Undoing the Migration Digital Black Box

Guest post by Martina Tazzioli and Lucrezia Canzutti.  Martina is Reader in Politics & Technology at Goldsmiths, University of London. Lucrezia is a Research Associate on the ERC-funded project SECURITY FLOWS at King’s College London.

In the last few years, a “data anxiety” has driven academic and non-academic debates about migration and digital technologies used at the border. Scholars, NGOs and journalists have paid increasing attention to how states produce and share information about migration, and which databases they use to store data about migrants’ crossings, arrivals and presence in their territory. Within this context, the metaphor of the black box has gained traction, pointing to the perceived ‘unknowability’ and opacity of the technologies used at the border, the way they operate, and the data they store. The black box conveys the idea of data which are kept secret and away from citizens’ sight. Consequently, calls for opening the migration digital black box underpin many works on digital borders; ‘opening the migration digital black box’ connects efforts to shed light on the internal workings of border technologies and claims for greater transparency and for unveiling private actors and states’ secrets.

The pitfall of the black box metaphor is twofold. Theoretically, the black box implies that there is something like a secret archive kept hidden from citizens’ view and presupposes a certain degree of coherence and consistency in states’ practices. Yet, we argue, this is often not the case: unevenness, technical glitches and states’ refusal to share data characterise police operations and border controls at the EU’s frontiers. Indeed, the metaphor of the black box restricts our ability to understand the relation between an operation’s internal workings and its outside environment, its conditions and effects. Methodologically, opening the black box involves unveiling what is kept secret and concealed about migration databases, making visible what is kept invisible. However, most of the times this turns into an impossible endeavour: restricted access and difficulties in interviewing state authorities make it impossible to trace how they operate at the border. Moreover, efforts to analyse the inner mechanisms of digital technologies risk ‘simply open[ing] more black boxes’ while displacing attention from bordering practices and the impact of these technologies on people on the move.

In this post, we shift away from the alleged secrecy of the database: rather than attempting to open the migration digital black box, we sidestep it and shed light around it. We do this by developing a methodology that looks at databases in light of the paper-based documents that migrants are given after identification procedures and of the different border practices that are enforced along migrants’ routes. That is, we study the use of databases by drawing attention to the imbrication of border practices, digital data and non-digital documents that surround it. In fact, databases and digital infrastructures cannot be abstracted from the materiality of practices and papers that form the “EU border work”. Importantly, our methodology also allows us to situate digital infrastructures of migration control within a contested politics of mobility formed by migrants’ struggles, technological glitches and political frictions among states.

Our research draws on data we collected at the Italian-French and Italian-Slovenian borders, which are two internal EU borders crossed by migrants coming from the so-called Balkan and Mediterranean Routes who are trying to reach France, Germany and the UK. However, the paper-based documents we analyse are from different locations across Italy, underlining the importance of taking a multi-sited approach. At the core of our investigation are questions raised by lawyers, NGOs, activists, and migrants in relation to authorities’ identification practices and the use of national and European databases. For example, at the French-Italian border, Italian police practices in response to France’s systematic pushbacks since 2015 (when France suspended Schengen) have changed over time: while in the past they did not fingerprint pushed-back migrants or only took two fingerprints, currently they take ten fingerprints. What does this change reveal about the data extracted from migrants? And what are the consequences for migrants themselves? NGOs and people in transit were especially concerned that ten fingerprints could be index of data being sent to EURODAC and, therefore, of migrants being subjected to the Dublin Regulation. Similar concerns were shared by both activists and migrants in transit at the Italian-Slovenian border.

Thus, we started inquiring about whether the fingerprinting process at the French-Italian and Italian-Slovenian border meant that Italy now systematically sends data to EURODAC at EU internal borders, and not only at external ones. Indeed, until recently Italian authorities uploaded into EURODAC biometric data of migrants who entered Italy by sea (external frontiers) or who claimed asylum in the country, but not those of migrants who were subjected to push-back operations. One way to answer this question would have consisted in opening the black box, that is finding a way to disclose police practices at different borders as well as the uses and technical details of their databases. Nevertheless, the unfeasibility of this research strategy quickly emerged as we were denied interviews with Italian police in different sites. Even in case of a positive answer, the possible gap between police’s declarations and actual practices would have remained. Thus, we opted for a different approach, based on the imbrication of digital and non-digital data about migration: with the help of Italian lawyers, we cross-checked the papers that migrants received after being pushed back from France and after being identified at the Italian-Slovenian border. Then, we analysed in which sites and in which occasions (push-back, “illegal entry” etc) migrants were given specific papers. Thus, we combined a multi-sited approach with a focus on of migrants’ journeys (asking them which procedures they went through in different places), and with an analysis of the different papers given by the police after the identification procedure.

