9.1 Introduction: Networks and Migration Trajectories

As discussed in the Introduction to this volume, much of the work on migrants’ social networks is on the effects which ethnic identity has on the composition and structure of migrants’ networks. In this chapter, in contrast, I discuss the effects which migration itself has on migrants’ networks. Even in times of electronic communication, social networks are shaped by space, so the fact that migrants have moved geographically has important effects. In addition, migrants vary in the networks they use to achieve their migration and the various types of migration within and across national borders have structural characteristics liable to affect social ties. In this chapter I draw on examples from international and internal migration, and from labour migration and skilled migration, to illustrate some of the effects of “migration trajectories”. I describe these migration trajectories not in terms of the individual traits of the migrants concerned (their human capital) nor in terms of their legal position (as I will show, there are many similarities between international and internal migrants following a labour migration trajectory) but rather in terms of the networks involved. As the examples in the chapter show, these various kinds of trajectories differ in the type of ties used to effect the move itself, but also in the relationships formed in the place of migration, in the work ambiance, the neighbourhood and outside work.

I also argue that the migration trajectories of parents have effects on the networks their children form: to take an obvious example, they affect where families live and this often determines the school attended, and the friends formed in the neighbourhood. These networks have effects on education and on the jobs young people obtain (cf. Eve & Ceravolo, 2016).

I believe it makes sense to differentiate in this way between migration trajectories because the network used to achieve migration itself has important effects on the relationships established in the place of immigration. If I am invited to apply for a skilled job in the place of immigration by a colleague with whom I have worked in the past, that colleague may form an initial basis for the network I establish both inside and outside work, a network which is liable to be different from that I establish if it is my sister who puts me up in her flat and tells me where I can find work. But the ability to form a tie – either with locals or with other migrants – is not a purely individual matter. The re-organization of one’s network e.g. around relatives and past acquaintances in the place of emigration, or around work contacts, or around a church or an association, is not just an individual choice: it depends on others in these ambiances being present and available to form new ties. So the reinforcement of a relationship with a sister depends not just on the relationship we had back home and on our individual characters and social trajectories, but also on the presence of a cluster of other people who provide a series of activities, subjects of conversation, interdependencies.

In the first section of this chapter I describe the reorganization of social ties which inevitably accompanies any migration, stressing that the social significance of a move is not just that it changes the forms of interaction with individual persons considered singly, but rather that it changes a whole social configuration. So, for example, a move does not just make it impossible for me to drop round to a friend for a coffee, but also takes away the whole context of local social interactions which made up the stuff of our conversation.

In the next four sections I use this configurational perspective to understand the difficulties internal migrants report in forming relationships with locals, and the strategies they develop to build a network in the new place. Among other things, I describe the social logic whereby some migrants form ties mostly with other migrants from “back home”. This pattern has often been described for international migrants, and is usually conceived in ethnic terms, as a result of preference for persons imagined to be similar culturally, combined with the diffidence of locals. In this chapter, however, I give examples from internal migration which cannot plausibly be interpreted in these terms and I offer an explanation in terms of the networks formed in particular migration trajectories.

This analysis of the specificities of migrants’ social networks in the place of immigration is the general framework I use for analysing the role of networks in migrants’ situation in the labour market. “Integration in the labour market” is a thoroughly social process, depending on information coming from other people about what jobs are available, conversations about what jobs are feasible or desirable for someone like me, the development of language and attitudes to present oneself as a suitable candidate, knowing someone who has influence over hiring decisions. The possibilities individuals are aware of, and the choices they make, are very much influenced by their local social network. So it is of considerable importance who this network is composed of: a network of other migrants from “back home” provides different opportunities from a network spun around a professional ambiance. More in general, the ties which may be of use in job search, or in constructing skills, are not divorced from other ties, so understanding the structure of migrants’ networks more generally is fundamental for understanding their position in the labour market. Also for this reason, I devote space to the logic underlying the formation of ties with other migrants from the same place of “origin”: the construction of a new network around “homeboys” has important effects on the information and opportunities in the labour market a migrant has access to. Whereas the link with the labour market is mostly implicit in the earlier sections of the chapter, it becomes explicit in the section on the link between the networks used to make a move and the ties which lead to a job, and a particular place in the labour market. The ties used to effect a move (e.g. ties with relatives in a classic chain migration, ties with professional colleagues as in many cases of skilled migration) are often de facto the ties which are crucial in getting a migrant their first job in the place of immigration. This is often recognized but its implications for how migrants’ networks develop are rarely reflected on.

Finally, the last three sections of this chapter deal with the effects of migration trajectories on the networks of children of migrants, and indirectly on the second generation’s position in the local labour market and system of social stratification. Here I focus on labour migration and on a number of structural effects associated with labour migration (whether internal or international): the frequent moves in the early stages of the migration cycle, the type of neighbourhood migrant families are liable to settle down in, the kind of jobs labour migrant parents are in, relationships with neighbours. I argue that these features have effects on the relationships children of labour migrants form, and that these in turn have effects on education and on the kinds of jobs they obtain.

9.2 Moving to Another Place: What Effects on Social Networks?

First of all, let us consider why space, and moving, should affect social networks. It is worth noting, at this point, that predictions of a space-less sociability have been frequent since the time that the telephone started to become widespread (Flichy, 1991). Yet telephone data from different historical periods and national contexts has always shown that the great majority of telephone calls are made between persons who interact locally and face-to-face, telephone calls forming part of the overall relationship. Research on more recent communication technologies do not seem to show a radical departure from this pattern of communication partners (Mok et al., 2010; Takhteyev et al., 2012; Ellison et al., 2007).Footnote 1 Research is, of course, continuing as communication technologies change, most recently with “social media”, but at present there is little evidence of social relationships becoming free of dependence on space.

It is also interesting to see how changes in an individual’s geographical location (i.e. migration of one kind or another) alters the network. If we take telephone calls as an indicator of an individual’s active social network, we see that a move to another town has a radical effect. Ties with kin survive moves rather well, with calls becoming less frequent but also longer, whereas calls to most friends disappear a few months after a move (Mercier et al., 2002).

