The question of labour turnover has been variously examined across labour and organization studies but it has not been studied systematically in relation to international migration. In this book we tackle the question of labour turnover (the churning of workers in and out of workplace organizations) from the perspective of migrant labour. Here we build on the critical strands of labour and migration studies to shift our gaze to the social composition of labour (Wright, 2002), focusing on the specific drivers and subjective and social dynamics that link the phenomenon of labour instability to international migration. We consider the relationship between labour migration and turnover as emblematic of the wider effects of the intersectional differentiation of work and employment on workers’ lives and action for change in capitalist societies.

Since the pioneering work of Hirschman (1970), the act of workers quitting their job, described as labour mobility or exit, has been countered to worker voice and presented as an individualistic, opportunistic behaviour taken autonomously by workers as opposed to engaging in labour collective voice over effort bargaining (usually expressed through trade union representation). The tendency to see turnover as a primarily individualistic behaviour can be found especially in the field of industrial relations, which privileges collective forms of action in the workplace, whether or not institutionally mediated by trade unions (see Smith, 2006; Beynon, 1973). In the field of organization and management studies, scholars have tended to favour a functionalist approach both to the question of turnover and the role of migration in flexible labour markets revolving around costs and efficiency issues for employers, while employment studies have concentrated on the impact of labour mobility on collective bargaining in the workplace.

Looking at turnover in relation to the transnational movement of workers is of paramount importance today. While the COVID-19 pandemic has left its mark on 2020 as an unprecedented time of low mobility, in 2019 there were 169 million international migrant workers globally (IOM, 2021). Migrant workers are differently distributed geographically, with the US and Europe constituting the largest share among destination countries, and with the Arab States1 representing the subregion with the largest quota of migrant workers as a proportion of the entire working population (about 41.4 per cent as compared to 20 per cent in North America) in 2019. Alongside migration for work, family migration should be considered directly or indirectly linked to labour migration, constituting one third of overall permanent flows. Still, among the migrant workers who entered Organization for Economic Cooperation and Development (OECD) countries in 2018 more than 5.1 million were temporary, which represents a continuous upward trend from 2018 into 2019.2 Temporary migrants are also among those often suffering poorer and relatively more insecure employment conditions as compared to the indigenous counterparts, working the longest hours and earning poverty wages in non-unionized jobs (Castles, 2011; ILO, 2004; OECD, 2019; Pun, 2016; and Strauss and McGrath, 2017).

Migrant working conditions must indeed be analysed in the context of market creation as well as wider regulatory frameworks. As scholars based in Europe, we cannot avoid noticing that the so-called free movement of labour for EU citizens, as well as migration from outside the borders of the European Union (Fiedler et al, 2017), have been and will remain at the core of the ongoing crisis of integration of the EU single market. The response to the recent pandemic has yet again seen nation-states reasserting their sovereign power to shut entry into their territory, even excluding European Union citizens from member countries within the Schengen area (Favell and Recchi, 2020), and yet certain types of mobilities have been allowed, and considered essential, in Europe and internationally, even during lockdowns (for example, Xiang, 2020). Initially, across the world, state governments made a momentous and brutal return to the scene of border controls to respond to the health crisis. Tens of thousands of migrant workers in the Gulf countries lost their jobs and have been either stranded and locked in overcrowded accommodations, detained, or deported back to their countries (IOM, 2021). Meanwhile, in response to the Indian government’s enforced national lockdown in March 2020, 7.5 million internal migrants rushed back to their home in the space of two months, with many undertaking their journey on foot, dying on the way due to lack of transport and poor travelling conditions (The Tribune, 2020). In contrast to these forced return migrations and the immobilization of labour, special programmes for the just-in-time supply of seasonal workers deemed essential for the reproduction of the population, such as in farming and fruit picking, were quickly introduced to respond to labour shortages during the peaks of the COVID-19 crisis (Nikolova and Balhorn, 2020; Tagliacozzo et al, 2021; Creţan and Light, 2020).

Against a backdrop of growing tensions or struggles around the question of human mobility and the politics surrounding it, we refocus the debate by wearing the lenses of labour mobility and worker power. Our aim is to unpack, through a historical and multilocational perspective, the relationship between the movement of people across geographical borders in search of work, their strategies of reproduction and survival as they do so, and the impact of their ongoing movement across labour markets, workplaces, unions, areas of residence, and origins. Far from pretending to offer a comprehensive reconstruction of the history of labour turnover and migration from the 20th century, throughout the book we present particularly illustrative events or moments (what we call vignettes) in the development of labour mobility regimes and in relation to the emergence of ‘turnover as an organizational problem’ to be managed by employers (March and Simon, 1958), but also: states, and unions.

The OECD has recorded data on labour turnover since the 1970s, looking at the ‘distribution of workers by job tenure intervals which is measured by the length of time workers have been in their current or main job or with their current employer’.3 The Employment Outlook by OECD (1996: 162) distinguishes between job and labour turnover: while job turnover occurs at ‘the level of an individual establishment or firm’ and corresponds to ‘the total number of jobs created less the number of jobs which have disappeared’ between two points in time, labour turnover is a more comprehensive figure which comprises ‘the movements of individuals into jobs (hirings) and out of jobs (separations) over a particular period’. In this sense labour turnover encompasses job turnover (OECD, 1996: 165). This comprehensive notion of labour turnover is more critical because it recognizes the nature of workers’ decisions to leave firms as being independent from firms growing or declining, and in turn being independent from employers’ decisions on whether and how to replace workers. Still, there are various limitations to the ways in which official bodies and statistical agencies understand and track labour mobility. Moreover, they do not consider migrant status in relation to labour turnover rate or job tenure. In addition, official measurements of labour turnover take the firm or the economic sector as units of analysis, collecting the data by country, whereby a harmonized dataset allowing for cross-national comparisons of labour turnover rates is still lacking.

