This book is about the cross-border movements of a specific group of people at a particular moment in their lives and trajectories. It is about migrants who left their country of origin, Somalia, settled in Europe and have since elaborated complex mobility patterns in order to improve their living conditions. Their lives have been shaped by a first, important move across national borders: the one that brought them to Europe, in most cases through asylum channels and following the collapse of their state of origin. This move has led them to being labelled “migrants” or “refugees” and treated accordingly. They are part of large waves of migration that European states have mostly tried to avoid. However, because of the international conventions these states have signed, among other reasons, they have accepted these newcomers.

Many studies about migrants from poorer countries focus on this particular cross-border movement (the “migration”) and its consequences for settlement and incorporation in the new country. The present research takes a step further: it explores other types of movements, those that Somali migrants may undertake once they have settled in their new country of residence. These “post-migration mobility practices”, as I refer to them, have not raised much attention among migration scholars or policymakers, largely because they do not fit into the common sedentarist narratives about migrants from less economically developed parts of the world to Europe or other powerful regions.

These cross-border movements might appear to be rather banal practices at first sight. It is indeed very common to cross borders, especially for people who live in in Europe, where countries are small and close to their neighbours. From the beginning of this study, however, I was persuaded that something more than just “travelling” or “moving around” was going on. Near the beginning of my fieldwork, I met with Awa, a Somali-Dutch woman based in London whose astute comments opened my eyes more than once and challenged me to think harder. We were discussing my research questions when she contested my initial assumption that she might be a “mobile woman”. When I asked her why, she suggested that Somalis’ tendency to cross borders frequently was sometimes treated too casually:

People move for reasons, right? And sometimes you hear, “Oh Somalis are nomadic, they just like to move around”. Like we have nothing else to do than move around; our ultimate aim is to move around [laughs]. Just to move. And you think: hello, I’m not moving around for the sake of moving around, you know! […] Nomads, they don’t move around because they just move around, but because the grass is not green anymore. They have reasons to move; it’s not exercise [laughs again]!

While I remain cautious regarding interpretations of mobility as a reproduction of nomadic habits, I believe that this quotation indicates the purpose of my research quite clearly. This study explores how, why and with what effects some settled migrants “move around”, or, in my terms, develop sophisticated “cross-border mobility practices”. If mobility is not the “ultimate aim”, then it is a means, a resource people use to gain access to certain advantages. This study assesses the varied cross-border mobility patterns that some settled Somali migrants develop, the conditions under which they are (or are not) able to develop them and the concrete processes that allow them to turn their mobility into an asset and obtain higher social positions within particular hierarchies.

Based on the qualitative data collected in this study, I have elaborated a novel typology of the various cross-border mobility practices migrants may develop. Star-shaped mobility, pendular mobility, secondary migration, temporary visits to the place of origin, definitive return and immobility are the six ideal types of the ways the people in this study may cross borders. The typology refers to – sometimes one-time but more often regular – movements that are significant enough to induce some changes in the lives of the people who undertake them. The typology is explained in detail in Chap. 2, which describes and analyses the ways in which the individuals I met cross borders, the frequency and regularly of their moves, the activities they undertake while on the move and the different places they link together through these activities and movements.

Taking migrants’ cross-border movements as a starting point further challenges transnational scholars’ tendency to assume that transnational ties connect migrants only to their country of origin and/or to people with a similar national or ethnic background. The findings set the ground for an empirically grounded argument that expands our understanding of transnational processes as potentially including places other than the country of origin and the country of settlement, as well networks other than those based on family and ethnic ties. However, I want to avoid the epistemological trap that those who adopt a transnational lens or focus on mobility sometimes find themselves in – that of over-emphasising fluidity and “deterritorialised” experiences and practices (see for instance Mitchell 1997; Adey 2006; Franquesa 2011). Strong attention is paid to migrants’ local contexts, the “localities” (Guarnizo and Smith 1998) that constitute specific hierarchies, create differentiated access to resources and shape the constraints and opportunities people face. This focus makes it possible to include relations and social inequalities at both the local and the transnational levels in the analysis of migrants’ lives and cross-border mobility practices.

This research builds not only on debates in migration and transnational studies, but also on academic discussions in the more recent field of mobilities studies (Sheller and Urry 2006; Kaufmann 2009). While studies on international migration have recently started to address, mostly cautiously and even critically, the idea of mobility, mobilities scholars have showed little interest in migration (Hui 2016). One of the most promising aspects of mobilities studies lies in the theorisation of mobility as an element of social differentiation involved in the production of power relations and inequalities. Some prominent scholars have defined this perspective as the “politics of mobility” (Cresswell 2010b; Massey 1994). This study builds on this perspective, arguing that mobility is an unequally distributed resource that can, under certain conditions, be mobilised and transformed into social and economic advantages. This central argument is developed primarily in Chap. 3. Illuminating the politics of mobility helps uncover how Somali migrants, as social actors with particular social positions, are able to develop a kind of “mobility capital” and convert it into further advantages.

The idea that cross-border mobility may constitute an important resource for some people is also central to a strand of the literature interested in “circulation” (see for instance Tarrius 1993, 2000; Tarrius et al. 2013; Morokvasic 2003; Morokvasic-Müller 1999; Dahinden 2010a; Schmoll 2005; Schmoll and Semi 2013). People engage in specific economic activities, such as (underground) trade, sex work or diverse types of odd jobs while being constantly on the move. They specifically opt for these practices as an alternative to more traditional forms of settlement migration, yet I demonstrate that settled migrants may develop similar strategies when they possess the ability to cross borders regularly.

This study challenges a (sometimes overly static) sociology of social inequalities because it examines the ways in which social actors’ social positions may differ depending on the places and hierarchies involved. It builds from recent research that analyses how wealth differences between different countries may inform migrants’ practices (see among others Nieswand 2011; Nowicka 2013; Kelly and Lusis 2006). The status inconsistencies these differences produce may pose constraints, but they may also create opportunities for those who are able to navigate in a transnational social field. In fact, while most of the respondents in this study encounter obstacles to their ability to valorise their assets in dominant social fields in their country of residence, they also find opportunities to mobilise their resources in other places and hierarchies. This study argues, empirically and theoretically, that capital is transnationalised through cross-border mobility practices, thus demonstrating that social inequalities need to be examined at the transnational level as well as at the level of the nation-state. Chapter 4 investigates these complex processes on the basis of the detailed life stories of three respondents who have developed different types of post-migration mobility practices over the years.

This focus on a neglected aspect of some migrants’ practices, i.e. their cross-border movements, opens new avenues for thinking about their lives and experiences. It makes it possible to explore the agency of people often considered helpless due to their situation as “refugees,” the protracted civil unrest in their country of origin or supposed problematic activities and practices there (female genital mutilation, religious fundamentalism or links with terrorism). While paying close attention to the local contexts (in plural) to which these migrants are connected, and to the specific constraints and opportunities that characterise them, I focus my analysis on the projects and activities these migrants develop thanks to their ability to cross borders.

The findings further challenge those dichotomies that usually consider mobility as the prerogative of a global, cosmopolitan, highly-skilled elite for whom borders have lost their meaning, while migrants are supposedly stuck in a sedentary stasis. Save their initial move (the migration), labour migrants and refugees are often considered, by both policymakers in immigration countries and many researchers, through the perspective of sedentarity and local incorporation and with the expectation of loyalty towards their new country of residence. While the transnational perspective has opened new ways of thinking about migrants’ multiple anchorages, this study provides empirical evidence of the concrete ways in which migrants may develop initiatives to build on connections with other places and improve their living conditions thanks to the specific activities they perform across borders. These alternative forms of “integration”, outside of the frames set by nation-states – or rather mostly in parallel to them – remain underestimated. Yet they can prove beneficial for migrants, and possibly for their countries of residence, as is discussed in Chap. 5 in particular.

While the mobility practices of highly qualified, circulating migrants are generally regarded positively in the academic literature, those of less privileged migrants have been of less interest. The study demonstrates that, under certain conditions, migrants who are not global elites may also benefit from the sophisticated and diverse mobility practices they have mastered.

1.1 Theorising Mobility Practices in Unequal Transnational Social Fields

1.1.1 Categories, Hierarchies and the (Re)production of Social Inequalities

The politics of research involves the theoretical premises that guide the design and the conduct of the research process. In order to make those premises explicit, questions related to the nature of reality (ontology) and the relationship between this reality and the researcher (epistemology) need to be answered (Denzin and Lincoln 2005; Guba and Lincoln 1994). Those premises can also be considered “belief systems” or “worldviews” that guide the research process (Guba and Lincoln 1994).

In contrast to positivist thinking, this qualitative study begins from the view that reality is not something “out there” to be discovered by an objective or neutral researcher, but instead socially constructed, and that the main goal of the research process is to understand how, by whom and in what contexts. The focus is on the perspective that people have on their activities and environments, the discourses they construct about them and the subjective meaning they give to them. I have adopted an interpretive, constructivist perspective to understand how some migrants travel, cross borders and develop transnational activities and social fields. The object of the study – what I call post-migration mobility practices – has thus emerged from understanding the meanings that respondents give to their own movements and/or sedentarity, and to those of others, at a particular moment in their biography, in particular countries of residence and from specific social locations.

According to this perspective, the “world of lived reality and situation-specific meanings that constitute the general object of investigation is thought to be constructed by social actors. That is, particular actors, in particular places, at particular times, fashion meaning out of events and phenomena through prolonged, complex processes of social interaction involving history, language, and action” (Schwandt 1998). The goal of research, therefore, is not to find causalities between events or to “explain” how the world functions, but rather to “understand” social phenomena from the point of view of those who live it. The term “understanding” is translated from Weber’s concept of Verstehen (Weber 1965): for Weber, social phenomena can only be understood by grasping the meanings that social actors attach to what they do (see also Aron 1967).

One of the aims of this research is thus to understand the meanings that the Somali migrants I met give to their situation. How do they see the country in which they currently reside? Are there other places that are meaningful to them, and in what ways? How do they manage to link those different places, and what do those links bring them? How do they relate to significant others located in different geographical places, but also in different social positions, in terms of gender, lifecycle, social class, ethnicity, level of education or legal status? And above all, how do cross-border movements play out in these women’s and men’s lives?

The organisation of society along various axes of differentiation – in particular gender, social class and ethnicity/race – partially shapes the distribution of resources among the population (see Scott 1986; Jenkins 1994; Bourdieu 1987). More specifically, the differentiated allocation of resources is related to the social and symbolic evaluation of those who possess those resources, depending on their social positioning along various axes of differences. In other words, individuals’ social location in ethnic, gendered and classed boundaries and hierarchies influences their access to resources, but also the value of the resources they possess.

Categories of social difference are representations and normative ideas that structure and organise social life. These organising principles and references are not neutral: they are allocated specific contents (for instance, women do certain things better than men, or ethnic groups have specific, bounded “cultures”) that are valorised unevenly. Women’s reproductive tasks related to the domestic sphere obtain less recognition (in symbolic as much as in monetary terms) than men’s activities in the labour market. Similarly, some migrants’ cultures are considered “backward” and implicitly inferior to the “modern”, European societies in which they live. The social construction of differences thus occurs through the hierarchisation of social categories (their relational valuation) and the related unequal allocation of economic and cultural resources (Anthias 1998b). Through these mechanisms, women come to be seen as different from men, migrants as different from nationals and refugees as a different category of migrants than expatriates: specific social meaning is given to each of these constructed categories. In turn, access to education, the labour market or specific social networks differs depending on who you are and how you come to be categorised in dominant discourses. What you are comes to legitimise how the resources you possess are evaluated, and therefore the ways in which you can convert them (Bourdieu 1986) into other resources and accumulate privileges and advantages.

