At the beginning of this book, I characterised the pervading view of the migrant workers who came to France after WWII as ageless and immortal. Yet as was just seen in Chap. 7, the all-too-frequent deaths which occur in migrant worker hostels up and down the land have put paid to this myth of agelessness in the starkest possible manner. Intersecting with the fiction of the everlasting youth of the migrant workers is a second myth, the myth of return (Anwar 1979). Return was a dream to which so many migrant workers in France aspired, but as survey data presented in Chap. 1 showed, most older immigrants now view France as the home where they will live out their days, surrounded by their partners and children (Attias-Donfut et al. 2005). In time, return became less likely for the straightforward reason of family reunification: France is the only home which the migrant workers’ children and grandchildren have known.

Less straightforward, however, is the non-return of the migrant workers who are the subject of this book: those living out their days in the hostels, who did not reunify their families in France. At the outset of this research, hostel residents’ non-return was presented as a puzzle and dilemma, firstly insofar as their mobility preferences at retirement call into question the assumptions of the ‘myth of return’ literature, which explains non-return on the basis of family localisation. In this case, though, the grounds for non-return cannot be family localisation, since the men’s families remain in places of origin. Secondly, older hostel residents also remain unmoved by the financial incentives of a return homewards, where their French state pensions, paid in euros and fully transferable, would have far greater purchasing power. What this book has sought to do therefore is resolve this paradox, by asking:

What explains the hostel residents’ preference for back-and-forth mobility over definitive return at retirement?

It bears reiterating that the hostel population is far from being representative of the older immigrant population in France. Indeed, the hostel system itself can be viewed as an outlier of history: a last bastion of the postwar guestworker model in concrete form, long obsolete yet somehow still standing today. If this is the case, what benefits can be had from taking the hostels as a contemporary focus of study?

The first point to note is that there is very little empirical work in languages other than French on the migrant hostels, and that which exists is somewhat misleading. Under any circumstances it is hard to concede that a total complement of over a quarter of million migrant worker hostel beds constituted a “restricted scheme” (Castles et al. 1984: 29), as was the case in the mid-1970s heyday of hostel construction (Lévy-Vroelant 2007). Moreover, the lion’s share of Francophone work on the hostels concentrates - not without eminent justification – on hostel residents’ difficult passage to retirement, in particular their acute health and welfare needs. Generally, researchers have overlooked the dilemmas of late-in-life return which confront residents at retirement and the outwardly puzzling preference for back-and-forth migration over definitive return.

Literature gaps aside, this book has also aspired to make a substantive theoretical contribution, cognizant of the rich potential for theory building in the ageing-migration nexus (Torres and Karl 2016). The older hostel residents constitute a population which can generate significant new insights for migration theory, despite their unrepresentative situation. Indeed, the theoretical value of this research lies in the very fact that it takes an anomalous case. Of all the post-WWII labour migrants to France, the hostel residents would appear at first sight to be the most likely candidates for return given their family ties to home and the economic incentives of return. That they do not return challenges the arguments of family localisation and neo-classical economics. This prompted me to ask whether other theories might be better able to explain hostel residents’ mobility strategies. Thus in Sect. 8.1 below, I assess the rival theoretical models which were identified in previous chapters as having potential to explain the puzzling preference for the va-et-vient at retirement.

Following this theoretical evaluation, in Sect. 8.2 I return to one of the key ideas developed in the book, namely the concept of ‘home’. In Sect. 1.2 it was shown that hostel residents do not appear to be well integrated in the main reference groups of French society – family, local community, nation – and instead retain ties to analogous groups in the place of origin. Yet they do not return there definitively. Does this mean that they are therefore ‘homeless’ everywhere – ‘doubly absent’ as Sayad (1999) put it – or have they found home in new and unexpected places? As the approach with the greatest untapped potential among the theories considered here, I draw on Niklas Luhmann’s social systems theory to ask what ‘home’ can mean for the hostel residents. Moving beyond orthodox conceptions of ‘home’ as connection to a particular locality and/or social group, Luhmann draws attention to the mechanisms of societal inclusion in functionally differentiated systems such as the economy, law, politics, health, education and so on.

Luhmann’s insights about societal inclusion are also extremely pertinent in relation to a recent piece of legislation which goes to the heart of this book’s concern with late-in-life mobility and home. The effect of this legislation has been to create a financial aid – available uniquely to older hostel residents – aimed at encouraging more durable return to countries of origin. Section 8.3 closes the book by surveying the prospects for this legislation and the men in the hostels which it targets.

8.1 Explaining the Puzzle: The Returns from Theory

The guiding aim of the research presented in this book has been to overcome the limitations of migration theory when faced with the outwardly puzzling return decisions of the hostel residents at retirement. The myth of return literature relies on family localisation as the principal explanation for the lack of return amongst the post-WWII immigration cohort in Western Europe, yet clearly this explanation cannot apply to those in the hostels living far from their families. Neo-classical economics was a further approach which had to be discounted at the outset. As was discussed in Chap. 1, despite the purchasing power of their Euro-denominated pensions in countries of origin, hostel residents remain unmoved by this financial incentive to return.

