1While, in recent years, large European cities have appeared both as typical places of settlement for exiles  and as key players in reception (Agier, 2016; Bontemps et al., 2018), small and medium-sized cities located outside of metropolitan and border areas also constitute spaces of arrival, passage and even settlement of exiled persons, in particular through dispersal policies. In France, in 2015 and 2016, these processes took place for the most part through the establishment of reception and orientation centres (centres d’accueil et d’orientation – CAOs) (Rigoni, 2018; Barnier et al., 2019). These new accommodation and administrative support structures were designed in November 2015 to “decongest” the Calais “jungle” and then reinforced in 2016  to shelter exiles evacuated from the Calais area or from the large street camps in Paris. They were presented as “respite centres” to allow exiles to reflect on their migratory journey and to take the necessary steps to apply for asylum. By autumn 2016, 259 CAOs had been created throughout France.  This national distribution shows an intention to promote the geographical dispersal of exiles in non-metropolitan urban contexts which are therefore not traditionally concerned by the settlement of these populations. Initially conceived as a “a temporary buffer zone” (“sas d’accueil”) receiving people for a one-month period, it is now clear that these structures are permanent and that exiles remain there for long periods (several months, even years).
2Although it is difficult to define the precise boundaries of a category encompassing these small and medium-sized cities in which the CAOs have been opened, certain urban and socio-demographic characteristics can be identified. Beyond the numerical thresholds,  economic and functional aspects provide a clearer definition, in particular a role as an urban centre with a radius of ten to twenty kilometres, without being the main hub of the departmental or regional territory (Mainet, 2011). In more specific terms, some of these cities are facing economic and demographic decline, combined with deterioration of the buildings in the city centre and significant residential vacancy. In these particular urban settings, significant population loss (Haase et al. 2016) is mainly related to a migration deficit (Beal et al., 2016): it is largely based on “the (non) capacity of cities to retain their inhabitants or to attract new ones” (Rudolph, 2017). In this context, national policies on the relocation of exiles can echo local development issues in these territories as well as their settlement policies (Jamet et al., 2014; Turpin et al., 2017).
3These specific criteria correspond to the two cities discussed in this article, Andouce and Rossac,  where CAOs receiving young single men mainly from Sudan, Eritrea and Afghanistan opened in autumn 2016. Beyond their socio-demographic and urban similarities, these two cities are both far from large metropolises and are not key centres of power at the national level, although each has specificities in terms of its geographical and administrative position - one is a sort of satellite sub-prefecture of a larger city, the other is the prefecture city at the centre of its conurbation. Which governance mechanisms, but also which of these cities’ specific interests, led to the opening of CAOs in these urban contexts? How did the actors in charge of setting up and managing the centres envisage and organise their establishment? What does the spatialisation of reception reveal about migratory and urban policies? These are the questions that guided our study of these two cases.
4The aim of our research was thus to explore how the settlement of exiles through CAOs was organised and what type of “reception”, to adopt the term used to designate this system, it could provide. While part of the research consisted of understanding why CAOs were set up in these cities, another part was devoted to analysing the structural dimensions of migratory and urban policies in relation to the most precarious exiled populations. This article is therefore based on an analysis of the practices and representations of those in charge of establishing CAOs. We thus consider cities as systems of actors formed and mobilised around these schemes, including public authorities - municipalities, local state representatives - but also actors in housing and institutional reception – providers of social housing, managing officers of CAO systems - and members of civil society who are mobilised to support the settlement of exiles. During our qualitative research, we met thirty-two people in these two cities through eighteen semi-structured interviews and about ten informal discussions during observation sessions, between January and April 2018.  This article focuses on the spatial dimension, at different levels, of the opening of CAOs. Our observations concerned, on the one hand, the typologies, spatial arrangements and urban locations of CAO structures and their environment, and, on the other hand, the interactions between respondents and exiled persons, within CAOs and within associations.
5Our investigation drew on several hypotheses. Firstly, the size of these cities, their administrative position and the distance from the administrative and economic resources that are essential for exiles could lead us to assume that the expertise in supporting exiles was weaker than in large cities. The unprecedented nature of these schemes in these specific areas, and the fact that they were presented as temporary, could therefore lead to more difficult conditions for exiled persons. However, these new experiences could also lead local actors to develop new ways of organising “reception”, based on the needs and expectations of the people received. Urban-planning contexts in which there is a major drive to reduce vacancy could lead local actors to strengthen reception measures in order to ensure the sustainability of their facilities. Although these elements tend to relate to local particularities, our survey was also intended to highlight more structural elements of the current governance of migration. Indeed, the setting up of CAOs in small and medium-sized cities can lead to an increased role for local actors and can therefore be seen as an indicator of the shift towards a “local focus” in migration policies (Scholten and Penninx, 2016; Zapata-Barrero et al., 2017).
6In the first part, we will analyse what the establishment of CAOs in small and medium-sized cities reveals about governance of migration. While it seems to be largely defined by a centralised and quantitative dispersal policy, carried out on an urgent and imprecise basis, and permeated by the norms of national and European migration policies, the positions taken by local authorities and local interactions nevertheless appear to be decisive in explaining the opening of structures in these cities. In the second part of the analysis, we will look at the cities’ attitudes towards the settlement of exiles, showing that the support of local actors towards these measures can be understood, in particular, by the residential vacancy rates, which lead them to regard the arrival of exiles as an “opportunity”, provided that it remains discreet. Finally, the third part of the article will present the top-down character of urban policy, despite the novelty of the reception schemes for these cities. Housing tenure and geographical distribution are decided according to conventional urban policy categories and ethnicising representations, rather than the wishes of the exiles.