Furthermore, social differences are rendered “natural”, which itself is a sign of the unequal distribution of resources (Anthias 1998b). Processes of naturalisation concern the content of the categories as much as the categories themselves, as well as the boundaries that are assumed to separate them from one another. Social order comes to be perceived as “an accommodation to the natural order” (West and Fenstermaker 1995). More importantly, naturalisation processes simultaneously justify the social hierarchies that constitute this taken-for-granted social organisation of differences. The (re)production of representations about the world takes place through “the subtle imposition of systems of meaning that legitimise and thus solidify structures of inequality” (Wacquant 2008). Hidden processes that tend to reify boundaries and hierarchies thus take attention away from the inequalities that they imply, even for those who are directly concerned. These processes participate in the (re)production of collectively shared representations that influence what people do and the meaning they give to their experiences and those of others.

To sum up, concurrent perspectives on social differences are not just a matter of different worldviews. They involve issues of power and have strong political implications in that they are involved in creating, maintaining and reproducing social inequalities. In this book, gender, social class and ethnicity/race, but also religion, age, stage in the lifecycle and legal status are considered to be not only outcomes of relational and dynamic social constructions, but also hidden dimensions of power relations. Social categories and their hierarchisation may differ depending on the socio-historical context in question. When it comes to mobile people, more than one context may be relevant simultaneously. Many of the people I have met navigate in different social and geographical environments: I have employed a transnational perspective that makes it possible to take into account the sometimes conflicting yet related contexts that influence these individuals’ lives.

1.1.2 A Transnational Perspective

Migration studies underwent a serious renewal during the 1990s with the development and subsequent refinement of scholarship on transnationalism. While the processes that were being studied, related to international migration, were not new, the perspective was, with scholars focusing on the sustained linkages that some migrants maintained with their country of origin after having migrated (see in particular Glick Schiller et al. 1992, 1995; Vertovec 1999; Faist 2000; Portes et al. 1999 for earlier definitions of transnationalism). This new approach modified researchers’ understanding of processes of migration and migrants’ lives in different ways, in particular because it challenged a number of assumptions that had guided studies on migration – as well as government policies – until then.

First, the new transnational lens challenged early assimilationist theories, which viewed migration as a process that would eventually and definitely lead migrants to abandon any kind of relationship with their country of origin, conform to the rules of their new country of residence and “melt” into it (R. E. Park 1928). Even though assimilation was understood in less normative terms by the 1990s, there remained a widespread view of migration as a rupture with previous identifications, loyalties and social networks. While it had earlier been considered natural for migrants to have to choose “between here and there”, it now became clear that some migrants had developed desires and abilities to live “here and there” at the same time (Tarrius 2002). Migrants’ economic, political, social and religious participation did not stop at the borders of their countries of immigration, but extended to other places, in particular their place of origin (Levitt and Jaworsky 2007; Vertovec 2009; Khagram and Levitt 2008).

A second and related assumption that these scholars worked to deconstruct was the idea (often found in social science research) that the borders of the nation-state constitute the natural delimiting container of society, what Wimmer and Glick Schiller (2002) have called “methodological nationalism”. Nation-states were created, in a specific historical context in the nineteenth century, as the units that delimit and separate territories from one another, based on the view that the people who live in those territories are nationals of the states in question (Wimmer and Glick Schiller 2002). This idealised vision of the world was soon challenged by migrants, i.e. people who settled in nation-states of which they were not citizens. Some scholars, in particular those interested in international migration, started to challenge this naturalised view of borders, which causes migrants to be seen as an anomaly that disturbs this taken-for-granted “national order of things” (Malkki 1995). Transnational scholarship has made it possible to go beyond the idea that migrants have to change their loyalties and identifications to fit this view of a unity between the state and the population located in its territory. It has opened the possibility of alternative perspectives on space and boundaries, in particular those of migrants themselves.

The literature on transnationalism flourished at a time of growing concerns about and academic interest in globalisation. Broader theoretical discussions acknowledged and analysed the increasing disconnections between space and time: while Beck was arguing for a “cosmopolitan turn” in the social sciences (see for instance Beck and Sznaider 2006), others analysed how flows, and in particular human mobility, had become fundamental dimensions of the globalised, postmodern world (Appadurai 2001; Bauman 2000).

A transnational lens also challenges the belief that the migration process is linear, a point that is central in this book. Empirical studies have made it clear that migration often does not involve only a single move from a point of departure to a point of arrival. The linkages that migrants strive to keep alive and further build on often coincide with different kinds of movements across borders. Migrants regularly move back and forth between their country of origin and their country of residence, go back to visit their families and friends who have not migrated, circulate between different locations and envisage and undergo temporary or definitive returns to their place of origin (Jeffery and Murison 2011). While it is true that transnational identities and practices do not require cross-border mobility, the transnational perspective has allowed researchers to analyse how the initial move and settlement in a new country is, for some migrants, intimately related to further types of movements.

I understand the transnational perspective as a lens rather than a new theory. Transnationalism does not explain things, actions or events. It is a way of looking at social reality without remaining trapped within taken-for-granted assumptions. Adopting such a lens makes it possible for a variety of practices to emerge from the data and, even more importantly, for the meanings that are given to those practices by respondents to be taken seriously. It is a vision that departs from the possibility of a transnational reality without assuming either its non-existence (as in assimilationist perspectives) or its necessary relevance for all migrants, as in some celebratory accounts of transnationalism.

However, when considering the transnational processes at stake, I have been careful not to privilege similarities based on a common origin at the expense of other types of differences that cross “ethnic groups”, such as those pertaining to gender and social class (Anthias 1998a). I have strived to avoid “groupist” (Brubaker 2004) perspectives on migrants’ transnational practices and identifications, and to get away from the “ethnic lens” (Glick Schiller et al. 2006; Anthias 1998a; Brubaker 2005; Dahinden 2012, 2009). Migrants’ enduring connections with their country of origin do not necessarily translate into identities and solidarities revolving around their ethno-national background. For all these reasons, I do not refer to Somali migrants as a diaspora. The history of the word “diaspora” and its social analyses is long and complex, and nuanced understandings of the concept are now widely accepted (Cohen and Fischer 2018; Dufoix 2003). For evident epistemological reasons, this book is not interested in determining whether Somali migrants should be defined as a diaspora or not. Emic rather than etic definitions of the term, however, would fit better with my constructivist perspective: such an approach seeks to illuminate how and under what conditions some social actors may define themselves as part of a diaspora. Diaspora becomes, in this view, a “claim” and a “category of practice” rather than an imposed definition of a predefined group (Brubaker 2005). Kleist (2008a, b), for instance, convincingly argues that identification with a diaspora among some Somali migrants – especially men – represents an important resource in their struggle for recognition. Horst (2013) similarly argues that European relief and development institutions work with such self-defined diaspora groups but have the effect of depoliticising their claims rather than empowering them (see also Sinatti and Horst 2015). The concept of diaspora is closely connected to ideas related to transnationalism (see for instance Faist 2010 for a discussion). I have opted to use the latter term because of its broader scope and because it makes possible the inclusion of cross-border activities and networks that have nothing to do with either the ethnic group of origin or the country of origin.

Scholars have used different concepts to refer to sustained cross-border linkages: “transnational social space” (Faist 2000; Pries 2001), “social formations” (Dahinden 2010b; Luis Eduardo Guarnizo and Smith 1998; Vertovec 2003) and “communities” (Goldring 1998; Kivisto 2001). Like others (for instance Goldring 1998; Nieswand 2011; Pessar and Mahler 2003; Portes et al. 1999; Vertovec 1999), I feel most comfortable with the concept “transnational social fields”, which was first developed and later elaborated by Nina Glick Schiller and other scholars (see in particular Basch et al. 1994; Glick Schiller 2010; Glick Schiller and Fouron 1999; Levitt and Glick Schiller 2004). This concept emphasises first the interconnections between people and institutions located in different places that might influence, even indirectly, migrants’ lives and practices. Transnational social fields are “networks of networks that link individuals directly or indirectly to institutions located in more than one nation-state” (Glick Schiller 2010). Second, although this concept should not be conflated with Bourdieu’s concept of “discrete fields of power” (Glick Schiller 2010), it allows for an empirical and theoretical acknowledgment of the multiple and interlocking contexts that shape the constraints and opportunities that migrants face. Social fields have no geographic materiality or coercive power, yet they are durable influential entities that affect the lives of both migrants and non-migrants (Nieswand 2011).

The concept of transnational social fields underlines the fact that “the transnational” stems from people’s practices and does not exist per se, in contrast to nation-states, their institutions and legal apparatuses. At the same time, however, it is constituted by interlocking networks and institutions that do exist materially and exert an influence on people. Moreover, social actors are socially located within those institutions and networks: they are gendered, ethnicised/racialised and classed, sometimes in conflicting ways. The transnational social fields and the local or national social fields and other hierarchies in which the migrants in this study are embedded interact to shape the frameworks of constraints and opportunities in which they move.

However, the “nation” remains meaningful and pertinent in transnational studies (Dahinden 2017; Glick Schiller 2015). Migrants’ practices, networks and identifications are not free from constraints, and most empirical studies show how much not only states, but also power relations between them, influence migrants’ lives. For a start, being able to cross borders depends on states’ issuing the necessary documents and their authorisation to cross their borders and enter their territories. Moreover, power relations between states exert a strong influence on who is and is not considered a migrant (Glick Schiller 2010).

Migrants who are transnationally connected are also locally anchored, and bringing those anchorages back into perspective is a way to investigate issues of social positions, inequalities and power relations. Guarnizo and Smith (1998) argue that transnational practices connect people across borders, but “are embodied in specific social relations established between specific people, situated in unequivocal localities, at historically determined times” (11). Migrants embedded in transnational social fields are always also geographically and socially located social actors. The concept of “locality” refers to the multiple specific contexts in which migrants are socially, economically and politically rooted and which constitute the constraints and opportunities they must square with (Dahinden 2010b; Guarnizo and Smith 1998).

The concept of locality – or rather its plural form, “localities” – makes it possible to include different geographic orders in the same analytical framework. While nation-states should not be left out of the analysis, they are not the only unit that matters in understanding transnational processes. Social actors are located in specific countries, regions, cities or villages and neighbourhoods at the same time, and they all participate in shaping the contexts in which they live. Legislation and policies in the place of residence, socio-economic and political contexts, local networks and opportunities to meet with other people and exclusion and inclusion processes all interact to create specific opportunities and constraints for migrants to become simultaneously locally anchored and transnationally involved.