Given that the approaches based on family localisation and neo-classical economics cannot account for hostel residents’ prolonged presence in France past retirement, I proposed to examine other theories of settlement and return, drawing on the new economics of labour migration, structuralism, transnationalism and social systems. To assess the explanatory potential of these theories, I will now return to some key points in the data.

Dependence-Assistance: The Distortion of the New Economics of Labour Migration

For a very few respondents, the neo-classical logic remains valid. As Djimé (Dembancané, Senegal) put it in Chap. 5: “When you are a pensioner here [Senegal], you live comfortably, you’re not the same as those who don’t have any resources, who don’t have a pension. It’s not the same life. Really, when you are a pensioner, you can live better here.” However, the lack of definitive return at retirement by most hostel residents calls into question these neo-classical precepts. Perceptions of individual wealth must be tempered by the size of the household in the pensioner’s charge. For my respondents, as discussed in Chap. 5, it did not make sense to talk of individual economic decision-making. Rather, the decision to migrate or to return is made on the basis of family circumstances, as remittances are a key element of household finances. This observation thus lends support to the theoretical approach known as the new economics of labour migration (NELM) (Stark 1991).

In the opinion of some residents, retirement means that there is less of an obligation to send remittances – “now it’s up to them to get by” (Fouad, 60s, Constantine, Algeria). This is consistent with the NELM assumption that an inter-generational transfer of responsibility in providing for the needs of the family occurs when the migrant breadwinner retires. For other respondents, however, the NELM logic becomes distorted and the assumption of an inter-generational transfer of responsibility is called into question. In these cases, the duty to provide for a significant number of dependents after retirement weighs heavily. Many residents I spoke to enumerated the large families in their charge.

A particularly oppressive state of dependence ensues if the general regime pension is not sufficient to meet the household’s needs. In this case, the only means for the family to attain solvency is for the pensioner to remain in France for more than 6 months of the year, in order to meet the conditions of eligibility for the old-age income support benefit (minimum vieillesse), topping up the pension to just over €700 per month. Among my own respondents, several explicitly mentioned household dependence on the minimum vieillesse as a constraint to more permanent return to the country of origin following retirement. It is worth reiterating that recourse to the minimum vieillesse can in no way be considered a manifestation of neo-classical rational actor behaviour: the extra money is channelled to household dependents, not retained by individual breadwinners. This brings to light normative questions over the limits and purpose of social protection in a transnational context. As Lindley (2009) notes, non-contributory income support benefits are a means by which the state guarantees a minimal standard of living for the least well-off in society. Yet some individuals may willingly accept a level of poverty below this standard in order to financially support those in yet more straitened circumstances in places of origin.

Losing One’s Bearings or Returning to Serve? Structuralist Interpretations

The structuralist approach stressed the difficulties of re-integrating to the home context, since migrants risk losing their place within the established order of “vested interests and traditional ways of thinking” when away (Cerase 1974: 258). This was a feature more of North African than West African accounts.

Many North Africans complained of having ‘lost their bearings’ in the place of origin. They had fallen out of touch with childhood friends and talked of how they would feel awkward when bumping into old acquaintances on the street. It was for this reason that they preferred to stay at home, among their family, when back in the country of origin. Others noted the obstacles to getting development projects off the ground. Respondents felt themselves to be blocked in these endeavours and blamed this on being outside informal local networks after the long years spent in France.

Such a narrative was not a feature of West African respondents’ accounts. The normative expectation weighing on respondents was to be of service to the community of origin at retirement, as Waly, an army veteran of Malian origin, noted in Chap. 6: “When you are retired, you should be of service over there; you return to better serve your village and your family” (Waly, 75, Kayès, Mali). For those who live at least some of the year in the hostel in France, such service takes place via hometown associations, which are often headquartered in hostels. Through such associations, hostel residents wield power in their local communities, despite their physical absence. As Catherine Quiminal notes, these associations have “transformed the absence of each migrant into a prominent political presence in the home village” (Quiminal 2002: 40; author’s translation).

West Africans, being the drivers of physical and social change thanks to their contributions to village development projects, were not disoriented upon return. Indeed, a virtuous circle operated whereby their political and social status increased as their hometowns gained in regional significance, which in turn was due to their ongoing development efforts. In the case of Dembancané, improvements to infrastructure there have seen it emerge as a local administrative centre and gain in prominence politically. Critical in this disparity between North and West African experiences of re-integration is the institutional dimension, namely the influence of transnational hometown associations. For West African respondents, these institutions lay the foundations for a political and social presence in places of origin despite long absences, and shape the possibilities for a successful re-integration. North African respondents, on the other hand, were confronted with difficulties when attempting to organise communal initiatives. Participation in hometown associations was far less significant for them.

The “Pressure of Communicability”: Evidence for a Transnational View

The importance of hometown associations in channelling migrant-homeland initiatives indicated that a transnational approach might be warranted. Indeed, the retired hostel residents are archetypal transnational migrants, engaging in frequent and durable economic, social and sometimes political ties between France and places of origin, managing two household budgets, communicating in two (or more) languages, and forever travelling back-and-forth across the Mediterranean. Furthermore, the motivation for mobility at retirement was often justified on the grounds of affective ties to home communities, as is stressed in transnational accounts (Cassarino 2004). It is only those hostel residents with a history of family conflict for whom return trips are not such a constant feature.

