Getting home during lockdown: migration disruption, labour control and linked lives in India at the time of Covid-19

ABSTRACT This paper uses the Covid-19 induced migration disruption in India as a lens to interrogate what this acute moment reveals about the precarity of India’s migrant workers and their experiences of work in ordinary times. Interviews with interstate migrants from north India employed in the Tiruppur region in the southern state of Tamil Nadu present their narratives of being stuck at work when lockdown began, their subsequent struggles to get home, and finally their plans to return to Tamil Nadu later in 2020. Migrant accounts of migration disruption shed light on (1) the local labour control regime at destination that routinely keeps interstate migrants locked into highly exploitative work environments, and that was intensified during lockdown, and (2) the ways in which this labour regime thrives on the spatio-temporal separation of productive and reproductive spheres in migrants’ linked lives. The Covid-19 disruption also reveals how this labour regime flexibly adapted to produce the simultaneous disposability and unfreedom of circular migrant workers. Drawing on critical literature on labour control regimes, and on the separation of productive and reproductive labour under contemporary capitalism, we show how the Covid-19 pandemic disruption was anything but a transformative moment for India’s vast circular migrant workforce.


Introduction
The imposition of a stringent lockdown in India on 24 March 2020 triggered the immediate movement of millions of migrant labourers across the country trying to get to their homes thousands of miles away. Estimates of more than 30 million short-term and circular migrants on the move constituted 'a movement of people dwarfing even that which took place at the time of Partition' (Harriss 2020, 5). While these numbers of return migrants may well be a gross underestimation (Breman 2020, 910), millions of others, however, remained trapped, unable to move due to a lack of transport and cash, and often forced to stay put by employers. The Indian government, Harriss rightly argues, had given no thought to the implications of the lockdown on the enormous numbers of working people, including the estimated 60-80 million internal migrants, whose jobs and incomes evaporated overnight (Harriss 2020, 3;Ahmed and Deshingkar 2020;Srivastava 2020a). The lockdown constituted a sudden disruption unparalleled in the history of migration in India.
This paper uses the Covid-19 lockdown as a lens to interrogate what this acute disruption to circular migration reveals about the precarity of migrant workers in ordinary times. We zoom in on a particular flow of return migrants to offer novel insights into two related spheres of migrant workers' lives: (1) the ways in which a tight local labour control regime at destination keeps migrants locked into highly exploitative work conditions, including through the Covid-19 pandemic, and (2) the ways in which this labour control regime is bolstered by migrants' precarity and their spatio-temporal separation from linked lives and reproductive labour in sending villages. The sudden 'care deficit' that migrant workers faced during the lockdown is merely a visible indication of how employers routinely reject responsibility for the social reproduction of their labour force. We demonstrate that employers not only washed their hands off any responsibility towards their migrant workforce, but deliberately activated strategies to tie in workers when needed and shed them when their labour became redundant. Imminent and forced return migration disrupted the 'linked lives' of migrant workers across hundreds of miles, while simultaneously forcing many to re-migrate despite risks to their own lives, exploitative work conditions, and uncertainties about future earnings (Kleist 2020).