Dispersal in small and medium-sized cities, revealing the influence of the state in the governance of migration
7In a context analysed from the perspective of multi-scalar governance of migration and a move towards a local focus of migration policies, CAO schemes emanate directly from a state organisation with a rationale of geographical dispersal throughout the national territory. What does the establishment of CAOs reveal about the “vertical dimension of the multi-level governance of migration” (Zapata-Barrero et al., 2017)? Can we read into this a continuation of state predominance, within the framework of a Europeanisation of policies reinforcing national security focuses (Guiraudon, 2010)? Yet CAOs only open in some small and medium-sized cities: local situations could therefore be determining factors in understanding the establishment of these accommodation and administrative support structures. From this point of view, what roles do territorial actors play? There are several types of local institutional actors, which we will group together under the term local actors, who are directly involved: the departmental representatives of the decentralised state, i.e. the departmental directorates for social cohesion and population protection (DDCSPP - in charge of implementing social policies, particularly on access to housing), local government elected representatives and technicians, as well as social housing providers and associations managing facilities, which often operate at departmental level.
Opening of CAOs: an urgent and silent imposition of the national level on the local level
8Andouce is a city of about 15,000 inhabitants in the Centre-Val de Loire region, whose mayor, who has been in office for about 40 years, made a name for himself on the national stage in September 2016 with a resolutely “refugee-friendly” speech. Shortly afterwards, a CAO with the capacity to accommodate fifty exiles was opened in the city. Located in the heart of the Midi-Pyrénées region, Rossac is a city of about 20,000 inhabitants. With the agreement of the mayor and the federation of municipalities (communauté de communes), a CAO was opened there in November 2016 to accommodate 23 exiles, while the neighbouring municipality received 18. In Andouce, as in Rossac, the various local institutional actors involved describe receiving an order to open a CAO from the state, from Paris, through the intermediary of the DDCSPP, which is responsible for monitoring the opening of facilities at departmental level.
9In November 2016, the state and the prefecture of Paris called on all French departments to find empty accommodation to house exiles evacuated from Calais or Paris. The municipalities were therefore contacted by the DDCSPP: mayors had the opportunity to make known their readiness to participate in the reception of exiles. Then the DDCSPP launched calls for tender to select managing associations of the CAOs that were to be opened: in the two cases examined, the local associations selected were those that managed the integrated reception and orientation service and the 115  scheme locally, as well as several emergency accommodation facilities in the department. However, they did not have specific expertise in dealing with exiles, most of whom were asylum seekers. Local social housing providers then rent accommodation directly to these managing associations, who are responsible for allocating the exiles to the accommodation.
10The various institutional actors we met all emphasised the extreme urgency with which this process took place. Local authorities and social housing providers spoke of “state pressure” to open these facilities. Moreover, it appeared from several interviews that information from the state on these new facilities had been extremely opaque. For example, an officer from the housing department in Rossac explained that the municipal authorities were not aware of the opening of the CAO until five days before it took place, while a social worker at the CAO stated:
“We knew a bus was coming when we saw the dismantling [of camps] on TV.” (Interview, January 2018)
12The state’s request came with an order to maintain confidentiality, or even complete silence, as explained to us by the Director of the Office of the Mayor of Andouce, who was very involved in the process:
“I was in a team with the representative of the DDCSPP, but in a blind partnership. It was a sensitive subject... well, there was no press... it had to be extremely confidential when we started working on these matters.” (Interview, February 2018)
14In the case of some social housing providers, officers only discovered that there were files for a CAO at the housing allocation committee, because the director could not give them the information beforehand:
“I was asked not to say anything, to create a blackout... because there is a fear surrounding the arrival of migrants.” (Interview with the director of a public housing agency (Office public de l’habitat – OPH), February 2018, Andouce)
16These elements seem to indicate a clear imposition by the national level on local authorities. The hurried and very directive nature of the process may suggest that the decision to settle exiles in small and medium-sized cities through CAOs is taken by the state alone, in particular by the centralised bodies which then send directives to the departmental directorates. While the haste and lack of transparency surrounding the first arrivals of exiles by bus may have been understandable given the scale of the evacuation operations in October and November 2016 in Calais and Paris, it seems that, prior to the routine operations which were carried out in the months that followed, there was still as little clarity and comprehensibility for local institutional actors, and even for the locally based state actors, the DDCSPP. Yet the DDCSPP keep records of the places available in the CAOs, which they report to the French Office for Immigration and Integration on a weekly basis. In the end, it is this institution that has control over the distribution of people throughout the country and which decides, according to evacuations or exits from reception centres, on the destinations of the exiles. Moreover, the various actors we met have little or no role to play in the “referral” of exiles, i.e. in their departure from the CAOs and the conditions of such departure, which depend on the administrative status of exiles managed by national bodies.
The fundamental role of municipalities and local interactions
17Nevertheless, it seems that the positions taken by municipalities and local interactions are very important in the decision to set up a CAO in one town rather than another. The views of local elected representatives are fundamental, particularly in influencing other local actors to become involved, in particular social housing providers and managing associations.