Other aspects which encourage a transnational reading are related to the empirical conditions for regular and sustained migrant-homeland interactions as discussed in Chap. 1. Admittedly, in terms of one of the factors stressed in this literature – the institutionalisation of transnationalism by government actors – there is little evidence to suggest that much attention is paid by sending states to elderly hostel residents.Footnote 1 According to one respondent, the consulates and homeland authorities take no interest in their ageing compatriots. “They don't give a damn… If they did there would be studies commissioned, with numbers and statistics on the people who have returned.” In terms of a second explanatory variable stressed in the transnationalism literature – ease of transport – this is important in explaining when and how hostel residents travel back-and-forth, but is not pertinent for the question as to why residents prefer the va-et-vient over definitive return, the main topic of analysis here.

However, a third transnational factor – ease of communication – does appear to be significant, although not in the manner one would expect. Chapter 5 highlighted that communication with family left behind is a central and often daily feature of life for hostel residents. Improved regularity in communications has been a recent development in residents’ lives following the expansion of access to mobile phone communications in origin countries. The transnationalism literature contends that regular communications can facilitate return, since individuals will be better informed about the environment they hope to return to. However for hostel residents, paradoxically, better communications appear to block return in some cases. Low-cost fixed and mobile telephony has resulted in new ‘pressures of communicability’, amplifying the burden of remittance dependency from left-behind relatives. In contrast to accounts which ascribe to ICT a pro-poor and developmental role which empowers people, these new technologies sometimes serve only to further ‘trap’ the men and impede their homeward return.

Your Papers, Please: Inclusion in Healthcare and Bureaucratic Organisations

Time and again, when meeting older residents for the first time, I was taken aback to hear them launch into a detailed explanation of why they were still in France. These justifications for their continued presence in France were quite unprompted on my part. What was striking was that they were usually based on one or both of the following elements – healthcare and administrative rationales.

With advancing years, health problems unsurprisingly become a more pressing concern. This was discussed in Chap. 4. Hostel residents, although not experiencing a greater incidence of ill health compared to the rest of the French population, tend to experience health problems earlier. As for paperwork, Chap. 3 detailed the many documents of personhood which the hostel residents are required to present in order to access social security benefits, both contributory (e.g. pensions) and non-contributory (e.g. income support, housing benefit). Valid proofs of identity, domicile, and income – such as residence permits, passports, annual tax declarations and rent receipts – are essential in this regard. However, this may be complicated if these biographical records do not conform to the standard assumed lifecourse institutionalised by the welfare state. This was seen, for example, in some respondents’ inaccurate birth certificates, erroneous or non-existent payslips, and duplicate social security reference numbers.

Proponents of economic theories might argue that healthcare and paperwork can be reduced to a simple utility calculation; the French healthcare system is heavily subsidised for all long-term residents, and free to the least well off. To benefit from an equivalent standard of care in the country of origin would require recourse to what can be a very costly private sector. Similarly in terms of administrative tasks, proponents of economic theories would contend that the only reason to keep paperwork in order is to prove eligibility for certain means-tested social security benefits.

In reply, I would argue that the prominence of such justifications for continued residence in France past retirement is not merely an economic rationale. Residence and mobility decisions are also constrained by a requirement to meet expectations of inclusion in French healthcare and legal-bureaucratic systems, so as to have continuity of care and to ensure that their entitlements are recognised. Such an interpretation does not seek to dismiss rational actor utility-based theories. Rather, the requirements for inclusion in the organisations of functionally-differentiated society, as theorised by Niklas Luhmann, broaden the scope of what a rational choice in migration decision-making can be.

For example with healthcare, respondents were concerned with ensuring a continuity of care and treatment. Over the years many hostel residents have developed relationships of trust with their doctors in France, and have strong faith in the superiority of the French healthcare system. These are powerful rationales for undertaking the va-et-vient. Furthermore, certain treatments and drugs are not available in the country of origin. In order to maintain a continuity of treatment, hostel residents plan their va-et-vient trips around appointments with doctors and consultants, and ensure that they have sufficient quantities of medication for the planned period of absence from France. Turning to paperwork, again, the rationale is not just a cost–benefit calculation for the hostel residents. What is at stake is their legitimacy to grow old in France: the right to a pension as a deferred salary, the right to a minimum standard of living, the right to home help services. In fact, the issue cuts to the very identity of the hostel residents. At retirement, a renegotiation of identity is necessary in order to compensate for the redundancy of the former raison d’être, which was employment (Bolzman et al. 2006). The passage to retirement is a transition whereby documents of personhood become critical tools of social inclusion.

Discussion

The foregoing has shown that at different points of the analysis there is support for all the theories discussed here. In short, no single theory accounts for all the patterns observed. However this does not imply that the theories concerned are too vague or inconsistent to be of any analytical use. Rather, the theories appear to be partly complementary (cf. de Haas et al. 2015). Furthermore, as Massey and colleagues maintain in a much-cited review of migration theory, the added value of each theory becomes most apparent when levels of analysis are kept distinct (Massey et al. 1993). Here I find Thomas Faist’s delineation of micro, meso and macro levels of analysis in international migration a useful typology by which to conceive of these dynamics (Faist 2000).