The fundamental reason, I believe, that many relationships drop away after a move is not that there is anything inherently “less real” about speaking over the phone or via a computer as against “face-to-face” interaction. But rather that the overall configuration which gives meaning to a single relationship, and thus to telephone conversation, is lacking.Footnote 2 If I no longer work in the same office as my friend, I will be less concerned by what the boss has said, less amused by the latest office gossip. If the friend who has a child the same age as my daughter no longer lives in the same town, we cannot get together to organize play activities, so there will not be calls to arrange these, but in addition we will not have the joking conversations about the trials of parenthood which sprung out of those exchanges. As I have argued (Eve, 2002b) it is wrong to see friendship as a purely dyadic relationship between two individuals attracted by their personal characteristics or by particular interests: we need to contextualize the speech and the material, emotional and intellectual exchanges which are fundamental to friendships socially. Even outside the ambit of friendship, when we are talking of social ties in general, a purely dyadic focus is rarely sufficient. The fact that telephone calls to kin are more resistant to geographical distance is understandable if we see social relationships as configurational. Certain kinds of interchange with the kin network become impossible when a person moves home, especially if active kin ties are spatially very concentrated in the former location, but other forms of interaction with a whole series of kin are likely to persist. So even after moving away, I will be concerned about my sister’s illness, and the effects this has on who can look after my grandfather. I may feel obliged to send money to help out also because others may “expect” me to do so. The overall structure of the network of ties between various kin means that a series of my relationships are in play (on the importance of reciprocity in networks, see Bilecen in this volume).

These dynamics are of course relevant for migrants and for non-migrants. But thinking in these configurational terms sheds light on the effect migration has on social networks in the place of arrival. In fact the configurational background which locals have supporting individual ties is not the same as that which migrants have. As Michel Grossetti (2005) points out in his study of “where relationships come from” few personal relationships grow out of casual contacts. In his study in Toulouse the great majority originated in a small number of contexts: school or university, work, various types of associations, neighbours and flat-shares, plus persons contacted via family and kin. Very similar findings were reported by Claude Fischer in northern California in the 1970s; and by Claire Bidart in her work on friendship. As Bidart (1997: 52) puts it: “One does not make friends on the street, in a crowd, out of nothing. Certain settings, certain places, environments are relatively favourable to the construction of interpersonal ties…”.

It should be clear that many of these standard sources of new relationships – what Grossetti, in Simmelian tradition, calls “circles”, what I have referred to as the “configurational background” of an individual relationship – are not available to an adult migrant, who will not have former school friends, neighbours or members of one’s former scout group or football team in the place of immigration. Or rather, they may have such contacts, but only among other migrants from the same town or village of origin. Likewise with kin. Obvious though this may seem, it is also fundamental when considering the networks migrants form when they arrive in a new place. We need to recognize that a part of the relational specificity of migrants lies in the way networks interweave with the life course (cf. Bidart & Lavenu, 2005: Garip, 2017).

A configurational approach of this kind to the formation of relations helps to understand that feature which has always struck observers about migrants’ social ties, viz. the fact that many migrants “stick with their own kind” and often have relatively few ties with “natives”, even after many years of residence in the place of immigration. This has traditionally been thought of in an ethnic key. Common sense interpretations have hypothesized a preference to live with others who are (supposedly) culturally similar, combined with diffidence on the part of locals impeding the formation of relationships. Scholarly approaches have not been unaffected, and a good deal of the interest of scholars in migrants’ networks has, as I have said, been directed to the ethnic composition of migrants’ networks. Ethnic dynamics are obviously important, but they need to be contextualized in the configurational background in which relationships in general are formed, and when we look at where relationships are formed, homophily no longer appears a purely individual trait (Wimmer & Lewis, 2010; Eve, 2002b).

Citizenship, and the lack of documented status which has become such a common aspect of international migrants’ experience in recent decades, as well as language and other aspects of “culture”, can obviously have very important effects on the ability to form social ties. However, it should be noted that even when such factors linked to “culture”, ethnicity and citizenship are lacking, similar patterns of clustering with persons from the same geographical origin may be present. Interviews I conducted with French young people in Rennes in 1998 (Eve, 1999, 2001) made it clear that many students and workers from outside Rennes clustered with others from their own local “origin”. This was particularly striking in the case of students: university is an organizational context exceptionally propitious for the formation of new ties since it brings together people, mostly at the same stage of life, with broadly similar future projects, in regular, repeated contact. Taken away from their previous daily round of sociability, students also have time and the desire for new contacts. Yet even in these exceptionally propitious circumstances, many students spent most of their time with others from the same town or village in Brittany. As one male student said “it’s the people of Saint Malo with the people from Saint Malo, Saint Brieuc the same, etcetera”. Several other interviewees made the same point: a young woman who did not entirely approve of this pattern, said “It’s true that we tend to stay in groups … always in a little micro-society … it’s easier … the people you know already … there are lots of us, so you bump into people you know … It’s true that we often go out together in a group of people who know each other … Slightly closed circles … we have our habits of life, because we’ve known each other for a long time … you don’t necessarily feel the need for others. I have a friend from Paris who doesn’t like it here because we’re a bit closed in, a bit hermetic”. Numbers are clearly a precondition for this kind of pattern: as the young woman cited earlier said, “There are lots of us here [from the same town]” – a feature I will come back to later.

9.3 ‘Home Boy’ Networks

Manchester SchoolFootnote 3 anthropologists studying rural-urban migration in Africa (e.g. Epstein, 1958; Harries-Jones, 1969; Little, 1974) used the term “home boys” – an English translation of words present in various local languages in southern and western Africa – to describe groups of rural migrants from the same village or region in urban contexts. Their descriptions of relatively tight knit networks of persons identified as distinctive in terms of their geographical “origin” could be applied to many situations, including that in Rennes.

Brittany is of course known for strong regional identifications, but regionalist ideology might be expected to encourage rather than discourage relationships with students from other towns and villages in Brittany. And in any case, I believe this kind of pattern could be found in other French, Italian or British universities where there are significant numbers of people from the same town, or indeed the same schools or same social circles “back home”, so I do not believe Rennes is specific.