18Of course, there are differences between the cities studied: the first publicly claims to want to “welcome” exiles, while the other does not communicate on the subject at all. These differences can be understood in the light of the political characteristics of the two cities: one is on the left of the Socialist Party and its mayor is publicly very committed to the fight against social inequalities, while the other is close to the more centrist strand of the same party, taking a softer public stance on social issues. The contexts of local urban governance could also explain the differences in their positions: while the municipal authorities in Andouce appear to be central to the development of all initiatives in the city, the governance of Rossac is partly controlled by the federation of municipalities, and several services which are essential in relation to urban issues, such as housing, have been merged between the municipal authorities and the inter-municipal body. However, as the president of the federation of municipalities is the mayor of the city, he has a strong influence on planning decisions, as is the case in the other city. In both cases, the municipality’s stance was a prerequisite for setting up the CAO project: it would not have been opened without their agreement.
19This way of operating is often found in other cities, often within the same territories. We heard many accounts of municipalities near those studied in which a CAO had been planned but had not been set up because of opposition from the city - either from the mayor directly, or from the municipal council, or even from local inhabitants leading to a retreat by elected representatives. One mayor of a small city near Andouce said that although it was an action that she personally supported, she would not have taken this decision without the clear agreement of the municipal council:
“Advocating a project like this at the municipal council is not easy. I didn’t have to ask for their opinion, I have the authority to decide, but it had to be a project supported and championed by the elected representatives. If I had felt that there was too much opposition, I wouldn’t have done it.” (Interview, March 2018)
21However, agreement of the cities is not always synonymous with a full commitment to the reception cause since proactive discourse is not present to the same extent in each city. Indeed, the order conveyed by local state actors may be strong enough for municipalities not to dare to oppose it. In Rossac, although there was an agreement from the local authority, its main function seems to have been to facilitate the installation of the CAO by not opposing it. Moreover, in terms of setting up the project, the actors we met all said that the distinctive feature of a “small department” is that “everyone knows each other”, which greatly facilitates the development and management of such a programme. Thus, not being a metropolitan area is presented as an advantage, as it allows for greater familiarity and easier exchanges between the institutional actors involved. This geographical proximity, which “forms the basis for the local involvement of actors” (Bertrand and Moquay, 2004), can also make it more difficult to refuse a scheme run by other actors in the territory, with whom the local authorities regularly collaborate directly.
22Ultimately, the study of the positions taken by the local authorities and local interactions reveals that the territorial dimension of the action goes beyond the municipalities chosen for the survey and lies more at the inter-municipal or even departmental level. Indeed, within the same geographical and administrative area, we found that it was often not only the city studied that was concerned by the deployment of the schemes, but also others around it. These kinds of local networks of cities involved can be found both in the case of those that have accepted the centres and those that have refused them.
The internalisation of categories used in national and European migration policies
23Despite these contextual specificities, the fact remains that the effects of the propagation of state categories can have a strong impact on local practices. Indeed, the administrative classifications, constructed at European level and which serve as a basis for national migration policy, seem to have been well assimilated by the various local institutional actors we met.
24Firstly, DDCSPP officers, local representatives of the state, who are not in direct contact with exiled persons, approach them through the official categories that emanate from the directives issued by national and European authorities: asylum seekers, refugees and even “déboutés” (rejected asylum seekers) or “dublinés” (“Dubliners”).  After all, they are in charge of disseminating these categories to other local institutional actors. Local government elected officials and technicians, social housing providers and the directors of associations in charge of CAOs also use these categories to describe their actions and the rationale behind them, primarily from a management perspective: administrative status determines (at least in theory) the possibility of being accepted in a CAO. Social workers at the facilities, who have more interaction with exiles, may develop more pragmatic adjustment practices regarding categorisation: however, several of them also display an internalisation of administrative and legal categories in their discourse.
25In this context, the most difficult situations for exiles are found among those whose asylum applications have been rejected and Dubliners, because from an administrative perspective there is no longer any need for them to be supported in and by the CAO. However, if they leave the official facilities, they have no further access to housing and are left to fend for themselves, possibly in emergency institutional accommodation, but often taken in by private citizen volunteers. While this is a fundamental issue, there was little criticism of the “Dubliners” category or of the application of the Dublin Regulation by those interviewed, although the social workers and the director of the CAO in one of the two cities studied seemed to be more aware of having room for manoeuvre in their management of these cases. While maintaining a legalistic discourse, they can thus occasionally circumvent the legal directives through personal initiatives and procedural tinkering. In the other city, the managing officers present themselves as having absolutely no choice in the (non) application of state directives. A form of denial of responsibility for their actions appears in their discourse, as they insist that “the state and OFII [Office Français de l’Immigration et de l’Intégration - French Office for Immigration and Integration] are in charge”, which, they say, leads them to focus their actions on the issue of emergency accommodation.  These officers, who are between thirty and forty-five years old and had never before worked with asylum seekers, seem to have integrated certain legal standards so well that when we asked them about the future of rejected asylum seekers and Dubliners, on several occasions, shielding themselves behind a strict approach to legality, they replied that “this is not an issue, as these people do not have or no longer have the right to apply for asylum in France.” Yet, one of the specificities of these legal and administrative categories is that they evolve, sometimes very rapidly from one law or directive to another, and that they cover situations that are very volatile - a Dubliner may escape application of the Regulation (“dé-dubliné”) and thus become eligible for refugee status in France, and a person whose asylum application has been rejected may file an appeal.
