Despite numerous migrant waves through the decades, the current influx of refugees and asylum seekers into Europe has been framed by very specific narratives. From humanitarian calls for action to warnings of impending collapse, Europe thinks of itself under a crisis, at a political breaking point that justifies extreme discourses and measures.
The Refugee Spaces data project aims to stimulate and demystify the phenomena through examining the evidence rather than speculating on the so-called crisis. Through mapping and analysis of the openly available data provided by institutional and governmental sources, the platform attempts to spatialise the political and security measures designed to contain migration and the mobility of refugees.
We understand that migration and refuge are in a permanent state of flux, so this platform can only represent a snapshot of a specific period, in part constrained by reliability and availability of the data. Since we started this project, migration has played a more influential impact on political issues across Europe and the rest of the world, becoming sometimes the centrepiece of polarising campaigns and radical partisanship.
Brexit, the surprising success of populist agendas in some important national elections across Europe and elsewhere are just a few examples of how migratory issues have been used, and manipulated, for radical change. Security borders and sovereign intromission have expanded to Africa and Asia; the policing of the Mediterranean is now an established security regime; and humanitarian initiatives, to help refugees in peril, have been often criminalised.
In the following maps, the project shows a cartographical analysis of spatial responses and the administrative infrastructure brought by migration and refugees, stressing on the territorial relationships that associate mass movement with urban hotspots in four selected countries: France, Germany, Greece and Italy. Further countries can be added to the platform in the future. At the urban scale, the project identifies urban clusters/regions that are integral to current migration influxes, exploring their different strategies for reception and control.
Refugee Spaces has been funded by the 2016 Bartlett (UCL) Materialisation Grant. The project is a collaboration between the Development Planning Unit (DPU), Space Syntax Laboratory (the Bartlett School of Architecture), and the Centre for Advanced Spatial Analysis (CASA). The information presented on this platform is not intended to be a conclusion, but a departing point to track the spatial and economic impact of migration on European territories. We hope and anticipate that the output of this project could be used as a base for further research and collaborative work on European refugee and migration phenomena in future.“ The report, available to download, contains reflections and preliminary work done in preparation of the platform.
This project was created through a collaboration between the Development Planning Unit (DPU), Bartlett School of Architecture (BSA), and the Centre for Advanced Spatial Analysis (CASA) at University College London. The project was funded through a Bartlett Materialisation Grant.
The project team included Camillo Boano, Giovanna Astolfo, Ricardo Marten, Falli Palaiologou, Keyvan Karimi, and Ed Manley, with Gala Nettelbladt, Tahmineh Hooshyar Emami and Asimina Paraskevopoulou.
The platform is a systematic repository of data on demography and cost made available by governments, international agencies and NGOs. Such data are integral to policy making, and strategic decision in humanitarian intervention. They posit plenty of methodological and ethical challenges around accessibility, consistency and validity. What follows is a sketch of the challenges in establishing a transparent methodological approach and data collection strategy for the project - and eventually, for an open access digital platform.
Data accessibility Easiest data categories to be accessed are national governments or NGOs statistics on asylum requests because they are listed in public records; however, data availability is not consistent across countries. For instance, there is a general lack of comprehensive updated data on the location of shelters, reception and accommodation centers in Europe. France has the most accurate compilation of information regarding location and governance of the centres – around 700 centres mapped within the national borders. In Italy we mapped a bit less than 300 centres, as the information regarding other centers is not collected nor disclosed. In Germany we mapped some 100 centres; despite wide accessibility of data, information available proved difficult to be compiled as migration management is highly decentralised.
Consistency in data collection methods. There is little consistency on the data nature, sampling and collection process in each country. Data on arrivals are collected at border; data on asylum requests and number of refugees are collected at the border and in the centres. Data have different time periods; they are used to interpret patterns in migration, but they can refer to change in border policy. Double counting is possible, as well as undercounting. Besides the inherent fragmentation of data sources and the heterogeneity of the systems of classification, these type of data sets are in constant flux, and such oscillations are not easily captured.
