1.1 Introduction

The global map of migration continuously evolves, sometimes with a steady rhythm whereby existing patterns are reinforced or slightly changed, at other times punctuated by sudden, epochal shifts. Several key events over the last two decades, such as the 2008 financial crisis, the so-called refugee crisis in 2015–2016 and, latterly, the Covid-19 pandemic, have fundamentally reconfigured previous human mobility dynamics. For the UK, the loss of ‘free movement’ as a result of the Brexit referendum is having more regionally specific impacts on migration flows. In addition, we will likely feel the impacts of other enduring challenges that will further diversify and fragment global mobility and migration patterns, such as the worrying spread of xenophobic and racist attitudes over recent years or the longer-term impact of climate change. The vilification of immigrants has led to travel restrictions for certain ethnic groups as well as an increase in harassment of and attacks on visible minorities. The effects of global warming and associated extreme weather events on migration and mobility are less easy to predict but will surely be made manifest. After the recent COP26 meeting in Glasgow in November 2021, the signs are already there.

Many of these unsettling processes have been framed as crises, both real and perceived, and thereby also offer an opportunity for critical reflection (Collyer & King, 2016). This has led to the questioning of the relevance and appropriateness of certain categorisations and conceptualisations which are commonly applied in the study of migration. Who is considered a migrant? What are origin countries, transit countries and destination countries? Who gets to decide which migration flows are to be stimulated and which are to be restricted or even banned? Why do migrants stay and why do they leave?

Estimates that predate these recent epochal events have already indicated that between 20 and 50 per cent of immigrants leave their destination country within 5 years of arrival (OECD, 2008). This means that a high proportion of migrants depart from their destination country in order to either (temporarily) return to their origin country or to onward migrate to a new destination country. There is an established literature on return migration which has provided new insights into the motivations and experiences of returning migrants and their various forms of ‘return’.Footnote 1 In contrast, onward migration remains a relatively under-researched migration trajectory. We argue that this is because migration has generally been conceptualised as a bipolar process between an origin country and a single destination country. The (possible or actual) moves to any further destinations have, therefore, often been disregarded.

Onward migrants have experience of staying in two or more destination countries for extended periods of time. Their trajectories are excellent illustrations of the fact that migration decisions are not one-off, single-dimension impulses taken at a given moment in time. Rather, they are, as Erdal et al. eloquently express in the opening paragraph of the next chapter, processual and relational processes which evolve over time. In parallel to this, migration journeys are found to be fragmented, incomplete, open-ended and often unplanned (Collyer, 2007; Crawley & Jones, 2021). The linked notions of a complex migration trajectory (Erdal et al. Chap. 2) or a migratory career (Martiniello & Rea, 2014) both look at how people navigate a range of potential pathways that are ongoing and unfinished. To reinforce this crucial point, we can do no better than to quote from Jung’s Chap. 8 in this book:

A growing number of scholarly accounts call for a consideration of the fluidity and non-linearity of migration processes … and the dynamic nature of migration intentions and decisions… Due to structural factors such as immigration laws and economic opportunities, migration has become increasingly complex, often including long and perilous journeys, transiting through a variety of countries and regions and consisting of phases of mobility and immobility, emplacement and displacement. In this context a clear distinction between transit and destination country becomes less evident and the origin–destination model reveals itself as increasingly unable to capture migration strategies and the complexity of migration processes.

This extract lays a convincing foundation for our focus on onward migration in this book. It also prefigures an acknowledgement of the importance of time in the sequencing of moves and stays. In particular, we need to appreciate how linear, chronological time intersects with life-stage events – both planned and unexpected – and with other more-structural changes, like the aforementioned global economic crisis or geopolitical events such as a refugee-producing civil war or the migration-loaded decision of the UK to leave the EU.

Yet onward migration is difficult to quantify, because current migration statistics and population register data only offer an incomplete picture. Large-scale surveys like the Migrations between Africa and Europe (MAFE) longitudinal survey or the Swiss-based NCCR Migration–Mobility Survey provide some more detailed insights into the prevalence and patterns of onward migration, respectively amongst specific African migrant groups or in particular destination countries like Switzerland (Toma & Castagnone, 2015; Zufferey, 2019). In addition, there are several recent empirical studies which suggest that onward migration is a fairly common migration trajectory amongst a wide range of migrant categories and in different geographical regions (see, for instance, Ahrens et al., 2016; Della Puppa et al., 2021; Paul & Yeoh, 2021).

In public and media discourses about highly mobile individuals – who might or might not include the specific category of onward migrants – it is usually cosmopolitan businesspeople from the Global North who dominate imaginations (Hannerz, 1990). At least before the Covid-19 pandemic, these elites were seen as jet-setting around the world for business meetings and moving from country to country at will and without restrictions (Beaverstock, 2005; Ossman, 2013). In her research with working-class Filipino and Indonesian migrants, Anju Mary Paul (2017) challenges this notion that cosmopolitan knowledge of the world is the preserve of elites. Paul shows how her ‘capital-constrained’ research participants undertake strategic stepwise migrations as migrant domestic workers to Singapore, Hong Kong, Italy or Cyprus which enable them to accumulate further skills, knowledge and resources, so that they can reach Canada and the US as their ultimate favoured destinations.