Besides national datasets, some limited regional and international level analyses exist. An OECD study from 2009 included statistics on labour turnover for 22 countries, using harmonized data. The annual average labour turnover rate was found to be 33 per cent of total employment between 2000 and 2005 (European Commission, 2010). Differences between countries are usually explained with reference to the degree of employment protection legislation and unemployment insurance (OECD, 1996; EC, 2010). The analysis of the EU Labour Force Survey (LFS) by the EU Commission in 2010 also shows that labour turnover rates change, critically, according to the group demographics, whereby young people tend to experience significantly more ‘transitions both in and out of the labour force and between jobs than an average worker’,4 which is interpreted also as the result of the extensive use of temporary contracts for this group. Beyond OECD countries, labour turnover data from the Global South are scarce or irregularly collected, and this omission may be also explained by the strong link between high turnover level and incidence of precarious work in those countries: as Professor Akua Britwum, a Ghanian scholar, responded to an International Labour Organization presentation on precarious employment, at the 2011 Global Labour University Conference, ‘What you are calling precarious work [in the Global North] sounds like what we in Ghana call work’ (cit. in Mosoetsa et al, 2016: 8).5

Knowledge of statistical and institutional definitions is certainly important to identify organizational monitoring processes and institutional responses to the labour turnover. Considering the official national and international statistics, we can identify at least three definitions of labour turnover: as the movement of a single worker between jobs in the labour market; as the churning of workers in a single workplace or industry; and, more generally, as the overall rotation of the workforce in global production processes (as international labour mobility). Besides government and internal bodies, different academic disciplines tend to refer to some of these three aspects or levels neglecting the others. However, in what follows, we move beyond these quantitative measurements of labour mobility as they are mostly concerned with labour market or cost/efficiency measures. In this book, we take into consideration these three different definitions or aspects of turnover in their connection with international labour migration, problematizing the question of labour turnover historically.6

One of the central arguments of the book is that the relation between migration and turnover is critical to understand worker struggle over mobility: migrant labour may be used by employers to access docile labour and reduce high turnover, but under different conditions it can also increase voluntary or worker-led turnover, as in the case of migrants who gained mobility rights through freedom of movement and are able to quit their job without losing residence rights or welfare entitlements.

It is indeed because of the differentiation of labour on the one hand (through state regulations of migration, citizenship, or residence status) and on the other, the precariousness generated by intense work and the differential access to welfare and labour protections, that turnover becomes a terrain of social contestation. In other words, we argue that labour turnover constitutes for migrants a first, rough response, not only to degrading and unacceptable working conditions, but also to the forms of racialized and gendered segregation that they are subject to inside and outside their workplaces. These forms of segregation are put in place by management but are sometimes also (directly or indirectly) reproduced by trade unions when they are bargaining and representing the ‘vested interests’ of their members in the workplace. In fact, labour unions are a source of worker power but in some cases they are also imbricated in racial politics and can be considered themselves as racialized formations (Lee and Tapia, 2021). In other words: the more migrants find hostility in the workplace, and are unable to count on the support of local unions or other political groups, the more they will express their dissatisfaction and seek alternative strategies in the form of labour turnover. Selecting particular cases where tensions around labour mobility are predominantly stark, we show that labour turnover and migration have been historically related: either when employers have used racialized immigrant labour to tame high turnover (at Ford in the early 20th century), or because immigrant labour has caused high levels of turnover and required a substitution of local workers with new groups or with new groups of migrants (from the Gastarbeiter in Germany and the post-World War 2 Bracero programmes in the US, to the kafala system in the Gulf countries).

The shifting maelstrom

The problem of turnover has existed since employment itself, but it has only been explored academically in the last hundred years7 when excessive labour instability appeared as the result of ‘the heavy loss of employees to competitors whose labour policies are more enlightened, who pay higher wages, and who provide more attractive working conditions’ (Brissenden and Frankel, 1922: 4).

Since the early 20th century, from the perspective of management studies, labour turnover has been understood ‘strictly in reference to the extent of shift and replacement necessary for the maintenance of the workforce’ (Brissenden and Frankel, 1922: 7), that is, from the point of view of management-led turnover or retention needs. In the field of management and human resources research, Bolt et al (2022), based on a systematic review of 1,375 articles on voluntary turnover (the largest to date), have called for a holistic theory of turnover to develop a finer understanding of what still remains ‘an enigma’. Bolt et al (2022: 569) propose a framework that is ‘strategically differentiated, multi-level, and longitudinal’, able to overcome important limitations such as the lack of application of existing turnover theories to qualitative case studies, the emphasis on individual rather than contextual factors, and the lack of attention to sector and occupational differences influencing the behaviours of different groups of workers. While remaining primarily concerned with the relative (dis)functionality of voluntary turnover and costs for the firm, this recent review makes the important point of integrating different theories from the social sciences as well as considering more carefully the macro and meso levels: the labour market and more specifically ‘the context of changing employment relationships and flexible careers that render loyalty and commitment largely redundant’ (Bolt et al, 2022: 571), and the organizational and occupational level respectively. While we do not aim to contribute to management theories, we build on these suggestions by bringing the interdisciplinary insights of migration and sociological research into the study of the labour process and worker mobility as turnover/exit.

From a labour studies perspective, turnover can indeed be measured concretely as the circulation of labour as variable capital in the process of production. In this sense, labour turnover concretely refers not to quantities of variable capital that have completed their churns within a certain period of time, but to the modes and effects of the use of that labour force vis-à-vis capital accumulation (Marx, 1993). While labour scholars have observed these dynamics of labour mobility within the workplace, we extend our gaze to include cross-border mobility.

