The moral dilemmas, legal contradictions, and social problems resulting from the interplay between migration and migration control are felt and managed by a wide range of actors across various administrative levels and institutional contexts. I have chosen a flexible approach that allows my analysis to shift between these different levels while keeping the overall research design as simple and ‘elegant’ as possible (Hakim, 2000). It is a combination of in-depth case study and systematic comparison. This allows me to not only contrast the formal entitlements and effective access of irregular migrants to various public services provided in different places, but also highlight how certain contextual features of each field site shape the roles that public institutions and individual street-level bureaucrats play when local service provision overlaps with immigration control. By looking at how individual actors experience, interpret, implement or contest the underlying policies I gain a better understanding of what this overlap means in practice. This chapter gives a brief overview of the units, levels and dimensions of my analysis, the variety of data sources and research methods I have used, as well as some of the ethical concerns and methodological difficulties I thereby encountered.

3.1 The Study: Research Design and Case Selection

I compare three different spheres, or sectors, of public service provision (healthcare, education and social assistance) across two locations (London and Barcelona) that are each embedded in specific regional (England and Catalonia) and national (the UK and Spain) contexts. Accordingly, the investigation covers a total of six cases, as represented in Fig. 3.1, which also provides an overview of the empirical chapters to follow:

Fig. 3.1
figure 1

Case overview and comparative design

Full size image

I have selected Britain and Spain primarily because they represent very dissimilar national contexts for the integration of irregular migrants and their access to public services. A secondary reason was my familiarity with the broader policy frameworks and fluency in the languages of both countries. They differ significantly in terms of their geographical location, constitutional structure of the state, the size of their informal economies, as well as their immigration histories and current legal-political contexts for the broader management of migration. Most importantly, as I discuss in Chap. 4, Britain and Spain follow very different approaches to reduce the number of people residing unlawfully within their territories: the former based on explicitly hostile policies, the latter relatively accommodating towards irregular residents and many of their claims. The respective policy frameworks reflect different problem perceptions and significantly determine the institutionalisation of migration control at the local level, creating distinct conditions for the micro-management of irregular migration.

While the two selected countries thus represent rather dissimilar contexts for the provision of public services to irregular residents, the selected cities have several things in common. Both London and Barcelona are located within parts of the country that are economically dominant and grow faster (demographically and economically) than the national average. They are home to some of the largest and oldest immigrant communities present in each country, and (unlike other cities and regions) continue to experience a further expansion and particularly rapid diversification of their foreign populations. They are also estimated to harbour particularly high concentrations of irregular foreign workers and residents (Pajares et al., 2004; Gordon et al., 2009). Within both cities my fieldwork concentrated on, but was not strictly limited to, districts where the effects of past and current immigration are (perceived as) particularly strong, namely the London Boroughs of Hackney and Lewisham; and the districts of Ciutat Vella and Sant Martí in Barcelona. I deliberately chose Barcelona instead of Madrid, which may seem the more obvious comparator to London. Barcelona is Spain’s second-largest city and in comparison to other European municipalities particularly outspoken about the need to provide services to its irregular migrant population (Spencer, 2020, p. 199). It is also the capital of Catalonia, which has for many years been one of the country’s most autonomous regions and strongest opponents to central government power and policies.Footnote 1 Looking at Barcelona thus more explicitly highlights the significant influence of not only local but also regional administrations when it comes to providing public services to irregular migrants (Price & Spencer, 2014; Spencer, 2018). At the same time, it helps me to demonstrate that my research design and analytical framework can be operationalised across multiple levels of government and allow comparisons to be drawn between cases embedded in rather distinctive administrative structures.

The second comparative dimension is between different public services: healthcare, education and social assistance. Each of them is provided to the local population through a specific set of institutions which operate within a certain legal framework and fulfil a particular function for society. Many of these institutions are becoming more and more involved in processes of everyday bordering and more and more of their employees help implementing immigration policy. Depending on the sector, however, the logic of immigration control thereby intersects with different combinations of institutional logics, functional imperatives and professional norms. What is important for the structure of my analysis is that the various ‘street-level bureaucrats’ who are involved in these ongoing negotiations occupy similar kinds of organisational roles within their respective sectors and institutions. While some roles (like that of a university lecturer or doctor) are typical for one particular sector or organisation, others are more generic and can be found in many different institutional settings (like receptionists or accountants). In order to facilitate systematic comparative analysis across the three sectors, I differentiate between three broader role-categories: administrators, professionals, and what I call the managers of irregularity. The first category includes workers who mostly carry out clerical duties, like receptionists, accountants and other administrative staff; whereas the second one comprises professional roles and responsibilities, such as those performed by doctors and nurses, teachers or social workers. The third category includes those actors who more specifically manage migrant irregularity within the respective organisation. As I will show in the empirical chapters, the various actors within each of these role-categories tend to perform similar gatekeeping functions, have equivalent degrees of power and discretion, and are comparable in terms of their proximity to immigration control or enforcement.