At the micro level, neo-classical economic explanations taking the individual as the unit of analysis were not validated since they predicted definitive return for retiring hostel residents: purchasing power differentials mean that a pension drawn in euros buys more in the country of origin than in France. Yet for retired hostel residents, it did not make sense to talk of individual economic decision-making, given the number of family members dependent on their remittances. The new economics of labour migration, with its focus on the utility-maximising household as the locus of decision and unit of analysis, made much more sense of the crucial role remittances have played and continue to play at the end of working life.

Beyond the micro level of individual decision-making, hostel residents wishing to return also have to negotiate the wider “web and content of ties on the intermediate level” (Faist 2000: 30). This includes the complex relations implicated in kinship and village ties. The structuralist focus on vested interests and traditional ways of thinking ‘back home’ illuminated the data here, not least the contrasting fortunes of North and West African respondents as they attempt to re-integrate socially in places of origin. Many North Africans alluded to having lost their ‘bearings’ in the home community. West Africans, on the other hand, mentioned the normative expectation of return at retirement in order to ‘serve’ their home communities. They have been able to convert a physical absence into a political presence via their hometown associations.

The significance of the hometown associations suggested that a transnational approach was warranted. This premise was further supported by the prominence of transport and communications technology as themes in hostel residents’ accounts. However, while transport factors are useful in explaining how and when the men travel, they do not address why they prefer back-and-forth migration over definitive return. Cheaper and more regular communications, on the other hand, had an unexpected effect. Instead of hostel residents being better prepared for return, the pressure of communicability they experienced by being ‘available and reachable’ via mobile phones (Tazanu 2012) sometimes impeded return. In sum, the transnational approach seems better suited to explaining the regularity and duration of back-and-forth trips. When confronted with the justifications for why residents keep on coming back to France, the explanatory variables identified by the transnationalism literature seem insufficient. Thus it was necessary to move up to the macro level of social systems.

Retired hostel residents repeatedly justified their presence in France on the grounds of healthcare and administrative reasons. It was shown that mobility decisions are constrained by requirements for inclusion in healthcare organisations, so as to maintain relationships of trust with medical professionals and to assure a continuity of medication and treatment. The same applies to administrative agencies. By keeping in good order their travel documents and other officially recognised proofs of identity, domicile, and income, hostel residents were able to ensure that their claims were recognised by the relevant authorities (cf. Mbodj-Pouye 2016).

A further insight which follows from this perspective is that different social systems become important to people at different stages of the lifecourse, with corresponding implications for migration decision-making. In Chap. 3 it was shown that hostel residents’ documents recording certain key deviations from the ‘standard’ expected lifecourse only become critical at retirement. Prior to this time such documents do not have a major influence on life chances and outcomes, although they are set in motion from birth and continue into adulthood and employment. While earlier in life it might be more appropriate to analyse return decisions in terms of economic success or failure (Gmelch 1980), later in life other factors become prominent. The idea that motivations for migration vary over the lifecourse is a rather straightforward and uncontroversial point. However, it is a point which is easily overlooked when the gaze of those doing and funding research is focused on younger people who move, be that workers, students or refugees. The growing body of literature on international retirement migration provides a useful and necessary counterpoint, and with this book I hope to have built on earlier conceptual and theoretical advances.

In terms of theoretical advances, I wish to highlight in particular my contribution to the literature on return decision-making at retirement (see Chap. 1 for an overview). While many migrants do return definitively, the available quantitative evidence indicates that those wishing to do so nonetheless constitute a minority, with the circular migration strategy of bi-residence being more popular (Attias-Donfut et al. 2005; Bolzman et al. 2016). The location of close relatives in the country of immigration – particularly a spouse or children – is highlighted as a major brake to definitive return (Ciobanu and Ramos 2015; Liversage and Mizrahi Mirdal 2017). A notable example is the work of Katy Gardner (2002), whose findings resonate strongly with the analysis presented in this book. Gardner (2002) discusses how her elderly Bengali respondents in London no longer felt at home in their communities of origin, and emphasised the value they placed on the British healthcare and welfare systems. Although she does not reference Luhmann’s theory of social systems, the parallels between our respective studies are clear. However, in Gardner’s study, definitive return was also shown to be blocked because her respondents had strong attachments to the UK through their British-born children and grandchildren. The added value of the present study is that by analysing the anomalous case of the hostel residents, who have not achieved family reunification, I have been able to isolate the critical importance of access to healthcare and pensioner allowances. This is an important contribution to theorising in this area, since it gives rise to a general proposition that accessing such resources is a key factor in return decision-making later in life.

8.2 A New Approach to Home: Domicile and Inclusion in Social Systems

Of the different theoretical paradigms evaluated above, social systems theory is the least well-known or applied in migration studies. One reason for this is that Luhmann wrote for the most part in German and his work is yet to be translated widely into other languages. A second reason is that Luhmann deliberately used abstract language in his theorising of society in order to minimise the risk, as he saw it, that readers would over-simplify his analysis (Moeller 2006). Nonetheless a handful of scholars have grasped the salience of Luhmann’s work in the migration field, foremost among them Michael Bommes (2000, 2012), Christina Boswell (2011, 2009 with Oana Ciobanu) and Jost Halfmann (2000).