Validity of data visualisations. Data do not have a true fundament of scientificity that ultimately lead to a clear scenario. Researchers and activists have spent considerable effort in showing how miscalculations and double counting often occur, for instance in the case of data on sea arrivals. Monthly arrival figures are calculated based on Frontex data, the EU border agency that detects border crossing at sea (and land). Border crossing can be multiple (as people might attempt to cross a border more than once) resulting in an inflated account of the number of arrivals. An excess of faith in the scientific authority and unquestionability of data might end up in the uncritical reproduction of a biased narrative and the perpetuation of the myth of a “migratory invasion”.
Time Data and analyses presented in the platform cover different periods of time, depending on data availability and sources. In visualisations that examine the phenomenon looking at Europe by using national data (main sources: Eurostat, UNHCR), the datasets extend from 2010 to 2015, except Frontex which goes till 2017. In visualisations that focus on the regional and/or urban level of analysis and which refer to country-specific data for Germany, Greece, France and Italy, the periods covered by the datasets vary and are case-specific. For the reception systems of Germany, France the most complete dataset is collected for year 2015 (for Italy is 2017). Data from 2015 are far more consistent, accurate and complete compared to either 2014 and 2016 – for different reasons. For reception in Greece, the most complete dataset available is for year 2017.
Scale Data were collected at four scales (continental, national, regional and urban) depending on the availability and combining a mixture of wider databases and specialised sources. Most of datasets are aggregated at national scale reflecting an abstract centralised representation. The privileging of the nation as the natural scale of analysis is one of the inbuilt biases of statistics that generates the greatest level possible of simplification. When possible, the project attempts to disaggregate national data at the urban level to better understand the actual impact of migration and refuge.
Centres Given the diversity of administrative devices and technical and humanitarian measures aimed at containing and – most of the times - immobilising refugees and migrants, the platform wishes to comprehensively include all premises where refugees and asylum seekers are accommodated, detained or in transit. Denominations are different in each country and reflect each country’s policy. The centres shown in the platform are permanent ones, open and operative at the time of the research.
Location Not all the data gathered is suitable to be spatialized at the urban level – either due to sensitivity of the information (and related ethical implications) as well as inherent complexity of disaggregating national data at the small scale. Tracking down the location of the centres for reception or detention, can expose refugees to risk of being persecuted. To avoid this, and as a fundamental measure, the exact address of the centre is not disclosed in any maps shown in the platform.
Statistical data tell us how many people have arrived or are hosted in a country; the higher or the lower number over a specific period; they never say why they’ve arrived in first place or where they want to go, what are the individual migration projects, or the level of vulnerability of the single subjects. The use of statistics overlooks singularity and specificity, reducing complexity and fluidity of migration into manageable, comprehensible facts and figures.
By simplifying migrant population through specific indicators, it’s possible to circumvent the need to acquire broader detailed local and qualitative insight. This blindness to local variability is precisely what should make statistics potentially offensive. Turning migration into numerical aggregates and averages dehumanises individuals, and ultimately strips them of their rights. In this sense, data are useful to depoliticise migration and refuge, creating a distorted image and perception of migration.
For this reason, the project considers essential to include descriptive and qualitative documents (see download area) that allow, as far as possible, to add information layer that is often missing in maps and graphs.
Finally, the use of data analytics raises several questions related to the responsible management of personal information, especially of vulnerable groups, such as refugees.
The project is based on a composite multi-source analysis. The information stored in the platform is taken from diverse sources: official reports from Interior Ministries, the European Commission and international bodies including the Office of the United Nations High Commissioner for Refugees (UNHCR), reports by civil society (at the international, regional, national or local level), field investigations (interviews, visits, etc.) and press articles.