The concept further makes it possible to understand that transnational practices and identifications do not stand in opposition to mooring, anchorage and incorporation into possibly multiple localities: rather, both types of processes interact with each other. Since the beginning of the 2000s, there have been discussions about the link between them, with a (sometimes normative) concern regarding the impact of transnational practices and identifications on migrants’ incorporation into their country of residence (see Bivand Erdal and Oeppen 2013 for a detailed discussion). In this book, “incorporation” refers to migrants’ “processes of adaptation” (Bivand Erdal and Oeppen 2013) to their new place of residence. These processes include migrants’ inclusion in the labour market or the educational field as much as their connections to local institutions and people (associations, neighbours, other parents, friends and so on). In contrast to the idea of “integration”, it does not make normative assumptions about local or national policies aimed at regulating migrants’ adaptation.

Most of the literature agrees that transnationality and incorporation combine in different constellations rather than compete with each other (Morawska 2003; Levitt 2003; Bivand Erdal and Oeppen 2013), and that high levels of transnational involvement do not directly translate into low levels of local incorporation (see for instance Itzigsohn and Giorguli-Saucedo 2005; Snel et al. 2006). A closer look at the different empirical studies at hand, however, reveals a nuanced picture in which social position at the local and transnational levels is an important differentiating element. Various studies concur that migrants with a high level of qualification and a good socioeconomic position in their country of residence are the most likely to develop transnational networks and activities (Portes et al. 2002; Levitt 2003, 2009). Others demonstrate that migrants who are the most marginalised and suffer the most from exclusion and discrimination also maintain more transnational links than others (Dahinden 2009; Itzigsohn and Giorguli-Saucedo 2005). Itzigsohn and Giorguli-Saucedo (2005) refer to the first form of transnationalism, which involves a high degree of incorporation made possible by the investment of sufficient capital at both the local and transnational levels, as “resource-dependant transnationalism”, and to the second, which develops in response to exclusion, as “reactive transnationalism”. In both cases, the conditions encountered in both the country of residence and other places, including the place of origin, intervene in encouraging or hindering transnational practices. Furthermore, there are linkages between the two processes, as the biographies developed in the coming chapters will demonstrate. For people whose lives take place in transnational social fields, different localities interact to constitute complex social realities that may benefit social actors – a central argument of this book – but which may also cause them anxiety, discomfort and suffering.

1.1.3 Bringing “Mobilities” into Migration Studies

The term “mobility” is employed in different academic domains (from biology to tourism studies), various policy fields (from urban planning to migration management) and everyday conversations (from “mobile” phones – before they became smart – to ethical personal choices of “soft mobility”). As a result, one can question the usefulness of such a widespread and fuzzy term for understanding any social phenomenon. Nonetheless, for the reasons outlined below, I find mobility relevant and useful in examining, empirically and theoretically, the movements that migrants undertake after they have settled in their country of residence. However, while European migration scholars have recently started to show a critical interest in issues and conceptualisations of mobility (Favell 2007; Favell and Recchi 2011; Faist 2013; Glick Schiller and Salazar 2013; Pellerin 2011; Amelina and Vasilache 2014), this new preoccupation is often closely linked to political concerns and categorisations.

I will first discuss the concept as it has been understood in migration studies, generally based on legal and political distinctions and contrasted with “migration”. This understanding is not the one I endorse in this book, but it must be acknowledged because of its recent prevalence. My own definition of mobility is inspired by mobilities scholars, and will be discussed further down.

1.1.3.1 Mobility Versus Migration: Legal and Political Distinctions

In recent years, mobility has been studied in the context of the new modalities of cross-border movements, border control by states and globalised economic contexts and power relations. In Europe, mobility has been contrasted with migration based on a political and legal distinction: the movements of people from third countries entering Europe (migration) are opposed to the free movements of European Union citizens moving within Europe (mobility) (Bauböck 2012). The EU actively promotes the mobility of its citizens: contrary to other types of human mobility, this specific type of movement is thought of as a cornerstone of European integration and European citizenship (Favell and Recchi 2011). This political project has led European policymakers to value mobility positively and promote the movement of people within Europe. In parallel to this new logic, border controls have increased vis-à-vis migrants from outside Europe. “The freedom of mobility for some (citizens, tourists, business people) could only be made possible through the organized exclusion of others forced to move around as illegal ‘aliens’, migrants, or refugees” (Verstraete 2003). Policy distinctions go together with legal categorisations, which are built on and further produce social differentiations between different types of people, based on nationality, ethnicity/race, gender and social class.

A second understanding of mobility has gained increasing acceptance in both policy and academic circles in recent years. According to this understanding, mobility is a short-term, temporary form of economic migration: workers cross borders for limited periods of time with no intention to settle (contrary to “traditional”, permanent, forms of migration), returning to their place of origin once the contract is over (Pellerin 2011). This kind of movement is also often called “circulation” or “circular migration”, because the people involved might move back and forth on a regular basis. While such temporary migration schemes are not new, they are regularly reintroduced as a way to respond to the changing economic needs of both countries of destination and countries of origin (The Gobal Commission on International Migration 2005; Castles 2006; McLoughlin and Münz 2011; Vertovec 2009). While the circulation of highly skilled migrants generally does not create significant concerns for states, the mobility of less qualified workers is the object of political debates and policies aimed at preventing the settlement and ensuring the return of these workers to their country of origin (Doomernik 2013).

Some scholars have argued for the need to analyse the discourses and social categories related to this recent dichotomisation of people’s movements into long-term, mostly state-controlled, migration, on the one hand, and short-term, freer and economy-led mobility, on the other (Pellerin 2011; Faist 2013; Glick Schiller and Salazar 2013). Different forms of mobility and migration are evaluated differently and are subject to very different legal and political treatment. As Faist (2013) argues, “the movement of persons is dichotomized in public debate into mobility and migration, with mobility connoting euphemistic expectations of gain for individuals and states, and migration calling for social integration, control and the maintenance of national identity” (1640). Sedentary migration is increasingly perceived negatively, as being problematic because of its impact on the receiving society’s collective goods, while mobility – either as short-term migration or the circulation of highly skilled specialists – is presented as a solution from which everyone benefits. However, the distinction between the two types of movements is shaped by unequal power relations, and it further reproduces global inequalities. A perspective that examines the “geopolitics of mobility” (Verstraete 2003) or the “regimes of mobility” (Glick Schiller and Salazar 2013) is necessary if we are to understand the relationships between different types of movements and the ways in which they are promoted or limited by states within global power hierarchies.

1.1.3.2 A Definition Derived from Mobility Studies

My conceptualisation of mobility is not based on legal or political distinctions, unlike those presented above. It builds on radically different epistemological and theoretical foundations and draws on developments in the field of mobility studies. This strand of scholarship “call[s] attention to the myriad ways in which people become part, in highly unequal ways, of multiple translocal networks and linkages” (Salazar 2016).

Mobility studies emerged from the observation that the concept of (spatial) mobility was applied in a variety of disciplines in the social sciences to describe and analyse very diverse phenomena at different scales: from daily mobility in the urban space to transportation systems, from residential mobility to transnational migration trajectories. In this light, mobility scholars have notably called for a broader multidisciplinary approach to mobility, and for a deeper theorisation of the concept itself (Kaufmann 2009; Söderström and Crot 2010). The so-called “new mobilities paradigm”, or “mobility turn”, is the clearest attempt to bring together different types of movements into a single analysis, and to challenge the tendency of the social sciences to treat stability as normal and mobility as problematic (Sheller and Urry 2006; Urry 2007). One of the most promising aspects of this recent literature on mobility lies in theoretical developments focusing on the conceptualisation of mobility as an element of social differentiation. The “politics of mobility” (Massey 1994; Cresswell 2006, 2010b) emphasises how mobilities participate in the (re)production of unequal relations of power alongside other markers of differentiation.

The mobility turn is constituted by a three-pronged agenda (Söderström and Crot 2010; Söderström et al. 2013). First, it draws new epistemological foundations by challenging the sedentarist premises that pervade the social sciences in general. It views mobility as a fundamental fact of social life and gives it the analytical credit that it has often lacked. This perspective coincides with critiques of methodological nationalism: the mobility turn undermines perspectives that emphasise bounded and rooted notions of place, regions and nations as the fundamental basis of identity (Cresswell 2010a). Second, the agenda aims at an ontological reconceptualisation of mobility that includes in the same analytical framework a vast array of interconnected forms of mobility. Third, the mobility turn calls for the development of appropriate methods of inquiry that do justice to the mobile social realities that researchers want to understand. These mobile methods include conducting (participant) observation of movement, exploring people’s imaginative or virtual mobilities (by analysing blogs, websites or email exchanges, for instance), creating mobile ethnographies, examining transfer points (airports and train stations, for example) and undertaking multi-sited ethnography (Sheller and Urry 2006; Urry 2007). While I did not move myself (except between the different places where respondents reside), the different types of interviews that I conducted focused on people’s past, present and possible future mobilities and their relevance in some migrants’ lives.

A critique that has been made of this change of paradigm is that treating mobility as pervading every aspect of social life undermines the theoretical and analytical power of the concept itself (Söderström et al. 2013). As Adey (2006) states in the title of an article, “If Mobility is Everything then it is Nothing”. A second, related critique, is that this emphasis on movement and change might result in a celebratory vision of mobility (Faist 2013). Just as some transnational scholars have been overly enthusiastic, treating this newly appearing reality as necessarily empowering to migrants, mobility studies may tend to praise movement and conceive of it as intrinsically good.

Against those two criticisms, a significant strand of recent scholarship has analysed how mobility and power relations are related (Söderström et al. 2013) and addressed the complexity of the experiences of (im)mobility (Van Hear 2014). Analysing mobility in these terms also makes it possible to go beyond preconceived ideas of mobility as positive and its association with freedom and emancipation (Franquesa 2011) and acknowledge the pressures and constraints that often go with being mobile. “De-romanticising” mobility further means attending to spatial dynamics and frictions (Schapendonk et al. 2018). These aspects are particularly relevant for the arguments I will develop in this book.

Mobility scholars are interested in the meanings of movements, their representations and valuations within specific contexts, and their embeddedness in power relations. “Understanding mobility thus means understanding observable physical movements, the meanings that such movements are encoded with, the experience of practicing these movements and the potential for undertaking these movements” (Cresswell and Uteng 2008). Mobility can be distinguished from simple movement because it is a catalyst for change in the lives of the people who undertake it, their identity or their social position (Kaufmann 2009). The social effects of mobility are its fundamental theoretical characteristic. Naturally, people do not move free from constraints: the political categorisations discussed above participate in shaping the types of mobility practices that social actors are able to develop.

In line with this perspective, I do not consider mobility a specific type of movement (contrasted with other types of movements defined a priori) undertaken by specific categories of people, but instead use a constructivist stance that views mobilities as historically constructed through social interactions in specific socio-historic contexts. The movements I am interested in exert an influence on people’s biographies, which is why it is necessary to let them emerge from the empirical data. I thus include any type of geographic movement significant enough to be acknowledged as having created some kind of change in the respondents’ lives. These empirical considerations allow me to build a typology of the various cross-border mobility practices undertaken by respondents in this study.

I will go further and argue that international migration can be considered one mobility practice among others, or – more interestingly – related to others (Favell 2007; Amelina and Vasilache 2014). What I describe as international migration involves a cross-border movement followed by long-term settlement in the new country, i.e. the action of taking residence.Footnote 1 Its specificity comes from the social and legal consequences of such a move, since those who undertake it are considered – by the state they have entered as well as by others – “migrants”.