Several studies have been carried out on the onward migration of asylum-seekers and refugee populations. Seeing that asylum-seekers usually cannot freely decide which country they flee to in search of protection and many European receiving countries disperse asylum-seekers to remote parts of the country, onward migration allows them to move to a place of their own choosing (Lindley & van Hear, 2007; van Liempt, 2011a, b; Kelly, 2013; de Hoon et al., 2020; see also de Hoon and van Liempt, and Serra Mingot, respectively Chaps. 3 and 6 in this volume). Another example is the United Nations High Commissioner for Refugees (UNHCR) resettlement programme, where onward migration is used as one of the ‘durable solutions’Footnote 2 that are open to a small proportion of refugees who face protracted displacement and cannot return to their origin country. Some resettlement refugees also engage in further internal secondary or onward movements after arriving in their new country of residence (Ott, 2011).

Furthermore, migrant families often establish multi-sited households and their geographical configuration can evolve over time and during their migration project (Siu, 2005; Das Gupta, 2015; Moret, 2018; Ramos, 2018; Jolivet, 2020). The multiple moves of different family members can occur in a temporarily lagged manner, often in response to migration restrictions or individual needs at different life-course stages (Ahrens, 2022; Carling & Erdal, 2014; Ramos, 2018). Family migration can also involve complex movements to different destination countries that span across generations (Bhachu, 1985; Voigt-Graf, 2004; Francisco-Menchavez, 2020). These circumstances, too, are illustrated by many of the chapters that follow.

What the aforementioned examples show is that, even though all these migrants live in multiple destination countries, their trajectories comprise different configurations of migrations and mobilities that change and evolve over time, often across generations. Contextual factors that further pattern these mobilities include the acquisition of citizenship and enhanced mobility rights (Lindley & van Hear, 2007; Ahrens et al., 2016; Ortensi & Barbiano di Belgiojoso, 2018; de Hoon et al., 2020), unemployment due to economic crises (Ahrens, 2013; Mas Giralt, 2017; Della Puppa, 2018; Ramos, 2018), experiences of discrimination and racism (Kelly, 2013; Das Gupta, 2015; Ahrens et al., 2016) and the expectation of better opportunities in another country (Kelly & Hedman, 2016). Furthermore, the lived experiences of onward migrants vary greatly, often dependent on the migrants’ legal status, socio-economic background, educational level, gender, ‘race’ and ethnicity. It may be even safe to say that hardly any of these migrants would label themselves as an ‘onward migrant’. So what value is there in conceptualising this migration process as onward migration?

The overall aim of this introductory chapter is to examine how onward migration and transnational mobilities are interconnected. For this purpose, we draw on the empirical chapters in this edited volume, as well as on other published research on onward migration.

We start off with a definition of onward migration. We then introduce the term ‘multi-sited transnationalism’ with the specific aim of broadening the analysis of migrant transnationalism beyond its common bipolar framing – restricted to the origin country and one destination country. Given that onward migration is an unfolding and open-ended migration trajectory, this also calls, by extension, for a broadened multi-sited (and multi-directional) conceptualisation of transnational mobilities, ties and practices. Subsequently, we examine how transnationalism can shape onward migration intentions and experiences. For instance, short visits to new destinations can enable individuals to better prepare for onward migration. Reciprocally, we explore how different onward migration patterns may result in varying forms and directions of transnationalism. Onward migrants, for example, may not only stay in touch with friends and relatives in their origin country but also maintain transnational ties with individuals in a previous country of residence or other destinations. Thereby we draw attention to the role of onward migration in shaping complex global mobility patterns and transnational links. Finally, this chapter offers some methodological reflections on how to research onward migration in a multi-sited transnational context.

1.2 Defining Onward Migration

Onward migration can be defined as a spatial trajectory that involves extended stays in two or more destination countries. Acknowledging that any migrant can be a potential onward (or return) migrant allows for a more open-ended and processual understanding of migration. After living in one destination country, migrants may decide to move to one or more new destinations. Countries and places thus can change from being destinations to becoming points of departure.

Furthermore, this longer-term view of migration processes over the lifecourse of individual migrants or even across generations of a migrant family, encourages us to question how migrants are categorised (as students, asylum-seekers, labour migrants, trailing family members, etc.). When onward migrants move through various stages of mobility and settlement in different destination countries, the motivations for their moves change, as do the categorisations that are attached to these migrants. Therefore, it is necessary to envisage a more dynamic understanding of migration categories that takes into account the fluidity (and potential multiplicity) of migrant categories and the social positions that migrants inhabit over time and throughout their ‘migration project’ or ‘migration career’ (Martiniello & Rea, 2014).

Despite growing scholarly interest in recent years in trajectories that span multiple countries, we would like to stress that complex multinational migrations are nothing new. Onward migration has always formed part of a scholarly understanding of global mobility patterns. Early theorisations of migration by Ravenstein (1885, 1889) and Mabogunje (1970) already described migration processes that occurred in stages along particular pathways – from smaller villages, via towns, to the bigger cities. Writing at the time of large-scale population movement following the Industrial Revolution, Ravenstein took the example of Irish migrants living in London. He argued that they were unlikely to have travelled directly from their Irish rural community or hometown but, instead, first arrived in port cities like Liverpool and then worked in several other parts of England before settling in the British capital.