26These positions can be partly explained by the lack of specific training for these officers on migration issues, asylum applications and the immigration law, as the establishment of the CAOs was not backed up by any provision from the state in this respect. The local management structures, which wanted to seize the funding opportunity presented by the state’s calls for tender for the CAOs, have little or no previous experience of dealing with these specific groups. Their geographical location, far from central power, can also reinforce these effects of transferring responsibility “elsewhere”, according to rationales of “moral subcontracting” (Williams et al., 2016), the spatial dimension thus reinforcing the “interface of distance” suggested by the delegation of public tasks to private structures (Alberti, 2019).
27However, these positions are not necessarily specific to small and medium-sized cities; they can be understood in the same way as that which has been analysed in terms of the “ethos of civil servants” in research on immigration offices (Spire, 2008; Clément, 2018), showing the ways in which officials adopt a discourse of adherence to state categories and policies that is reflected not only in positions of principle, but also in discriminatory practices. These positions thus echo more structural phenomena of “legal regulation of complex social situations” (Barbou des Places, 2008): legal categories, apprehended as an objective reality, allow institutional actors to implement classifications that prevail over integration in the city (Belmessous and Roche, 2018).
When residential vacancy turns reception into a hidden opportunity
28This local support for the establishment of CAOs driven by the state can be understood in relation to the urban characteristics of these small and medium-sized cities: in the context of residential mobility, their conurbations are unable to retain their inhabitants and their populations have been decreasing steadily for the last fifteen years, while their residential vacancy rates, in slack housing markets, are higher than the national average.
29The low residential attractiveness of the area led several local institutional actors to consider the arrival of exiles as an “opportunity for a department in decline like ours”.  Yet, what are the conditions for this settlement of exiles to be positively apprehended by local actors? Do efforts to reduce vacancy give rise to the implementation of specific measures to “welcome” exiles? Which actors ultimately bear responsibility for reception in these small and medium-sized cities?
Hide these exiles from my sight: rendering invisible a subject construed as sensitive
30In the two cities studied, the first partnership approach between DDCSPP, local authorities and social housing providers consisted of identifying vacancies in the housing stock, in particular by means of very precise mapping. Thus, while actors in the housing sector most often use the argument of reducing vacancy in their stock as a motivation for opening reception centres, public state and municipal actors play a strong role alongside them in the choice of places for exiles to live. The very fact that the housing markets are slack was mentioned by all the institutional actors we met to explain the ease with which it was possible to open a CAO. Residential vacancy is therefore both what makes it possible to meet the state’s demand and what explains the investment of local actors, in particular social housing providers. Indeed, reducing vacancy in their stock is a strategic priority, as it allows them to have access to increased economic resources through additional rents. Population policies are major issues in local policies (Desage et al., 2014), with the declining socio-demographic contexts of the cities studied leading local actors to seek newcomers.
31However, this process of mobilising vacant dwellings is repeatedly justified: local actors underline that these dwellings “have not been taken from anyone”. This expression, used by several interviewees, is often followed by an explanation of the criticisms that could be expressed by inhabitants about exiles occupying housing that could be used by others. Local actors fear a nationalist discourse and pitting the most disadvantaged against each other: the fear of being accused of “only giving to migrants” and the rise in votes for the National Front  in the presidential and legislative elections are often cited. 
32For this reason, in addition to the high degree of confidentiality surrounding the setting up of CAO facilities, institutional actors emphasise that they do not want to make the presence of exiles too visible. In this context, during discussions, the local authority is often given the role of promoting social acceptability. For example, a deputy mayor of Rossac, in charge of social affairs in the city since 2008, explained:
“It’s a daily task to get this population accepted. At first, people are afraid of foreigners. I don’t know if it’s linked to the rural environment, but it’s said that people are afraid of those they don’t know.” (Interview, January 2018)
34Likewise, several local elected officials explained that they regularly have to respond to questions from their constituents about the presence of exiles in public spaces:
“People say to me, ‘What are they doing here all day? We see black people walking down the street, they’re just there, they’re not doing anything!’” (Interview, January 2018)
36Indeed, the example of Rossac revealed strong hostile reactions from inhabitants living near the accommodation occupied by the exiles. As in Andouce, the Rossac CAO strictly speaking corresponds to offices in the town centre where social workers can receive the exiles to provide administrative and social support. However, the exiles’ accommodation is located in a residence, close to the railway station, which had been largely unoccupied, as it was being renovated. From the outset, the installation of six flat shares gave rise to numerous complaints from former tenants about noise pollution and disturbances caused by people coming and going. The CAO’s social worker was involved in a daily attempt to calm these conflicts by mediating on site.
37During an observation session at the end of January 2018, a very tense scene unfolded, during which an inhabitant in her sixties shouted down the stairwell to the social worker accompanied by a young male exile, complaining about noises throughout the night that prevented her from sleeping. Later, in an exchange with another social worker, another young male exile explained that he too was unable to sleep the night before because a neighbour had been tapping a broom against her floor. He wondered why this woman had made such an effort to wake up the whole building, because, as far as he was concerned, he and his roommates were sleeping peacefully at the time, and it seemed to him that she could have simply asked the person who was disturbing her to stop making noise.