In these final pages, I would like to build on these pioneering efforts by singling out a term which features prominently in migration research, including in the title to this book, namely the idea of ‘home’. There is a growing literature on the relation between home and migration (Ahmed et al. 2003; Al-Ali and Koser 2002; Boccagni 2017; Rapport and Dawson 1998). However, ageing migrants’ experiences of home and home-making were marginal to this body of work until recently (see Walsh and Näre 2016). More generally, home is a concept which is often deployed rather uncritically in migration studies, being “equated with a country of origin even if migrants might actually feel at home elsewhere” (Hatfield 2010: 245). One manifestation of this is the customary distinction between ‘home’ and ‘host’ societies.

The starting point for this book was the observation that geographically single hostel residents did not seem to be well integrated in the conventional reference groups of the French ‘host’ society – the family, the local community, the nation - and instead retained ties to analogous groups in the ‘home’ society. Paradoxically, however, they do not return definitively. These reference groups –family, local community, nation – represent various sites and scales in the literature on home. To conclude, I would like to explore what a Luhmannian concept of home could be.

As I have sought to stress throughout, I was often struck by the fact that hostel residents felt compelled to justify their presence in France on the basis of healthcare needs and administrative requirements, revealing the influence of medical and bureaucratic organisations on the men’s lives. This gave me pause for reflection: could inclusion in the organisations regulating access to different social systems also constitute a type of home for the men in the hostels, valued for ‘instrumental’ reasons insofar as it procures healthcare and a minimum standard of living?Footnote 2

If one takes a Luhmannian view, the answer is yes. What it means to be ‘included’ and to be ‘at home’ no longer necessarily implies only an emotional attachment to a territory (place) or social group (household, ethnic group, nation), as home has conventionally been conceptualised (Porteous 1976; Rapport and Overing 2007). Luhmann’s theory of social systems challenges the assumption that society is formed primarily of groups of individuals distributed across distinct, demarcated territories. This has been the dominant conception of society across history, encompassing kin groups, tribes, ethnic groups, and in the modern era, nation-states. Instead, Luhmann proposed an alternative model of society based on individuals’ acts of communication in distinct social realms differentiated by their function, such as politics, law, the economy, religion, science, and so on (Luhmann 1995). In pre-modern societies, social inclusion and life chances were determined by social strata and rank. But in the social systems which make up functionally differentiated society, social inclusion operates according to the codes and logics of each autonomous system. These criteria are fundamentally more inclusive than in pre-modern times: discrimination on the basis of inherited or ascribed characteristics such as ethnicity or class is no longer relevant. In principle, social systems tend to openness and inclusion: what matters is whether the “rules of access” specific to each system are met or not (Luhmann 1990: 35). For example, whether a job applicant has the required qualification profile, whether a patient has symptoms needing urgent medical treatment, whether a political candidate meets the constitutionally-defined conditions for holding public office. Functionally differentiated systems are “indifferent” (Luhmann 1990: 30) to all attributes other than whatever the relevant expectations happen to be.

Nonetheless, in practice, it is organisations in society (such as courts, hospitals, firms, schools, and so on) which regulate inclusion in social systems, and individuals may in fact be exposed to considerable risks of exclusion (e.g. unemployment). The role of organisations in functionally differentiated society has been most astutely analysed by Michael Bommes, who led the way in applying a systems theoretic approach to migration questions (Boswell and D’Amato 2012). Bommes argues that organisations in functionally differentiated society develop expectations about the typical person profile capable of being included as an organisational member (e.g. as an employee, a patient, a student etc.). To counter the risk of not being included, welfare state institutions have designed elaborate interventions in education, health, labour markets and so on, to enable individuals to meet these standardised expectations and thereby to participate fully in society. The effect of these interventions is to accompany and institutionalise the modern life course or career, based around three sequential stages: education in the formative years, employment and founding a family during adulthood, and retirement in old age.

The welfare state backs up the likelihood that careers can be built. (…) This means participation and access to education, work, the family, as well as to economic, legal, political and health resources. Participation in each single context implies the fulfilment of certain preconditions that are provided elsewhere (Bommes 2000: 93).

In such a conception of society, it is one’s observable and “biographically accumulated” record of inclusion and exclusion in different systems which determines future options for membership (Bommes 2000: 91). Hence the critical importance in the modern era of proofs of personhood such as birth certificates, passports, medical records, diplomas and training certificates, tax declarations, payslips, and social security numbers. As Bommes stresses, migrants are likely to be disadvantaged in this scenario as they have not benefited from the life course interventions institutionalised by the welfare state in the formative years (healthcare, education, contributions to social security and pension funds). Instead, their educational qualifications may not be recognised by the relevant organisations, and their later entry to the labour market means they have accumulated fewer contributions in social security funds. Likewise, to turn to the specific case of the hostel residents, previous chapters have underlined the problems arising later in life from inaccurate dates of birth on proofs of identity, and the consequences of unscrupulous employers not declaring migrant workers on their payrolls.