What does history tell us about the interrelationship between labour turnover and migration? Our review of the literature shows that governments and employers are constantly attempting to coordinate a joint state-capital response to the ongoing uncertainties around labour mobility, whether across borders or workplaces, uncertainties that are expressed by the increase of labour turnover. These efforts have been carried through policies to manage migrant labour across borders and occupations, including contemporary point-based and temporary schemes.

Indeed, if the primary objective of capital is to speed up the churning of labour so that the extraction of surplus is maximized and the cost of labour reproduction minimized (Burawoy, 1976), then the labour force in its concrete form adopts a relatively constant behaviour: that of refusing the intense pace of work tentatively dictated by management. Such individual, and sometimes collective, action of refusal only appears as the turnover of labour force (Gambino and Sacchetto, 2014), while in fact it represents the immediate form of protest and resistance against the command of capital over labour and life. Indeed, vis-à-vis the mobility of capital, labour is not merely following or remaining inert. As argued by Guglielmo Meardi (2012b: 105) in his ground-breaking analysis of both the social consequences of EU enlargements to incorporate new Eastern European member states, and the power imbalances between capital and labour’s differential mobilities, ‘the common assumption that mobility is a prerogative of capital but not labour has been shown to be inaccurate, and this has not failed to impact on power relations between capital and labour’.

The analysis of the churning/rotation of the labour force brings to light the need for a continuous mobility of living labour in order to sustain the labour process. This mobility has often been subject to a variety of controls by a range of private and public actors: states, local authorities, recruitment agencies, formal and informal brokers, employers, migrants’ communities, and networks. The ways in which capital has historically attempted to entrap or bridle wage labour is certainly instructive of such relationships, an aspect that is considered throughout the book as we look at the actions of multiple agents and their infrastructures in the management of migration (Xiang and Lindquist, 2014).

Historically, the endless processes of primitive accumulation and rapid industrialization have often generated high labour mobility to support the needs of enterprises (Fisher, 1917; Douglas, 1959). Since the 16th century, capital has sought to continuously access and replenish sources of labour, both mobilizing workers and/or expanding across differentiated geographies of production and starting new processes of accumulation (Harvey, 1989). We may understand these processes in terms of options or strategies available to capital, whereby the first option to access labour is precisely by obtaining workers through migration, as labour mobilization under capitalism is intimately connected to the movement of workers across borders for the sake of capital (Hofmeester and van der Linden, 2018; van Rossum and Kamp, 2016; Linebaugh and Rediker, 2000). In the fascinating reconstruction by Bridget Anderson (2013), the disciplining of the vagrant population and of the poor, the codification of the slavery of Amerindians and African Americans, and the reintroduction of forms of servitude in Eastern Europe were all mechanisms that allowed for a better surveillance and confinement of the labour force between 1600 and 1700. And yet this history was not the only one of forced migration and subjugated racialized people.

One of the first and most significant examples of labour mobilization can be found in the forced labour of African migrants in the transatlantic slave trade from the 16th century.8 The slave trade from about 1500 to the mid-19th century encompassed about 12 million people, mainly from western Africa to the Americas, and developed particularly to expand the availability of the necessary workforce and to substitute Amerindian and indentured European workers.9 The turnover for Amerindians and indentured European workers had higher costs than mobilizing African slaves, because the former held power to quit thanks to their good knowledge of the surrounding environment which facilitated escape, while European indentured labourers enjoyed greater freedom and rights than slaves (Moulier-Boutang, 1998). However, also for the slaves, the first and foremost form of resistance, organized ‘from sundown to sunup’ (Rawick, 1972), were runaways, individually or in groups, although for them the chances of successful escape were lower. Planters gradually learned about the importance of the composition of the labour force to manage their plantations since workers’ common socio-economic conditions provided a source of elemental unity among them – at a time when forms of racialization were not yet refined to maintain rigid internal hierarchies (Breen, 1973). Thus, the strong institutionalization of slavery at the end of the 17th century in the so-called New World may be seen as a response to the forms of labour resistance and turnover of the Amerindians and later of the European indentured workers and African American slaves. The resistance and revolts of African American slaves expanded the demand for Asian indentured labour as well as for European migrants, indentured or not.10 After the abolition of the slave trade by Great Britain (1807), this request increases, and gets stronger after the abolition of slavery in the British colonies (1833) and in the United States (1865). Labour turnover represents, in this sense, the possibility of escape from the lynch law, flee from exploitation or social oppression, and even the struggle for a better life. Against the multiple labour regimes built around the world, workers put in place forms of mobility by building maroon communities, moving to other countries, or turning to other identities and jobs. Sooner or later capitalists and governments learn that workers, even in the more coercive situations, resist and organize themselves.

It was the pioneering work of Moulier-Boutang (1998) that highlighted that as soon as waged labour was created from the 16th century, it immediately appears in its form with a Janus face, that is, half free, half constrained. In fact, ‘free’ wage labour is channelled from its birth through the legal mechanisms enforced by colonial powers (and capital) to govern the mobility of the labour force. This oscillation points to the historically blurred boundary between free and unfree labour (Steinfeld and Engerman, 1997). The ambivalent nature of wage labour in this context is articulated by Moulier-Boutang in his understanding of labour as a continuum from the slave labour of the modern era, through the indentured labour of the colonies, to the free waged labour of the industrial era until today. According to this interpretation of the history of the commodification of human labour, indentured labour should not be considered as a form of pre-capitalist labour as compared to the contemporary form of waged employment, but as both being part of the same continuum, of the same process of exploitation.