However, my interest is limited in three ways. First, I only consider those movements that were performed after people had migrated to Europe and were settled for many years, what I call “post-migration” movements. Second, my focus is on cross-border mobility, which excludes all types of movements within nation-states. This focus could be considered to be in contradiction with one of the foci of mobilities studies, which is to include varied types of movements without a priori delimitations. Studies that focus only on international movements have been critiqued for their pervasive methodological nationalism on the grounds that they reproduce states’ categories and interests (Favell 2007; Kalir 2013). However, my limitation is not a theoretical one: I do not pretend that cross-border movements are more or less relevant to people’s lives than internal ones, or that they are essentially of a different kind. This methodological choice stems from my interest in the relationship between migrants and states. While transnational scholars have emphasised the need to go beyond a naturalised understanding of the nation-state, they have also shown that states shape migrants’ lives in important ways, especially when it comes to crossing national borders. As Hui argues, “migrants do not exist a priori” (Hui 2016): they become migrants because of states’ regulations and border control. Third, I have not totally endorsed the “mobility turn”, as I focus on people’s physical cross-border movements more than on virtual forms of mobility (Urry 2007). While I describe post-migration mobility practices as closely interlinked with the circulation of objects, knowledge and ideas, the mobility of these other entities cannot be explored in detail within the framework of this study.

Despite these limitations, I see many advantages that justify the use of the concept of mobility to examine the cross-border movements of Somali migrants from their country of residence. First, the concept of mobility makes it possible to bring different types of spatial movements that are often considered separately into a common analytical framework. By acknowledging the diversity of the forms of cross-border movements that people may undertake, the concept makes it possible to challenge linear visions of migration as a unique and unidirectional movement followed by settlement in the new country of residence. It also includes in the picture ways of moving other than permanent migration.

The second advantage of using the concept of mobility is that it makes it possible to look at migrants’ biographies and life trajectories in the long term. It adds temporality to spatiality. Initial migration is often considered the only relevant movement in migrants’ lives, a border crossing between two moments of sedentariness. But mobility in its multiple forms may remain an important aspect of some people’s lives over the years. It is also subject to changes, depending on what has been experienced, the resources that have been accumulated, the obstacles that have been encountered and the opportunities that have arisen in different places. A certain type of mobility may be developed at some point in someone’s life, dropped later, replaced by another type of mobility practice, followed by a period of sedentariness and so on. As Schapendonk and Steel (2014) argue, “by providing in-depth insights into how these trajectories evolve (i.e., how they are produced, facilitated, slowed down, and blocked) and how they are experienced in differentiated ways, the mobilities perspective helps us to move away from bipolar and frictionless conceptualisations of transnational migration” (263). Studies concerned with individuals’ mobility trajectories highlight these different phases in a life course, often emphasising how the lifecycle determines what kind of mobility is possible and potentially profitable at what moment in a life trajectory (Van Liempt 2011a; Engebrigtsen 2011; Walker 2011; Ley and Kobayashi 2005; Schapendonk and Steel 2014; Kalir 2013). Mobility practices, in this light, appear as fundamental aspects of strategies that migrants develop after their initial migration. If mobility is – in some cases – strategic, that means it can become a resource for some people.

This leads me to the third advantage of the concept: while “movement” remains a descriptive term, “mobility” implies more than just the physical move because it is also infused with meanings and power relations (Cresswell 2010b). As the coming chapters will show, people move in different ways, with different aims and different consequences. Their mobility is constrained or facilitated by the other types of assets they possess and the ways these assets are valued in different contexts. The idea of mobility as an unevenly distributed resource reveals processes of social differentiation in which the potential to be mobile as well as the actual fact of moving play a part (Kaufmann et al. 2004). This perspective brings power relations into the analysis of migrants’ (im)mobile lives. On the one hand, it illuminates how for some people to be able to deploy specific strategies of transnational mobility, there need to be others who are less mobile, not mobile at all or mobile differently. On the other hand, it emphasises that mobility and immobility are strongly shaped by social actors’ relationship with nation-states, based on processes of gendering, ethnicising/racialising and classing.

1.2 Methodology and Methods

A qualitative, interpretive and constructivist approach has guided the methodological choices in this study, which has no ambition to measure or quantify the processes under study, or to generalise findings to all migrants, or even to all Somali migrants. On the contrary, the goal here is to get sufficiently close to the research partners to grasp the meanings they give to their experiences and practices. Lengthy fieldwork, a significant number of qualitative and narrative interviews, a few informal conversations or other communication opportunities and (limited) observations have enabled me to create an “intimate relationship” (Denzin and Lincoln 2005) with the respondents, and to understand the processes I am interested in.

1.2.1 Methodological Transnationalism

This study is interested in the cross-border movements of migrants, and in the ways in which respondents interpret the potentially multiple meaningful places that are interconnected in their lives. The transnational lens that has been adopted in this study has methodological implications, in particular the need to attend to the multi-scale and multi-sited contexts in which respondents might physically, virtually or emotionally move. “Methodological transnationalism” (Nieswand 2011; Khagram and Levitt 2008; Amelina and Faist 2012; Amelina et al. 2012) avoids the traps of methodological nationalism and its inherent methodological flaws. It makes it possible to conceptualise the physical and social space in which migrants evolve as not bounded by the national boundaries of their current country of residence (see also Levitt and Jaworsky 2007). In other words, methodological transnationalism is “a methodological programme that aims to enable the study of transnational mobilities and transnational formations by avoiding a nation-state centred methodology” (Amelina and Faist 2012).

My research has been designed and adapted in the course of the study to take into account the transnational practices and identifications that respondents may develop. Interviews emphasised the possibility of transnationalism, the multiple and sometimes contradictory contexts that shape people’s lives. Although the interviews took place in a single national setting, they were designed to include the transnational aspects of the migrants’ identities, beliefs and activities (Khagram and Levitt 2008; Salazar et al. 2017; Barglowski et al. 2015).Footnote 2 To better understand the (potential) transnational dimension of the research participants’ lives and experiences, I also included, at a later and experimental stage of the research, some visual methods to grasp more comprehensively the places they come to see as relevant, and the meaning they give to each of them. The analysis also includes the connections between those places.

Since data are always located and situated, methodological transnationalism constitutes a real challenge for researchers, because it multiplies the backgrounds against which data must be interpreted and understood. This study is interested in the multiple contradictory social locations, in terms of gender, ethnicity/race and social class, that migrants might experience depending on the context in which their resources are evaluated. It analyses their mobility practices as strategic responses to these differentiated experiences. At the methodological level, this focus requires an “analytical framework allowing for the description and analysis of multiple and simultaneous forms of inclusion of migrants and non-migrants in different socio-spatial contexts and institutions within a global society without prejudging the primacy of one of them” (Nieswand 2011). Even when people do not actually move to other places, they may be involved in transnational social fields that might strongly influence their experiences, practices and identities (Levitt and Glick Schiller 2004). In the end, methodological transnationalism is what offers the researcher a reminder to open her eyes and ears to things that might happen or have meaning in a place other than that of the fieldwork.

1.2.2 Reflexivity: On the Researcher’s Positionality in the Field

A crucial aspect of the research interaction is that the researcher is personally involved in the relationship and thus in the production of knowledge. Such a stance requires reflexivity regarding the context from which the empirical data emerge. The research encounter in itself is an unequal relationship that develops upon the request of the researcher, who has had to convince (sometimes with the help of others) respondents to participate. The researcher thus initiates the encounter and sets the basic rules (Bourdieu 1993), in particular regarding the topic of the discussion and the goals of the research. However, each of the people involved has different explicit or implicit expectations when entering the relationship that might shape the direction the interview takes. Respondents might also have suspicions about the “real” aim of the interview, for instance with migrants who have had or are still in contact with the authorities regarding their legal status (Carling et al. 2014; Bloch 1999). A few respondents were in the process of obtaining the citizenship of their country of residence: I was particularly careful, when analysing their data, to include the possibility that they had constructed a discourse able to fit both types of encounters, one with an independent researcher and one with an (undercover) representative of the national authorities.

The last example demonstrates that the basic asymmetry pertaining to the research encounter is strongly shaped by the social positioning of those who enter the relationship, along the lines of nationality, ethnicity or race, gender, age, level of education, religion and so on. I entered the field as a white female researcher with a Swiss name, an academic background and a university affiliation. The people I was interested in meeting and talking with were all first-generation Muslim Somali migrants. Their national origin and their migration history has obliged them to deal with external categorisations in European contexts, focusing in particular on their colour, religion and history of migration. In the British context, Somalis have come to occupy the status of a “suspect community”, increasing the need for reflexivity (Ryan 2011; Ryan et al. 2011).

Several migration scholars (see for instance Carling et al. 2014; Thapar-Björkert and Henry 2004; Ryan et al. 2011; Baser and Toivanen 2017) have discussed lines of differentiation other than ethnicity or religion that constitute the background for (sometimes conflicting) commonalities and differences between researchers and research participants: gender, age, socio-professional status, migration experience and so on. The politics of research involves the ways in which researchers introduce themselves and their study when entering the field, but also the ways in which research participants actively take part in shaping the encounter, based on their own representations and assumptions about research and researchers in general, and this researcher in particular (see also Ryan et al. 2011). I introduced myself both professionally and personally. At the professional level, I presented myself as a researcher based at a Swiss university but also, when asked or when I felt it was appropriate, as a doctoral student and/or as an anthropologist. I also emphasised that my interest in Somalis had been a long-term one, since I had chosen my research topic after having participated in previous research on Somali refugees, demonstrating what Carling et al. (2014) have referred to as a “sustained commitment”. This long-term commitment could also be perceived from the duration of the fieldwork, my regular return to places where I had done interviews (in both Switzerland and Britain), and my requests for supplementary meetings with those I had already met and for new contacts.

I also introduced myself on a more personal level. Personal characteristics definitely shape the research relationship, although it is often difficult to assess in what manner (England 1994; Sanghera and Thapar-Björkert 2008). Being a female in my 30s, for example, allowed me to participate in a quite informal discussion with three women about my age that touched upon intimate topics, including sexuality. However, accessing women, in particular those who were not involved in some kind of visible associational activity, proved difficult, demonstrating that being female is not a sufficient commonality to create trust and interest. I did find, however, that disclosing that I am a parent and, at one stage of the fieldwork, that I was pregnant, had an effect on some women, who suddenly became more open towards me, building on this commonality, since we were all mothers (see also Carling et al. 2014).

Feminist scholars have been critical of researchers who try to enter the world of more marginalised people and the neo-colonial implications inherent in social scientists’ appropriation of the voices of these marginalised “others” (England 1994; Reay 1996). Similarly to England (1994), I do not feel totally comfortable with being a member of a “dominant” group of society who wants to capture the experiences of more marginalised and stigmatised others: England was a straight woman studying lesbian women in the early 1990s, while I am a privileged white Swiss national interested in Black Muslim Somali migrants who mostly belong to the lower or lower-middle class. Even though reflexivity does not remove the power differentials that inform the research relationship (England 1994; Sanghera and Thapar-Björkert 2008), it renders the researcher visible and requires them to acknowledge where their analysis comes from.

