Decentering Return Migration Research: Are we on the same page?
Struggling with numbers
by: Joris Schapendonk | Radboud University
The issue of terminology and definitions, and more particularly, the issue of translating terminologies and definitions across different disciplines and geopolitical settings appears to be a constant challenge in large interdisciplinary research projects that cover broad geographical areas. Our GAPs project that seeks to decentre the study of migrant returns and return migration policies is no exception. An ambitious project for its geographical scope, interdisciplinary character and methodological design, the project involves 17 partners, across 12 countries, and includes scholars (with multiple backgrounds) as well as non-academics.
A lawyer and social scientist might have very different answers to questions such as: Where is, in the end, the fine line between coerced and voluntary returns? Such questions are at times broad, and at times specific, but in any case they do form a constant challenge for the research teams. There is a need to build in mechanisms to explore whether we are still on the same page in terms of languages, conceptualizations, operationalizations and expectations.
In addition to the question of research preparation and operationalization, we should critically engage with the question of whether data are ‘on the same page’ (i.e. do they measure and communicate the same item across political settings and across data sets, and for what reason?). This issue of data is particularly important because datafication is an inherent part of the politicization of irregular migration, hence it is part of the very centered storyline of migration governance. Critical engagements with migrant-related labels have long been mainly a matter of study within critical political science or qualitative research on the changeability of statuses. However, this has changed with the intervention of Savatic and colleagues who point to the following ironic situation: States design migration and asylum policies based on the distinction between categories of travelers, yet they themselves produce data that allow for the triage of, for example, refugees and irregular migrants. This became particularly clear with the 2015 migration management crisis in Europe. While Frontex – and with them mainstream media in the EU – were counting so-called irregular migrants crossing Europe’s borders, the analysis of Savatic and colleagues shows that an estimated 75% of all the people who entered are likely to have received asylum status soon after their arrival, so they are wrongly framed as irregular in the very first place.
Similar questions emerge on data on return migration and deportation. The question of which data are presented is about epistemological politics. Whereas seemingly transparent data sets are openly available on asylum procedures, return procedures (including coerced returns), NGO reports indicate that there are many unseen and uncounted, and often illegal and violent, border returns in the world. This includes pushbacks (as highlighted in Pushbacks Evidence – Refugee Rights Europe), but also the sudden deportation of migrant communities between Morocco and Algeria or from Tunisia as well as the forced return of Syrians from Turkey and Lebanon.
Moreover, there is the issue of articulation of only some returns and return orders. If I were to ask my students to construct a top 5 of ‘Third Country Nationals’ with the highest number of return orders, I would not be surprised if they would go to the Dienst Terugkeer & Vertrek website (DT&V = Dutch Return and Repatriation services) and reproduce the list with the ‘usual suspects’ of Algeria, Morocco, Nigeria, Albania, Iraq (appearing in different orders in the last five years on the DT&V website as the highest numbers as nationalities that are removed/returned from Dutch territory). However – and this is rather remarkable – the citizenship group receiving the highest number of return orders from the Netherlands does not appear on this list. This group is formed of US citizens. In 2022, for instance, 1,065 US citizens received a return order, compared to 165 Iraqi citizens and 900 Moroccan citizens. This information does not appear in any report on return migration in the Netherlands. Some might say that this is because the files of US citizens do not reach the DT&V, and they do not count as ‘assisted’ or ‘coerced return’, or because US citizens can easily prolong their visa. That might all be the case, but leaving this particular number out of statistical overviews and reports does say a lot about why these statistics are produced in the first place. In the end, the decision about what data to hide and what data to reveal and reproduce in reports and academic research on return migration is a question of knowledge and research politics.
Next to the question of visibility/invisibility, there is the issue of consistency. When there is a strong consistency of data – when data is constantly ‘on the same page’ – this should not automatically be translated to the issue of accuracy. Instead of triangulated accuracy checks, coherence might emerge because the same data travels between different data sets (as copy/paste). Similarly, if data appears to be inconsistent over data-sets, this should also raise concerns. For the Dutch case, it is striking how data from Eurostat on irregular migration, return migration and asylum are difficult to match with national data. Inconsistencies might be explained by deviating definitions. I experienced this myself through my involvement in the formation of a GAPs data repository. To enter data about return decisions issued for irregular migrants into the database, I looked at the DT&V figures that I thought might be most accurate (see also report of the Dutch case written by Tineke Strik and Sherry Ebrahim). These data are described as including “the inflow of files” of migrants that are to be returned. Files are handed over by the immigration authorities (IND), national police, the marechaussee, and these files may also come from 'other cases' (excluding asylum-related). In some cases, the latter might relate to migrants not living in irregularity, as the data description discloses. I am still wondering whether all the other inflow of files relate to people living in irregular situations. Furthermore, we also know that irregularity itself has different dimensions (irregular entry, stay, irregular work, etc) that are often mixed in public and academic debates. Next to deviating interpretations, there are also data with rather straightforward definitions that still show discrepancies. For instance, when the database seeks for #TCNs/foreign nationals refused entry at the border, Eurostat counted 3,070 cases, and the Dutch authorities counted 4,720 cases for the Netherlands in the year 2022 (similar discrepancies emerge in other years).
It might be me, but if the datafication question is so entangled with the centered narrative of migration governance, I wonder whether we should not look for other data. For instance, how many hours do border guards spend on the border without refusing any person to enter? Or, how many hours does a qualitative researcher need to really be able to make sense of quantitative data on return migration? With regard to the latter, I am still counting. I just opened another detailed data dictionary (that of DT&V). Rather than getting ‘on the same page’, I risk getting lost in many pages soon.
Contact:
Joris Schapendonk | Radboud University, Nijmegen | joris.schapendonk@ru.nl