This approach allows a nuanced understanding of the relationships between certain structural or contextual features that characterise each case (including laws and regulations, the institutional setup, informal guidelines, or political rhetoric) and the perceptions, decisions and behaviours of individual actors, which in turn significantly determine policy outcomes. At the institutional level, comparisons can be drawn between actors occupying different organisational roles within the same sectors or institutions, as well as between the same organisational roles in different sectors and institutions. Looking at these issues in two different national and local settings thereby helps to distinguish the contextual particularities of each sector from those that are specific to each environment.

3.2 The Methodology: Data Sources, Data Collection and Data Analysis

Irregular migration and residence are intrinsically characterised by a lack of official data and reliable statistical evidence (Singer, 1999; Black, 2003; Düvell et al., 2010). My study therefore mostly relies on original, qualitative field data that I collected in London (from July 2014 until the end of February 2015) and Barcelona (between the beginning of March and mid-October 2015). The core of this data comes from 86 semi-structured interviews with a range of local actors more or less directly involved in negotiating the provision of public services to irregular residents and thereby also the implementation and effectiveness of internal control measures. They include representatives of civil society, members of the local administration, street-level bureaucrats with various professional backgrounds, as well as irregular migrants themselves.

In London I conducted 33 interviews with a total of 35 ‘non-migrant’ respondents.Footnote 2 Half of them were representatives of relevant civil society organisations, including charities, advocacy groups, migrant associations and church communities. The other half were professionals and administrative staff working in the fields of healthcare (5), education (6) and social assistance (4), as well as representatives of the Greater London Authority (GLA) and different local Councils (3). I also interviewed twelve migrants in irregular situations – six males, six females; aged between 20 and 50 – who had migrated from various non-EU countries and been living in London for between 1.5 and 20 years, experiencing different kinds and degrees of irregularity. At the time of the interview none of them had a formal right to reside in the UK but some had previously held a residence permit while others were awaiting the outcome of outstanding applications. In Barcelona, I conducted 32 interviews with a total of 35 ‘non-migrant’ informants and nine interviews with migrants in irregular situations. The latter included six males and four females, aged between 19 and 42, who had arrived between 1987 and 2013. Of the ‘non-migrant’ respondents, eleven were civil society representatives (mostly NGO workers); five were members of the local and one of the regional administration, the rest were professionals and administrative staff working in public healthcare (7), education (6) and social assistance (5).

In both cities, interviews typically lasted around 45 min (all of them between 30 min and 2 h) and those with equivalent respondents followed the same rough structure. Also my strategy for selecting and approaching potential respondents was largely the same: First I contacted relevant NGOs, advocacy groups and other civil society organisations (mostly by email) and usually managed to arrange a formal interview within a couple of days, sometimes weeks. These interviews mainly served to better understand the context and to identify the major difficulties and barriers in relation to irregular migrants’ access to various public services. Questions were adapted on a case-by-case basis to fit the particular area of work or field of expertise of each organisation. Since this access strategy proved rather ineffective in the case of institutions like hospitals, health centres, schools (in particular), social service departments or other Council offices, I had to approach these in person and often rely on personal referrals from civil society and other intermediary actors. It often took several weeks and a lot of persistence to identify suitable respondents and obtain their participation.

The resulting interviews with street-level bureaucrats focussed on three aspects: (i) their perception of the meaning and relevance of (irregular) immigration status and the treatment and rights of irregular migrants within their respective institutional domain; (ii) the extent to which they felt implicated in implementing specific policies of internal control or immigration legislation more generally; and (iii) potential contradictions or convergence of the primary function of their organisation or their own role and responsibilities with immigration-related duties, rules and regulations. Meetings and interviews with local politicians and representatives of the local administration were arranged by email and mainly examined the role of the city or district authority with regards to public service provision in general and in relation to irregular residents; the localised effects of immigration policy, and particularly their concerns about internal control measures.