At the same time, it is equally important to highlight that (onward) migrants are not necessarily constantly ‘on the move’, nor should we assume that they want to be. The moves of onward migrants have been compared to those of footloose ‘cosmopolitans’ or ‘nomads’ (Ossman, 2013) but they nevertheless often build deep and meaningful connections with the places and people where they live. Following Hägerstrand’s insightful definition of migration as a change to an individual’s ‘centre of gravity’ (Hägerstrand, 1957, 27), one may argue that onward migrants can have multiple centres of gravity over time and throughout their lives. Thus, it is important to study the spatial mobility of onward migrants – alongside the desires and strategies of non-migrants (and one-step migrants) to ‘stay in place’ – as another possible response to changing contextual and life circumstances (Schewel, 2020).

We propose onward migration as an overarching concept that can bring together the growing literature on this topic. This literature has remained rather disparate, partly due to the many terms used to label this type of migration. Such terms include transit migration (Collyer, 2007; Schapendonk, 2013, 2020), fragmented migration (Collyer & de Haas, 2012), secondary movement (Ott, 2011), stepwise migration (Konadu-Agyemang, 1999; Paul, 2017; Francisco-Menchavez, 2020), post-migration movement (Moret, 2018), multinational migration (Paul & Yeoh, 2021), twice migration (Bhachu, 1985; Das Gupta, 2005, 2021; Della Puppa & King, 2019), triangular migration (DeVoretz et al., 2003), sequential migration (Faggian et al., 2006), third-country migration (Tan & Hugo, 2017) and serial migration (Ossman, 2013; Zufferey, 2019). Along with others (e.g. Lindley & van Hear, 2007; Ahrens et al., 2016; Ortensi & Barbiano di Belgiojoso, 2018; Della Puppa et al., 2021), we privilege onward migration as a clear and unambiguous term and build this book around this claim.Footnote 3 Moreover, we believe that onward migration is also more comprehensive than the other concepts that describe particular types of spatial trajectories of specific migrant categories. Importantly, onward migration implies a more open-ended trajectory because it does not limit the migration project to a certain number of moves (i.e. twice migration) or indeed suggest continuous relocations (i.e. serial migration).

1.3 Multi-sited Transnationalism

Transnationalism is part and parcel of globalisation; at the same time this dual theoretical lens offers distinct insights. During the latter years of the twentieth century and throughout the early twenty-first century, globalisation has given rise to new social formations and networks constituted by rapidly developing information and communication technologies, which have resulted in more interactions and interconnections between different countries and world regions. Meanwhile, transnationalism is concerned with the cross-border connections and practices of individuals, communities and institutions that transcend the boundaries of the nation-state (Glick Schiller et al., 1992; Basch et al., 1994). In this edited book, we focus on migrant transnationalism, sometimes referred to as ‘transnationalism from below’ (Smith & Guarnizo, 1998).

One of the key contributions of the pioneering literature on transnationalism during the 1990s was to challenge the ‘sedentary bias’ that tended to dominate the social sciences, including migration research. Glick Schiller et al. (1992) argued that, after leaving their origin country, transmigrants can become firmly ‘rooted’ in a destination country and, by splitting time and resources, can simultaneously maintain multiple linkages to their origin country. The cross-border activities and relationships of migrants can be cultural, social, political, economic, etc. Furthermore, the linkages which migrants forge are not fixed but are continuously evolving and ‘becoming’ – meaning that they can wax, wane or fade away completely over time. These engagements across international borders can mean that migrants also develop dual, hybrid or conflicting identities and senses of belonging.

By highlighting migrants’ cross-border connections, transnationalism also intends to overcome ‘methodological nationalism’, an implicit theoretical stance which took the nation state as a self-evident unit for the analysis of social and historical processes. In their seminal book Nations Unbound, Basch et al. (1994, 7, emphasis added) state that transmigrants ‘take actions, make decisions and develop subjectivities and identities embedded in networks of relationships that connect them to two or more nation-states’. Nevertheless, empirical research on transnationalism has tended to focus on a particular destination country whilst giving some consideration to the transnational moves, connections and identifications with the migrant origin country. In so doing, such research has inadvertently fallen into the trap of swapping methodological nationalism for ‘methodological bi-nationalism’ (Sperling, 2014), which does not adequately reflect the complex lifeworlds of (onward) migrants.

One further shortcoming of theorisations of transnationalism is the assumption that the transnational ties and mobilities of migrants by default are solely directed at their origin country. This supposed uni-directionality of transnational ties can arguably be linked to understandings of diasporic communities, which have traditionally been defined by their strong ‘homeland orientation’ (see the critique by Brubaker, 2005). More recent studies offer a more diversified picture of transnational ties and activities. Mazzucato (2011), for instance, highlights the ‘reverse remittances’ that non-migrants in Ghana send to migrants abroad in order to support their journeys to and the initial arrival phase in Germany. Similarly, research about the second generation has identified the ‘inter-destination transnationalism’ that connects young Latinos in Europe with cousins and other family members living in the US (Sperling, 2014). Based on empirical work conducted by the contributing authors in this edited volume, we conceptualise transnationality as potentially both multi-directional and multi-national, reflecting the staged trajectories of onward migrants.