Ethnographic studies of internal migration often provide cases which are interesting because cultural differences, citizenship or foreign-ness cannot be evoked to explain the patterns found. Michel Bozon’s (1984) account of social relations in Villefranche-sur-Saône (a town of around 30,000 inhabitants in east central France) provides interesting accounts of the difficulties of migrants in forming relations with locals (Caladois, as they are called), and the strategies migrants use to adapt to this situation. The majority of the migrants – especially among the manual workers who came to work in the town’s textile and engineering industries – had come from the surrounding region, so there were particularly few differences in “culture” between “migrants” and “locals”. Yet many migrants gave an interpretation in cultural terms, accusing the Caladois of being “cold” and “closed” (Bozon, 1984: 50–51). A woman from the Haute-Loire says “The Caladois mentality doesn’t suit me. People live in closed circles, very cool with outsiders … I have friends in Villefranche but they come from outside, from the Haute-Loire, Bayonne, Grenoble. I have very few contacts with real Caladois”. A man from the nearby Beaujolais country contrasts the superficiality of the conversations he has in Villefranche to those back home: “In the country when you ask how are things you get a real answer. Here if someone asks how you are before you have time to say anything he’s already gone” (ibid., 50–51). This man admits that one reason for the difference is that in his home village he often talks about family matters whereas he cannot enter into those conversations in Villefranche, but he finds the situation unsatisfying anyway.

A teacher from Lyon is also interesting on the micro-interactions in play. This man says he sometimes feels irrevocably a foreigner because people born in Villefranche “often mention a whole lot of people who I’m not sure I know… they talk about what they did, or about things they did together when they were at school… I was having a drink last Wednesday with the [rugby] trainer and two or three men all born in Villefranche, I didn’t join in the conversation much because they were asking themselves what had happened to a group of friends with whom they did masses of things when they were young. So there is a way of relating to people which isn’t the same, you see; there I felt very much an outsider… And yet I’ve lived longer in Villefranche than in Lyon” (ibid., 63–64). Having arrived in Villefranche as an adult this man lacks the background set of ties which is fundamental to participate in some forms of interaction and conversation.

Many local migrants in Villefranche adapted to these difficulties in the same way as many of my interviewees in Rennes by socializing mainly with other “migrants”, or by using Villefranche purely as a work place and going back to their “home” town or village nearly every weekend (ibid., 52). Like many international migrants, many of these very local migrants did up a house in the place of origin, or built a new one (ibid., 65).

It is also interesting to see the ambiances where ties with locals were formed in Villefranche. The teacher from Lyon says he knows a lot of people through rugby and his political activity. As Bozon says, there are certain ambiances where it is easier for migrants to make friendships and other social contacts: associations, churches, political parties have an intrinsic need for members and so may constitute a channel for migrants willing to give their time (and perhaps willing to fit in to the existing hierarchy). The biographies of many internal and international migrants would probably bring out the role played by organizations of this kind in the re-organization of their social network. In a well-known article Hirschman (2004) sees religious bodies as offering migrants respect and social recognition as well as resources. Other associational ambiances have a similar role, and I would argue that the significance such ambiances have for migrants has similar configurational roots. Sports clubs, too, obviously have need of talented members who can help win local competitions, as well as volunteer trainers or helpers. Like Bozon’s interviewee, many other migrants, internal and international, could probably trace ties with a number of individuals to the “circle” of sports or cultural clubs.

Work is, of course, another ambiance where migrants form relationships. I will come back to the specificity of the work ambiance as a source of social ties for migrants; here, I will simply point out that interviewees in Villefranche who did not work outside the home particularly stressed their isolation and the “coldness” of local Caladois (Bozon, 1984: 52).

International migrants also sometimes explain the difficulties of forming ties with locals in the same kind of terms as Bozon’s teacher, emphasizing the effect of arriving as an adult. An Indian manager in Canada, interviewed by Suhair Deeb and Harald Bauder (2015: 58) says “People who have grown up in Canada have very strong ties to the school they went to and to the summer camps they went to. We do not share those ties (…) access to those natural networks. [Immigrants] have created new networks, but since they are new, they don’t run that deep, and they don’t have the depth, and people in the network don’t have the attachment to you.”

In her references to “depth” this interviewee seems to be thinking of personal friendships, but the ease of communication which other interviewees refer to (and which they lack) has consequences in the workplace well beyond close friendship. As Deeb and Bauder point out (2015: 52) “social talk is intertwined with work talk”. As I have mentioned, the specificity of migrants’ networks has consequences on careers, on absorption of what is seen as the “right language”, and even on ability to perform work tasks.

In the comparison with Bozon’s material, it is also interesting that some of Deeb and Bauder’s (2015: 57) interviewees (foreign managers in Canada) stress the role sports contacts had in “integrating” socially into Canadian offices, and acquiring cultural capital alongside social capital through participation in sport (cultural capital and social capital are of course intertwined).

I hope it is clear that, even independently of ethnic identity, migration has effects on social networks which can be understood reflecting on the configurational logic by which new relationships are formed. Even though many migrants themselves use an ethnic-type framework to explain the over-representation of other migrants and under-representation of locals in their personal networks, the kind of similarities I have referred to in the experience of internal and international migrants suggest that we should also look at factors such as the social temporality of migration and the inability of migrants to draw on a series of relationships made in the past.

This “configurational” logic is clear also when we look at the practices necessary to maintain any social tie – what Wellman (1985) with a play on words called “net work”. Also in migration studies, many writers (e.g. Schapendonk, 2015) have stressed the importance of such a focus. As I have said, when we do focus on practices of meetings, activities, talking, writing, the subjects of spoken and written discourse, we see how – even when the immediate interaction is between just two individuals – many other people are involved in the background.

9.4 Ties with Co-nationals, Ties with Locals

There is obviously a lot of variability in the tendency of migrants to form a “home boy” network in the place of immigration – variability on individual but also social lines. In my attempt to understand how the migration process affects social networks I want to compare skilled migrants and labour migrants. At least since the time of the Chicago School, descriptions of labour migration have stressed the fact that most migrants have relationships prevalently with other migrants, primarily from the same place of origin. This seems true also of internal labour migrants. We may cite some evidence from what is by far the largest migration flow in the world, viz. internal migration in China.Footnote 4 Yue et al.’s (2013) study of the social networks of migrants in a city in the province of Fujian found that over 99% of interviewees had ties with other migrants (the average number of such ties documented was 20), whereas only 39% had any significant ties with non-kin locals (Yue et al., 2013: 1715). Yue and colleagues point out (ibid., 1706) that migrants are not different in ethnicity or religion from locals and that “obstacles of language and culture shock” are less than those faced by most international migrants. So the massive predominance of other migrants in these migrants’ networks might seem striking.

