Contextualising Frontex: A Long-Term Perspective on Database Monitoring of Migrants
Population monitoring through data collection has increased and become part of the everyday life in Western liberal states since the 1980s and 1990s. Whether we make a phone call, a bank transfer, log in to work, or visit the hospital, a variety of state and non-state actors gather and process our information to enable and constrain access to various goods. In what has been convincingly described as a ‘surveillance society’, migrants, especially non-EU citizens, are under particularly close scrutiny in Europe. In this light, the most recent reform that again expanded mandate of Frontex in the area of data analysis and exchange is perhaps not surprising. However, for the agency to carry out these new tasks of monitoring migratory flows and performing risk analyses, Frontex requires extensive, reliable data supply. This, in turn, revives the question of the role of databases in the monitoring of migration.
Databases as a key population control device
Databases enable the storage, search, and retrieval of information. Advanced electronic databases are believed to be highly effective, both by critics and proponents, in facilitating population control on a large scale, enabling institutions to identify and constitute individuals, especially those attempting to evade the state’s eye such as undocumented migrants. Indeed, during the 1990s and 2000s, the European Union established three major databases (namely Eurodac, SIS II, and VIS) that should provide information on all asylum and visa applications filed, as well as irregular entrants and residents apprehended, in Member States. Frontex has now access to this large pool of data as its revised mandate allows Frontex members to run searches across all three EU-wide databases.
And yet the effectiveness of monitoring techniques such as databases is far from evident. Data protectionists warn of unchecked power and the potentially insatiable greed for data that puts migrants (and their sponsors and spouses) under general suspicion. On the one hand, critical voices from migration and surveillance studies fear that electronic databases represent a ‘superpanopticon’ maximising state control. On the other hand, anthropologists and historians have pointed out that state monitoring schemes largely failed. How effective is database surveillance by governments? We know little about the implementation of the database approach and about national legacies of state monitoring preceding Eurodac, SIS II, and VIS.
In this polarised debate between critics from various sides it is worth taking a step back and looking at the emergence of data collection practices of migrants in Europe. This blogpost contextualises Frontex by discussing a pioneer database in Europe, the German Central Foreigners Register that presumably served as a model when establishing the European databases, and by drawing some lessons from the German case for the European context regarding the effectiveness of database surveillance.
The lure of surveillance through database: build-up and design of the German Central Foreigners Register
The Central Foreigners Register (Ausländerzentralregister, AZR) is a comprehensive database established in 1953 and operated by the German federal interior administration (Bundesverwaltungsamt) in Cologne. It stores detailed personal information of all migrants who apply for a visa or a residence or work permit in Germany, including any previous application results, warnings issued, or search warrants. This central reference index is consulted prior to any permit being granted, extended, or rejected. The volume, usage, and accessibility of the AZR were only comprehensively regulated in 1994—that is after it had existed for almost four decades. Today, it contains approximately 26 million records, accessible to over 14.000 authorities and agencies. Therefore, the AZR is considered by the German authorities as the ‘central pillar’ of Germany’s migration control system.
Germany pioneered the database approach to migration control in post-1945 Europe. Following the end of the Second World War, Germany was a hub of people on the move. While migration through post-war Germany was highly regulated with a multiple permit system, the implementation of this system was decentral and discretionary. The Allied Forces were concerned about the inconsistent implementation of the registration system by local Foreigners Authorities and requested a central register to be set up in order to standardise the issuing of entry, work, and residence permits. Due to international pressure, the build-up of the Central Foreigners Register began in 1953 in West Germany.
The AZR was designed to provide a constant, reliable overview of all migrants entering, residing in, or leaving the Republic. As previously mentioned, the AZR should be consulted by local Foreigners Authorities prior to any permit being granted, extended, or rejected. The centralised database also had a blacklist function allowing differentiated searches for undesired migrants, such as Communists in the global context of the Cold War. Finally, the database was believed to be the only reliable means of obtaining a comprehensive statistical overview of migrants in Germany. In 1967, the operation of the AZR was computerised and changed over from a paper-based index card system to electronic data processing with the aim to speed up the register activity and make it more efficient. However, the archival records in the Federal Archive Koblenz suggest a series of unintended and counterproductive effects.
Migrant monitoring in the 1970s and today: the striking similarity of pitfalls and malfunctions
The AZR struggled to provide the perfect gaze on migrants it was intended to deliver for three reasons. First, the creation of the AZR suffered from a lack of political willingness and from diverging interests. As an early example of a hybrid database, the AZR combined the international expectation of migration monitoring juxtaposed with existing state and local foreigner registers originating in 19th century police states. The request in 1953 that all Länder should submit their registers (or a copy thereof) to the German Federal Interior Ministry provoked a power struggle between the Federal Interior Ministry and the Länder Interior Ministries. Not every Land was willing to hand over their index cards, some considered this a loss of competences and significantly delayed their submission. Likewise in the EU today, there are diverging interests and power struggles impeding the effectiveness of EU databases. Consider for example, that EU Member States at the southern Schengen border may have little interest to take fingerprints of asylum seekers as long as the Dublin Regulation requires asylum applications to be made in the Member State of first arrival without there being a system of redistributing the share of asylum seekers across the EU.