38Faced with such hostility expressed day and night, the exiles placed in this residence frequently ask to be moved, as they cannot bear the living conditions and the psychological pressure to which they are exposed. In this context, the mayor’s office stands out for its lack of public discourse. According to a local authority employee, there is, on the contrary, a determination that “it should not be too visible”. Neither hostile nor publicly hospitable, it seems that the representatives of this city prefer not to raise the subject too much in the public arena, for fear of negative reactions from their constituents - which nevertheless already exist.  The opportunity presented by the settlement of exiles is thus recognised, provided that it remains discreet. Although public discourse on reception is more explicit in the other city studied, the desire not to make the settlement of exiles too visible is also present. Spatially, this is reflected in the location and layout of the CAO: the office space is located on the ground floor of a small street in the city centre but is not marked by any external signs. This is where the fifty exiles meet individually and on an ad hoc basis with the managing officers to follow-up on their administrative file. The location of the shared accommodation in the city is kept highly confidential, according to the staff, in order to “protect the users of the CAO”, who are described as a “vulnerable group”  by the staff of the managing association, who have previously worked in traditional emergency accommodation structures.
39In the same vein, a point often made by those interviewed in Andouce is that all the social and cultural services to which exiles can have access are aimed at those who have recently arrived in the area as well as the “Andoucinois” (native population). The deputy mayor in charge of social affairs insists that there is no question of “making a distinction”:
“There is not and will not be any specific action for migrants. We see them as a population like any other. The best welcome we can give them is not to make any distinction.” (Interview, February 2018)
41Local actors thus adopt a position that corresponds to a traditional French public policy approach by emphasising the notion of equality rather than diversity (Escaffre and Lelevrier, 2019).
The role of civil society in reception: initiatives by inhabitants and institutional mistrust of volunteers
42In different proportions depending on the city, civil society intervenes to support and guide the settlement of exiles. These local volunteers may form a collective or act on an individual basis.
43In a town such as Andouce, where the municipality publicly claims to welcome people, it would be expected that local public actors would offer activities for exiles, especially as the young men in the CAO are inactive most of the time, since they have no possibility of working while their asylum application is being processed. Apart from very occasional French classes, no activities are specifically set up for the CAO population.  One of the explanations often mentioned in interviews is that it is a city with a highly developed cultural and social policy, aimed at all groups with socio-economic difficulties and not specifically at exiles. This is due to a certain working-class characterisation of the municipality  and the political values upheld by it. In fact, there are many cultural facilities in the city: a multimedia library, a cultural centre, a concert hall, a swimming pool, an ice rink, and several social actions aimed at the most socio-economically deprived people. However, events run by volunteers, such as food distribution, represent important opportunities for socialising and activity for the exiles in the CAO. In Rossac, as of December 2016, inhabitants formed a collective to support exiles. Its members organise numerous activities, in particular French lessons, but also football matches, games evenings and film screenings, administrative services, visits to the multimedia library.
44Support and guidance of exiles therefore seems to be largely provided by supportive inhabitants, in a collective or individually. Yet this commitment often aroused fear among the institutional actors we met. A deep mistrust of volunteers was expressed, particularly by CAO managing officers and municipalities. Volunteers are suspected or accused of being either too invasive or too activist, but in any case, of complicating the task of the institutional actors, or even undermining their legitimacy:
“Inhabitants should not be too intrusive, as they may interfere with the work of the social workers or give contradictory messages.” (Interview with DDCSPP representative, February 2018, Andouce)
“They aren’t ‘their’ users either! They have a tendency to play the professional and think they are better than the social workers. You have to manage them and set the rules.” (Interview with a managing association officer, February 2018, Andouce)
47This apprehension - often based on difficult experiences of interactions with volunteers who criticise the way CAOs are run - has given rise to the drawing up of volunteer charters or regulations to provide a framework for these activities and, in so doing, avoid making them too visible. Mistrust is particularly reinforced when the mobilised inhabitants explicitly position themselves as activists opposed to national migration policies. In this case, the problem lies in the fact that there can be a direct confrontation between what is expressed by these inhabitants and what is carried by the managing officers, since the former often oppose the state’s migration categories and policies while the latter are in charge of applying them. In the department of Andouce, housing rights activists were involved in particular, as an extension of their criticism of the failings of housing policies in general.
A top-down policy imbued with ethnicising representations
48The establishment of CAOs is a tool in the local effort to reduce residential vacancy but remains a new experience and is presented as exceptional in both cities. The local actors have to invent the modalities of installation and location of the centres. In this context, while the potential usefulness of the facilities could lead to particular attention being paid to the wishes of exiles housed in CAOs, do local actors take their needs into account? What are the rationales behind the spatialisation of accommodation within cities and the distribution of exiles in housing? How do the official reception objectives combine with the traditional categories and practices of urban policies, which do not involve the inhabitants and are based on specific population strategies?
Forced sharing of accommodation to teach exiles to “live as they do in France”
49The vacancy in the social housing stock in both cases studied concerns mainly large types of accommodation, with four or five rooms. This largely explains why CAO accommodation is organised in shared flats, in this instance with four or five people. However, the managing officers have discretion in setting up the shared accommodation. In discourse and practice, there is never any mention of the wishes and expectations of the exiles themselves. Moreover, in addition to the imposed typology of the dwellings, it is forbidden, as in all social housing, to carry out alterations, which can make it even more difficult to appropriate the accommodation (Deschamps et al., 2017). This is a characteristic of what Bernardot (2005) defines as forced residencies, “neither designed nor fitted out according to the needs of its occupants, but according to administrative and security principles decreed by the authorities”.