If free labour nowadays represents the prevalent form of migration, various systems of coercion and segregation continue to debase the conditions of migrant workers, through temporary recruitment schemes, postings, and other migration regimes as state-managed circulatory labour migration (Buckley, 1995; Lillie and Greer, 2007; Pun, 2005; Morrison et al, 2020; Parreñas, 2022). Institutional controls on the right of free residence are still a primary factor in lowering the conditions of migrant workers across all countries. And yet, long distance recruitment remains the main device to compensate for high turnover in generally low-appeal sectors, such as agriculture, construction, and in-person services, that cannot be relocated overseas (Stalker, 2000).

Together with the mobilization of labour, a second option or strategy for capital to access (cheap) labour is through the relocation of production (or spatial fix, see Harvey, 1989; Silver, 2003).11 Although indirectly linked to migration, this strategy of capital accumulation has been widespread in particular since the 1950s, when US investments started to aim towards less unionized areas within the country, such as Indiana and Tennessee, and later towards Asian and Latin American countries (Cowie, 1999), where a young and low-wage workforce, low taxation, and freedom from pollution controls could breathe life into capital. The maquiladora system based on cross-border foreign direct investment (FDI) in Mexico provided a case in point (Sklair, 1989; Hutchinson et al, 1997). Work in outsourced plants did not deter migration; on the contrary, it stimulated it by disrupting traditional local economies while imposing at best modest pay and working conditions, and thus familiarizing workers with the motivational and economic structures of the US (Williams and Passe-Smith, 1989; Sassen, 1996). Since the 1990s, a similar process of relocation has developed in Europe; it was characterized by the dislocation of distinctive economic activities from Western to Eastern Europe and a corresponding segmentation of labour market conditions (Ellingstad, 1997). As has emerged in the first thirty years of political and economic freedoms in Eastern Europe, workers have elected exit as their primary form of resistance to the post-1989 restructuring (Meardi, 2007), prompting employers to design tighter forms of control, such as subcontracting and new migration policies. At the beginning of November 2022, the Romanian government, for example, approved new legislation restricting the rights of non-EU workers, including the obligation for migrants to obtain the written consent of their first employer in order to change jobs during the first year of their labour contract (Dumitrescu, 2022). And yet, as in the past, migrant workers are unlikely to accept the place assigned to them by employers and destination country authorities (Castles, 2006).

The two options for capital are interrelated, since relocation of production also necessitates the mobilization of new labour, making the process far from smooth for the state and for corporations. These relations of mobility do not correspond with unidirectional or regular patterns. As also argued by Harvey (2018: 385) there is an irregularity springing from the intrinsic contradiction between and within capital and labour as opposite forces in capitalist social relations, as they are rather ‘forced into curious patterns of struggle and compromise over the geographical mobility of labour’. Rather than clear directionalities, it is the fluctuating and mutually influencing movement of workers and capital that seem to characterize the past decades of globalization of both labour and investment flows (Alberti, 2019).

However, we suggest that in addition to the institutions that manage these mobilities, attention should be drawn to workers’ ways of life and struggles that tend to constitute themselves as social spheres on the basis of which they renegotiate their conditions of existence. In the public arena one still notices the political and social response by those who resist this vortex (the ‘maelstrom’ of capital), that is, the process where capital tends to reduce the workforce to a simple bearer of labour to be consumed: from the fugitive slave in the Americas to those who still migrate against the dictates of capitalist accumulation. As in the past for fugitive slaves, ‘the strongest expression of individuality coincides with the most powerful manifestation of collective action. It is within this nexus that the possibility of overthrowing the barriers of discrimination is situated.’ (Gambino and Sacchetto, 2014: 118–19).

Contradicting the simplistic binary view between individual and collective action, Lucassen and van Voss (2019) stress that between 1600 and 1850, not only deserters ran away in groups, but individual runways also depended on collective networks. Furthermore, the experience of one person could support others ‘to vote with their feet’, because every worker running away represents someone who was able to break dominant constraints, while directly or indirectly supporting the spread of radical ideas (Linebaugh and Rediker, 2000). In this way we stress how mobility may be considered itself as a form of protest: in the words of Marcel van der Linden (2008: 179), ‘the transition between ‘running away’ and ‘fighting for better working conditions’ is in reality rather fluid’. Following this approach, we are interested in exploring the continuum of mobility and worker power through the lenses of international migration.

Beyond nationalistic perspectives

This book tackles the question of labour mobility and engages with theoretical debates across different fields of the social sciences, with sociologists of work and labour researchers as primary interlocutors. The starting point of this book is that understanding the struggles around labour mobility and immobility today is a necessary endeavor for scholars of work and migration alike. In our effort to bridge labour and migration studies, one of the key insights that a labour approach to international migration may adopt is the critique of methodological nationalism. Far from indicating simply nationally-based research, we follow Çağlar and Glick Schiller’s definition of methodological nationalism, understood as:

an intellectual orientation that approaches the study of social and historical processes as if they were contained within the borders of individual nation-states. … [M]ethological nationalists confine the concept of society within the boundaries of nation-states and assume that the members of these states share a common history and set of values, norms, social customs, and institutions. (Çağlar and Glick Schiller, 2018: 3)

As we will show, moving from methodological nationalism to an approach that takes into consideration how specific localities and local labour regimes (Baglioni et al, 2022) intersect with global dynamics is an essential step to move away from many of the biases and myopia of the social sciences and employment and labour studies, specifically vis-à-vis international migration. The studies of human geographers Wimmer and Glick-Schiller (2003), and more recently Glick Schiller and Çağlar (2011), have been path-breaking in illuminating the persistence of the ethnic lens and of the privileging of transnational communities, rather than transnational social fields, as fundamental categories in migration studies.12 Indeed, the point of view of the destination country and the nation-state has remained dominant in Western migration studies, constraining research in the field and attention to processes of social change (see also Castles, 2010).