In the empirical chapters that follow, I have opted for an account of social reality that gives a great deal of space to the research participants, through either their stories (as reconstructed by me) or selected quotations related to a particular argument (or both). The extensive use and valorisation of my empirical data renders explicit the processes through which empirical data are transformed into research conclusions. Those data “reflect the empirical fieldwork, guarantee its validity, and enable its critique” (Olivier de Sardan 1995). In turn, this use of the empirical material allows, at least partially, for a clearer distinction between what respondents say about their stories and what belongs to my interpretations of the social reality I have been given to see and hear. To be sure, I would not dare to claim that I speak on behalf of the research participants: even though my analysis is based on the ways in which they perceive their lives and environments and I give them a voice by including excerpts of our discussions and partial accounts of their lives and experiences in this book, I am solely responsible for my theoretical statements and conclusions.

1.2.3 Delineating the Field: Conceptual Issues Related to the Population Under Study

The terminology used to describe the population under study is far from neutral or unequivocal. The “Somali migrants” I have decided to study are women and men who lived in Somalia, left it at some stage in their lives and are currently residents of a European country. Some of them migrated directly to Europe, while others lived in other places in between. The national borders of Somalia are contentious, and I have opted here to use “Somalia” to denote the territory that has been recognised by the international community to belong to the country since its independence in 1960. People in this study thus originate from different parts of the now officially named “Federal Republic of Somalia”, including Somaliland and Puntland. My decision should be understood a methodological choice rather than a political stance. Accordingly, the term “Somali” does not preclude the possibility that respondents may have alternative identifications with their place of origin.

Contrary to other studies on Somalis who live outside their country of origin, I refer to my respondents as migrants rather than refugees. Although “migrants” can also carry implicit associations, I use it in a descriptive manner to define people who have moved from one place to another, have crossed at least one international border and have settled in a new place. A significant number of Somali international migrants in general, and of the respondents in this study in particular, have, at some stage during their migration trajectory, entered asylum systems and been legally recognised as refugees. However, this legal category should not, in my view, be confused with an analytical category, as it does not adequately take into account the nuanced identification processes that are at stake (Malkki 1992, 1995; Crawley and Skleparis 2017). That I do not emphasise the refugee aspect of the identity of my respondents – except when it emerges as relevant in their own discourses – does not mean that the context in which their migration has occurred has no impact on their lives and practices.

Another important issue arises from the fact that this study takes as the unit of analysis persons defined by their country of origin, Somalia. Nationality becomes the common marker of the respondents I have included in my fieldwork. However, I follow the trend in the field of migration studies to “de-ethnicise” research designs (Wimmer 2007; Fox and Jones 2013), avoid “groupist” interpretations of social processes (Brubaker 2004) and avoid the assumption that nation-states are the natural framework of analysis (Wimmer and Glick Schiller 2002; Amelina and Faist 2012). Therefore, the ethnic background of my respondents does not constitute an explanatory framework for their practices and identifications: it is rather one possible outcome of interactions in particular contexts and within particular power relations. The analysis focuses as much on what may differentiate people with the same ethnic or national origin (for instance, along the lines of gender or class) as on what may lead them to identify with a (constructed and contextual version of the) “ethnic group”.

1.2.4 An Ethnographic Approach Based on Interviews

I met respondents in different cities in Switzerland, as well as in varied neighbourhoods in Greater London. These countries were chosen as sites of the study because they constitute two very different contexts regarding, on the one hand, the characteristics of the Somali migration and, on the other, national and local legislation, settlement conditions and employment opportunities (see further in this chapter).

National contexts constitute important historical and social settings, as they help determine the constraints and opportunities that migrants face. However, my aim is not to produce a systematic comparison between Britain and Switzerland. National regulations have an impact on the circulation of people (policies regarding immigration, asylum or naturalisation in particular), but also on the circulation of objects (trade regulation). However, migrants’ practices are affected by a wide array of national regulations and policies that do not directly relate to border crossing, but to educational landscapes, family policies or access to the independent labour market. Other geographic scales may also be relevant in the lives and experiences of the people under study. In this sense, Britain and Switzerland should be understood as a geographic delineation for my study rather than national contexts to be compared.

The fieldwork took place in different stages between May 2009 and December 2011, a strategy that proved to be advantageous. Coming back to respondents after some time was a way for me to prove my interest in their lives and experiences, and to build trust. Carrying out “multiple sequential interviews”(Charmaz 2001) over an extended period of time also allowed me to explore and address changes that had occurred in respondents’ lives since our last encounter. People got married, had children, moved, changed jobs, started new activities and reinforced or diminished their transnational relationships, involvements and identifications. Meeting on more than one occasion constituted an opportunity for both me and the respondents to see things from a different angle, since time had passed and things had changed. Coming back to respondents also helped me follow up on issues that had not been fully explored in the first interview (Olivier de Sardan 1995), ask new questions that had arisen through preliminary coding (Charmaz 2001) or carry out a new type of interview.

I conducted different types of interviews. With most research participants, I conducted semi-structured interviews. Later in the fieldwork, I started conducting biographical interviews as well as spatial interviews, using visual material. A few group interviews were also carried out, while my data also consist of limited moments of observation as well as more informal communication that took place during the research. Details on the different interview techniques, as well as a map of all interviews indicating the number of encounters and type(s) of interview conducted with each respondent are included in the Appendix to this chapter.

In line with my methodological stance, the sample was not constructed a priori, but was developed and refined over the course of fieldwork. As basic criteria, I had decided to include in my study first-generation migrants from different parts of Somalia whose main country of residence was, at the time of the interview, Switzerland or Britain. Since having secured a stable legal status constitutes a condition for the development of cross-border mobility, I also created a sample primarily constituted of people who had resided in Europe for a relatively long time (at least a decade).

Needless to say, I have not aimed for a statistically representative sample in any way. I have opted instead for a theoretical sampling strategy constructed during the course of the fieldwork (Glaser and Strauss 1967; Charmaz 2006). For instance, I started to hear about businesswomen who would regularly undertake cross-border movements in the frame of their informal (and thus hidden) activities, which led me to look for this kind of persons to include in my sample. This sampling strategy is based on categories that are not constructed according to demographic criteria (for instance gender, legal status and so on), but on the relative positioning of those actors in terms of hierarchies or functions (Bertaux 2005). In short, I developed my sample by including, gradually and alongside early coding and analysis, people who had developed different types of relationships to mobility (or immobility). I have looked for people who move, or move in different ways, and who do not move in order to understand the circumstances that led them to develop those various practices related to mobility or immobility.

To access potential respondents, I multiplied and diversified the “entry points” – the contact persons or organisations – as much as possible, as well as, to a lesser extent, “spontaneous meeting places”Footnote 3 (see also Dahinden and Efionayi-Mäder 2009). I used personal contacts as well as gatekeepers active in diverse types of associations in both Switzerland and Britain. I also used (transnational) snowball techniques, asking research participants to refer me to other potential respondents (see Fig. 1.1 in the Appendix).

A pitfall that researchers in migration studies in particular need to avoid when constructing their sample is that of including only respondents for whom their national, ethnic or religious background is a central identification category. In order to avoid this pitfall, I adopted a sampling strategy that “resist[ed] the temptation to (only) go looking for (and finding) ethnicity’s most visible manifestations amongst those migrants (perhaps a small minority) who are predisposed to displaying and performing their putative ethnicity, often in rarefied (and reified) forms” (Fox and Jones 2013). Even though my sample was based on the national origin of research participants, I took care to also include people Wimmer (2007) refers to as “lost to the group”: for instance men and women who have no contacts with ethnic associations or NGOs, do not live in areas with a large Somali population (in London) or are married to people with a different national, ethnic or religious background.

Despite this precaution, the criteria according to which the sample was constructed (early arrival in Europe, possession of a stable legal status and – for most respondents – experiences of mobility) create a bias. While my interest in cross-border mobility practices justifies these criteria, I am aware that the people I met certainly have more economic, cultural and social capital than most Somali migrants who live in Europe.

An ethnographic approach based on different types of interviews has thus guided the research process and has aimed to create respectful and, when possible, on-going relationships with research participants, allowing for exchanges of views and for meanings to appear (Sherman Heyl 2007; Kvale 1996). In my attempt to ensure respondents’ comfort during the interviews, I mostly let them choose the moment and the place of the encounter, suggesting meeting in a public place (cafés and restaurants, for example), at their homes or anywhere they chose. A majority of interviews were carried out in various cafés, restaurants, cafeterias, McDonald’s, pubs and hotel bars. Some also took place in the interviewee’s home or in other private places such as a friend’s home (either the respondent’s friend or mine). I met a few respondents at their workplace (office or shop) and, on two occasions, while on the move: one female respondent in London took me on a walking/bus tourist tour for a few hours before we went to have dinner together, while a man had unexpected business at the time of the meeting and offered me to join him in his car and carry out the interview while he was driving to the various places where he had to pick up or drop off things. This freedom I gave research participants sometimes resulted in their deciding to include other people in the encounters, which turned into group interviews. I never opposed these additional presences and adapted the interview to the new conditions.

Most interviews were recorded, but some were not because a few respondents did not feel comfortable with being recorded. All recordings (or notes) were fully transcribed. Participation in my research was voluntary, of course, but not financially rewarded. In most cases, however, I resolved issues of reciprocity through small acts that indicated my gratitude to the respondents for their time (see also Ryen 2007): I generally paid for the coffees or drinks that we consumed in public places and brought a small gift to people’s homes (a box of chocolates or a gift for a new-born baby, for instance).

While I met the majority of respondents only once, I conducted “multiple sequential interviews” (Charmaz 2001) with a selected number of them. The data gathered with respondents I met more than once are particularly rich: through those multiple encounters (from two to five, see Fig. 1.1), I could achieve a wider and deeper understanding of their trajectories and experiences as well as the meanings they accord them. I could enter their worlds at different times, sometimes in different places, and in different ways (the different types of interviews). I selected the people with whom I met more than once for a number of reasons that are not mutually exclusive: my scholarly interest in their situation, experiences and (im)mobility practices; their willingness and motivation to participate in the research; and practical reasons such as, in the case of London, their physical presence in the country when I was there myself.

In total, 24 interviews were carried out in Switzerland with respondents who live in Geneva, Zurich, Lausanne, Neuchâtel and Sion. The Swiss sample is made up of 18 people, half women and half men. On three occasions, the interview turned into a group interview, for at least part of the time, as another person joined the discussion.

In Britain, I conducted 17 individual interviews and four group interviews – two rather long discussions with three women each, one with three men and another one with a man and a woman. Except for three respondents I met in Bradford (a city 2 h north of London), all research partners live in different parts of central and suburban London. The British sample includes a total of 19 people, of whom 12 were women and 7 were men.

All interviews have been included in the analysis and thus inform the results. However, the empirical chapters consciously focus on some people more than others. These more prominent respondents have been chosen because of the richness of “their” data (for instance, because of multiple encounters) and because their stories reflect other respondents’ practices particularly well. Since the temporal dimension of respondents’ lives constitutes an important element of my analysis, this detailed attention to particular stories allows for a fuller account of a few trajectories rather than a superficial glimpse of many.