50These decisions, justified by the determination to promote “integration”, in the words of the interviewees, are imbued with ethnicising (Jounin et al., 2008) and culturalising (Guillaumin, 1972; Giraud, 1985)  representations, as shown by the explanations given by the managing officers of a CAO concerning their reflections prior to the allocation of individuals to accommodation:
“Is flat-sharing a way of avoiding isolation? There is the issue of isolation and then there is also the idea of taking responsibility for a flat with several people and sharing tasks, living with others on a daily basis, and then there is the... let’s say... the religious and ethnic situation which also comes into play, to say that here we are placing them with other users who may have a different religion, culture, ethnicity... and that the idea is also to get along together.” (Interview with CAO officers, February 2018, Andouce)
52This ethnicising and culturalising view of the practices, interactions and lifestyles of exiles is found in all our interviews with institutional actors. It is very often accompanied by the idea that exiles must be taught to live “as in France”. This attitude refers more broadly to the very top-down conceptions of urban planning and its institutional actors, who often think that they have to teach inhabitants how to live, in particular the working classes living in social housing, by disseminating “living standards” (Gardesse and Valegeas, 2016). Like other immigrant populations or those perceived as such, CAO residents are the object of an ethnicising and culturalising assignment, according to which, because of their perceived racialised otherness, exiles are unable to live normally (or even properly) in accommodation:
“They need to be taught French, they don’t even know what the CAF [Caisse d’allocations familiales – French family allowance agency] is! You have to teach them, just like you have to teach them how to live in housing. To close the window because heating is not free. You have to teach them how to live in France.” (Interview with a DDCSPP representative, January 2018, Rossac)
54While it is undeniable that people who are unfamiliar with the French administrative system benefit from being guided in their discovery of certain systems, the rhetoric used is based on an overbearing and infantilising stance (Pautard, 2015), which is often customary among institutional actors in urban planning, but also in the field of social support. Indeed, this top-down and restrictive approach to allocation of accommodation and occupancy rules can be found in other areas of urban policy on emergency accommodation and housing aimed at integration, despite the establishment in 2010 of consultative councils of people receiving accommodation and support, and declarations of intent to involve those “affected by these accommodation policies” to a greater extent in their design.  In fact, social workers often develop a “housing pedagogy” that aims to instil a “culture of habitation” in residents (Lanzaro, 2014).
55Ethnicising and culturalising representations are also systematically found in the factors used to explain problems between neighbours: according to the various institutional actors we met, if there are disputes about use, they are understandable because they are foreigners, non-Europeans, who inevitably have lifestyles that are far removed from those of their “French” neighbours - these are people who are “used to living outside, so they often leave their doors open”, according to a CAO managing officer, or “who often eat later than other tenants”, according to an OPH representative. Here again, what seems to be decisive for institutional actors, even though they do not work with exiles on a daily basis, is geographical origin and/or nationality and/or religion, more than social affiliation, age, gender or household typology. These arguments were used even when, shortly before the interview, the respondent had expressed disapproval or even indignation about the hostile reactions of neighbours, who were often considered intolerant or even ignorant:
“There are instances of intolerance, of people complaining. There were undoubtedly problems with the behaviour of migrants, who lived according to the rhythm of their countries. So what they were asked to do was to conform to French habits, they were asked, very politely, to pay attention, to meal times, to comings and goings.” (Interview with a social housing provider, director of OPH, February 2018, Rossac)
57These representations are based on stereotypes associated with geographical origins but are not combined with a consideration of cultural specificities directly expressed by exiles. It is not the particularities of the individuals that are considered, but the supposed gap between them and the norms of living thought to be French. The modalities of urban policy that govern the establishment of CAOs thus remain very typically top-down.
The strength of the “social mixing” category in urban policy: the success of “dispersion” versus the fear of “communautarianism”
58As regards the location of the flats where the exiles reside, once again it is not their wishes that prevail. Here, as in other medium-sized cities in France and elsewhere in Europe, people’s needs in terms of proximity to public facilities (Deschamps et al., 217; Brault et al., 2018), or their location in relation to each other, are not taken into account in the allocation of housing.
59In Andouce, the mapping work carried out in partnership, as mentioned above, led to exchanges between the municipality, the public housing provider and the managing association in order to “geolocate” the future arrivals and to choose, among the vacant dwellings, those that seemed to be well situated to receive them. It was then a determination to avoid grouping them together that led to so-called “dispersed” settlement. This is even presented by the municipality as an essential requirement:
“There were conditions set by the mayor, and he would have said no [to the establishment of the CAO] if all fifty people were housed in the same place. We spread out, we distribute, we ensure a mix throughout the city.” (Interview with the Director of the Office of the Mayor of Andouce, February 2018, Andouce)
61In Rossac, the situation is a priori different since all the CAO accommodation is located in the same residence, which was chosen because it was due to be renovated. Despite this effect of opportunity, discourse on the preferred nature of dispersed settlements for exiles can be found, in particular among the social housing providers and local elected representatives. While the management of collective accommodation can be presented as a more practical solution by the managing officers, allowing them to work more easily on a daily basis and to be closer to the users, the fear of the spatial concentration of immigrant and racialised populations, in what some actors apprehend as a potential “ghetto”, remains a determining factor:
“It risks concentrating and ghettoising migrants. That’s why we’re much more in favour of dispersed reception: when you have all the residents in the same place, it’s less conducive to integration.” (Interview with a social housing provider, OPH director, February 2018, Rossac)
63A local urban policy that aims to avoid groupings of people viewed according to their supposed ethnicity and otherness appears to be consistent with the category of social mix and the way it is commonly invoked by urban policy actors to avoid forms of “communitarianism” (Rhein, 2002; Epstein, 2013). Ultimately, the spatial distributions of exiled people within cities echo the technocratic and centralised management rationales and modalities of dispersal at the national level.