As we illustrate in the book, if migration studies have reproduced the ethnic lens, this is even more the case for industrial relations scholarship, which has maintained a national prejudice in terms of selecting the boundary of the nation-state. In this way, the nation-state remains both a primary terrain of joint-regulation with national-level actors, and a key player in social dialogue on work and employment (Giles, 2000; Meardi, 2012a). Paradoxically, even the more enlightened and migration-sensitive literature in industrial relations has remained concerned with the questions of governments, employers, and union control over the borders of national labour markets, and the ‘impact’ of migration on national markets and national union strategies (Penninx, 2017; Marino et al, 2017a). More broadly, most of comparative industrial sociology and global political economy (see Afonso and Devitt, 2016) maintain the nation-state, its institutional system of employment relations, and the national welfare state (Alesina et al, 2004) as the main unit of analysis, whether or not they are considered victim or under threat from the effects of international migration flows on local workers and citizens.13 Further, methodological nationalism in migration research widely intertwines with Eurocentric or Westernized approaches to migration and labour research, whereby empirical studies appear largely concentrated in the Global North (King et al, 2011). Denaturalizing the categories of ‘nation’ and ‘space’, we support an approach that ‘considers the simultaneity of the transnational practices of individuals, organizations and institutions taking place in multiple localities’ (Amelina and Faist, 2012: 2).

Whether or not a country’s government represents its own national borders as more or less porous according to the needs of the time, we argue that their labour market has never been purely national, in the sense that at least since the origin of the capitalist era, any market has been structured by the possibility for employers to meet their labour demand by mobilizing workers worldwide. As Tilly and Tilly (1998) argue, labour markets are not homogeneous or defined entities, but open systems that can be affected by various factors, and particularly by mobility.

Labour markets today are certainly less prone to slavery and serfdom than their predecessors in the 18th and 19th centuries and up to 1920 when high rates of labour mobility were the norm (Jacoby, 1983: 261). It is no coincidence that the first explicit treatment of the question of labour turnover is found at the beginning of the 20th century, when large enterprises in the US were confronted by exponential growth (over 300 per cent) in the rates of turnover (Alexander, 1917; Slichter, 1919), but also by increasing levels of cross-border migration.14 In response, they started developing new personnel strategies to counter such movements, and retained or attracted new labour forces in a context of increasing inter-firm competition over labour supply.

In contrast to methodological nationalism, we believe that any treatment of the topic of migration demands a global perspective (Hofmeester and van der Linden, 2018), for the simple reason that the migrant workers move from and to an extremely wide range of countries according to historically accumulated differences and inequalities (Federici, 2004). In short, they operate in highly differentiated local contexts, bringing with them a multiplicity of experiences not only of the different working regimes, but also of the strategies of their mobility (exit), of resistance (voice), and of self-organization in the terrain of social reproduction.

Migration studies have developed a critical research agenda on human mobility by looking at the variety of subjective and objective factors moving people across the globe, and at the socio-economic and cultural implications of such inter/transnational movements in terms of policy, identities, everyday practices of reproduction, family, and ethnic relations (Miller and Castles, 2009; King et al, 2010; Hoang and Yeoh, 2015; De Genova, 2017). However, the field of labour studies has only recently appreciated the importance of studying migration as a distinctive topic. In the words of McGovern (2007), it is quite ironic if not paradoxical that Western industrial relations have remained disinterested in migration for many decades, if we consider that one of the primary reasons for people to move is linked to work. In other words, the implications of migration concern all workers, regardless of their citizenship.

The contribution of migration studies has been essential to throw some light on the complex factors shaping the interlocking of mobility and immobility (for example, Xiang, 2020). Differently from migration scholars, however, we privilege the point of view of labour mobility rather than human mobility, qualifying our field in terms of our interest in the history of worker struggles (Silver, 2003; van der Linden, 2008) and the question of labour indeterminacy (Edwards, 1990; Smith, 2006). We distance ourselves from ‘classless studies’ of migration and transnationalism, particularly those that assume a falsely universal ‘middle-classness’ of all individuals on the move (Conradson and Latham, 2005a; Favell, 2008), problematizing the relations between class and migration historically (van der Linden, 2008; Papadopoulos et al, 2008).

Furthermore, we take a critical stance towards the labour and political economy scholarship that reproduces a structuralist or functionalist approach towards migration, whereby migrant workers tend to be depicted as merely economically driven and self-interested individuals, ‘greasing the wheels’ of the labour markets of destination countries (Ruhs, 2006a; McCollum and Findlay, 2015); or unilaterally moved by sending states as sources of remittances to benefit the economy of the home country; or as ‘hopping’ from country to country with the aim of financial maximization, as atomized individuals detached from their wider communities and households, ‘dancing on the tune of the wage differentials’ (Rogers, 2009: 45).

Overall, in terms of the relationship between labour turnover and migration, it is without doubt that the very nature of work and the specific working conditions experienced by the workers – whether repetitive, precarious, with low wages, and highly disciplined, are key factors in explaining why workers decide to quit their jobs across different historical periods and degrees of labour market fluidity. However, in what follows, we endorse Jacoby’s (1983) approach highlighting the importance of the compositional analysis of the workforce, by paying attention to the specific race, legal, and gender differentiation of workers at a certain time and place. Our aim is to understand the dynamics of labour mobility today and how it interacts with historically specific forms of ‘race management’ of different groups of workers (Roediger and Esch, 2012), as well as with its gendered dimensions, hitherto seldom considered (Chapman and Prior, 1986; Theodossiou, 2002). Indeed, the history of controls over the movement of populations is a long one, because mobility modifies the composition of the workforce and critically impacts on the management of workplaces and on the forms of worker agency. This emphasis on the labour composition of turnover across jobs, industrial sectors, and global production processes allows us to explore the specific relationship between the recruitment and management of migrant labour, the control of labour mobility, and worker struggles.