Thus, we propose an explicitly multi-sited conceptualisation of transnationalism because it helps to facilitate a more comprehensive understanding of transnational mobilities and ties. We thereby follow calls for a broader definition of transnationalism that captures the ‘multiple ties and interactions linking people, organisations or institutions across the borders of nation-states’ (Vertovec, 1999, 447). Our conceptualisation of multi-sited transnationalism contributes to the two schools of thought within the research on transnationalism in different ways (cf. Itzigsohn et al., 1999). For scholars who subscribe to the ‘social field’ interpretation of transnationalism (Levitt, 2001; Levitt & Glick Schiller, 2004), focused on creating communities that transcend borders, multi-sited transnationalism could be interpreted as an expansion of the transnational field to include multiple and possibly shifting centres. While, for those scholars who favour a practice-oriented interpretation of transnationalism (e.g. Guarnizo et al., 2003), focused on specific transnational ties and actions, multi-sited transnationalism can be understood as nodes of contact in different destinations which mirror the spatial configurations of social networks – either at the individual or the community level.

There are many ways in which this preliminary discussion on multi-sited transnationalism could be extended. One obvious implication is methodological: how do we study migrants who are repeatedly ‘on the move’ through a sequence of origins and destinations which are themselves often reversed, i.e. when destinations become origins and vice versa? We reflect on this in the final section of the chapter and invite readers to examine the methodological foundations and field techniques of the ten chapters that follow. For now, we briefly take the discussion forward on a more theoretical and epistemological plane.

First, it is clear that onward migration, along with return migration, re-emigration and circular, transit and other forms of migration and mobility, complicates the single-origin-single-destination approach to conceptualising migration and transnationalism; however, in what way do such complex migration trajectories ‘speak back’ to migration theory? A recent paper by Paul and Yeoh (2021) gives some useful indications, shifting the debate from the standard bifocal lens of transnationalism – ‘here’ and ‘there’ – to underscore the complexity of what they term ‘multinational migrations’, which include onward migration. This complexity has many dimensions, starting from the inherent spatio-temporal variety of multiform migrations. Onward migration takes place for a variety of reasons, not just economic improvement; such motives may be mixed and change over time, with each move and according to the migrant’s lifecourse circumstances. Onward migration comprises migrants of different social classes, occupations and legal statuses and not just cosmopolitan elites. The final target destination may be planned from the outset – but only be attainable via intermediate stages – or the migration trajectory may take a more organic, evolutionary form or be based on pure serendipity and random events. As the chapters of this book illustrate, migrants may or may not move up through a hierarchy of increasingly ‘desirable’ destinations; ‘lateral’ or even ‘downward’ moves may occur due to forces majeures or to strategic, intermediate decisions (see, especially, Chap. 8 by Jung on Senegalese migrations for a good illustration of this). Jung’s chapter is in line with a recent paper by Schapendonk (2021) which critiques the ‘grand narrative’ of onward migration which frames it as a staged process involving a South-North directionality and a progression up the ‘hierarchy’ of targeted destinations.

A second key point about multinational and onward migrations made by Paul and Yeoh (2021, 6–8) is that they are increasingly shaped by what Xiang and Lindquist (2014) have termed ‘the infrastructures of migration’. Defined as ‘the systematically interlinked technologies, institutions, and actors that facilitate and condition [migration and] mobility’ (Xiang & Lindquist, 2014, S124), migration infrastructures comprise five dimensions: commercial (recruitment agencies, smugglers etc.), regulatory (state and other apparatuses and regulations governing international movement), technological (transport and communications), humanitarian (NGOs and international aid to migrants) and social (migrants’ networks and social capital). Again, the chapters that follow provide ample evidence of the critical role of these aspects in the way that onward migration is facilitated, shaped and also sometimes blocked.

Third, and reflecting to a certain extent the infrastructures of migration, onward migration is often related to changes in migrants’ categorisation and their relationship with place and transnational ties. Often, migrants are not only moving onwards between countries but also shifting their status between that of tourist, visitor, student, working migrant and attached family member and from one legal status to another, more or less ‘legal’. Migrant categorisation also intersects with country of origin, citizenship, ‘race’ or ethnicity, gender, education and other variables. Such an intersectional categorisation is also present in some of the chapters that follow, most notably in Roberts’ Chap. 7, where intersectional identities in multiple onward migrations are narrated through the perspective of the female partner in heterosexual migrant couples. Roberts unveils gender-differentiated experiences of transnationalism, career progression and life satisfaction, mediated by nationality, ‘race’ and access to different citizenships in various global locations.

1.4 The Effects of Transnationalism on Intentions and Experiences of Onward Migration

Migrants wishing to relocate can make use of their evolving transnational connections in order to visit and (re-)familiarise themselves with their origin country or other destinations. Carling and Erdal (2014) originally coined the term ‘exploratory transnationalism’ to refer to the return visits of migrants to their origin country. These shorter return trips can be useful in enabling migrants who have spent a considerable time abroad to re-acquaint themselves with the local context and prepare for an eventual return migration (Duval, 2004). For onward migrants, however, ‘exploratory transnationalism’ can be arguably even more crucial, as it enables them to test the waters in a place or country where they have not lived before and helps them to convert their migration intentions into more concrete plans (Ahrens et al., 2016).

Onward migration is often the result of an interplay of particular drivers and obstacles. For some migrants, the desire to onward migrate is formed when they have secured more-stable residence papers or become naturalised citizens. There are various reasons why migrants may decide to migrate again at this point. First, stable residence papers or citizenship usually also afford them enhanced mobility rights. Second, having a stable residence in one destination country can also be perceived as a type of insurance should the next move not turn out as planned. Third, having spent many years in one destination country equips migrants with more detailed knowledge about the opportunities and shortcomings of their current place of residence – and maybe even of other countries. Researching Polish and Filipino nurses in Norway who had lived and worked in various countries before their latest move, Erdal et al. (Chap. 2) describe how, through their prior moves, migrants acquire ‘tangible transnational capital’ in the form of life skills, know-how and comparative advantage in the global market for healthcare workers.