Our main sources for statistical datasets are, so far, Eurostat and UNHCR. The use of different sources to analyse a single variable can present difficulties, for the coherence of calculations done based on data registered to develop the maps. For example, in relation to the same variable, the figures published by the Eurostat agency on the number of refusals of entry per year for a given country sometimes differ from those provided by the Interior Ministry of that country or those gathered by an NGO. While it is sometimes difficult to harmonise different sources, we have agreed on a hierarchy of sources for the database.
Regarding the information recorded on the “country” data-sheets (annual figures on arrivals, refugees/asylum seekers in the centres, etc.), the first sources are figures from European bodies (Eurostat, European Commission, etc.), since in principle efforts to harmonise data have already been undertaken by these bodies. Where these are not available or enough, we use the figures provided by national bodies (Interior Ministries, etc.).
The spatialization of refuge in France as portrayed by centres mapped for year 2015, reveals efforts to decentralise the management of reception via greater dispersion of centres in rural areas. Earlier reception centres (centre typology CADA - Centres d'Accueil des Demandeurs d'Asile), which were set up since the first creation of a nation reception system (DNA) in 1973 and onwards, are found to be in their majority located in urban centres (54% of mapped CADA) and in their periphery in functional urban areas (11%). For centres which appeared in 2015, the policy scenario shifts towards the establishment of new orientation centres in rural France. Over half (58%) of the mapped centres for reception and orientation (typology CAO - Centres d'Accueil et d'Orientation), which opened in response to the needs of the relocation process of the Calais camp population, are in rural areas. In the same year, the system promoted the provision of one-stop services for asylum seeker registration/welcome in urban centres (GUDA - Guichets Uniques pour Demandeurs d'Asile). Out of 33 GUDAs mapped, 28 are located in cities, namely approximately 85% of first reception takes place in urban centres. Effectively, since 2015, reception process begins in cities and redirects refugees and asylum seekers in rural France as much as possible. The overall picture in 2015 shows how historical and new reception centres are found to distribute reception almost evenly between urban (44%) and rural (43%) areas.
Île de France, the region of Paris, is evidently the one showing the highest number of mapped centres. It is a historical reception core, with 33 CADAs (the oldest of which is active since 1904 according to action-sociale.org), 8 GUDAs, only 1 CAO and 10 detention centres (out of 30 mapped in total for France). In general, mapped data show a relatively uneven regional distribution with standard deviation from the mean number of centres (which is 34) being approximately 17. Overall, year 2015 shows how new policies in response to high influx create a transformation in the spatialization of migration and refuge.
The map shows centres’ location - urban or rural - for year 2015, based on Eurostat classification of spatial units (see here): ‘cities and greater cities’; ‘functional urban areas’; and remaining centres as rural areas.
The overall capacity of centres operating in the reception system is not fixed and has adapted over the past years. These figures follow the national structure of France’s national reception system (Dispositif National d'Accueil - DNA) for asylum seekers, which operates across two processes:
Le dispositif de premier accueil: A network of information, guidance and support services for asylum seekers which is managed by the French Immigration and Integration Office (OFII) since 2010, or by operators (private or public organisations) liaising with OFII under the supervision of the Ministry of Interior.
Le dispositif d’hébergement: Accommodation arrangements for asylum seekers in reception centres (CADA).
Between 2008 and 2012, the number of asylum seekers in France increased by 73%, putting pressure for a reform process of the French asylum policy. The first meeting of the national consultative committee took place on 15 July 2013 and led to the adoption of Act No. 2015-925 of 29 July 2015. This Act provisioned the strengthening of the national reception system (DNA) network by increasing the significantly accommodation capacities of both regular and emergency reception centres. Since the reform, CADAs have created their own management and quality control tools, with a view to improve their service and track their costs through budget reports. The policy of increasing the accommodation capacity for asylum seekers continued in 2016 at unprecedented pace. As of December 31st, 2016, the cumulative capacity of CADA, AT-SA and HUDA was 54,145 compared to 43,895 on the same date in 2015.