Contact with irregular migrants themselves was mostly made or at least facilitated via a number of dedicated NGOs and migrant associations, but also with the help of friends and other personal contacts. These interviews mainly focused on when, where, and under which conditions they perceived their ascribed irregularity as a particular barrier or risk factor; and more specifically, how it affected their position vis-à-vis the local authority and public welfare institutions. I thereby generally focused on their (irregular) residence in the city, rather than why they had decided to migrate and how they entered the country. The non-representative samples of (irregular) migrant respondents only comprised people who were not in detention nor facing immediate deportation at the time of the interview, and who had been living within the respective city for more than a year (typically between two and five years, but in some cases for decades). While in some cases their presence in the country was officially known to the immigration authority,Footnote 3 others had never been in contact with any state authority and were thus living ‘under the radar’ of the immigration regime.

With some of my respondents I have maintained contact (via social media, email or telephone), which allowed me to conduct several follow-up interviews and additional informal conversations during subsequent visits to both cities. This helped me to complete my understanding of their accounts and increase the overall consistency of the data and information obtained. Throughout the interview process I tried to triangulate the data obtained from the different sets of respondents, in order to make sense of and corroborate their various accounts (see Ellermann, 2006). This exposed several (if minor) inconsistencies between some of the accounts of welfare bureaucrats and professionals and the migrants’ own experiences; and between representatives of the local authority and those of third sector organisations.

Wherever possible, and only given the interviewee’s explicit consent, interviews were voice-recorded and fully transcribed.Footnote 4 All but two interviews in Barcelona were conducted, transcribed and coded in Spanish, and only those segments I quoted subsequently translated into English. The analysis of close to 1.000 pages of interview transcripts was done with the help of the software NVivo and involved several rounds of thematic coding. Both my analytical framework and the structure of the empirical chapters arose in the course of this coding process, which took two and a half months to complete. A series of more and more systematic coding queries based on my research questions helped me to identify and make sense of relevant patterns and relationships that emerged from the data (Bazeley, 2007). Throughout the subsequent empirical chapters I draw extensively on original interview data in order to illustrate the processes and mechanisms that help or hinder the internalisation of immigration control in various settings and from the perspective of different actors.Footnote 5

Apart from interviews I also collected other kinds of information and contextual data: A review of the existing academic literature, policy documents, legal provisions and relevant reports from government as well as non-governmental sources helped me to comprehend and start to compare the overall scope, predominant forms and causes of migrant irregularity in each context, and the way both national governments portray and have reacted to the presence of irregular migrants. Media reports and press statements gave me an idea of the dominant political discourses surrounding the issue of irregular migrants’ access to public services. In addition, I gained crucial insights and understanding through non-participant and participant observation in a variety of social and institutional settings. On one hand, I spent a lot of time in places where mainstream public services are provided (like hospitals), or relevant information about where and how to access them (or potential service alternatives) can be obtained, including official contact points, local community centres or migrant advice agencies. This led to many informal conversations with receptionists and other members of staff, migrants in more or less precarious situations and the people who accompanied them to these sites, as well as other service users. Several interviews resulted from these by-chance encounters. On the other hand, I was a regular volunteer (once a week) at a local migrant advice and support centre in London, and an active member of a migrant collective and advocacy group in Barcelona. This proved very helpful in terms of facilitating access and establishing trust with potential research participants (Staring, 2009; Gonzales, 2011; Hasselberg, 2012), including not only irregular migrants themselves but also their support networks, legal advisors and other members of the local community. The conversations and connections that arose from these longer-term personal engagements thus often constituted the start of ‘referral chains’ that would eventually lead to crucial information, further contacts or interviews with migrant as well as ‘non-migrant’ informants.