Migration disruptions caused by natural disasters, changing border policies, economic recessions (Martínez-Buján 2019), military coups (Nurick and Hak 2019;Bylander 2018), conflicts (Kleist 2017) and, more recently, epidemic outbreaks are neither unusual nor new. Bylander introduced the term 'migration disruptions' to capture the dynamics of such moments and move away from the framework of 'crisis' and instead 'consider how large-scale migration disruptions may offer telling windows on everyday dynamics within migration systems ' (2018, 3). A framework of crisis suggests that 'precarity, vulnerability, and exclusion are produced through acute changes to, or crises within, otherwise stable migration systems ' (2018). The concept of 'disruptions', by contrast, captures the ongoing challenges and interruptions at multiple levels that are inherent to the migration experience of labour migrants across the globe. In similar vein, we take the Covid-19 lockdown as a disruptionalbeit an extreme oneto, first, provide insight into the everyday dynamics of a labour control regime that managed to retain a tight grip on migrant workers throughout the pandemic and, second, to reveal how the spatio-temporal interconnections of productive and reproductive labour in the linked lives of migrant workers sustained this labour control regime to perversely deepen migrants' vulnerability. Our findings also unsettle assumptions about crises as catalysts for transformation or meaningful change for migrants (2018,5). Instead, disruptions often not only reproduce the status quo but consolidate the precarity and vulnerability of migrant workers. Here, we examine how the disruptions caused by the pandemic and lockdown largely served to strengthen an existing regime of labour control, intensified an ongoing process of externalisation of the costs of reproduction, and entrenched the precarity of migrant labour in India. While studies have explored the impacts of disruptions in the context of transnational migration (Nurick and Hak 2019;Bylander 2018), there is considerably less scholarship on the impacts of migration disruptions on the lives of labour migrants within countries, which is the focus of this paper. This paper provides ethnographic evidence of the return journeys of a group of north Indian migrant workersmainly from the states of Bihar and Uttar Pradeshemployed in the rural powerloom industry around the city of Tiruppur, in the southern state of Tamil Nadu. 1 The recruitment of interstate migrant workers in Tamil Nadu is shaped by a particular labour control regime rooted in the exploitation of migrants' precarious economic and social position at home, as well as their separation from kin back home and a lack of social networks at their places of work. We document how the disruption created by the lockdown intensified this labour control regime and migrants' precarity during the lockdown (Lewis et al. 2015). We present workers' narratives of the lack of care and support they experienced in Tamil Nadu, their struggles to get home between March and June 2020, and ultimately their evolving plans to return to Tamil Nadu in July and August 2020.

Labour control regimes and the linked lives of circular migrants
In what follows, we integrate two bodies of literature whose intersection helps reveal the extreme forms of migrant labour exploitation that were further intensified by the migration disruption during the Covid-19 lockdown in India. First, an extensive body of scholarship on migration in India has shown that ever more migrant labour is recruited on a circular basis (Breman 1996(Breman , 2020Srivastava 2020a). Circular migrants are not recruited to fill local labour shortages as is commonly believed (Breman 1996). Instead, employers deliberately employ circular migrants to tap into a flexible, informal labour pool that can be recruited and dismissed at will without having to shoulder responsibility for its longer-term reproduction. Employers also prefer replacing increasingly demanding, unreliable and recalcitrant local workers with circular migrants who are perceived as docile, compliant and cheap. Moreover, the contemporary exploitation of internal migrants is rooted in uneven development across the country, which results in circular migration from less developed, often agricultural, regions in the north and the east to industrial growth centres in the west and south (Srivastava 2020a, 165). Unequal regional development produces a form of 'internal colonialism' (Shah and Lerche 2020, 2), in which rapidly developing regions, such as the Tiruppur industrial cluster in the south, thrive on the recruitment of labour from poorer regions in northern India.
While internal migrants do not experience a change in formal citizenship status, as is the case for transnational migrants (Lewis et al. 2015), they nonetheless lack substantive citizenship rights in their places of work, given their limited access to state welfare at destination (Srivastava 2020b). In India, all social welfare delivery is contingent on local registration, including food rations under the Public Distribution System (PDS) and work under the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS). Moreover, social identity infuses internal migration, with migrants experiencing discrimination along caste, gender, religion and regional lines, among others, resulting in a system of 'conjugated oppression' (Lerche and Shah 2018). Such conjugated oppression is particularly brutal for long-distance migrant workers who are far from their families and deprived of full citizenship in the states where they work, thus making them particularly vulnerable at the hands of employers, as revealed during the Covid-19 pandemic. Circular migrants in India are routinely submitted to harsh labour control regimes, involving long shifts, dangerous work, low wages, and various forms of unfreedom (Breman, Guérin, and Prakrash 2009). Local labour control regimes have been described as the localised 'forms of domination and exploitation within labour processes that shape labour's material and political conditions' (Pattenden 2016(Pattenden , 1814. The concept of a local labour control regime, Fudge argues, 'is a useful way to analyse how workers' freedoms are constrained in the process of capital's attempt to extract value from workers ' (2019, 117). Indeed, with reference to Chinese industrial zones, development studies scholars coined the term 'dormitory labour regime' to capture the 'highly paternalistic, coercive, and intense production system, in which workers' lives are dominated by employers, and working time is more closely under the control of employers than in systems where working life and home life are separated' Pun 2006, 1456;Pun and Smith 2007). In China, the primary aim of such a spatial production politics was to facilitate 'the continuous access to fresh labour reserves from the countryside, depress wage demands and inhibit collective organization' Smith 2007, 1457). In India too, circular migrants are subjected to local labour control regimes that curtail their freedom, coercively control their time and labour, and maximise surplus extraction (Fudge 2019;Guérin 2013;Pattenden 2016). Researching Indian garment workers, Mezzadri (2017) introduced the concept 'sweatshop regime' to capture both the control and coercion exerted over workers within the labour process and the wider forms of oppression that shape workers' lives, including in their world of social reproduction (see also Pattenden 2020). Drawing on this literature we unpack a local labour control regime in Tamil Nadu that, building on the dormitory logic, sought to enlist North Indian circular migrants for employment in the rural powerloom industry.