By connecting the fingerprinting procedure with the papers given to migrants, we were able to piece together the circumstances in which biometric data was sent to EURODAC or kept exclusively in the Italian national database (AFIS). For example, at the Italian-Slovenian border, we came across a paper inviting people to ‘go to the Police Headquarters in order to formalise their asylum claim’. This was given to migrants who had been subjected to the complete fingerprinting procedure (10 fingerprints), had stated their intention to claim asylum in Italy, and had therefore been registered in AFIS and EURODAC as ‘asylum seekers’ (EURODAC CAT 1). By contrast, after having their fingerprints taken, pushed-back migrants at the French-Italian border were given a paper inviting them to “go to the Police Headquarters in Turin to define [their] status”. This paper, as confirmed by immigration lawyers, had no consequence for migrants and it differs from the ones given to them in hotspots, where their data is systematically sent to EURODAC. Indeed, when migrants are registered in EURODAC as CAT 2 (“irregular entry”), they are given an expulsion order instructing them to leave the national territory within 7/14 days. In this case, migrants are also registered in the SIS (Schengen Information System) database.

A similar document to the invitation letter mentioned above was also given to people who were identified and did not state their intention to claim asylum at the Italian-Slovenian border. Among the latter group, some were given expulsion orders, whereas others were invited to go to ‘the Police Headquarters in Trieste to determine their case’. We were not able to establish the exact criteria according to which one form was issued over the other, but anecdotal evidence suggests that it might be linked to people’s stated desire to reach other European countries. Regardless of this, in both cases migrants’ data could not have been sent to EURODAC CAT 2 (“irregular entry”), as Slovenia is an internal EU border and CAT 2 concerns external borders only. It is also worth noting that Italian authorities have no interest in registering migrants in EURODAC since, in accordance with the Dublin regulation, this would establish Italy’s responsibility for examining migrants’ asylum claims as a country of first entry. Further evidencing this is Italian authorities’ frequent use of the SPAID system (Peripheral Fingerprint Acquisition System), which allows them to check people’s fingerprint record through a portable device using only two fingerprints, but does not enable a full identification procedure (fotosegnalamento). The SPAID system has routinely been used to check people’s identity against different databases without having to register their data; at the time of writing, the system was suspended due to a combination of software issues and changes in guidelines and legislation, which explain why authorities currently take 10 fingerprints instead of 2. 

In the picture below, we summarise the different identification and fingerprinting practices employed in Italy and their potential implications for asylum seekers.

This picture, of course, is a snapshot of the current situation in Italy. However, bordering practices are continuously (and often abruptly) changing and evolving, giving rise to inconsistencies that can help strengthen the perceived inaccessibility of the migration digital black box. Thus, to push against the black box narrative it is not sufficient to take a single ‘photograph’ of different practices at a given moment in time. To be meaningful, research must follow police and identification practices across time and space, keeping track as far as possible of changes and their possible meanings and implications, and trying to bring together the knowledge of different actors working in support of and in solidarity with migrants.

Conclusion

The metaphor of the migration digital black box distracts from the actual bordering practices, as it evokes linear and consistent operations and de facto corroborates the image of a seamless border regime, in which the data shared is kept hidden from citizens’ view. The image of the black box corroborates ways of seeing migration “like a state” by legitimising state’s own narratives about the functionality of the technologies deployed at the border. Instead, border controls at the EU frontiers are characterised by uneven changes in practices, as well as by disruptions and refusals in sharing data. Most importantly, by putting databases at the core and by narrowing down the analysis to that, we end up extrapolating these latter from the silent or more blatant political conflicts among states, as well as from the imbrication of digital and non-digital border technologies.

We have proposed to retrace the use of databases in light of the papers that migrants are given and of the places where identification procedures happen. In so doing, we have sidestepped the black box and shifted outside of it. This requires paying attention to migrants’ journeys and to their reconstructions of what they experienced at different borders. In order to understand how migration databases function, we cannot focus exclusively on the digital: a scrutiny of paper-based documents and an understanding of migrants’ trajectories are essential for excavating digital infrastructures at the border.

Undoing the migration digital black box is key for rethinking political and legal interventions at the border. Indeed, the digitalisation of migration governance has pushed lawyers, activists and NGOs to magnify the difficulty of understanding how borders work: the image of the black box, we suggest, haunts political imagination. Against this background, it is essential to re-inscribe databases and digital technologies into the “contested politics of mobility”, formed by uneven police practices, technical glitches and failures, and migrants’ struggles for movement.

Any comments about this post? Get in touch with us! Send us an email, or post a comment here or on Facebook. You can also tweet us.