As has further been underlined in previous chapters, for hostel residents it is vitally important to observe minimum periods of residence in France in order to access certain forms of social protection, be it cash or in-kind benefits. In effect, they have to prove that they are domiciled in France. The etymological connection between home and domicile is evident given the latter word’s Latin roots, from domicilium ‘dwelling’, in turn from domus ‘home’. Domicile in modern English usage can signify “a person’s home” as well as “the country in which a person has permanent residence” (Concise Oxford Dictionary 1999), giving it purchase on the multi-scalar nature of home upon which geographers have insisted (Blunt 2005). Nonetheless, the potential of the term domicile as a conceptual tool for unpacking the notion of home has barely been explored by social scientists.Footnote 3 Instead, domicile is discussed above all in legal and fiscal contexts (Schwarz 2012).

I argued earlier that explaining back-and-forth mobility at retirement in systems-theoretical terms cannot be reduced to a simple cost-benefit calculation for the hostel residents. What is at stake is their legitimacy to grow old in France: the right to a pension as a deferred salary, the right to a good standard of medical care, the right to home help services. As the epigraph to Chap. 3 eloquently evoked, having one’s ‘papers’ in order is fundamental to upholding these rights. Papers also need to be in order when contemplating international migration itself: for many of my respondents, the realisation that they needed proofs of identity in order to travel was the prompting factor in registering (i.e., registering their identity, their personhood) with the authorities. Many of my respondents did not procure birth certificates, let alone passports, until the time when, as young men, they thought about emigrating.

This underlines the initial transition the men had to make from a society of origin where such modern proofs of identity were unnecessary for day-to-day living, to a new society where such documents of personhood are essential and everyday tools of inclusion. In Luhmannian terms, this individual act of migrating to Western Europe represents the transition to a functionally-differentiated society in microcosm. The men had spent the first twenty-five years of their lives quite happily without a piece of paper stating who they were. This changed abruptly when they began to be confronted by state officials demanding certain documents in exchange for certain rights being recognised. These demands have only grown stronger as the men make the transition to retirement, a testament of their life-long struggle to achieve social inclusion and uphold their rights.

In summary, home can be about belonging to a territory, a localisable space, just as it can be about belonging to a social collective (household, community, nation). But I have argued that to be ‘at home’ can also mean upholding one’s claims to domicile and personhood in order to achieve inclusion in different social systems. To repeat, this is not to deny the importance of place-belonging and group-belonging as constitutive of home. Instead, it is to recognise the growing complexity of social relations. If Luhmann’s claim that society has become more complex and differentiated is accepted, then it becomes interesting to explore the implications of his theory for our thinking about home. Luhmann’s insights about societal inclusion are particularly relevant in light of a recent piece of legislation which uniquely targets older hostel residents and their relationship to home. The final section of the book considers the prospects for this legislation and the people it is addressed to.

8.3 Concluding Prospects: Steering for Home?

On January 1st 2016, as I was finishing what I thought was the final draft of this manuscript, a new piece of legislation came into effect in France. Applicable only to people living in migrant worker hostels over the age of 65, this new law had a direct bearing on this book’s central question about return at retirement. The primary object of the legislation was to institute a financial aid to encourage ageing hostel residents to return to their countries of origin on a longer-term basis. This new benefit is called the Aid for Familial and Social ReinsertionFootnote 4 and given the topic I could hardly have submitted the manuscript to the publishers without writing about it.Footnote 5 A further round of fieldwork in France therefore followed (January–March 2016), interviewing several individuals with close familiarity of the drafting of this law. What was particularly interesting about this new law was that it appeared to contradict the principles of territoriality and nationality which have long shaped the institution of the welfare state (Böcker and Hunter 2017).

The nationality principle results from the fact that welfare states evolved historically as part of a wider process of nation-state formation. Nation-states, therefore, are said to have special responsibilities towards their own nationals and may refuse social protection to non-nationals. In the latter part of the twentieth century, nationality conditions have increasingly given way to residence conditions as a means to define eligibility for social protection (Soysal 1994). Nonetheless, the closure reflex of welfare states may still operate at an earlier stage insofar as immigration policies regulating entry to state territory may deny entry if a prospective immigrant is more likely to be a net recipient of welfare rather than a net contributor.

The territoriality principle requires that social protection is restricted to those persons residing on the state territory. By extension this implies that the consumption of welfare benefits should be restricted to the state territory and that benefits should not be exportable should the recipient move his or her residence (Halfmann 2000). Regarding exportability, an important distinction is to be made between contributory benefits (i.e. those financed by an individual’s social security contributions) such as pensions, and non-contributory benefits such as old-age income support (financed by general taxation). This was discussed in Chap. 3, where it was noted that thanks to bilateral social security conventions between France and countries of origin, hostel residents are free to transfer their state pensions to countries of origin. By contrast, non-contributory benefits such as the minimum vieillesse (old-age income support) may not be exported. As was pointed out in Chap. 3, due to the biographical specificity of their careers (employment in low-paid sectors; fraudulent employers not declaring employees for social security purposes etc.) many hostel residents did not contribute much to their pension funds and therefore claim the minimum vieillesse to supplement their incomes at retirement. The fact that this benefit is non-exportable and requires at least 6 months’ presence in France was the stated rationale why many did not return to their families at retirement on a longer-term basis. In effect, the temporal and territorial requirements of inclusion in the welfare state timetabled their presence in France.