Social reproduction is central to an understanding of migrant workers' experiences of the Covid-19 disruptions too. Indeed, migrants' working lives are also shaped by the separation of productive and reproductive spheres that marks the functioning of neoliberal capitalism (Fraser 2016). Migrant workers experience a fragmentation of the sphere of production ('the scattering of wage-work across sites and sectors') and a fragmentation of the broader sphere of reproduction ('a separation of workplaces from homeplaces') (Pattenden 2020, 24;Srivastava 2020c). That fragmentation is both spatial and temporal: migrant workers' homes and kin are often hundreds of miles away and may take days to reach. Yet, these fragmented spaces are not necessarily disconnected. Migrants remain closely linked to their household and kin networks back home. Their decisions to migrate are usually the outcome of familial considerations and reproduce interdependencies. The concept of 'linked lives' helps us make sense of the ways lives are linked across the life course, even when household members live and work away from each other (Elder 1994;Kõu, Mulder, and Bailey 2017;Bailey et al. 2021). These 'linked lives' tend to have a degree of regularity to them, in that they are usually planned and rooted in a set of expectations about how migrant journeys fit into long-term life course trajectories, intrafamilial exchanges, and aspirations for a better future (Bailey et al. 2021;Bryceson 2019). The idea of linked lives is also encompassed by the notion of 'transnational families', described by Bryceson as 'familial groups with members living some or most of the time separated from each other, while nonetheless feeling a sense of collective welfare, unity and familyhood' (2019, 3043).
Our informants similarly are part of what can be termed 'interstate families', in which family members dispersed across states rely on each other's material exchanges and emotional support and care. Young men usually migrate for work, while older and female family members stay behind in their home state. Typical flows of exchange include the family's initial sponsorship of the migrating member and subsequent expectations of remittances in return (Bryceson 2019, 3047). While scholarship has focused on material flows of goods and money, less attention has been given to the care and emotional support that circulate between migrants and their families (Bryceson 2019, 3051;Nurick and Hak 2019). This paper examines how migrant workers' linked lives became disrupted during the lockdown with consequences for their well-being at work and back at home. The Covid-instigated crisis disrupted the regularity of these exchanges as stranded migrant workers struggled to reach home, send remittances, or access family care and emotional support. While being part of spatio-temporally divided households makes it challenging for migrant workers to access family support at the best of times, being forcibly separated from their families during the lockdown enhanced their vulnerability at the hands of employers and strengthened the labour control regime they were subjected to.