The map shows information on centres’ capacity, where available, for the year 2015.
The variety of centres, programmes and institutions reflects France extensive, and complex, system of reception, which navigates between national strategies and local initiatives:
The definitions of reception, detention and expulsion centres are different according to each country and reflect each country's nomenclature and policy. Given the diversity of administrative devices and technical and humanitarian measures aimed at containing migrants, the research wishes to comprehensively include all premises where refugees and asylum seekers are accommodated, detained or in transit.
So-called reception centres appear to be designed to provide assistance and shelter, although their "residents" - migrants and asylum seekers - obviously have no option other than remaining there. Most of the centres shown in the platform are permanent ones, open and operative at the time of the research.
The map shows reception/detention centres mapped and their typology, for year 2015.
The main source of regional funding comes from the general budget of the State, taxes assigned, local authorities and European funds. The court audit report (February 2015) highlights the difficulties in tracing the precise costs of asylum policy, beyond the direct budget allocations (p.25) and gives a list of limitations and assumptions associated with the attempt to collect data on cost (appendix 4 in the report). Amongst the issues, the audit picks up on the problem of missing data on the cost of the personnel who look after people who have been denied asylum in the prefectures, as well as of security forces involved (p.26). The audit associates problems of data collection with the non-synchronised archiving mechanisms between the prefectures who file actions on applications in Telemofpra (the application system which allows a connection to the central database of OFPRA) and the slow central-government processing of those applications (pp.17-8). Telemofpra links to updates on applications’ status, such as approvals and rejections and subsequently, to the management and distribution of the temporary waiting allowance (ATA).
The visualisation shows total regional budget contributions (in € euros) for first reception (PFA: plate-forme d'accueil) by the French Office of Immigration and Integration (OFII) and the European Asylum Migration Integration Fund (FAMI: Fonds Asile Migration Intégration), based on figures identified by the court audit report published in February 2015.
The figures identify the following 4 regions with highest budget allocation:
The efforts to mobilize and establish an organised system for the reception of migrants in France began in response to the influx of refugees due to the coup d'état in Chile in September 1973. The national reception system (Dispositif National d'Accueil - DNA) for asylum seekers operates with two processes:
Complementary to this process, is the provision for emergency accommodation for asylum seekers (les dispositifs d’urgence) in emergency centres (AT-SA and HUDA).
During their reception, asylum seekers mainly interact with NGOs within a reception platform (plateforme d’accueil), which is usually co-managed by the regional prefect and the NGO on site. There are three dominant NGos who operate in the accommodation sector for asylum seekers: Coallia, France Terre d'asile, and Forum Réfugiés-Cosi. In 2015, these represented 7,206 places in the CADAs, namely, approximately 39% of the overall places which are managed by NGOs. (Cour des comptes, 2015, pp.15-60).
The financing of the French Office for the Protection of Refugees and Stateless Persons (Office Français de Protection des Réfugiés et Apatrides - OFPRA) is mainly provided by a subsidy for public service charges paid by the Ministry of the Interior under Program 303 ‘Immigration and asylum’. Other resources come mainly from European co-financing under the European Refugee Fund (ERF) & the Asylum, Migration and Integration Fund (AMIF). (OFPRA, Rapport d’Activité 2014, p.70).
Translating from the Court of Audit: “There is quasi-structural under-budgeting for Program 303 - Immigration and Asylum. The Court of Auditors has had the opportunity to note in its budget execution analysis notes that the forecast of the evolution of asylum applications was unrealistic, so that the budget for the 303 program did not appear not sincere. The under-budgeting of the 303 program mainly concerns the emergency housing of asylum seekers and the ATA.” (Cour des comptes, 2013, p.22).
All the data gathered by this project are available to download. Please select a dataset from the dropdown below to see its description.