The difficulties arising from this situation have not gone unnoticed by policymakers. Indeed, already in 2004 French Prime Minister Jean-Pierre Raffarin wrote to the High Council for Integration (HCI), the agency tasked with monitoring migrant integration issues, to request a written ruling on the issues facing France’s ageing migrant workers. The HCI report drew special attention to the 6-month residence condition governing access to old-age income support. This was labelled a “de facto inequality” facing older migrants, since foreigners were more likely than French-born to claim the minimum vieillesse and the residence rule obliged some to remain in France rather than return home as they wished (HCI 2006: 126; author’s translation). Relatedly, the report noted the burden on public funds occasioned by providing medical care and long-term elder care to this ageing population, particularly those living alone in the hostels who do not have family members nearby to provide informal support.

It was in this context that the idea for an Aid for Familial and Social Reinsertion was born, championed by Jean-Louis Borloo, Minister of Employment, Social Cohesion and Housing from 2004 to 2007. In essence, the Aid provides a lifelong income to all foreign nationals (including EU citizens) aged over 65Footnote 6 who choose to return to their countries of origin for more than 6 months per year. Several restrictions apply. Importantly, to be eligible one must reside in a migrant worker hostel (or a former hostel now operating as a résidence sociale). It is also stringently means-tested, meaning that those with annual incomes above €6600 are not eligible. As regards the practicalities of payment, beneficiaries receive a yearly sum which is calculated differentially so as to top up their income to €6600 (equivalent to €550 per month). This income threshold may seem shockingly low, but it should be recalled that due to exchange rates and purchasing power differentials €550 goes much further in Mali or Morocco than it does in Marseille or Montreuil (see Sect. 1.2).

While the numbers eligible for the Aid are strictly circumscribed, its wider significance in radically challenging the foundational principles of the welfare state are hard to overstate. This in large part explains the long delay between the passing of the legislation in parliament (2007) and its implementation (2016).Footnote 7 In effect, the principles of nationality and territoriality are turned on their head. The Aid is akin to a non-contributory benefit funded by general taxation, yet it is only open to non-nationals and, what is more, only to those non-nationals who have moved their permanent residence away from the national territory. This prompts the intriguing question as to why French policymakers would open up the welfare state in such a radical and potentially destabilising manner. This risk was made explicit by Yannick Imbert, director of the French Office for Immigration and Integration (OFII), during his appearance before the Parliamentary Commission on Older Migrants: “Would the French public accept that people no longer residing in our country could continue to receive benefits from French agencies?” (Bachelay 2013: 201; author’s translation) The issue is all the more intriguing when one considers that this legislation was developed at a time when in other respects European welfare states were in sharp retreat.

Close analysis of publicly available documentation, as well as ten interviews with politicians, civil servants, union officials and migrant rights activists well-acquainted with the dossier, shows that policymakers had two main motivations in drafting this legislation. The first chimes with a broader tendency that I have elsewhere called ‘institutional repentance’ (Böcker and Hunter 2017), that is to say an admission that many hostels are ill-adapted to an ageing clientele, coupled with recognition of the “sacrifices made by these workers for the economic development of France” (République Française 2007; author’s translation). In an interview, Rachid Bouzidi, a ministerial adviser, outlined why Minister Borloo was committed to the policy:

Quite simply the fact that in the cabinet of Jean-Louis Borloo we believe that everyone has the right to live with his family in a decent and dignified way. We believe that it is abnormal that chibanis Footnote 8 remain in France against their will, only to retain their modest incomes and access to health. These men primarily arrived in the 1960s to work, because France asked them to. At that time, not a road, not a bridge, not a building was built without their help. So today, to permit them to live as they wish is not special treatment. It’s simply the right thing to do (Raouf 2007; author’s translation).

A second, more cynical, interpretation is that the true intention of policymakers was to empty the hostels once and for all of a public which poses increasing burdens on public services (Dimier 2007). The fact that the law is targeted exclusively at those living in hostels, rather than older foreigners in other types of housing who may well harbour similar wishes to return, rather reinforces this suspicion. On the one hand, the legislation could be interpreted as a means to accelerate the transformation of the hostels by replacing the older residents with younger publics. This has been a long-term aim of Sonacotra-Adoma’s stakeholders in government since the mid-1990s (Bernardot 2008). As detailed in Sect. 2.2, the company has made efforts to diversify its clientele to include young workers on short-term contracts, single mothers, homeless people and – increasingly – asylum seekers. This interpretation was strongly shared by one of my interviewees, ex-director of a prominent migrant rights organisation: “[The aim of the legislation] is uniquely to empty the hostels.” On the other hand, emptying the hostels of their ageing clientele would also save resources in public services, notably medical and elder care. The same interviewee concurred: "For the government it's a win-win situation, 100%. They empty the hostels, and they make budgetary savings.” As hinted in the HCI report, and elaborated elsewhere in this book (see Chaps. 4 and 7 especially), caring for older hostel residents constitutes a non-negligible drain on resources in certain municipalities. Borloo himself raised this point in his testimony to the Parliamentary Mission on Older Migrants in 2013, describing the Aid as “a gesture of Republican dignity which moreover would cost France nothing … One might even regard [the hostel residents’] return visits home as generating savings for our public services, notably in health” (Bachelay 2013: 445–446).