Four lenses to understand the turnover of migrant labour

Our approach to the question of labour turnover moves away from mainstream economic and management/human relations approaches, mostly concerned in measuring the flows of the individual movement of workers in and out of workplaces and industries, and their costs for organizations. We rather adopt a holistic and critical approach that considers the mobility of workers multi-dimensionally (in terms of individual job changes, churn within an industry, and the rotation of labour in global production processes), which takes into account both the wider social factors and subjective drivers of such mobilities. In particular, starting from a wider consideration of the structural constraints that capitalist accumulation imposes on labour mobility, we privilege four crucial lenses across migration and labour studies to understand the dynamics and effects of the turnover of migrant labour, namely rising from the labour process theory (LPT), the autonomy of migration, social reproduction theory(s), and industrial relations. Such an interdisciplinary framework is necessary, we believe, to investigate the question of labour mobility across the cited dimensions, and to explore the workplace, national borders, the community, and the institutional spheres of mobility negotiations respectively.

The first lens follows the assumption of LPT that acknowledges the existence of ‘structured’ antagonism as constitutive of the capital–labour relation (Edwards, 1990; Thompson and Smith, 2001; Elger and Smith, 2005). As Chris Smith (2006: 392) put it: ‘labour power has two elements, the power over work effort and the power to move between firms’. In his work, Smith (2010) developed a flow approach to labour power that understands it in relation to wider issues of labour mobility and its management by states and other actors. The flow approach emphasizes the importance of the impact of free wage labour on capital controls and the freedom of movement of individual workers in and out of work (see also, Smith, 2015; Pun and Smith, 2007). We think this approach should consider not only free wage, but also unfree or semi-free labour (slavery, indentured, and tied to temporary work permit; see Sarkar, 2017) since workers developed agency and were able to open spaces of negotiability even in harsh situations (Strauss and McGrath, 2017; Coe and Jordhus-Lier, 2011).

The second lens builds on Chris Smith’s theorization of labour mobility power grappling with the problem of the transformation of living labour power into expendable labour effort (Deutschmann, 2002), from the point of view of migrant labour strategies of mobility as an instance of freedom and rebellion from capital controls (for example, Alberti, 2014; Andrijasevic and Sacchetto, 2016; Morrison et al, 2013). These latest approaches have drawn from a view of migration as a social force that is larger than mere economic motivation and that follows an autonomous logic beyond capital needs, by nodding to the autonomy of migration perspective (Mezzadra, 2004; Bojadžjev et al, 2007; Papadopulos et al, 2008). Our study shows how labour turnover can therefore be understood as a manifestation of this antagonistic and unequal relationship, where the subjectivity of migrants alongside capitalist dynamics of production and reproduction plays a critical role.

Throughout history, the work considered outside the parameters of production has often been associated with marginalized and oppressed subjects, such as slaves, children, colonial and postcolonial migrants, and women. Instead, our third perspective pays attention to social reproduction, namely all the activities that facilitate the maintenance and reproduction of a labour force (Picchio, 1992), and capitalist society and life as a whole (Katz, 2001), as strictly connected with every worker experience. In so doing, we emphasize that social reproduction is not just another way of conceptualizing domestic or care work. All these activities are interlinked with labour processes, both when they are shaped by the system of production in a specific location, and when they in turn shape workplace relations. As the different welfare regimes highlight, class relations are developed in the double dimension of the labour and reproduction processes. Capitalists are aware that the localization of their investments have to take into account the practices related both to work processes and to social reproduction, alongside global supply chains (Tsing, 2009). Reproductive practices can in fact constitute a crucial element in the forms of organization and worker mobilization as they develop at a political and economic level (Laslett and Brenner, 1989).

Finally, with regard to the industrial relations lens, in the past decade we have noticed a growing interest in the question of labour mobility and international migration across industrial sociology and employment relations (Pun, 2016; Bal, 2016; Martínez Lucio and Connolly, 2010; MacKenzie and Forde, 2009), including the working experiences of asylum seekers and refugees (Bakker et al, 2017; Tören, 2018), trade unions’ efforts to integrate migrant workers (Adler et al, 2014; Tapia and Turner, 2013; Marino et al, 2017a; Ford, 2019), diversity and equality strategies in organizing workers (Holgate, 2005), and emerging theorizations on how labour systems of regulation are related to international migration (MacKenzie and Martínez Lucio, 2019). Despite this renewed interest in migration, however, very few have looked at the different dimensions of migrant labour regimes worldwide and at their implications for employment relations. We deem it necessary to span the fields of production and social reproduction as criss-crossed by tensions and struggles (Cravey, 2004). Furthermore, some fundamental biases and erasures remain in the field of industrial relations that create opacity in fully understanding migrant labour and migrants’ constraints to access voice in the workplaces, as well as to articulate their multifarious forms of resistance. This erasure needs to be traced back to the ways in which turnover has been negatively associated with migration and variously racialized (Roediger and Esch, 2012). Even within the more radical tradition of industrial sociology, while turnover and quitting have been recognized as valid expressions of conflict over the ‘frontier of control’, of capital over labour (Edwards and Scullion, 1982), there is a tendency to stigmatize all forms of mobility as individualistic behaviour primarily responsible for a race to the bottom in working conditions (Cremers et al, 2007; Krings, 2009). Such mobility power has also recently been considered in contrast with the experiences of indigenous workers, who are rather deemed lacking such capacity (Thompson et al, 2013).