Research on skilled migrants, in contrast, often finds people who have relatively few ties with co-nationals. In their interviews with Polish migrants in London, Ryan et al. (2008) found that while those in manual jobs often lived in almost exclusively Polish networks, professionals had much more mixed networks, including substantial numbers of British and in many cases few Poles. Gill and Bialski (2011) found a similar difference in the networks of “lower” as against “higher socio-economic groups”, with the former having very few ties with British locals, and the latter few ties with co-nationals.

Ryan and her colleagues attribute the difference to “cultural capital” (Ryan et al., 2008: 683) and Gill and Bialski stress the importance of fluency in English and of education. Command of the local language is of undeniable importance in the ability to form ties with locals (the professionals mentioned in both studies seem to have had a good command of English before coming to Britain) and no doubt other forms of cultural capital have a role.

However, other factors less linked to pre-migration attributes of individuals, and more linked to the channel of arrival also seem to have been important. Gill and Bialski (2011: 246) refer to the role of “institutional” ties in shaping the networks of their professional interviewees in Britain. Likewise, Ryan et al. (2008: 683), describing the ties of a medical professional who “had contact with several British professionals in his specialist area of medicine who advised him to come to London [for further study]” described a network “shaped by employment and the contacts [formed] at work rather than by specifically ethnic or migrant circles”. Numerous other studies of skilled migrants (e.g. Poros, 2001, Harvey, 2008) in various jobs and various countries document similar recruitment paths, and it seems likely that these personal recruitment channels via employees or employers with whom the migrant has already worked or had contacts, serves as an important basis for the construction of the personal network in the place of immigration.

9.5 Mobile Occupations

But how does this channel of arrival affect the ties skilled migrants establish in the new context? Clearly, a workplace where the majority are locals makes the formation of relationships with locals easier than a workplace where the majority of employees are other migrants, possibly from the same place of origin. As is well known, many labour migrants work in niches where the majority of the workers are migrants, the migration chains used to effect migration overlapping with chains of recruitment into a specific workplace or set of workplaces (cf. Grieco, 1987 for an exceptionally clear description of the pattern in a case of internal migration to unskilled jobs in a steel company). Although a certain number of skilled migrants are recruited to an organization via friends or even relatives (the literature on skilled migration gives a number of such cases) it is probably exceptional to find themselves in an office where co-nationals are the majority – a situation which is common on many building sites or among seasonal labourers on farms.

As I have argued, the lack of other sources of ties may make work a particularly important ambiance for migrants. In another research project, on skilled Irish migrants in Britain, Ryan (2015) found that several of her interviewees did stress the pleasures of associating with other Irish in Britain. However, inquiring into the origin of these friendships, it became clear that cultural similarity and shared nationality was certainly not the only basis of these relationships. Relationships with Irish migrants, like those with British people (which were also numerous), were made at university or at work.

Several interviewees were teachers, and had made many of their friends at work. One woman said “any new friends I make these days are teaching friends”; another “my main social network here [is] with teachers in school. I socialise with them quite a lot, like once a week, twice a week”. A third woman, asked how many of her friends were teachers said “It’s easier to say how many aren’t teachers” (Ryan, 2015: 13–14). Another, similarly, said that “all” her friends were teachers (ibid., 15).

Teaching is a profession associated with geographical mobility (for evidence from Italy, see Collucci & Gallo, 2017): on the one hand, teachers are needed wherever there are children, on the other hand young teachers at the beginning of their career may have to move in order to find a job. Ryan’s interviewees formed friendships with other teachers partly because they provided practical advice and emotional support in what could be a stressful and demanding environment, and reinforced their identity as a teacher (Ryan, 2015: 14–15). But the formation of a friendship presupposes that both parties are prepared to give their time. It seems likely that there may also have been other teachers in the same school, or a nearby school, who were relative newcomers to the area, and thus particularly open towards the formation of new relationships.

It is likely that some work ambiances are more favourable than others to the formation of relationships outside work, and the extension of a migrant’s network via the formation of contacts with other people, friends of the colleague, so forming a configuration of ties which is a basis for friendship. It may be that this kind of extension of a work relationship to the leisure sphere is more frequent in occupations where many persons are “migrants”.

Careers in certain occupations certainly have an intrinsic relationship to mobility. In his discussion of middle class occupations Watson (1964) made a distinction between “burgesses” and “spiralists”. The former were in occupations where it was important to consolidate a local reputation and clientele, perhaps building on the basis of relationships established prior to commencement of the business; Watson saw shopkeepers and artisans, but also professionals such as solicitors as burgesses. Spiralists on the other hand were in occupations where advancement in a career involved moving to another geographical location in the same large organization, but at a higher level, or to another similar organization again in a different place. Many managerial careers in the private sector had this kind of structure, as did many public sector jobs.

Bozon (1984: 48) makes a more general distinction between two very different strategies of affirmation sociale, on the one hand affirming an essentially local identity and life-style, on the other hand one predicated on non-local ties. Like Watson, he sees some occupations, such as cadres, as necessitating and valuing mobility, others, such as shopkeepers, but also ouvriers, as valuing and cultivating relationships with a local ambiance. Bozon does not develop this idea, but it might be fruitful to think of what we normally see as the style of life, family relations, and attitudes characteristic of particular occupations and social classes as the product not only of the education necessary to enter the occupation, and the organizational constraints and conditions of work, but also as a product of trajectories of geographical mobility.

In any case, when considering the difference between the networks which skilled migrants and labour migrants form in the immigration context, I suggest that we need to take into account the mobility trajectories of locals in the occupational ambiance where the migrant finds work.

9.6 Not Taking Links with Co-nationals for Granted

I have spent a lot of space analysing the difficulties which migrants of various kinds find in forming ties with locals, and the specific contexts in which ties are formed with locals.Footnote 5 But what can we say about the conditions in which migrants form ties with co-nationals? The common sense assumption that nationality per se is a basis for a relationship has been shown to be empirically unjustified by a host of research, including some I have mentioned. Accounts in the literature often describe relations formed before migration “back home”, but are mostly implicit on the way such pre-existing ties form the basis of a wider network established in the place of immigration including co-nationals not known before – almost as though it were natural for relations to be extended to other co-nationals, and therefore not worth explaining. As I have mentioned, I believe that we need to examine the practices which maintain ties, and the configurational context in which such practices are embedded.