1.2.5 Data-Analysis Methods

Data were analysed during the research process. Empirical data were analysed through two main methods: one, inspired by grounded theory, involved coding and memo writing and aimed to let categories emerge from the data at hand; the other, inspired by biographical case reconstruction, aimed to regain sight of individual life stories that transversal coding largely neglects. Relatedly, Bilge (2009) developed a “two-step hybrid approach” that combines inductive thematic analysis and a more deductive approach to her empirical data, in her case in order to reconcile grounded theory methods with an analysis of intersectionality.

The first step was inductive and based on theoretical coding. I felt it necessary to let my empirical data “speak for themselves”, and opted for an inductive type of analysis inspired by grounded theory methods (Charmaz 2001, 2006; Strauss and Corbin 2008; Glaser and Strauss 1967). The concept of “post-migration mobility practices” has been constructed through multiple phases of coding and discussing categories. Furthermore, at some stage I was able to build a typology of the mobility practices that the women and men in this study were undertaking, as well as of the activities that were taking place simultaneously. I was also able to trace relationships between mobility practices and, on the one hand, transnational engagement and, on the other, the processes of local incorporation in which people were involved. Finally, I could grasp the complex relationships between different types of mobility, and between mobility and sedentarity, at the level of the individuals themselves, but also at the collective level, i.e. between people who are (im)mobile in different ways. All this caused me to develop the idea that mobility amounts to a sort of capital in the lives of some respondents, and that it is necessarily related to other types of resources that respondents do or do not have access to. These results brought me to the need to engage with the data through a second method.

The second step of data analysis is thus a more deductive, theory-driven biographical reconstruction. There were two shortcomings of the initial, data-driven method that I needed to deal with. The first was that this transversal analysis had made me lose sight of my respondents’ individual trajectories and biographies as coherent wholes. The second was that the focus on people’s accounts of their lives left too little room for the ways in which they were related to broader social contexts and inserted in relations of power (see also Bilge 2009).

I thus engaged in a second reading of my empirical data that aimed to understand and analyse each respondent’s biography and illuminate the larger contextual and relational aspects of these coherent “stories”. To do so, I separated my material into “cases” and organised the information I had on the life of each respondent chronologically in order to obtain a (necessarily partial) biography, following a biographical case reconstruction method (Rosenthal 2004; Bertaux 2005). By doing so, I could analyse people’s “life stories” by focusing on processes of mobility/sedentarity and local anchorage that could be observed, in light of the varied resources that people could mobilise at different times in their lives. Furthermore, as other scholars have discussed (Erel 2007; Apitzsch and Siouti 2007), biographical analyses are particularly suited to attempts to understand the transnational dimension of (some) migrants’ lives. In this study, biographical analyses allowed me to delve more deeply into one of the problematic aspects I had encountered when trying to understand the practices of the people I had met: the fact that their resources were not valued equally depending on where they had acquired them or where they were mobilising them. Only through their personal accounts of their individual trajectories in general, and their mobility practices in particular, could these discrepancies become apparent and be dealt with analytically.

In this second analytical step, I therefore re-embedded my data in their larger social and historical context by reconstructing whole biographies, their evolution over time and their relationship to one another.

1.3 On Somalia and Somali Migrants: A Contextual Introduction

1.3.1 Recent History

I remember, when I was in Somalia, they used to call Somalia “paradise” at that time. Now it’s changed to hell. The Italians used to call it “paradiso”, so it was beautiful; especially Mogadishu was a very beautiful city. […] Life was nice, weather is 28 all year, you don’t need a jumper, just a t-shirt, that’s it, you know. Life was beautiful. […] My only wish is to see Somalia better, because you’ve never seen a better country in this world. If Somalia were safe, to be honest – because everything is there; all it needs is peace. No peace, no life. You can have millions but if you don’t have peace, you stay in your home; you can’t go out. What are you going to do with your money? Nothing. […] I remember the blue ocean, and the sun, so beautiful. […] I remember the beach, the beach called Lido. [That’s in Mogadishu?] In Mogadishu, yeah. I was born in Eastern Somalia, Puntland, but I grew up in Mogadishu. So the sea – we used to play there every Friday, playing, running, the beach. White, very white, and the blue ocean. You’ve seen the image anyway. That’s a proper holiday place. So we were on holidays, actually, every day. Now the beach is dirty, because of so many Western companies. They come to the beach, that’s why the pirates started. You know, in Somalia we’ve got big problems with pirates. The pirates, they started their things because the ships, they come through the Somalian sea, they dump dirty chemicals, nuclear waste, everything, their unused things, they dump there, and leave, because nobody controls the area; it’s open sea. So they don’t care. [Biographical interview, recorded]

Shariif, a man in his mid-30s who lives in London, remembers the country he left as a teenager, before the civil war: he draws a sharp contrast between the “holiday place” of his romanticised childhood memories and the “hell” that Somalia has been for the last 20 years. His description introduces many of the contextual dimensions that will be discussed in this section: the colonial past and the various foreign interests or interventions in the country, the collapse of the Somali state and the long and steady destruction of the country, the different regions that comprise Somalia (the centre and the south of the country, where Mogadishu, the capital, is located; Puntland in the northeast; and Somaliland in the northwest) and finally the recent international concerns raised by piracy and links with terrorism. The quotation also reveals the complex and sometimes ambiguous relationship that many of the respondents have kept with a country they had to leave and have sometimes not visited again, and yet with which they have often maintained various links.

The term “Somali” can refer to both the nationality of people currently internationally recognised as nationals of the Federal Republic of Somalia and the ethnicity of people who share a belief in a common Somali ancestry. However, the two definitions do not systematically overlap and are subject to conflicts, as this brief history illustrates. Although Somalis are commonly presented as a culturally homogeneous group of people who share a common language, religion and ancestry (Lewis 1994), this image does not reflect the complexity of the relationships between the different clans, regions, ethnic groups and political projects that constitute Somalia.

The Somali Republic was born in 1960 from the union of two former colonial zones that had belonged and been managed by Britain and Italy. But the Somali region covers a larger territory: in the nineteenth century, colonial powers divided it between Great Britain, Italy, France (which possessed what is now Djibouti) and Abyssinia (now Ethiopia – to which the still disputed region of Ogaden belongs) (Kleist 2004). Decolonisation in the region was intended to eventually reunite the entire Somali-speaking territory into one nation-state, but that never happened.

At the beginning of the 1960s, the new nation-state’s government was not strong enough to promote economic development and good political governance. Moreover, the concentration of all institutions in Mogadishu caused discontent in other parts of the country. The government was overthrown in 1969 by a coup that installed General Mohammed Siyad Barre as the new president. Barre, backed by his then-ally the USSR, declared “scientific socialism” and prohibited all references to clan and clan-related genealogies, while at the same time promoting people from his own family in all spheres of power (Gundel 2002; Lewis 1994). He did not abandon his nationalist dream of a big Somalia and, in 1977, invaded Ogaden, an Ethiopian Somali-speaking region, which led to a bloody war, large population displacements in the region, a change in allies, with Barre turning to the US for support, and a military defeat a year later. This “marked the beginning of an evolving crisis throughout the 1980s, which involved an economic downward spiral, political marginalisation, repression, migration, and manipulation of external aid” (Gundel 2002). The conflict escalated until, in 1988, armed opposition groups from northern Somalia (currently Somaliland) – and soon joined by southern groups – started attacking the government’s forces. Civil war spread throughout the country, but the conflict did not end when Barre was removed from power in early 1991. The opportunity to create a unity government was missed, and the various armed factions started fighting among themselves, leading to a violent conflict (Menkhaus et al. 2010). Around half a million people died during the 2-year period that followed, due to violence and the dramatic famine that struck the region; according to estimates, an additional million people left the country, while two million more were internally displaced by the end of 1992 (Kleist 2004).

While the southern and central regions of Somalia sank into instability and recurrent violence for the next 20 years (see below), Somaliland and Puntland took separate paths. In 1991, the Republic of Somaliland proclaimed its independence, and, although it has not been recognised by the international community, it has been able to create a relatively stable political apparatus and administration (Meyer 2010). Puntland was created in 1998 as a regional autonomous (federal) state: in contrast to Somaliland, it does not have nationalist aims but, because of the absence of a functioning central government, nonetheless acts de facto independently from south-central Somalia (Meyer 2010).

Despite the international intervention aimed at restoring peace between 1993 and 1995, most of Somalia was not able to re-establish a functioning government. What took place instead was the “the rise of ‘governance without government’, in which informal governance arrangements, involving various combinations of customary law, sharia courts, municipalities, business leaders, neighbourhood watch groups, and civic movements, worked to provide a modicum of law and order and services to communities” (Menkhaus et al. 2010). The collapse of the state nevertheless left the population and institutions of Somalia in a situation in which insecurity and fragility prevailed. Various peace conferences were organised with the support of the international community. In 2004, the formation of a new Transitional Federal Government (TFG) – first based in Kenya and then in Somalia – gave some impetus to hopes for peace and stability. But efforts to establish the new government in Mogadishu failed, and Somalia witnessed the emergence of radical Islamist groups and the intervention of Ethiopian military forces, leading to one of the bloodiest times in Somalia’s civil war and new massive population displacements at the end of the 2000s (Meyer 2010; Menkhaus et al. 2010). Since then, political processes have resumed, leading in 2012 to the establishment of the first permanent central government in the country since 1991 (the Federal Government of Somalia). Under the new constitution, the country became a federation, the Federal Republic of Somalia, and a period of fragile reconstruction began.

While the international community has paid limited attention to this region despite the disastrous humanitarian situation caused by war and periods of intense famine, recent links with terrorist groups and piracy seem to have increased its interest.

This short summary does not do justice to the complex history of Somalia, but it suffices as a basis to understand the heterogeneous types of relationships that the respondents in this study have developed with it over the years. Depending on their personal history and that of their families, their position before or at the beginning of the war, their movements and migration trajectories, the regions in which they lived and the clan affiliation they do or do not claim, their relationship to, and perspective on, “Somalia” can take various forms. A particularly important point in this regard is the place of origin within Somalia. While most of the people I met are originally from Mogadishu or other regions in central-south Somalia, a significant number (especially among respondents who live in Britain) are from Somaliland. The distinction is very important to some people in this latter group, who relate to an unrecognised country where economic, social and political development is a reality, and who do not want to be associated with their war-torn neighbours in the south.

1.3.2 Mobility and Migration from Somalia

Although the internal and international displacement of people fleeing the conflict has been very important over the last 30 years, migration and other types of mobility practices have been common among Somalis for a long time. Nomadic pastoralism and trade, two important livelihoods in Somalia even today, are based on circulation, which has led some authors to emphasise the role of mobility in the discursive construction of “Somaliness” (Horst 2006; Lewis 1994; Kleist 2004). Interestingly, although most of the migrants who left Somalia in the early 1990s were urban and sedentary, many still refer to “the figure of the nomad” as an important part of their identity (see also Fink-Nielsen et al. 2004; Engebrigtsen 2011). Many of the people I met in this study frame their mobility practices (secondary migration, back-and-forth movements, regular visits to other countries where they have family members) through an idealised version of nomadism. I repeatedly heard respondents explain their travelling habits through their background as pastoralists.