While the effects of immigration controls on dividing workers and making migrants more vulnerable are increasingly recognized by national union confederations (for example, Crawford, 2020), the emphasis of the trade union literature remains on the negative effects that mobility, turnover, and immigration policies have on ‘unionizing potential’ in the workplace. Fixated on the role of unions in controlling demand and supply in the labour market and monitoring state borders, or in integrating by servicing migrants according to their specific issues, labour scholars do not seem to pay enough attention to the hindering effects of immigration controls on worker voice and on their opportunities to resist in the workplace. The extensive research by Anderson (2010), as well as Fudge and Strauss (2014), have clearly shown how statutory and regulatory systems of labour mobility and citizenship produce ‘institutionalized uncertainty’ for migrant workers, where the temporariness of the work permit, and dependency on the employers, limit migrant labour chances of unionization or any other form of collective and individual voice (see also, Novitz, 2009; Rigo, 2005). The issues of migration, worker mobility, and quitting practices by workers continue therefore to represent a contested terrain and a somewhat unresolved issue, especially in the sociology of work and industrial relations, which are specific areas of our research. We believe that labour scholars can profit from research in migration studies and human/labour geography. We illustrate below how we tackle these distinctive issues in the book and the specific arguments we develop across five chapters and the conclusion.

Structure of the book

In Chapter 1 we explore the theoretical debate, reviewing the primary approaches to the study of labour mobility across migration and employment studies. Alongside the critical review of wider theory of capital and labour mobility, the aim of this chapter is to reinterpret labour mobility through the lenses of migration research. We show how underpinning most of the debates on migrant labour in both the political economy and labour studies literature is a tendency to reproduce methodologically nationalist frameworks, that tend to ostracize migrants and their mobility as the external and dangerous outsider, from which to protect a romanticized and sedentaristic view of the working class. In contrast, we point out the contributions of both critical migration and feminist studies in human geography, and the autonomy of migration (Mezzadra and Neilson, 2013; Karakayali and Bojadžijev, 2007; De Genova, 2017; Papadopoulos and Tsianos, 2013), to overcome the functionalist framing of migration labour, as well as homogenizing views of worker identities and solidarity. Rather than seeing the migrant as ‘the other’, we embrace mobility as foundational for both migrant and non-migrant workers (Çağlar, 2016). In this chapter we show the ambivalence of turnover in terms of the relative advantage for either workers or employers according to the state of the labour market, the sector, or other circumstances. We also show that, contrary to mainstream readings, voluntary turnover as worker escape from poor conditions can be understood as an expression of conflict, as well as and in particular by migrant workers, moving beyond binary views of individual vs collective forms of resistance (compare Hirschman, 1970; Edwards and Scullion, 1982).

As we highlight in the logistics of living labour (Chapter 2), labour turnover may assume a relatively orderly or disorderly character under different economic and social circumstances, since the relation between migration and turnover may be variable as they highlight different interests for the management and exercise of mobility by multiple actors. Migration controls and the state regulation of mobility have in fact often been aimed at limiting migrants’ mobility across jobs and occupations – and ideally, forcing them to go back to their countries when they are no longer needed: from the temporary migration programmes of the 1970s in Europe and their reappearance in contemporary migration policy in the Global North (Ruhs, 2006b), to the spread of temporary programmes in East Asia and the Gulf today (Liu-Farrer and Yeoh, 2018; Bal, 2016). Since the 2000s, the debate on international migration has stressed how employers, labour brokers, and intermediaries move workers ‘from above’ (Xiang and Lindquist, 2014), thanks also to the support of states that enforce migration controls (Castles, 2011) through visa and residency regulations. However, labour mobility cannot be considered only from the point of view of the constraints, as migrants build their own infrastructures to gain and sustain their mobility (Ciupijus et al, 2020). Similarly, the mobility practices of migrants need to be understood in relation to their experiences of the labour process and the nature of work (Alberti, 2014).

Labour scholars have tended to highlight how it is often the intense nature of the labour process (Chapter 3) and poor working conditions that cause high labour turnover, and how such situations require the ‘expansion’ of the labour market through the import of foreign labour or labour from a different location in the same country. High salaries and welfare policies may be alternative strategies to retain the labour force at the enterprise level, and yet such strategies may be limited to a certain period and can be practiced only when an enterprise has a monopoly of a certain technology or skills.15 Moreover, past research has shown that is it difficult to identify a certain level of wages by which it is possible to make up for repetitive and demanding work. Peña (1997: 57) argues as well that the employer may not be interested in improving working conditions in the first place, whether because technological levels may be difficult to improve in that particular sector, or because the costs of machinery outweigh that of higher wages for workers. Etienne Balibar (1974)16 highlights how technological innovation, the ‘parcelling out’ and intensification of labour, are strictly linked with labour mobility, influencing each other, because they are part of the same process. As other historical sources have shown (Gambino and Sacchetto, 2014), the spatial fix (Silver 2003; Harvey 1989; 2018), operated by multinationals through the developing of long supply chains sometimes located in Export Processing Zones (EPZs), employs migrant workers in repetitive, routine, and simplified labour processes that require a constant daily renewal of the workforce. Scholars stress how young unmarried women in the Global South are often preferred because they are cheap, docile, and come with a low level of labour turnover (Fernández-Kelly, 1983; Elson and Pearson, 1981). However, as we argue, migrant and (often) women workers, employed in what we call enclaves of differentiated labour (EDL) dotted along global supply chains and concentrated in the Global South, put in place ‘oppositional tactics’ (Ong, 1997) and struggles that emerge through a multi-sited form of collective and individual resistance (Pun, 2005). With the notion of EDL, we therefore contribute to a labour-centered approach to the functioning of Global Production Networks (Newsome et al, 2015) by introducing an intersectional analysis of the migrant workforce put to work in these ‘free’ zones, and by providing a deeper understanding of how a differentiated regulation of labour mobility shapes the unique character of these areas in the Global South and their control regimes, and partly in the precarizing migrant sectors of the North, as well as workers’ resistant practices.