Even after becoming citizens of a destination country, some migrants may feel that they do not have access to desired opportunities which may result in ‘reactive transnationalism’ (Itzigsohn & Saucedo, 2002). Such a reactive pushback might normally be expected to be directed at their origin country, but it could equally pull them towards another new destination or previous country of residence, especially if their own country continues to offer them little hope of a sustainable future livelihood. Formenti (Chap. 11) argues that, because of the discrimination and the language barrier that Portuguese-Guineans face in the labour market in Peterborough, they develop a greater sense of belonging and identification with ‘Portuguesehood’ following their onward migration from Lisbon. This new-found Lusophone identity enhances their social networks locally in Peterborough, but it also stimulates transnational ‘return’ visits in another direction, namely to their previous country of residence, Portugal (see also Ahrens, 2022).

Yet not all migrants have stable residence papers or citizenship to facilitate an onward migration. Ahrens (2013) found that Nigerian residents in Spain were severely affected by the austerity measures introduced following the 2008 financial crisis. They engaged in semi-legal moves to other EU member states, where they had the right to stay or visit but not always to work. In contrast, in countries like the United Arab Emirates, circular and onward migration is a way of life for the large and highly diversified migrant community. The kafala system means that all migrants, including those born in the country, are issued with a short-term, renewable visa that ties them to a sponsor – usually their employer (Ali, 2011). Once their employment contract is over, or they reach retirement age, they are generally compelled to leave the UAE, and many such migrants are creative in using their accumulated transnational ties – for instance to their other family members or to places where they have previously lived or been educated – to fashion trajectories of return or onward migration (Ali, 2011).

1.5 The Variegated Effects of Onward Migration on Transnationalism

In the following subsections we discuss the different ways in which onward migration can affect migrants’ experiences of transnationalism and their emerging transnational identities.

1.5.1 Inter-generational Transnationalism

In her pioneering book on international onward migration, Parminder Bhachu (1985) documents the ‘twice migration’ of East African Sikhs to the UK. They were part of a wider group of South Asians in East Africa, comprising wealthy Gujarati entrepreneurs (who had lived in East Africa for centuries) and Punjabi indentured workers (who built the Kenya–Uganda rail lines). When the newly independent nation states introduced Africanisation policies from 1968 onwards, these groups were an obvious target. However, they found that, despite their official status as British subjects, they were unable to enter Britain freely. The 1962 Commonwealth Immigration Act subjected Commonwealth citizens to immigration controls for the first time. Alarmed by the influx of thousands of East African Asian families, the British government further tightened controls in 1968. On the whole, the settlement of substantial numbers of East African Asian families in places such as West London and Leicester, cities which already hosted large Indian-origin migrant communities, was successful, partly because of these co-ethnic links and partly because of the professional and entrepreneurial orientation of the migrants.

Voigt-Graf (2004) illustrates the diverse nature of transnational networks by developing different spatial models for the three Indian migrant communities which are part of her research. She argues that, where cultural links to the original diasporic ‘hearth’ are lost, migrant communities might come to regard a place where they have lived for an extended amount of time as the ‘new centre’. This was the case for some Indo-Fijian twice migrants who, after migrating to Australia, regarded Fiji as their ‘new centre’ because they no longer maintained strong kinship ties to India.

In this volume, Della Puppa and Sredanovic (Chap. 9) find that Italian-Bangladeshis and their children who have onward migrated to London from Northern Italy become homesick not for their Bangladeshi homeland but for Italy, where many have spent most of their lives and which evokes a newer sense of home pending an improved integration in London. Hence, they visit Italy frequently and Bangladesh rarely, not just for logistical and financial considerations but also for affective reasons. They miss the climate, the coffee, the ice cream and the more relaxed social environment of Italy, which they contrast with the hard work, low pay, stress and general grimness of life in London (see also King & Della Puppa, 2021). The main motive for their onward transfer to London, facilitated by the acquisition of Italian (and hence ‘European’) citizenship in pre-Brexit times, was to advance the educational and career prospects of their Italian-born children, which can be considered another form of inter-generational transnationalism – passing the benefits of onward migration to the next generation.

Several other chapters likewise focus, mainly or partly, on the inter-generational dimension of onward migration, where the future of the children, in what de Hoon and van Liempt (Chap. 3) call the ‘transnational educational field’, becomes the main rationale for the move. Within Europe, the UK features as the onward migration destination of choice, for a series of reasons which resonate across different cases – Italian-Bangladeshis, as noted above (see Chap. 9), Dutch Somalis (de Hoon and van Liempt, Chap. 3), Dutch Sudanese (Serra Mingot, Chap. 6) and Portuguese-Guineans (Formenti, Chap. 11). Acquiring, respectively, Italian, Dutch and Portuguese citizenship facilitated these onward migrations to the UK but the pull factor was the British comprehensive education system, seen to be more inclusive towards ‘visibly different’ migrant- and refugee-origin children and to offer them better life-chances and careers, including access to higher education. In contrast, the Dutch system, for instance, tends to ‘condemn’ refugee-origin children to lower streams and qualifications, not helped by teachers’ low expectations of such children. Finally, the importance of children pursuing their education in the globally relevant English language is seen as highly advantageous over being taught in Italian, Dutch or Portuguese.