I do not endorse scholars’ explanations of Somali migrants’ mobility as being related to a cultural heritage of nomadic practices that are reproduced in the context of migration (see for instance Bang Nielsen 2004; Engebrigtsen 2011 for such explanations). While skills and strategies of risk diversification may have been acquired through past experiences of pastoralism, I opt for a specific theoretical stance that emphasises the political dimensions of mobility rather than their cultural embedding in a (sometimes romanticised) nomadic heritage. I agree with authors who argue that Somali migrants’ identification with the figure of the nomad is a way of framing forced exile in positive terms (Fink-Nielsen et al. 2004) and of emphasising “continuity and […] mobilis[ing] images of being adventurous, tough and independent, rather than marginalized, displaced and helpless” (Kleist 2004).

Apart from traditional practices of circulation, Somalis also have a long history of cross-border migration based on colonial and other geopolitical relationships. Seafarers working for British companies in the late nineteenth century started settling in the UK when they obtained sedentary jobs and were joined by their families during the 1950s, (Kleist 2004; Tuck 2011). Higher-class youths went abroad to Italy, Britain, Egypt and other countries that offered good educational opportunities and/or scholarships to Somali students. In the 1970s, the oil boom drew hundreds of thousands of unskilled Somali workers to the Arab Peninsula, while highly skilled Arabic-speaking Somalis could find good employment opportunities in the Gulf states (Kleist 2004).

Jamac, a man in his early fifties whom I met in London, illustrates those kinds of movements: born to a wealthy family in Mogadishu, he received a scholarship from the then European Economic Community to study in Italy. After returning to Somalia, he worked for the government for a couple of years before starting a business that imported Italian pasta to Somalia. Shortly after the war started, he used family connections to settle in one of the Emirati states, where he worked in different international trade jobs for some years. In 2002, he finally moved to Britain, where his wife and children had been living for some years.

Jamac’s brief biography indicates that mobility and migration were part of larger strategies before the war in Somalia, often based on links with former colonial powers or other economic and political partners. However, the largest movement of people, comprising several million individuals, has occurred since 1988, the beginning of the civil war. Pérouse de Montclos (2003a, b) estimated that up to four out of five individuals had to flee because of the conflict. It is difficult to know exactly how many people have moved either within Somalia or across borders because of the significant numbers who have never officially registered as refugees, preferring to remain undocumented or to count on other kinds of support than those expressly targeting forced migrants. However, the UNDP estimates that a million Somalis currently live outside the country (Sheikh and Healy 2009). To give an example of this wide dispersal, Van Hear (2005) states that by the end of the 1990s, Somali refugees had filed asylum applications in more than 60 countries. Events following the intervention of Ethiopian troops in 2006 have prompted more people to leave their place of residence: the UNHCR registered more than a million refugees as well as another million internally displaced persons for 2011–2012, making Somalis the second-largest population under its responsibility for these years (UNHCR 2012).

Most Somalis have settled in Somalia’s neighbouring countries, Kenya, Ethiopia, Djibouti and Yemen, either in refugee camps or in various cities and villages. A second group resides in the Gulf states, often living and working there as undocumented persons, while a third group has settled in industrialised countries, with large numbers of people in Britain, the US, Canada, the Netherlands, Denmark and Norway (Sheikh and Healy 2009; Pérouse de Montclos 2003b; Hammond et al. 2011).

However, mobility often does not end with settlement elsewhere, contrary to simplistic ideas of a “refugee cycle” (Lindley and Van Hear 2007). Somali migrants’ trajectories often include subsequent movements: studies have demonstrated that those who left at the beginning of the war often tried to return, left again (maybe for another place), moved back and forth between cities and refugee camps or left a first country of settlement to find better conditions in another one (Moret et al. 2005; Zimmermann 2009; Horst 2006). Return is another type of movement that shapes the lives of some migrants, particularly those who originally came from the more stable Somaliland. Whether returning from neighbouring countries’ refugee camps in the context of UNHCR-led programmes of “voluntary repatriation” or from industrialised countries to settle permanently or for regular visits, the number of “returnees” is significant (Hansen 2007, 2008; Fink-Nielsen et al. 2004; Kleist 2007). Other migrants decide to enjoy the citizenship they have acquired in the European or North American country where they had settled and legally move to a new country where they feel they can benefit from a more advantageous situation: Al-Sharmani (2004, 2006, 2010) has studied families that moved from the US to settle in Cairo and acquire a higher social status, while several authors have observed the secondary moves performed by European Somalis from various Continental European countries to the UK (Van Liempt 2011a, b; Bang Nielsen 2004; Lindley and Van Hear 2007; Ahrens et al. 2014). These different types of movements point to the concrete reality of what I refer to as “post-migration mobility practices”: Somali migrants’ migration trajectories are complex, often comprising consecutive moves in different directions.

It has become clear now that migration shaped the lives of many Somalis, even before the outbreak of the civil war in the late 1980s, but it has done so in considerably larger numbers since then. However, it should be equally clear that Somali migrants do not constitute a homogenous category of people. Their varying endowments of economic, cultural and social capital have been important in determining when they could leave Somalia and where they could go. As Jamac’s example above illustrates, the education he received in Europe and his transnational kin networks allowed him to leave Somalia when he felt the need to and to find employment and security in the Arabian Peninsula, only to later move to Europe once again.

1.3.3 Political, Legal and Discursive Contexts: Britain and Switzerland

Although Britain and Italy had previous colonial links with Somalia, it was only in the early 1990s that European states started to witness the arrival of, and to host, large numbers of Somali migrants. Except for those who had arrived before the war as students or workers, the vast majority of Somali migrants arrived in Europe through asylum channels. Over the years, family members joined them through family-reunification schemes, and Somalis now constitute one of the most important groups of recent non-European migrants in Europe.

Different waves of Somali migrants can be distinguished – the first migrants who arrived for different reasons before the war (in Britain more than in other European countries), the wave of refugees who left the country early in the conflict, and people who have arrived more recently (Fangen 2007; Eyer and Schweizer 2010). Most of the respondents in this study belong to the second group. Having arrived in the 1990s, they have obtained a stable legal status, in most cases have been naturalised and have more or less successfully integrated into their country of residence.

Some European countries have attracted more Somali migrants than others over the years. Together with the UK and Italy, the Netherlands, Norway, Sweden, Denmark and Germany are the main host countries, each hosting more than 10,000 Somali migrants (Hammond et al. 2011 based on unpublished estimates by Carling and his colleagues). According to the same estimates, Switzerland, France and Finland are home to between 1000 and 10,000 Somali migrants each, while smaller numbers are found in other European countries. In the early 2000s, the first reports emerged about Somali migrants being perceived by the local media, policymakers and the population as causing problems and being poorly integrated (Farah 2000; Pérouse de Montclos 2003a, b). Difficulties in entering the labour market, high rates of unemployment and reliance on social welfare and other specific “social problems” such as the consumption of khat (a herbal drug), high divorce rates, youths’ low educational achievement, continuing practices of female genital mutilation and – in some places – their appearance in criminality statistics and increasing interest and participation in religious fundamentalist groups, have been noted in more recent reports about this population (Pérouse de Montclos 2003a; Hammond 2013; Eyer and Schweizer 2010; Harris 2004; ICAR 2007; Open Society Foundations 2014).

All of these studies, however, demonstrate that, despite some truth to the claims outlined above, the “Somali population” in Europe is highly diverse and cannot be summarised by reference to a few problematic aspects. Its heterogeneity comes from dimensions related to gender, age, legal status, educational background, history of migration and mobility.

The women and men in this study have settled in either Britain or Switzerland (or, in a couple of cases, in both). Although the region, city, village or neighbourhood in which they reside also greatly influences their lives, this chapter focuses on the national contextual level of their “main country of residence”. Britain and Switzerland’s historical relationship with the rest of the world and with migrants, their respective construction of “otherness” and their immigration policies and perspectives on citizenship have an impact on respondents’ experiences and on the meanings they give to those experiences. The strategies Somali migrants develop, in particular when it comes to (im)mobility, need to be understood against these specific backgrounds.

1.3.3.1 The British Context

Migration from Somalia to Britain has a long history: seamen from British Somaliland employed by British merchants at the end of the nineteenth century and Somali industrial workers after the Second World War were among the first migrants from the region to settle in the UK, later bringing their families as well. These early links led to the arrival of workers, students and, since the late 1980s, significant numbers of refugees, making Britain the European country with the largest Somali population (Hammond et al. 2011) and Somalis one of the largest ethnic minorities there. Their actual numbers are difficult to estimate, because of the large number of naturalised people and those who have moved from other EU countries to settle in Britain. Estimates range between 95,000 and 250,000 (Hammond 2013). Census figures from a decade ago indicate that almost 90% of them lived in London, but recent refugee-dispersal policies have probably led to a wider dispersion across the country: Bristol, Manchester, Birmingham and Leicester, among other cities, now host large Somali populations (Hammond et al. 2011). Hopkins (2010) estimates that around 60,000 Somalis currently live in London.

Somalis who arrived as asylum seekers since 1988 have generally been granted either full refugee status (with an indefinite leave to remain and access to similar civil rights and duties as UK nationals) or a provisional admission (exceptional leave to remain before 2003, humanitarian protection of discretionary leave since then, with more limited rights, in particular regarding family reunification and travel documents) (Lulling 2007; Harris 2004). It is also difficult to find figures that break down the Somali population by legal status, but statistical evidence (Office for National Statistics 2013) and the comparatively generous naturalisation regime that prevailed until 2005 (Sawyer and Wray 2012) suggest that a majority of all people of Somali origin who live in Britain have become British citizens. Almost all of the people I met in my study who live in Britain arrived there as asylum seekers or through family-reunification schemes and have become British citizens.

Official British figures include Somalilanders among Somalis, despite the recent claim, by some of them, that they constitute a distinct group, as some of my respondents in London told me. Because of the historical connection with what was once British Somaliland, migrants from this region constitute a significant share of the Somali population, yet it is difficult to determine whether they constitute a majority.

More recently, an important wave of secondary migration has occurred that consists of Somali migrants who had settled in Continental European countries – and became citizens of their country of residence – who subsequently moved to Britain: the number of these European citizens enjoying the right to free mobility within the EU is unknown, but various studies indicate that they are far from negligible (Van Liempt 2011a, b; Bang Nielsen 2004; Lindley and Van Hear 2007; Ahrens et al. 2014). These “European Somalis”, as they call themselves, are part of my British sample, as I have met women and men who live in different parts of London with Dutch, Danish, French or Norwegian passports.

The imperial and colonial background of Britain’s “traditional” migration flows has led to a “racial” perspective on diversity (based on the divide between “whites” and “Blacks”). However, what had appeared as a natural delineation between “races” has become more complicated in recent years, in particular because of the differing socioeconomic paths of the varied populations initially considered “Black” (Alexander 2002), the immense diversity of the migrant population in terms of countries of origin and other characteristics (Vertovec 2007) and the emergence of Muslims as a relevant category of differentiation (Modood and Salt 2013b).