3.3 Methodological and Ethical Challenges for Research in the Context of Irregularity

One of the initial assumptions behind this book is that irregularity represents a barrier and creates more or less imminent risks or challenges not only for the people who are rendered ‘irregular’, but also those who must (or choose to) interact with the former. The same irregularity thus also poses a problem for any research that draws on the perceptions, experiences, and interactions of both kinds of respondents. Even though my focus is on the perspective of street-level bureaucrats, answering some of my research questions also required insights and information that could only be gained through face-to-face interviews with irregular residents themselves. Mainly due to the precarious nature of their relationship to the state in which they live, irregular migrants clearly constitute a ‘hard-to-reach’ or ‘hidden’ population (Singer, 1999; Atkinson & Flint, 2001; Gonzales, 2011). An obvious challenge thus consisted in locating, identifying and gaining access to potential interviewees and to establish a relationship of trust with this particular group of informants. In addition, people who live in “a dependent or unequal relationship” (ESRC, 2010, p. 8) or in situations characterised by a general lack of choice and self-determination (Bilger & Van Liempt, 2009) must be regarded as ‘vulnerable’. Irregular immigration status thus clearly renders a person vulnerable and in-depth research into their lives and interactions will almost certainly touch upon particularly sensitive topics like their ‘illegal’ entry or employment and might reveal their involvement in related criminalised practices such as the falsification of documents. This gives the researcher a fundamental responsibility to carefully balance the potential social benefits of the research project against any potential harms for those directly involved in the investigation (Düvell et al., 2010).

My research proposal underwent full ethical review by the University of Sussex’ Social Sciences Research Ethics Committee, which required me to specify appropriate strategies for identifying and approaching vulnerable informants, to conduct all interviews in places that were confidential, safe and familiar to them, and to make sure that any personal data obtained was stored securely (see Cornelius, 1982). While all interviewees were anonymised, non-migrant participants are identified by either the organisation they work for or their particular function within it. The identities of migrant participants were additionally protected by not even recording their real/full names, and by changing any recognisable locations or other details which might allow their personal identification (Singer, 1999; ESRC, 2010). Their particular vulnerability also had implications for the process of obtaining informed consent: While every potential respondent received a copy of the information sheet and had the opportunity to ask additional questions regarding the project, I asked only my non-migrant participants to sign a written consent form ahead of the interview. In the case of migrants in irregular situations I instead obtained their verbal consent after making sure that they had read and understood the information provided. As experiences from previous studies have shown, written consent could risk to undermine (or be perceived as undermining) the anonymity of vulnerable interviewees and create additional suspicion (Finch & Cherti, 2011).

Suspicion can also arise in relation to potential informants who are not in an irregular situation themselves, but (try to) help, work with, or otherwise encounter people who are. My experience with public welfare workers has shown very clearly that such encounters often cause serious doubts or even anxiety about the limits of legitimate commitment and possible engagement. The question of ‘am I acting against the law by helping a person who I know (or suspect) has no residence permit?’ was quite pervasive throughout many of my interviews – and became central to my analysis. In such cases I always made clear that I was trying to understand my respondents’ role and thus to put myself into their position within a certain institutional context. Interviews and conversations often started from the specific difficulties that (potential) irregularity creates for them. In posing my questions I always treated them as representatives of a certain professional group, since what I wanted to get at were not their personal views and opinions, but their perspective as a teacher, social worker or health centre receptionist. Making this clear from the outset often facilitated their agreement to participate and seemed to enhance their confidence.

At the same time, the widespread uncertainty among potential research participants in combination with the relative ease of refusal – by denying any knowledge and previous experience or contact with irregular migrants – also creates an undeniable selection bias. On one hand, those respondents who agreed to be interviewed probably tended to be more aware of and/or concerned about irregular migration and its control than some of their colleagues – which arguably helps the investigation. On the other hand, and given the quite explicitly critical orientation of my research, it might have attracted participants who are themselves more critical towards their government and its policies. This might have distorted the overall picture in terms of their collective views and perceptions, but not the basic mechanisms and difficulties underlying their behaviour. The latter kind of bias might have been reinforced by the fact that also several of my ‘non-migrant’ respondents were contacted via referral chains leading back to my participation in migrant support organisations and advocacy groups. In general, it must of course be kept in mind that in any case, “a narrative may not simply be the story of a life but rather a conscious or unconscious strategy for self-representation and legitimisation of projects for the future” (Bilger & Van Liempt, 2009, p. 135). This is true for both the accounts of migrants and of street-level bureaucrats.

Ultimately, care must also be taken in relation to possible misuse of the research findings once they are published. For Bilger and Van Liempt (2009, p. 131), the researcher’s “power over the distribution of knowledge” i.e. the power to decide on how findings will be used and disseminated, further aggravates the already unequal relationship between researcher and researched. This power imbalance reaches far beyond the end of the research process, since significant harm to both participants’ privacy and their broader interests can also arise at a later stage. I have made, and am still making, every effort to prevent such harm.