Some literature stresses the very instrumental or casual bases of relationships with co-nationals. Gill and Bialski (2011) describe Polish unskilled migrants forming relationships with fellow Poles in Britain simply because they were taken to a house where other Poles lived, and because they worked with these co-nationals; but these relationships were described as superficial and soon dropped out of interviewees’ networks, also because they saw them as competitors who could not be trusted. Many other migrants recount similar turnover in the early months and years of migration, as also conflict-ridden and instrumental relationships (Cvaijner, 2019).

The pattern of casual or instrumental ties soon lost is perhaps particularly common when migration chains have not become consolidated. For when there is a certain number of migrants from a particular origin, a new migrant is more likely to be given a bed by a cousin or a friend, who will perhaps also present the new arrival to an employer, or at least give advice as to where to look for a job. This initial contact may then become an important basis for the whole network formed in the immigration context. What is important in this context is not just the fact that the tie may be strong at a dyadic level (after all, relationships between siblings can be conflictual, and many relationships with relatives or acquaintances back in the emigration context may have been superficial). But if there is a certain number of relatives and friends in the local immigration context, and there are also contacts with relatives and friends back home, the relationship is just one part of a wider configuration involving several people and various activities, exchanges and conversations. Thus the kind of practices which are a basis for trust and control, and the creation of shared ideals of behaviour. Hence a more feasible basis for a future network in the immigration context than the ties described in the previous paragraph.

It should be clear that a certain number of relatives and friends or acquaintances is a necessary condition for this kind of pattern. Many skilled migrants probably do not have a sufficient number of kin or others from their pre-migration network to make them the basis of their network in the place of immigration. I suggest this is one of the reasons why unskilled migrants are more likely than skilled migrants to have networks dominated by persons from the same geographical origin. Of course, what is necessary is a certain number of people from one’s network back home, not the number of persons of the same nationality.Footnote 6 A Romanian in the Italian city of Turin not arriving via a migration chain could have no ties at all with co-nationals even though Romanians form nearly 6% of the city’s total population. However, since most labour migrants do arrive via a migration chain, there is liable to be a certain number of people who know each other and others back in a specific place of emigration (in the case of Romanians in Turin, and the small town of Marginea, see Cingolani, 2009).

In this chapter I have not wanted to go into questions of ethnic identification. But it should be noted that the kind of rather dense networks I have described as produced by labour migration are also those which tend to be associated with clear ethnic identifications (Lubbers et al., 2007).

9.7 Migration Networks and Recruitment Networks

These different patterns of social networks established via different migration trajectories have important effects on the type of access to the labour market migrants have. I say “migration trajectories” not migrants with particular types of skills (skilled migrants vs. unskilled migrants), because we know that many migrants experience downward occupational mobility, so that even well-educated migrants who follow a labour migrant trajectory may remain stuck in “immigrant jobs”. I argue that this is a question of networks. As is clear from the examples I have given, and innumerable others in the literature, directly or indirectly, the social ties used to effect the move from a place of emigration to a place of immigration are often the social ties used to obtain a job. Some migrants move to a particular place precisely because they have obtained a job there. Many others do not have a job fixed up but do have information from their contacts that jobs are available in a particular workplace or a particular segment of the local labour market. To generalize, it can be said that “migration networks” in the sense of the ties used to effect change of place, are in one way or another often also “recruitment networks”. It is this which makes it sensible to reason in terms of what I have called “migration trajectories”. And to argue that migration trajectories structure both the social lives migrants create in the place of immigration (for example, around co-nationals, around kin, or around colleagues) and the jobs they find.

Focusing like this on networks rather than on the skills of migrants helps to understand not only the de-skilling and downward mobility which is common in labour migration, but also the fact that even migrants who do not have language barriers, and have marketable skills, may remain in a situation of blocked mobility.

Examples are provided by McGregor (2007) and Thondhlana et al. (2016) in their accounts of Zimbabwean migrants in Britain. The majority of these migrants had worked in professional, semi-professional or managerial jobs in Zimbabwe, or had been university students; but in Britain they worked in poorly-paid, low status jobs, often with inconvenient hours. More specifically, the majority worked in one particular “unskilled” job as carers of the elderly or disabled. As McGregor (2007: 810) says, The fact that so many Zimbabweans have gone into care partly reflects the availability of work, but also the clustering of Zimbabwean social networks around the industry. As the Zimbabwean community has grown, its members have passed on their experience to newcomers, and many entered care work through personal introductions or efforts on the part of friends and relatives already working as carers”. Thondhlana et al. (2016: 584) note that, on arrival in Britain, most of these highly qualified migrants had “exclusively Zimbabwean” networks. This gave them access to the care niche in the labour market established by earlier migrants (often non-degreed nurses or teachers) but not to other jobs. For some, confinement to this niche was due to being undocumented or having a visa (e.g. as a student) which limited the hours they could work “legally”. But many who were documented, could work full-time, and had the qualifications to do better-paying and more prestigious jobs, were also in the care sector. Thondhlana and her colleagues stress the role that Zimbabwean networks have in lowering the aspirations of new arrivals, convincing them that they would only be able to obtain unskilled work, and offering contacts to jobs in the care sector, and I suspect that many network mechanisms were in play.

The school and university qualifications of these Zimbabwean migrants were usually those of British examination boards, and all interviewees spoke fluent English, so the concentration in work in unskilled niches cannot be explained in terms of human capital. A minority of Thondhlana, Madziva and McGrath’s interviewees had professional ties in Britain even before emigrating and thus took a different “route of migration” (2016: 576) and went straight into professional jobs. But the majority did not have such ties, and thus arrived in Britain via family reunification, as students or asylum seekers, and formed ties with a Zimbabwean network in London or other parts of Britain which provided “no ties to immediate entry into professional jobs”, and indeed had an “immobilizing” effect (2016: 584).

This Zimbabwean case seems a demonstration of the pattern where a certain “mass” of migrants from a particular origin forms “community” networks which channel migrants who have no alternative ties into a niche in the secondary labour market. In this particular case, the protagonists are downwardly-mobile skilled migrants. But it is worth thinking of the implications also for similar patterns among less educated migrants, who, like these skilled Zimbabweans, form a network around co-nationals. It is in fact an important aspect of labour migration that (with the exception of self-employment) there is little occupational mobility to better jobs, an immobility which is not explicable solely in terms of education or work experience. It seems possible that the mechanisms I have described whereby networks are re-organized in “the migration process” are also the reason why so many labour migrants remain blocked in “immigrant jobs” for all their lives.