A different kind of cross-generational transnationalism is described by Patterson in Chap. 10. South Sudanese refugees housed in the vast Kakuma camp in Northern Kenya see education as their only escape from protracted displacement; however, this is available only to a fortunate few who can win scholarships to study abroad and be supported by remittances sent by older relatives who have already onward migrated to North America, Australia and the UK.

1.5.2 Split Transnationalism

Onward migration often takes place by different family members in a temporally lagged fashion, because certain family members are unable or unwilling to relocate together or because the move takes place in a strategically staged manner (Ahrens et al., 2016). The emergence of multi-local households is often determined by the infrastructure of immigration regimes, which can make it difficult to bring family members to destination countries in the Global North because the migrants applying for family reunification or family formation often have to comply with an income threshold, as family members often do not have recourse to public welfare funds in their first few years after arrival.

In other scenarios, one member of the family – often the male household head (cf. Della Puppa and Sredanovic, Chap. 9; Formenti, Chap. 11) – makes the onward migration first in order to ‘test the water’ and prepare the ground for family reunion. However, in the Dutch-Somali case (de Hoon and van Liempt, Chap. 3), it is more common for women to be the onward-migration pioneers to the UK. Often separated or divorced, with their (former) husbands remaining in the Netherlands or returning to Somalia, the women are both the carers and the breadwinners of their families relocated to the UK. Dutch register data reveal that around half of Somali transnational families are ‘scattered’ due to the combination of onward migration and ‘absent fathers’. This case study also puts up for scrutiny the commonsense notion that children onward migrate with their parents; in the Dutch-Somali case many children move either on their own or with a relative who is not their biological parent. This extension of the standard model of the nuclear family implicit in much Western migration scholarship is also found in the case-study material from other chapters in the book, notably Chap. 6 (Serra Mingot) which takes the Sudanese extended family as its unit of analysis and Chap. 10 (Patterson) where successful South Sudanese refugees in the US and elsewhere send remittances to support the education of younger pupils and students in Kenya who may not be their close relatives.

Finally, Tobin et al. (Chap. 5), using the conceptual framework of ‘transnational figurations of displacement’, examine the family and kin-based networks mobilised by Syrian refugees seeking onward mobility within and beyond Jordan as their first place of refuge. Here, as with the South Sudanese in Kakuma, onward migration is an aspiration, an imaginary, rather than a realistic prospect, for the vast majority of refugees, which often results in families being split because of the rigorous legal restrictions on entry to ‘third’ countries. This is summed up in the description of interviewee Um-Hosam who ‘is now alone with one son in Jordan while [the rest of] her family are flung across the globe’.

1.5.3 Widening and Re-routed Transnationalism

Onward migration self-evidently widens the transnational range beyond its conventional bipolar form to a multi-polar expression of an origin country and two or more destinations. Whilst some chapters in the book explicitly deal with one onward-migration destination, sometimes defined as the ‘third country’ (de Hoon and van Liempt, Chap. 3; cf. also Serra Mingot, Chap. 6; Della Puppa and Sredanovic, Chap. 9; Formenti, Chap. 11), other chapters detail a series of more complex spatial trajectories, including Erdal et al., Chap. 2; Flikweert et al., Chap. 4; Roberts, Chap. 7; and Jung, Chap. 8.

Jung (in Chap. 8) examines the widening transnational practices of male Senegalese onward migrants in Brazil.Footnote 4 Some of his research participants arrived in Brazil via Cape Verde or Argentina. Their trajectories and plans to settle are often shaped by chance encounters that make them consider new destinations. They tend to remain mobile between a growing number of destination countries, which Jung likens to a ‘widening horizon’ of possible destinations. Nevertheless, Senegalese migrants maintain strong transnational ties with their immediate family members who stayed put in Senegal and exhibit the transnational behaviour of target earners, as they make considerable investments in their origin country in preparation for their envisaged permanent return to Senegal.

Erdal et al. (Chap. 2) present a study about the complex multi-sited transnational lives of nurses from both Poland and the Philippines who were working in Norway at the time of interview but who had previously lived in other destinations like Belgium, Denmark, the Netherlands, Saudi Arabia and the UK. The authors emphasise the importance of temporality in understanding the ways in which the transnational ties and networks of their research participants evolved over time and how their perspectives on destinations also changed in response to macro-level processes and their own life-stage. Even though the research participants had not necessarily planned to go to or settle in Norway, nonetheless many of them had stayed in the country for several years. Therefore, Erdal and her colleagues refer to this process of familiarisation with different places as Norway ‘becoming destination(s)’.

The notion of ‘becoming destinations’ is also applicable to the two biographical case studies presented by Roberts in Chap. 7. Two married couples, one of Bangladeshi origin (Nahid and Amit) and the other a mixed partnership between white Australian Sophie and black Ghanaian Daniel, navigate their complex single and coupled trajectories around the world, taking in multiple destinations, some of which are resided in more than once at different life-stages. In this way, the two couples develop strong transnational ties with several different countries; some of these ties are shared by both partners in the couple, while others are advantageous to one partner but not to the other for intersecting reasons of ‘race’, religion, life-stage and career and income-earning opportunities.