Second, the establishment of the AZR was complicated by limited state capacity. The maintenance of the initially paper-based index card system represented a major challenge even in a highly bureaucratised state such as Germany. The early years of the AZR were characterised by a repeated shortage of storage space, index cards, and staff to manually attend and search the register. The change to electronic data processing did not speed up the AZR but produced new problems of data supply. For example, during the 1970s the local Foreigners Authorities of Munich completely stopped sending update cards to the AZR as they were preoccupied with the computerisation of their own records. Officials were highly concerned that this compromised the reliability of the AZR, especially since Munich was considered the main gateway to Germany from the South, leaving the register out-of-date and prone to providing false information. In 2015, Munich was again a central gateway for refugees arriving on the Balkan route and the German authorities were completely overwhelmed and incapable to register all new arrivals. Despite comprehensive computerisation and international connectivity of the European databases, data supply and data reliability continue to be the weak points of the system nowadays.
Third, a major pitfall of the AZR was that it effectively produced the fallacy of the perfect gaze and nurtured the aspiration of comprehensive surveillance. When the database was still in the establishment process, it already received a large volume of unexpected data requests by a variety of state and non-state actors. This included international state governments who sought information on emigrants suspected to be ‘Communists’, private companies looking for details on runaway migrant employees, as well as religious associations planning pastoral care for migrant faith communities. These additional information requests occupied the register staff and slowed down other essential activities such as answering the steadily growing backlog of requests by local Foreigners Authorities to clear individual permit applications. The fallacy of the perfect gaze still prevails today, when politicians call for hightened registration following terrorist attacks or when the European Commission advocates for ‘stronger and smarter information systems’ to ensure a high level or internal security. But knowing who is in the country does not prevent criminal acts; a tighter registration system neither provides water-tight surveillance nor can it guarantee security.
Current EU developments: towards a ‘smart’ and ‘interoperable’ network of databases?
Arguably fuelled by the aspiration of the perfect gaze on migrants, Germany has pushed for the introduction and expansion of the database approach at the EU level since the 1990s. The three major EU databases (Eurodac, SIS, and VIS) are now located in a bullet-proof building in Strasbourg and managed by a special agency called eu-LISA (European Agency for the operational management of large-scale IT systems in the area of freedom, security and justice). A ‘European Search Portal’ (ESP) is currently being set up to make these separate databases ‘interoperable’, that is, allowing enhanced search functions across all three of them to a range of actors, including Frontex. The ESP also has a ‘smart’ sorting function: It is designed to facilitate the free movement of EU citizens and frequent travellers across Europe while identifying those considered a security threat.
Today, technology is more advanced than in the 1950s and 60s, facilitating ever bigger and faster databases. For example, compared to the AZR’s approximately 26 million records, VIS alone stored almost 65 million visa applications, some 64 million facial images, and roughly 57 million fingerprint sets at the end of 2018. The establishment of databases at the EU level certainly enhanced migration monitoring insofar as it centralised information and made it easier to access and use on a larger scale in terms of data volume and users, including supranational actors such as Frontex.
However, it would be misleading to think that databases allow an impeccable overview of migrants. Technology may be more advanced today but the effectiveness of databases still depends on the same basic issues as in the 1950s to 70s: state capacity and political interests. Reportedly, the AZR count of the overall migrant population resident in Germany is still inaccurate as the register contains duplicate and out-of-date files. Moreover, data supply requires the cooperation of data suppliers whose interests often diverge.
Concluding reflections on the symbolic character of databases for migration management
In light of the problems highlighted here, the primary function of migration databases seems to be a symbolic one, that is (re)producing the image, belief, and authority of the modern state as capable of exercising migration control, both to an internal and external audience. From a political sociological perspective, the database is highly effective in producing a specific self-understanding of the state as a rationalised, sovereign power. The archival records show that public officials were highly anxious that the AZR should be working. In this sense, the database performs a ‘ritual of verification’ as found in accounting and auditing practices. The AZR rarely functioned according to the official blueprint but it provided a ‘framing’ and a certain ‘style’ of migration control from which (West) Germany emerged as a legitimate efficient state power among the international league of Western capitalist states in the post-war era.
From a current global perspective, the popularity of the database approach is not ceasing but expanding, see for example China’s social credit system. Would database monitoring be more effective if it was based on algorythms and artificial intelligence, as practiced with ‘predictive policing software’ in the United States, United Kingdom, South Africa, Brazil, Italy, Switzerland and the Netherlands? This would require further research. This blogpost suggests that contemporary surveillance, security, and migration scholars and policymakers could learn a lesson from the past. The German case shows that the database approach to migration control nurtures external expectations and self-expectations that the state should be able to monitor migrants. However, it remains highly questionable whether the state’s ambition to grasp increasingly mobile populations can ever be achieved through state registers and databases as long as there is human (and technological) agency.
**Acknowledgements: Research for this blogpost was conducted as part of the ESRC-funded project ‘Seeing Illegal Immigrants: State Monitoring and Political Rationality‘. For research assistance I would like to thank Lukas Märtin. For very useful comments on earlier drafts of this post I would like to thank Carolyn Moser.