In Chapter 4, we argue that the realm of social reproduction (Katz, 2001) is a necessary field of analysis of migrant practices if we want to understand the tensions of migrant labour and labour turnover, but that it has been overlooked by labour scholars. Our overview of cases shows a tendency whereby the lower the cost of social reproduction for the individual worker (for example, a migrant worker without a family to support in loco, or without the need to send remittances back home), the higher the labour turnover. The other side of this argumentation is indeed the one developed by the labour sociologist Michael Burawoy, who as early as 1976, while researching the system of migrant labour in South Africa and at the Mexican frontier with the US, showed how the geographical separation of reproduction and maintenance of labour produces a reduction of the costs for employers. In other words, the family of the migrant back in the country/area of origin was responsible for reproducing the life of the worker and future workers. While these insights have been path-breaking in understanding the continuous exploitation of workers across the two spheres, our perspective moves beyond structuralist approaches to social reproduction that have hitherto concentrated on the role of migration in saving employers’ labour reproduction costs. We point to the frictions that migrant labour can create on the terrain of social reproduction, that is both within and beyond the labour process. Social reproduction indeed needs to be understood transnationally through the everyday practices of migrants. The control of reproduction emerges again in the way migration policies prevent migrant families from reuniting or exclude migrants from access to welfare benefits.17 By bridging the spheres of production and reproduction, our historical and multilocational analysis shows that migrant mobility practices and their political potential should be understood beyond a narrow focus on the workplace and the point of production as the only one shaping worker action.

On migrant organizing (Chapter 5) we show how the ‘problem of turnover’ has influenced the ways in which temporary migration has been variously stigmatized, including within the labour movement, through processes of racialization and gendering, with migrants often left at the margins or outside of union structures, or at best incorporated in special committees while unions ignored any appeal to an in-depth understanding of migration and racial issues. Our starting point is that because of the persistence of nationalist frameworks of integration and the ideology of the universalism of working-class unity, trade unions in Western countries, but also in other world regions, have reproduced the marginality of migrant labour. Despite some initiatives by traditional unions to support migrant workers’ empowerment, migrant and precarious workers have organized to create new forms of representation inside and outside union structures by exposing major cracks in the system of labour representation, already in crisis. Migrants have developed industrial action or have organized their own resistance both in collective and individual forms, despite their temporary status as casual workers, through either union internal structures, or via NGOs, or by setting up their own unions (Alberti, 2016; Iannuzzi and Sacchetto, 2020; Però, 2020; Peano, 2017; Weghmann, 2022). With increasing mobility across borders in different world regions, some local unions as well as global confederations have become more sensitive to migration, by accepting the need to tackle the realities of mobility and transnationalism, and by carrying out more or less successful attempts at creating migrant branches (Rogalewski, 2018) or transnational union networks (for example, Meardi, 2012a). Non-union actors in these efforts appear to have played a critical role particularly in the Asian regions, where migrants experience extreme forms of precariousness (Ford, 2019), opening up industrial relations to a multiple- and hybrid-actors perspective, with the potential to radically renew labour organizing. Their attention has gone to bottom-up approaches, highlighting democratic participation rather than mere greenfield recruitment (McAlevey, 2016; Holgate et al, 2018). Overall, recent grass-roots initiatives across Europe, the US, and beyond, in poorly unionized and fragmented sectors, show new patterns of the autonomous self-organization of migrants within and outside traditional union structures, indicating the rise of migrant-led radical unions (Moyer-Lee and Lopez, 2017; Cillo and Pradella, 2018) amidst the wider crisis of institutional unionism and the decline of collective bargaining. We note that despite the theoretical developments with the incorporation of mobility into the labour process framework, and despite the increasing interest of industrial relations in migration, the consideration of the concrete effects of migration controls over the working lives of migrant workers remains limited for both industrial relations scholars and union practitioners. It is time for industrial relations scholarship and practice, we suggest, to take the issue of labour mobility fully into account, analysing more carefully migrant worker organizing.

Our theoretical argument that is developed in the final chapter (Conclusion) suggests that the critique of the trade union approach to migrant workers needs not only focus on migrant collective action or agency (Arnholtz and Refslund, 2019; Milkman, 2020), but on a broader rethinking of our understanding of worker power (van der Linden, 2008) through the lens of migration. With this aim, in our concluding chapter we reinterpret Wright’s (2000) forms of power, including new insight on mobility bargaining power, and discuss it in relation to the question of migrant labour turnover. Considering the relevance of migrants’ own mobility bargaining power (Strauss and McGrath, 2017: 203) as one form of labour market power (Silver, 2003), we renew Erik Olin Wright’s traditional framework showing how, in the context of the complex logistic of migrations, the ongoing crisis of social reproduction (Mezzadri, 2019), and the tightening of racialized and gendered (im)mobility regimes across an increasingly uncertain world, turnover emerges again as a critical practice of escape and resistance for workers. As immigration controls and growing precariousness critically hinder migrant voice and autonomy in the workplace, the ‘power’ of turnover, and what we identify as social labour mobility power transcending individualistic views of migrant mobility practices, may encourage unions to fully confront the question of temporariness and mobility of the workforce by building multiple and hybrid forms of solidarities also involving the reproductive sphere. We therefore present our theorization of labour turnover through the lenses of migration by looking at the continuum of migrants’ mobility struggles, encompassing both the individual and the collective realms, as well as the fields of production and social reproduction. The question of unions’ organizational renewal to strengthen internal democracy, sustain their dynamic aggregation with other forms of worker organizations, and truly open up industrial relations to the realities of transnational mobilities, constitutes the pragmatic horizon of our inquiry.