This seems to be true of both international labour migration and internal labour migration. In the case of internal migration to Turin in the mid to late twentieth century, Eve and Ceravolo (2016) show the concentration of regional migrants in “typically migrant” sectors, from building to domestic service, even many years after their move to Turin; Ramella’s (2003) qualitative interviews also show the lack of occupational mobility of these migrants. Cases like this from internal migration suggest that there are other factors in play than knowledge of the language or citizenship. I believe it is more likely that the networks formed in this classic kind of chain migration channel information about job opportunities towards very few niches, without giving information about possible alternatives which might be feasible. In fact the kind of network constructed in the place of immigration via chain migration linking together kin and friends from the place of origin in a new pattern of ties creates a very occupationally homogeneous network, which has the benefit of offering numerous links to jobs within a few niches of the labour market, but only those.

It is also interesting to consider migrants who, notwithstanding an initial lack of a network connecting them in any way to a professional job eventually do manage to obtain one. Louise Ryan’s chapter in this volume (also Ryan, 2016) gives examples of highly-educated Polish migrants who came to London without any professional ties. In spite of their qualifications, and often good English, the lack of any connection with local professional networks made it impossible to get skilled work in their own field, or indeed any skilled job. However, through ties with other people met at work, in the neighbourhood or at parties, these persons eventually came into contact with people in the professional ambiance they were interested in. In other words, the network they gradually built up in London provided information, allowed the interviewees to acquire skills, confidence and appropriate language to be able to present themselves as a plausible candidate, sometimes orienting them to opportunities they would never have thought of. Cases like these seem to show first of all, that networks are essential in one way or another: human capital is not enough on its own. But it is also interesting that migrants who are socially mobile in this way do not seem to come via a classic chain migration or to have a network in London of relatives or friends doing low-skilled work. Lacking the kind of network which the Zimbabweans referred to earlier had, they create a more heterogeneous one, forming ties with persons met at work, through their children’s school, in other words in the type of contexts where some of the internal migrants I referred to earlier in this chapter made ties if there were few people from “back home” in the place of immigration. So educated migrants who have no professional ties in the immigration context may eventually get a qualified job. But a migrant who emigrates via a network which provides no connections with a professional ambiance needs to build up ties within that ambiance. If kin, friends and acquaintances are inserted in a low-skill niche in the secondary labour market, they will not be able to provide the information, encouragement, or financial support which may be necessary, nor provide contacts with other persons who might be able to. For migrants, as for other people, “integration into the labour market” is a social process, not something effected totally on one’s own. For this reason, the structure of one’s social network, including the occupational ambiances it gives access to, may be crucial. Someone who arrives via a network which has nothing to do with their skilled profession may take years to construct ties within a professional ambiance – and not all migrants have that time. Hence many educated migrants stay in low-skilled jobs.

Analysing migrants’ careers in the labour markets of the country of immigration in terms of networks thus sheds light on factors known to be common in migration such as blocked mobility and over-qualification. But also on the very nature of labour migration.

9.8 How Does the Migration Process Affect the Networks of Children of Migrants?

As Boyd and Nowak (2012: 88) point out in their review of the literature, there is not an enormous amount of work on the networks of children of migrants. It might be added that what exists is mostly on the question of the ties “the second generation” forms with children of locals as against young people of the same national “origins”. The concept of “second generation” has always been unclear. While a few scholars have rejected it in toto (Le Bras, 1998), most continue to use it, and the concept does have sense in circumstances where children of migrants are classified by others in ethnic terms, have an ethnic auto-identification, and where this classification is not just an “optional identity” but one which shapes social destiny in important ways. However, in this chapter I am trying to examine the way migration itself, rather than ethnic identification, affects social networks. So what effect could parents’ migratory trajectory have on children’s social networks?

As noted earlier, studies which ask for the origin of social ties find that a considerable number come from family and kin. So to the extent that migrants’ social networks have specific traits this is liable to have effects on the traits which children of migrants “inherit”. In this section I will focus on a few ways in which children of labour migrants have networks shaped by their parents’ migration trajectory. I argue that the specificity of these networks has important effects on educational trajectories, and on contact with particular occupational ambiances. Directly or indirectly, therefore, these networks have important effects on the types of jobs children of labour migrants end up in.

Many of the friendships which children and adolescents form are local, being based on the neighbourhood or a local school. So it is worth noting that the housing trajectory of labour migrants is very specific, even compared with local working-class families. In an early period, many migrant families make several moves within the area of settlement, in an attempt to get out of run-down housing, to get a better job or to move closer to relatives or friends. This initial period of numerous moves has also been documented for internal labour migration by Ramella (2011). These moves of home naturally often cause changes of school, which may have an effect on educational attainment (most research on the effects of school changes shows that it damages school attainment: see e.g. Gasper et al., 2012). And on friendship networks.

Even apart from this initial period of high mobility, the neighbourhoods where migrant families end up living are specific, even compared to the non-immigrant working class. Once again, this comes out in twentieth century labour migration to Turin, where regional migrants were distributed much less evenly over the various parts of the city, more concentrated than the local working class in new areas in the south and the north of the city (Eve & Ceravolo, 2016). A series of mechanisms connected with mass migration itself probably contributes to this specificity of the neighbourhoods where migrant families settle, and where children of migrants grow up. The local working class, via their network of ties to relatives and to institutions, is liable to have more information about available housing, and more ties to owners of property to let (Badino, 2018). The very fact that, in an initial period, many migrants are in housing which is in poor condition or overcrowded puts them in a good position with regard to the criteria adopted for access to public housing. So it is not surprising that in many countries, and in different historical periods, public housing seems to have gone prevalently to the families of internal or international labour migrants. Mass migration eventually leads to the construction of new housing, and once again, due to their lack of ties with existing property owners, migrant families may move into this new housing built by the private sector – in Europe, on the outskirts of cities, less well-served by services, and by established schools.