By contrast, Flikweert et al. (Chap. 4) focus less on which countries are navigated via onward migration and more on what happens there in terms of migrants’ ability to send remittances to their African home countries – Ghana, Senegal and the Democratic Republic of Congo. Intriguingly, the authors find that being ‘on the move’ in onward migration increases migrants’ propensity to be gainfully employed and to send remittances, compared to migrants who are settled in their final destination.

In the book’s other remittance-focused study (Patterson, Chap. 10), the remittances are sent not to the origin country – South Sudan – but to young family members and even sometimes to non-kin who show strong academic promise and reside in Kakuma refugee camp in Kenya – an example of what might be called ‘re-routed transnationalism’, like the Italian-Bangladeshis in London (Della Puppa and Sredanovic, Chap. 9) whose transnational ties to Bangladesh fade away, to be replaced by Italy as their nostalgic reference-point.

1.6 Onward Migration and Multi-sited Transnationalism: Implications for Research Methods

Throughout this book, onward migration is framed as a part of complex, non-linear migration trajectories which, in some cases, are planned and stepwise and, in other cases, are subject to unplanned and unexpected turns of events. As well as the spatial multi-locality which such onward migrations produce, often resulting in the splitting and scattering of transnational family structures, a temporal perspective is also important, as Erdal et al. stress in Chap. 2. Time is itself a complex construct in the various ways in which it intersects with onward migration. Amongst the many temporalities involved are the simple, chronological, linear time against which all (im)mobilities are recorded; biographical or life-course time, especially the way that onward migration is related to key life events for an individual or a family; what we might call ‘structural time’, comprising economic cycles of boom and recession or geopolitical processes such as EU integration policy and ‘free movement’; and, finally, the temporal ruptures of more sudden events such as the Brexit referendum or the onset of the Covid-19 pandemic. For onward migration, all these temporalities are potentially valid; however, particularly important is the life-course perspective, which sees onward migration as a multi-sited longitudinal family trajectory to fulfil changing aspirations where, however, not all family members can benefit equally. As Serra Mingot cogently states (Chap. 6), examined through a life-course temporal lens, ‘migration can lead to new conceptualisations of the temporality of the life course’ as well as the inverse: ‘personal conceptions of the temporality of the life course can often lead to [onward] migration’.

How, then, to capture methodologically the space–time specificities of onward migrations and their associated multi-sited transnational effects? The chapters in this book offer some guidelines, as do the papers collected by Collins and Huang (2012). In the present book, Tobin et al. (Chap. 5) suggest that ‘translocal figurations of family and kin’, which they see as both ‘classical sociological configurations’ and ‘ambivalent entanglements’, are an appropriate frame to document the dynamically changing socio-spatial constellations formed by a people (Syrian refugees in Jordan) desperate to escape to another country where their lives will be better. Serra Mingot (Chap. 6) mentions multi-sited matched-sample ethnographic methods for a better understanding of the ‘bigger picture’ of the onward migration of Somali refugees into and within Europe. Both Erdal et al. (Chap. 2) and Jung (Chap. 8) also propose more explicitly graphical methods to portray the spatio-temporal routings of the successive migrations of transnational individuals and their families – in the first case (Filipino and Polish nurses in Norway) in the form of conventional cartographic mapping and, in the second (Senegalese onward migrants in Brazil), in the form of ‘migration history charts’ (cf. Carling, 2012).

We want to extend these guidelines to a more solid methodological statement. Our starting-point is the landmark paper by George Marcus (1995) on ‘multi-sited ethnography’ in which he encouraged field researchers to ‘follow the people’ (1995, 106). Building on King (2018), we take forward Marcus’ entreaty to consider the extent to which multi-sited transnationalism can indeed be studied through the technique of multi-sited ethnography, whilst incorporating some of the critiques of Marcus’ formulation originating from more ‘place-bound’ anthropologists.

Surprisingly, Marcus had little to say about migration in his 1995 paper: just a few lines, plus a passing reference to transnationalism, even though the defined object of study of multi-sited transnationalism is ‘ultimately mobile and multiply situated’ (1995, 102). Instead, his paper takes the broader remit of examining ‘the circulation of cultural meanings, objects, and identities in diffuse time-space’ (1995, 96). It has been for others to expand the discussion on the utility of multi-sited ethnography for studies in migration, transnationalism and diaspora (for instance, Gallo, 2009; Mand, 2011; Riccio, 2011; Beauchemin, 2014; Boccagni, 2016; King, 2018; Francisco-Menchavez, 2020; and see Hage, 2005 for a sharply critical view).

As Mark-Anthony Falzon notes (2009, 4), the success-story of multi-sited ethnography reflects an increasing awareness of globalisation, transnationalism and the social construction of space and place – and of the so-called ‘spatial turn’ in the social sciences since the 1990s. This has led to the perceived inadequacy of single-sited research and of single localities as units of analysis in an increasingly mobile world. Falzon (2009, 1–2) develops and expands Marcus’ definition of multi-sited ethnography in the following terms:

Marcus argued that multi-sited ethnography defines as its objective the study of social phenomena [migration par excellence] that cannot be accounted for by focusing on a single site… The essence of multi-sited research is to follow people, connections … and relationships across space… Research design proceeds by a series of juxtapositions… In terms of method, multi-sited ethnography involves a spatially dispersed field through which the ethnographer moves … via sojourns in two or more places.

