation as a part of ltering 'real refugees', reception is seen as an (in uential) part of the housing trajectory of refugees.Evident from the presented case study, the current location of several Belgian collective reception centres, in terms of accessibility (public transport), the absence of nearby public meeting places and with regard to the scale of the municipality (population and local services), is rarely ideal for providing social and spatial interactions.To organise reception structures, a geographical 'qualitative' distribution plan is needed in which asylum seekers are spread across municipalities, based on the presence of adequate social, economic and transport services.The number of asylum seekers to be accommodated should depend on the population size of the municipality.By doing so, a 'true opportunity framework' for social and spatial interaction would be provided that might be bene cial to the development of ties and networks.
In the absence of the conditions mentioned above, a strange paradox occurs.During their stay in asylum centres, refugees are not supposed to integrate into society.Immediately after their formal recognition, politicians expect refugees to integrate as quickly as possible.However, 'access to housing' no longer has an active presence in integration policies.For this reason, it is vital that, analogous to the Dutch 'housing rst' principle, this again gains greater prominence within regional integration policies.More permanent housing is preferable but, because of shortages in the (social) housing market, investments in transit housing or organisations providing these types of housing should be a short-term priority.Investment in transit housing extends the search time as well as providing an opportunity for the development of (social) networks.Emphasis should be on the possibility of providing housing in the vicinity of reception centres, so that ties that have already been developed (spatially, socially) are not lost as they may be relevant for further integration processes (e.g.work, and language).The non-permanent character of these transit residences is of great importance.Moreover, they are integral to the particular housing trajectory of refugees, since rst housing accommodation after reception is often temporary.This nding became evident from interviews within the 'Solidair wonen' project.Despite the temporary character of such accommodation, it is crucial to take into account location and typology.Container dwellings such as those in Houthaven (Amsterdam) are temporary units situated on derelict or underdeveloped sites and are unpopular with local communities.This type of accommodation is not the context in which residents can be part of a community.
These pressing interventions and investments concerning reception and initial (transit) housing are a vital rst step towards the development of housing pathways for social inclusion and integration, as well as the recognition of refugees' human dignity despite their temporary legal status.

Figure 1 -
Figure 1 -Trajectory from Reception to Housing in Belgium (By Author)