64The establishment of CAOs in Andouce and Rossac, small and medium-sized cities with poor residential attractiveness, is an interesting example of the link between the centralised nature of migration policies and their territorialisation in non-metropolitan contexts. Indeed, the multiplication and establishment of CAOs throughout the national territory can be understood as much by the dispersal policy carried out by the state as by the mobilisation of local authorities. However, migration policies developed at national and European level are still very important in determining the possibilities of reception. On the one hand, they govern the geographical and institutional orientations of exiles, according to their administrative status and the work of local state actors. On the other hand, they permeate the representations and practices of local actors, who use state categories to construct and justify their actions. CAOs in fact reproduce at the local level the sorting patterns of exiled persons, making the situation of rejected asylum seekers and Dubliners, who are numerous in these structures, particularly difficult.  The rationale for managing flows therefore seems to predominate over the issues involved in reception. Moreover, although the CAOs seem to mark a shift towards a large-scale systematisation of the spatial distribution of exiles, they are part of the continuation of dispersal policies that have been implemented for many years with cycles of installation and evacuation of camps, particularly around Calais since 2002 (Clochard, 2007; Guenebeaud, 2017). Furthermore, the example of these two cities shows how these forms of delegated reception implemented by the state lead local authorities to organise themselves very quickly, even though they lack resources and staff trained on asylum issues. This process of “informal decentralisation”, which can be analysed as a sign of the shift towards a local focus of migration policies, is thus imposed on municipalities that often have little capacity to adapt to new and urgent situations (Pauvros, 2014). More than simply updating national policies in non-metropolitan urban contexts, the spatialisation of these centres carries the risk of reinforcing the segregating and unequal effects of reception depending on the means and intentions of local authorities.
65While these results appear to be structuring elements of the governance of migration, others shed more light on effects specific to the contexts of the cities studied. The high rate of residential vacancy, particularly in the social housing stock, plays an important role in the deployment of CAOs. The mobilisation of vacant housing is not purely technical or random: local actors organise the location of accommodation within cities according to their population strategies, which are largely based on ethnicising rationales, without taking into account the wishes of exiles. The categories of urban policy are thus imposed on the populations concerned, with the success of the use of “social mixing” and “dispersion”, as opposed to the fear of “communitarianism”. Thus, in the specific contexts of the two cities studied, the opening of CAOs is in keeping with long-established practices. The novelty of these schemes in these cities could have given rise to innovation, but the lack of time, training and resources led local actors to reproduce traditional methods of action that were largely structured by national migratory and urban policies. Finally, what seems to be decisive in explaining the opening of CAOs in these cities is the encounter, or the opportunity effect, between a structural dimension specific to these urban areas (the residential vacancy of large dwellings and the needs of social housing providers to bring in rents), a political dimension (the interactions between local actors and the state in the context of a mobilisation of national reference systems) and a financial dimension (the strategy of the local managing associations, as the state’s calls for tender for the CAOs represent a substantial source of income for them, even if they are aimed at a group to whom they are not accustomed to providing services). Thus, we can detect what Belmessous and Roche (2018; 18) describe as “a form of violence in the reception of exiles for utilitarian purposes: (the exploitation of) their position of dependence for the benefit of the urban environment that agrees to receive them”. The reception that is announced in the name of the centres thus proves to be largely illusory, since it does not involve the setting up of specific activities or spaces to socialise.
66These policy practices reflect more broadly the usual modalities of urban policies, in metropolitan contexts as well as in these small and medium-sized cities, whether in terms of emergency accommodation (Lanzaro, 2014), population policy and social mixing (Lelevrier, 2010; Epstein, 2013), or even the participation of inhabitants (Gardesse, 2013). We observe a continuity in the use of “aid-based and repressive” management models for the working classes, involving the denial of their autonomy (Bernardot, 2008), as well as the idea, recurrent in many urban projects, that residents must be taught to live by controlling their residential behaviour (Pautard, 2015; Valegeas, 2016). These practices are coupled with an ethnicising and culturalising approach to the populations concerned, which refers to the production, both day-to-day and institutional (Sala-Pala, 2005), of ethnicity, and transforms these so-called reception facilities, like others before them, into spaces of waiting and confinement (Kobelinsky, 2015).
We use the term “exiles” to refer to the immigrant persons mentioned in the article, while most interviewees employed the category “migrants”. We therefore use the term “migrants” in extracts from interviews. Exiles in the CAOs have various administrative statuses and migratory paths, but the majority are dealt with under the national reception scheme (dispositif national d’accueil - DNA) for asylum seekers.