9.9 Numbers of Children and Young People

Another important effect of labour migration regards numbers. A “wave” of migration in response to a local demand for migrant labour brings a large number of mostly young adults to an immigration destination, and to specific neighbourhoods. After a few years, young migrants have children (or bring children from the place of emigration) creating a “boom” of children, and then of adolescents and young people in certain micro-neighbourhoods and schools. At a certain point in the migration cycle children and adolescents may make up a large part of the neighbourhood: in the huge estate on the outskirts of Paris where Lepoutre (1997: 32) did his research no less than 38% of the total population was under 20.

In his analysis of American inner cities in the 1970s and ‘80s William Julius Wilson (1987: 36) claimed that one “cannot overemphasize” the importance of large numbers of young people. Without suggesting that numbers always have the same effects, it would be wrong to neglect the numerical dimension and the influence it may have on social networks.

Certainly, at least since the time of Street Corner Society (Whyte, 1943), accounts of migrant neighbourhoods have noted the intense social life of adolescents and young people (especially males), in spaces around flats inhabited by migrant families. And many accounts of childhood and adolescent sociability (especially that of boys, generally less controlled than girls) stress the pull of groups hanging round outside the flat. Drawing on his interviews with the children of the Algerian-origin Belhoumi family in France, Stéphane Beaud (2018: 143) says of the boys “school was not their priority during childhood and adolescence. The group of friends was more important, and it was not easy for them to subtract themselves from their obligations to the group”, spending time at home doing homework or engaging in educational activities. This was just as true in the case of internal labour migration. Anna Badino’s interviews with children of migrants from the south of Italy, whose parents came up to Turin in the labour migration of the 1960s, make it clear that the pattern was similar, especially for boys: in the words of her interviewees “We were always in the squares in front of the flats” (Badino, 2012: 79) “We lived outside” (ibid.: 82).

Citing teachers from the time, Badino argues that this “street sociability” had a significant effect on commitment to school work, eventually leading to the pattern of early school-leaving which was very common (especially for boys) among children of regional migrants in Turin, net of effects of class as conventionally considered (Eve & Ceravolo, 2016). A similar connection between school achievement and the childhood and adolescent sociability of boys is made by one of the Belhoumi sisters who says: “Like most girls of Maghreb origin in the local neighbourhood, we did much better at school than our brothers… from a young age the boys went out and spent lots of time with their friends without bothering much about studying” (Beaud, 2018: 144). Other children of migrants report a similar conflict of loyalties between neighbourhood friends and the different style of life and commitments connected with high school and university (Amrani & Beaud, 2005; Beaud, 2002).

9.10 Relations with Neighbours in the Second Generation

As I have argued in earlier sections of this chapter, if there is not a re-grouping of relatives and friends from the place of emigration in the same neighbourhood in the place of immigration, neighbours will not be former school-mates or relatives. So instead of the multiplex relationship which exists in many places where most inhabitants have lived in the same place for two generations, neighbours will just be the people who live next door.

This has implications for adults (first generation migrants) but also for the relationships which the younger generation has with neighbours. In their analysis of social relationships in a small town in the English Midlands in the 1960s, Elias and Scotson (1994 [1965]) describe the consequences this has for social controlFootnote 7 over children and adolescents. In the area inhabited by families who knew each other, the authority of neighbours was recognized, whereas in the neighbourhood where there were few such long-standing ties between neighbours, attempts by adults to intervene in children’s or adolescents’ behaviour were openly scorned. Using Coleman’s (1988) term, in this latter neighbourhood, inhabited by families coming from other parts of England, there was little “intergenerational closure” in the networks outside the family because adults in the various families had few ties with each other. In the neighbourhood inhabited by local families present for two generations, in contrast, ties between the families were strong, and norms were easily enforced over all – in the way Coleman described in his theory of “closure of social networks” (1988: S105–S108). The autonomy and relative freedom of groups of adolescents (at least boys) which emerges from many accounts of children of migrants can be understood also in this relative lack of close adult control. 

As is well known, Coleman argues that such closure has important effects on parents’ ability to control children’s behaviour, on their ability to impose adult norms such as the value of schooling, and thus effects on children’s educational attainment.

To take another influential theory in educational sociology, the relatively autonomous peer-group sociability I have described as common in many neighbourhoods of internal or international migration is certainly a long way from what Annette Lareau (2011) calls “concerted cultivation”. In this latter style of upbringing, children spend a lot of time in home-based activities with adults present, or in organizational contexts, from swimming lessons to theatre, once again adult-supervised. This mode of upbringing is contrasted with a style of “natural growth” where children are given more autonomy and spend much of their free time with other children in activities organized spontaneously. Lareau argues that the close interaction with adults characteristic of concerted cultivation develops the kind of attitudes and language skills which schools reward.

Of course, Lareau’s distinction between styles of upbringing concerned class divisions not migration. But a series of conditions in many migrant families and migrant neighbourhoods make concerted cultivation difficult to put into practice. Many organized sports and cultural activities require money and perhaps a car, and time, to ferry children from one activity and event to another. The style also requires the collaboration of other parents willing to organize their time in similar ways, and many migrant families do not have that kind of relationship with neighbours. Home-based activities with friends invited home may require more space than the cramped flats many migrant families live in. On all these dimensions, labour migrants are liable to have less capabilities than the average local working class, although the situation may vary according to what a particular national education system makes necessary (Keskiner, 2015).

In other words, it seems likely both that the migration trajectories of parents have effects on the networks their children form, and that, at least in the case of mass labour migration, this probably has effects on educational results.