The question as to which of the two rationales dominated is open to discussion, and indeed Borloo’s testimony here indicates that both were at stake. But in any case what is beyond dispute is the outcome desired by policymakers, namely to facilitate the departure of the older men from the hostels on a more durable basis, through a combination of legislation and financial incentives. The Aid for Familial and Social Reinsertion is thus a clear example of ‘societal steering’, that is to say action taken by political or legal institutions to influence the behaviour of individuals in society. The capacity of such institutions to do so has been debated and theorised at length by sociologists, notably Habermas (1984) and Luhmann (1990), and this literature is generally very sceptical that the legal and political systems can directly and causally steer individual action (Boswell 2011). What then are the prospects for this policy and the men in the hostels which it targets? Will it finally result in the mass return of the ageing residents, and the emptying of the hostels?

It is admittedly early to offer a judgement, given the recent introduction of the Aid. Yet the available evidence indicates that a mass exodus from the hostels is highly unlikely any time soon, at least if the current low income threshold is retained. As of May 2016, 5 months after the law’s entry into force, a grand total of four people had filed an application for the Aid.Footnote 9 The body responsible for administering it, the Caisse des dépôts et consignations, had in addition fielded about a 100 telephone inquiries, and the website dedicated to the Aid had been consulted some 400 times. These figures contrast with the 35,000 potential beneficiaries cited on the government’s press release announcing the entry into force of the legislation.Footnote 10 One of my interviewees, a migrant rights activist, explained hostel residents’ lack of interest in the scheme. Firstly, most retired hostel residents have pensions between €500 and €600 per month. So they do not gain very much, if anything, by this measure. They prefer therefore to stay in France for more than 6 months per year in order to be eligible for the minimum vieillesse, which brings their monthly income up to €800 (as of 2016). Only those who have very small pensions, say €250 per month or less, may find the measure attractive: by taking it they could spend most of their time with their families and benefit from the relative financial security which a monthly income of €550 affords in countries of origin.

Nonetheless, it bears repeating that the decision to return is not only a question of economic costs and benefits, as this book has stressed at various points. Nor can it be directly incited by the actions of political and legal institutions. As Luhmann (1990: 84) succinctly puts it, people “cannot be steered causally and technically by means of law and money.” Due to the inherent inability of political or legal institutions to accurately model the impact of their interventions in other systems, money and law always lead to distorting and counterproductive effects. In the area of immigration policy, this leads to a “structural tendency to ‘short-circuit’ the complexity of the migratory processes [which policymakers] are attempting to steer” (Boswell 2011: 12). Indeed, the chapters of this book have shown that hostel residents’ decisions about return revolve around a series of complex issues: about bureaucracy and paperwork, about continuity of healthcare, about friendships and family quarrels, about the observance of religious ritual. In sum, these are questions about social inclusion in a wider systems theoretic sense, and about mobility as a strategy to achieve that.

Last but not least, return decision-making later in life also provokes questions about what Levitt and Glick-Schiller (2004) call transnational ‘ways of belonging’, via memories, nostalgia and imagined or vicarious connections to distant places. Conceiving of mobility in this way may be especially relevant to older people’s experiences insofar as memories, stories, photographs, food, music and so on can provide compensatory mechanisms when physical movement ceases due to ill-health or frailty (Ciobanu and Hunter 2017). This was brought home to me most memorably during my fieldwork in Senegal, and it is here that I will close, with a scene from my time in Dembancané in January 2009.

Late one sunny afternoon I was out for a stroll with my host Jaabé. He had wanted to show me some of the building projects financed by the expatriate villagers living in France. As intimated earlier in the book, such projects are numerous and although Dembancané is not a big place, we were gone several hours, inspecting the clinic, the post office, two schools, and the recently inaugurated market hall. After the tour, as we were walking back to Jaabe’s house, we came across a group of older men sat in shade by the side of the path. After the customary introductions one of them started to speak to me. It was an observation about the weather: “It’s cold chez nous”, he said. Being from a country where that topic is a conversational mainstay, I was almost relieved to be on familiar territory for a moment, as I joked: "Non, non, au contraire, it is hot here today, for me at least!" Yet my new acquaintance’s remark was far from idle chatter: he wished to make a point. "Non, non, chez nous, la France – there’s snow in Paris." And the penny dropped: I was in the presence of a fellow Parisian. What he said next struck a chord and will remain forever imprinted on my memory:

You know, after spending forty years in Paris, it’s the only thing in my dreams.

At night, the métro, the boulevards teeming with people, all of that.

I never forget. I am not able to forget.Footnote 11

How could it be otherwise? In their lifelong quest for inclusion and recognition – in France as worker, pensioner, patient; in places of origin as a person of rank – circulating between these two worlds became an existential imperative. I do the va-et-vient, therefore I am. When their bodies can no longer carry them, the voyage back-and-forth continues, in memories and dreams.