From this definition and these guidelines, the expectations are clear: the migration researcher ‘follows the people’ (onward migrants) to their next destination and carries out fieldwork in the various places (origin, plus two or more destinations) that the migrants inhabit.

Of course, in reality things are not so simple and this is where the various critiques of multi-sited ethnography and the responses to them come in. Falzon (2009), Coleman and von Hellermann (2011), Boccagni (2016) and King (2018) give balanced overviews of this critical debate, most of which touch on the dilution of ethnographic fieldwork when it is spread across multiple sites and the logistical challenges of carrying out research within a compact timeframe across places that may be very far apart in terms of geographical and cultural distance. The critical charge against Marcusian multi-sited ethnography has been led by Candea’s (2007) argument in defence of the single bounded field-site and Hage’s (2005) disbelief that multi-sited ethnography is even possible.Footnote 5

The criticism that in-depth ethnographic fieldwork cannot be carried out across multiple sites because of time and resource constraints is countered as follows. First, when studying migrants’ onward migration and multi-sited transnationalism, it is not so much a deep understanding of the sites themselves which is important but, rather, the ties, relationships, networks and movements between them, since these are the elements that make up the multi-sited ‘transnational social fields’ (Levitt & Glick Schiller, 2004). The key, then, is ‘mobile ethnography’ (King, 2018, 45). Like the onward-migrating research subjects, the researcher combines periods of static fieldwork with being ‘on the move’ (Falzon, 2009, 9). There is a growing literature on ‘mobile methods’ which is of obvious relevance to researching migrants who are moving repeatedly (see, especially, Büscher et al., 2010). Onward migration and other complex migration trajectories involve a dialectical relationship between mobility and immobility (since a migrant has to ‘stay put’ for a certain threshold of time in order to be recorded as a migrant and not as a tourist or visitor) and the multi-sited field approach offers the researcher the chance to study both stasis and movement across borders within a transnational or translocal perspective (Amelina & Faist, 2012; Amelina et al., 2012).

A second arena of debate centres around the definition of ethnography. If the term ‘multi-sited ethnography’ is replaced by ‘multi-sited methods’ or ‘multi-sited fieldwork’, many of the concerns of the more-purist anthropologists about the time necessary for a deep immersion and understanding of ‘the field’ disappear. In practice, as several of the following chapters demonstrate, most field research with onward migrants involves in-depth qualitative interviews, sometimes only in one place (where the migrants currently are), sometimes in two or more places. Such interviews are often combined, according to the methodological statements of the chapter authors, with participant observation. Whether such a combination is sufficient to qualify for the label of ‘ethnography’ is a matter of debate. Certainly, on the plus side, the combination of a multi-sited approach and mixed methods, often involving quantitative survey data, gives the possibility to triangulate the findings and more statistically rigorous results (see Beauchemin, 2014; also Chaps. 2, 3 and 4 of this book, respectively by Erdal et al., de Hoon and van Liempt, and Flikweert et al.).

Whilst the ‘lone’ migration researcher, ethnographer or otherwise, can follow people, things and themes throughout time and across space and borders, there are other research strategies that involve teamwork and collaboration. Such approaches are common in funded research projects with multiple partners employing mixed methods, several disciplines and a comparative dimension (Fitzgerald, 2006). In this way, the multi-sitedness of field research can be synchronous, being both ‘here’ and ‘there’ (and ‘there’; cf. Hannerz, 2003) at the same time, perhaps achieving the methodological rigour of matched samples (Mazzucato, 2009). The research équipe model avoids the risk of the single field researcher being in the ‘wrong’ place or engaging in ‘bouncing’ (Burawoy, 2003, 673) or ‘hopping’ (Hage, 2005, 465) between different sites in a yo-yo fashion (Wulff, 2007, 139–145).

A final set of perspectives on multi-sited research involves bringing in some less traditional aspects of the practice, reflecting new methods of collecting data and envisioning onward migration as imaginaries. First, Marcus’ (2009: 184) idea of the multi-sited imaginary is highly relevant to onward migration since migrants develop visions and desires of where they would like to (onward) migrate to. Migration thus becomes a ‘horizon of imagination’ (Graw & Schielke, 2012). Yet the actual move may never happen or may only happen much later when circumstances change or when the move can take place via one or more intermediate destinations. Several chapters in the book discuss, implicitly or explicitly, the imaginaries of ideal destinations (see, in particular, Erdal et al., Chap. 2; Tobin et al., Chap. 5; Jung, Chap. 8; Patterson, Chap. 10). Second, the advent of technologies of instant, cheap communication not only affects migrants’ ability to stay in touch with each other and with ‘home’ across transnational space but also facilitates researchers’ ability to ‘follow the migrants’ via telephone calls, Skype interviews and other instantaneous channels (as noted by Roberts, Chap. 7). Finally, in a move towards what Marcus (2008) and others (Coleman & von Hellermann, 2011) have termed ‘para-ethnography’, smaller-sample, in-depth field research with (rather than ‘on’) onward migrants can consist of working with selected participants who, in effect, become co-investigators in the co-creation of knowledge. This may also involve situations where the lead researcher is auto-ethnographically involved in the topic under investigation, either as a member of the migrant group being researched or through other aspects of their life experience. After all, how many researchers forging their own academic careers are themselves onward migrants?