9.11 First Jobs

Apart from the effects which children of labour migrants’ social ties may have on their school career, and hence indirectly on the jobs they end up in, there are also more direct effects: for children of migrants, as for other young people, parents’ and parents’ friends’ networks are particularly important for the first job (Kramarz & Nordström, 2014). Unlike some children of middle-class parents, children of labour migrants who do well in the education system are of course unlikely to have contacts via parents who could help them get a qualified first job; this is one (negative) way in which parents’ inappropriate networks may affect children of migrants’ insertion into the labour market (for these dynamics, see the chapters in the present volume by Keskiner and Waldring; Rezai and Keskiner; Lang, Schneider and Pott). But for those who did not do well at school and are applying for a manual job, parents – or parents’ friends and acquaintances – may be able to help. We know in fact that many unskilled jobs in the labour market generally are obtained through strong ties (Grieco, 1987, 1995; Hanson & Pratt, 1991, 1992; Morris, 1984) so we can ask whether migrant parents are as capable of helping their children as non-migrant parents, and what kind of jobs they channel their children into. If parents have a network of acquaintances who work mainly in similar niches in the secondary labour market – as we have seen, networks of this kind are common among first generation migrants arriving via the labour market trajectory – it seems likely that this may affect the jobs they could guide their children into. In her chapter in this volume Yaël Brinbaum suggests that the jobs which family and friends do provide to some children of migrants in France may in some cases limit their occupational mobility, giving them contacts in the non-skilled sectors where they themselves are employed. However, if children are not attracted by “immigrant jobs”, they may be able to offer no help for other kinds of manual jobs (Keskiner, 2017). So this aspect of labour migrant parents’ social networks – in combination with the social networks children of migrants themselves form in the neighbourhoods they grow up in and the schools they go to – may be one of the reasons why many studies have found occupational disadvantage of children of migrants even net of education (Heath & Cheung, 2007). The French survey data Brinbaum uses in fact suggest that the low rates of occupation among children of North African and sub-Saharan African origins is related to the fact that few young people of these origins obtained jobs via networks. High rates of unemployment and economic inactivity among the first generation, combined with concentration in work unappealing to young people, seem to have potent effects also on the second generation.

It might be thought that the situation of children of labour migrants was similar to children of the local working class. But the position of labour migrants within the local labour market is very specific, as is also the housing trajectory, and the neighbourhoods families settle in. So the network contacts parents are able to provide, and the information local neighbourhood youth provide, are liable to be different from those non-immigrant working class parents have. For this reason, it seems justifiable to talk of networks shaped by parents’ labour migration trajectory, not just a general class effect.

9.12 Conclusion

Much of the work on migrants’ social networks is focused on the maintenance of ties with co-nationals “back home” and in the place of immigration, and on the existence, or the lack, of ties with members of the “majority population”. However, if we want to understand patterns of social mobility of migrants or their children, other aspects of their social networks may be at least as important. I do not claim to have written a complete survey of the way migration shapes the social networks of migrants and those of their children. For example, I have touched only indirectly on contacts with organizations (from childcare centres to schools to sports clubs) yet as Mario Small (2009) shows, many social ties originate in organizations, so it would be important to bring out the specificities of migrant families’ relationships with organizations, which certainly exist. And I have not touched on the specificities of the networks of migrants arriving via the asylum trajectory, specificities I believe help to explain the lower employment rate of refugees compared with other migrants (what is called the “refugee gap”) (Perino & Eve, 2017; Eve & Perino, 2018). However, I hope I have shown that migration does affect migrants’ networks and those of their children. This in itself is important, because as I have said, effects of migration – rather than of being identified as ethnically different or a citizen of another nation – are surprisingly little studied.

I have argued that migrants inevitably reorganize their social networks with migration, and have analysed some of the ways in which their social relationships in the place of immigration are influenced by the fact of having moved geographically, and thus not having had a personal and family past in the place of immigration. Many social transactions fundamental to the maintenance of social ties – from childcare to commensality – are difficult or impossible to perform at a distance. Migration is age-selective and this has effects on ties formed in the place of immigration. Labour migration has strong effects on the neighbourhoods migrant families end up in, and this in turn affect the schools their children go to and who they play with outside the home. Most migration is class-selective, forming networks more class-homogeneous in the place of immigration.

I have also argued that there are important differences in the networks established by migrants arriving via different “migration trajectories”. As mentioned, I have not distinguished between labour migrants and skilled migrants on the basis of their skills or individual attributes, but rather on the basis of the kind of networks they tend to follow. Thus the core of the argument is the relationship between a migrant’s network at T1 prior to emigration and at the start of their migration, and at T2 some years later. Individual attributes like education obviously have effects on the ties people form, but I have focused on network factors such as the presence in the place of migration of kin, or a cluster of acquaintances from back home, and on the social ties which these other people have. After all, networks grow in large part out of existing ties (Eve, 2002b). And as Portes and Rumbaut (2014: 141) say, “isolating themselves from the influence of kin and friends is quite difficult for newcomers in the early stages of adaptation” – which makes the “characteristics of the ethnic community” of “decisive importance” in “moulding their entry into the labour market”.

Scholars have often distinguished between migrants who draw on the resources of an “ethnic community” and migrants who follow a more “individual” path. As Nee and Alba (2004: 91) put it, “In adapting to life in the United States, immigrants generally choose between two paths: ‘ethnic’ strategies (which rely on strategies in their own communities) and ‘mainstream’ ones (which involve the American educational system and the open labor market).” Portes and Rumbaut (2014: 141) distinguish on similar lines between migrants who rely on networks in their ethnic community and those – usually professionals – who “accept jobs away from areas of ethnic concentration and who compete primarily on the basis of their own scarce skills”. As essays in this volume and much other work shows, however, professionals often make use of networks, directly or indirectly: it is just that they are different networks. As I have argued in this chapter, the difference between a more “community” pattern of social relations and a more “individual” or “mainstream” one can be understood precisely in terms of the social ties migrants have at the beginning of their migration: for this reason it seems to me to make sense to talk of “migration trajectories”.

It will be clear, I hope, that comparing different forms of migration is fruitful to understand the network mechanisms in play. An important aspect of such comparisons in this chapter has been that between internal and international migration. I certainly do not want to deny the massive importance which citizenship, legal status and ethnic identifications have on international migrants’ lives, but I believe that lives are also influenced importantly by the systematic social processes due to migration itself – and to particular migration trajectories. And since we can exclude factors linked to citizenship, internal migrants and their children provide illuminating evidence of the effects of migration processes as such. I have shown the similarities in the networks of internal and international migrants, the difficulties in forming ties with locals, and the tendency to form ties with those from “home” if these are present. I have argued that the tendency to form networks composed mainly of people from “home” can be understood not just in terms of supposed cultural similarity but in terms of the network mechanisms in play. Arriving via a migration chain has a significant effect on a migrant’s network in the place of destination independently of ethnicity. The professional ties used by many skilled migrants to effect their move have different consequences, often leading to a network in the place of destination heavily based on colleagues and persons known through colleagues – especially where many of these colleagues are also non-locals, as is true in many occupations. Relationships with colleagues, but also with neighbours, with kin and friends are specific for migrants: as the examples I have given from internal and international migration illustrate, migration has structural effects on networks.