Article L 744-3 2 of the French Code on the Entry and Residence of Foreigners and the Right of Asylum (Code de l’entrée et du séjour des étrangers et du droit d’asile – CESEDA) — inter-ministerial directive of 9 November 2015, supplemented by the directives of 7 December 2015 and 29 June 2016 and the operating charter of July 2016.
Défenseur des Droits (French Ombudsperson) (2016), Rapport d’observation, démantèlements des campements et prise en charge des exilés Calais-Stalingrad (Paris) (Observation report, dismantling of camps and treatment of exiles in Calais-Stalingrad, Paris), December, 80 p.
Between 2,000 and 20,000 inhabitants for small cities, 20,000 and 100,000 for medium-sized cities according to the French National Institute for Statistics and Economic Studies (Institut national de la statistique et des études économiques - INSEE).
The names of the cities have been changed to preserve their anonymity to the extent possible. For this reason, the elements mentioned to characterise them are in some cases deliberately imprecise and we do not present any cartography.
Seventeen interviewees were institutional actors, local state representatives (departmental directorate for social cohesion and population protection [direction départementale de la cohésion sociale et de la protection des populations - DDCSPP]), local government elected officials and technicians, social housing providers, associations managing CAOs; eleven were residents who support the exiles, some organised in associations. We also conducted interviews with four exiles, who are not mentioned in this article. We are currently carrying out another research project, with Christine Lelévrier and Mariana Tournon, exploring questions that are an extension of those raised in this article. In this context, we met 134 additional respondents: this material is not directly used in this article, but it corroborates the results presented. Furthermore, this new research enables us to attempt to overcome the methodological limitations of our first survey on CAOs, which, since it focused on a relatively short period of time, did not allow us to properly identify the exit routes out of the centres.
115 or SAMU Social is a municipal humanitarian emergency service provided in several cities in France.
The term “Dubliners” (dublinés) refers to the Dublin III Regulation adopted by the EU on 26 June 2013, which establishes that the country responsible for an exile’s asylum claim is the first country in which that person’s identity documents were checked: in other words, persons whose asylum claim has already been considered in another EU member state cannot apply for it in another country. In practice, this gives rise to deportations and returns to the first country.
Interview with a CAO officer from Andouce, February 2018.
Interview with a social housing provider, OPH director, Rossac, February 2018.
Now called Rassemblement National.
Looking at the electoral scores of the last presidential election, at the level of the cities, the votes for the extreme right are in line with the national average or slightly below: 22% and 14.5% in the first round compared to 21.3% in France, then 35% and 25% compared to 33.9% in France. However, when compared to the results of the 2012 presidential elections, a local increase is clearly visible, given that the FN obtained 17% and 13% (17.9% nationally) - see the work of Carvalho (2013). The far-right party did attempt to use the establishment of CAOs as a campaign argument, as evidenced by posters on roundabouts in the departments castigating the arrival of “migrants”.
However, in this same city, a large solidarity network has been organised among inhabitants to develop activities for these same migrants: there are therefore also positive reactions to their arrival, but of which the institutional actors say little or nothing. This heterogeneity of positions corresponds to research showing the diversity of political opinions within a given territory and advocating “the primacy of an explanation of electoral behaviour by social positions in their contexts” rather than by reifying social or spatial categories (Rivière, 2014) and contesting the attribution of the FN vote to certain so-called peripheral territories.
Interview, January 2018.
The duties of the CAO itself only concern the administrative follow-up of cases, as well as other social work duties, such as medical support.
It is a town described by its elected representatives as a “relatively poor rural community”: there are several factories there and it is mainly workers who live in the municipality, with those holding management positions tending to live in the central city of the department, located some thirty kilometres away.
In line with Jounin et al. (2018), we understand ethnicisation as a process of “social classification based on origin”, supposed or real, in a theoretical and epistemological position that is at the same time “dynamic, constructivist and relational”. This approach integrates the processes of racialisation, which correspond to a radicalised mode of ethnicisation, by “rendering differentiation according to origin absolute” (Jounin et al., 2008). The issue at stake is the social relations of race and the naturalisation of difference. Therefore, the notion of culturalisation also seems important to us: it refers to what Guillaumin (1972) analysed in non-biologising forms of neo-racism, but which are still based on the naturalising absolutisation of supposed or real cultural differences - or, to put it another way, the rationale of naturalisation investing cultural differences (Scrinzi, 2008).
Definition given by the Conseil national des politiques de lutte contre la pauvreté et l’exclusion sociale (French national council for policies to combat poverty and social exclusion - CNLE). See the report by ASDO (Agence d’études sociologiques pour les décideurs publics - sociological studies agency for public decision-makers (March 2015) for the DGCS (direction générale de la cohésion sociale - directorate-general for social cohesion), Évaluation de la démarche de participation des personnes accueillies ou accompagnées au sein du CCPA (Evaluation of the participatory approach of people accommodated or supported within the CCPA [Conseil consultatif des personnes accueillies - advisory council of persons receiving accommodation]).
See Cimade’s mapped assessment (January 2017), https://www.lacimade.org/cao-premier-bilan-apres-le-demantelement/, as well as the OFII’s assessment (29 December 2016) which indicates that 54% of those in CAOs are “Dublin cases”.