Furthermore, “the gradual recognition that [migrants] were not temporary migrants but settlers and fellow citizens led to their being conceived of as ‘ethnic minorities’”(Modood and Salt 2013a), a notion that doesn’t exist in Swiss policies, for instance. The accommodation of these “ethnic minorities”, i.e. not only the migrants, but also their descendants, in Britain is strongly based on the sociological and political idea of multiculturalism. Policies of multiculturalism aim to foster equality for all individuals by focusing on the rights of ethnic minority groups – conceptualised as “communities” – to maintain their cultures of origin, based on the idea that solidifying bonds based on common ethnicity would then lead to bonds between communities. Since the early 2000s, discourses about the “failure of multiculturalism” have become increasingly prominent (Werbner 2009), and new policies have been designed to overcome the supposedly “problematic integration” of migrants and ethnic minorities. As in Switzerland and most other European countries, in recent years integration policies have increasingly focused on new requirements for migrants to “prove” their desire to belong, in the form of language and country-knowledge requirements in particular (Goodman 2010; Tuck 2011).

Public opinion towards migrants and ethnic minorities in general, and people perceived as “culturally different” in particular, has become more negative (Park et al. 2012). Descendants of migrants as well as recently arrived migrants are increasingly perceived as poorly integrated, self-segregated from the “majority” and a danger to social cohesion and national security (Alexander et al. 2007). Furthermore, like elsewhere in Europe, asylum seekers are increasingly subject to negative stereotyping in Britain. Asylum has become a politicised issue. The local population, which often confuses asylum seekers and refugees, associates both with disguised economic migrants who steal jobs and benefits from them and receive preferential treatment in housing and other public services (Aspinall and Watters 2010). Furthermore, since 9/11 and 7/7, Muslims have increasingly become a heavily stigmatised “suspect community” (Ryan 2011; Ryan et al. 2011). For instance, different polls show that the British population perceives Muslims negatively, with a majority tending to believe that they are a threat and do not make an effort to integrate and create problems, or they associate them with extremism (Moosavi 2012). This negative framing is strongly gendered, since Muslim women, in particular those who wear a headscarf or a hijab, are considered either victims or people who actively refuse to integrate (Ryan 2011).

In this context, Somalis are presented, in the media as well as in public and political discourses, under the rubric of extremist religious practices: terrorism, involvement in Jihad and female genital mutilation are typically associated with the Somali community. They suffer from negative stereotyping on multiple grounds: these perceptions may lead to discrimination, racism and exclusion practices.

1.3.3.2 The Swiss Context

Switzerland has also experienced the arrival of different waves of Somali migrants, although nowhere nearly as large or diversified as in the UK. Because Switzerland has no colonial links with any part of Somalia, the arrival of migrants from this region is rather recent: very few Somalis lived in Switzerland before the 1990s, and most who reside in the country have arrived since then, mostly as asylum seekers (Eyer and Schweizer 2010; Moret 2006).

More than 10,000 migrants of Somali origin are estimated to be living in Switzerland, and there were 8326 Somali nationals officially registered in the country in 2016. We can add to this number the 1623 Swiss citizens who were born in Somalia. These statistics do not include the naturalised children of migrants, who were born in Switzerland. Among non-naturalised individuals, just under half have an annual (B) or permanent (C) residence permit, while nearly 40% hold a provisional admission (F Permit) and 15% are still in the asylum process (N Permit for asylum seekers).Footnote 4 Furthermore, while Britain has long attracted Somali migrants and refugees, many women and men who arrived, sometimes by chance or accident, to Switzerland, eventually decided to leave the country and try their luck elsewhere, either legally or through irregular means (Moret 2006).

Although no official numbers exist, my impression is that only a minority of the Somali migrants in Switzerland comes from Somaliland. In my sample, only two people identified as Somalilanders, and I have found no organised group claiming this identity.

In contrast to the situation in many European states, in most cases Switzerland has not granted Somali asylum seekers full refugee status: however, having acknowledged their need for international protection, Swiss authorities have given them a subsidiary form of protection called “provisional admission” (F Permit), a permit they must renew every year (Moret 2006; Eyer and Schweizer 2010). Although the policies have changed in recent years, respondents in this study have experienced provisional admission when it was highly restrictive, in particular with regard to access to the labour market and rights to family reunification, social assistance (which is the same as for asylum seekers), choice of place of residence within Switzerland and cross-border mobility (prohibited except in very rare cases) (Moret 2006; Kamm et al. 2003). Provisional admission, despite its name, is often a long-term status, as people may remain subject to it for many years before it is either revoked or transformed into a better status.

Furthermore, in comparison with other European countries, Swiss laws and practices regarding naturalisation are particularly restrictive.Footnote 5 Based on jus sanguinis, the new Swiss Citizenship Act (which entered into force in 2018) states that foreign citizens may apply for Swiss citizenship after 10 years of residence in the country (the years between the ages of 8 and 18 count double), provided that they hold a permanent residence permit and can prove, to the satisfaction of the authorities, that they have become well integrated (Kurt 2017). Under these conditions, for most Somali migrants in Switzerland, acquiring a stable legal status and/or becoming naturalised has been a long and difficult process.

The political accommodation of diversity has been a different story in Switzerland than in Britain. Although it has de facto been a country of immigration for decades, it has been reluctant to admit it. Migrants – usually referred to as “foreigners” – are commonly considered a threat to Swiss identity, resulting in strong boundaries against them (Dahinden et al. 2014). This fear of “over-foreignisation” (Überfremdung in German) has influenced Swiss policies regarding the entry and incorporation of those migrants who have settled in the country. Based on the idea of cultural incompatibilities between migrants and the Swiss population, policies were designed with the explicit aim of assimilating those foreigners who stay (ibid). The “assimilation paradigm”, which was particularly dominant in the 1960s, gave way to an “integration paradigm” in the 1990s: the host country was perceived as holding part of the responsibility for promoting migrants’ integration and was to financially commit to it (Niederberger 2005; Duemmler 2015). However, the ideal of civic-integration policies that has spread throughout Europe (Goodman 2010) has recently reached Switzerland: this ideal focuses on migrants’ deficits and requires them to make the effort to integrate fully (through measures such as integration contracts or language and local knowledge tests), thereby reintroducing old assimilationist values (Duemmler 2015).

Switzerland’s historical construction of the “threatening others” is thus different from that of Britain. It explicitly builds on nationality rather than race or ethnicity. Yet, otherness in Switzerland also builds on racial representations, even though they are mostly hidden behind ideas perceived to be more neutral and legitimate, i.e. nationality and cultural difference (Lavanchy 2015; Michel 2015). Studies have also indicated that overt forms of colour-based racism and discrimination target Black people in general, and sub-Saharan Africans in particular (Fröhlicher-Stines and Mennel 2004; Efionayi-Mäder and Ruedin 2017).

As in Britain, religion has also become increasingly important as a marker of difference. Muslims increasingly generate anxiety among the Swiss population, which perceives them as a threat to the state’s fundamental values and gender equality (Helbling 2010; Duemmler et al. 2010; Fischer and Dahinden 2016). Populist parties also actively fuel fears of the supposed “Islamisation” of the country by designing initiatives to ban Islamic symbols (such as minarets in 2010, or the burqa in the Canton of Tessin in 2012). Finally, the politicisation of asylum over the last 30 years has strongly influenced how asylum seekers and refugees are perceived and treated: migrants who came to the country as asylum seekers and refugees are increasingly associated with profiteers, bogus refugees and criminals (active, in particular, in drug dealing).

Faysal, a man who has lived in Switzerland for 20 years, feels the deteriorating general climate with regard to attitudes towards foreigners. He volunteers at a nursing home and recalls having recently heard xenophobic comments by some of the residents, who, in his perception, repeat what they hear from others around them. He further relates these comments to recent anti-immigrant initiatives, demonstrating how having a Muslim name and being Black amount to a double difficulty when it comes to exclusionary and discriminatory practices. Like other respondents, he recalls instances of difficult border crossings in European airports, even with a Swiss passport:

I’ve started feeling those things since September 11. Because my name is [Faysal], the last few times I’ve travelled I’ve been checked when I returned […]. “Sir, please step aside”. They took my passport, my ID. But there are also Yugoslavs who have Swiss nationality, but they don’t look at them; they only look at your name. If you are Mohamed or Ahmed, all that, and then your colour. And we [Somalis], we have both together! […] Anyway, always, always you will be a foreigner, and that is a little difficult. [Semi-structure interview, recorded]

In contrast to Britain, where they are targeted as a “difficult” ethnic or national group, Somalis in Switzerland do not garner much attention from the media, politicians or the wider population (Eyer and Schweizer 2010): probably because of the relatively small size of their population vis-à-vis other migrant groups, they remain relatively invisible. However, despite their rarely being subject to collective stigmatisation based on their nationality, they cumulate other external ascriptions that may be difficult to handle, as Faysal and other respondents expressed.

1.3.3.3 Concluding Remarks: On Somali Migrants Who Live in Britain and Switzerland

The respondents in this study are not totally representative of the British- or Swiss-based Somali population. The women and men I have met mostly arrived in Europe 10–20 years ago: they belong to a privileged subsection of the population that were able to leave Somalia relatively shortly after the conflict started. They or their families belonged to the economic elite (for instance, traders and businesspeople), were members of the government or civil servants in the administration or were other professionals with a high socioeconomic status (professors, chemists, engineers, doctors and so on). As with most refugees, migration has meant the loss of many of their assets (savings, property, sources of income) and of their socioeconomic status through the non-recognition of most of their education and work experience. Despite the decline in their social status after their initial move to Europe, the socioeconomic position that was theirs before they left has influenced their later trajectories of migration and incorporation, as will be made clear later in this book.

Furthermore, because the people I met arrived in Europe in the 1990s or 2000s, they, unlike more recently arrived Somali migrants, have been able to obtain a stable legal status (usually citizenship of a European country) and benefit from less restrictive asylum, integration and naturalisation policies and a climate that was less hostile towards migrants.

In sum, the respondents in this study certainly occupy better legal and socioeconomic positions than the average Somali in their respective country of settlement. Yet they are still relatively unprivileged vis-à-vis the population as a whole, as they have had their access to many resources severely restricted since their arrival in Europe, for different reasons.

At the same time, several studies on Somali migrants in different places in Europe have demonstrated the significant transnational involvement of this population (again as a general trend). In particular, they are known as “exceptional remitters” in comparison to other migrant groups, in terms of both the proportion of migrants who send money and the regularity with which they do so (Carling et al. 2012): they send individual or collective remittances to different parts of Somalia or other places where family members live (see also Horst 2006, 2007; Lindley 2010; Pérouse de Montclos 2003a).

Like other migrants, many European Somalis are also actively committed to political transformation in their country of origin and engage in activities aimed at relief and development there. They create transnational associations, develop and implement projects to improve the situation there or try to get hired by international organisations or NGOs active in their region of origin (Kleist 2008a, 2008b; Hammond 2013; Horst 2017). Many of the people I met in this study emphasised their wish to contribute, participate in the reconstruction of their country and support those who have not have the chance to leave. Some were actively planning such activities, while others were already fully involved in political, humanitarian or development projects in their country of origin.

Migrants’ transnational commitment, however, has raised the interest of authorities in their countries of residence, which take a dim view on migrants’ focus on their place of origin. Against evidence that transnational commitments can often only be made by those who are solidly incorporated in the country of residence (see Hammond 2013 for the case of Somalis in Britain), there is fear that transnationality comes at the expense of local integration. Somalis’ secondary movements within Europe, and general patterns of transnationality and “supermobility” (Tuck 2011) among this group, have recently raised some concerns among policymakers who feel that these practices are not consistent with local incorporation and loyalty to the country of immigration.