Capitalists everywhere have long refused to recognise the unpaid and largely invisible care work that sustains and reproduces its labour force, and they have systematically externalised the costs of social reproduction onto workers' households (Mezzadri 2017;Fraser 2016). Capitalist employers not only invisibilise the contributions of reproductive labour to the sphere of production and value creation, but actively mobilise this separation of productive and reproductive realms to facilitate the intensification of labour exploitation (Mezzadri 2020a;Mezzadri and Majumder 2020;Carswell 2016). While this externalisation of the costs of social reproduction from the sphere of capitalist production is in many ways unexceptional, it tends to produce a crisis of care during moments of disruption, when the refusal of employers and the state to take responsibility for the welfare of labour is laid bare (Fraser 2016, 3;Rao 2021). The pandemic forms one such moment when capitalism's inherent crisis of social reproduction erupts, producing an acute 'care deficit' (Rao 2021, 40). In the context of India, where well over 80 per cent of the labour force is informally employed, the provision of basic sustenance during the pandemic had to be shouldered by workers' own households, many of which were struggling to survive and already 'overburdened with responsibilities of care and welfare provisioning, debt servicing, and payments' (Stevano et al. 2021, 277). Yet, this otherwise 'invisible economy of care', provided by households hundreds of miles away, became vital for stranded migrant workers during lockdown (Stevano et al. 2021;Shah and Lerche 2020).
We thus integrate the frameworks of local labour control regimes and social reproduction theory to make sense of the particular fate of a group of north Indian migrant workers stuck in Tamil Nadu during the 2020 Covid pandemic. We use this moment of exceptional migration disruption to demonstrate how capitalist employers thrive on a local labour control regime and on the spatio-temporal separation of productive and reproductive realms to shed all responsibility for the social reproduction of migrant labour. We also demonstrate how the lockdown was used to tighten employers' control over migrant workers with a view to secure their return, and thus their longterm availability for production and accumulation, post-lockdown.

Context and methodology
The material of this paper draws on long-term research on rural textile workers in the village of Mannapalayam, located the hinterland of Tiruppur, a large garment production and export cluster in Tamil Nadu, South India. Mannapalayam is located at the heart of a rural powerloom belt south and west of Tiruppur. Following water shortages and declining agricultural incomes, landowners in this region, mainly belonging to the Kongu Vellalar Gounder caste, invested in rural powerloom units from the 1970s (Carswell and De Neve 2013). As family labour became inadequate to run the looms, local labour was recruited and the majority of the powerloom workforce today consists of local Dalits (primarily Arunthathiyars) and members of intermediate castes, including migrants from neighbouring districts. Elsewhere we have described how Dalits became indebted to Gounder powerloom owners through cash advances, and how bonded labour intensified caste antagonisms between Gounders and Arunthathiyars (Carswell and De Neve 2013). While Arunthathiyars became tied to powerloom owners and subjected to exploitation and discrimination, Gounders in turn complained of workers becoming lazy, difficult and unreliable, not in the least because they were aware that employers needed them to keep the looms running and could not sack them given their cash advances.
The local labour control regime thus drew on caste, indebtedness and bonded labour to keep workers in check and extract labour power. However, when we first conducted research in Mannapalayam in 2008, it was clear that this regime was already under pressure. Employers had become increasingly frustrated thatdespite having distributed large cash advances to bind labourersthey struggled to maintain control over their local workforce. Indeed, local workers regularly switched factories (carrying the debt with them), asked for larger advances, took days off, or sought to escape demeaning powerloom work altogether by looking for work in Tiruppur or relying on expanding state welfare support (Vijayabaskar 2011). As their grip over the local Dalit workforce loosened, Gounders switched to an altogether different labour force from the early 2000s, when they started buying expensive automated looms, known locally as Sulzer looms. Having invested heavily in these units, often with extensive bank loans, factory owners started recruiting interstate migrants from Bihar and UP to keep their machines running round the clock. These young men were lured with the promise of free accommodation on factory premises and agreed to work 12-hour shifts, 7 days per week, 30 days per month. No local Dalits were willing to work at such an intensity and at levels of pay well below the local rates.
Migrant workers were provided with on-site accommodation, usually 10 by 10 feet rooms shared by 4-6 men, as well as in-room cooking facilities, but were kept under 24/7 surveillance by watchmen and CCTV. Most production units are located on farms, with migrant workers living isolated from each other and from the local population. Migrant workers are recruited through friendship, kin and village networks, with none mentioning formal intermediators. Such chain migration meant that many are closely related as kin and friends, and most of our informants belonged to a cluster of villages in one district of Bihar. They can best be identified as long-term circular migrants given that many have worked in the region for several years, only returning to their home states for annual breaks, festivals, harvests, and family ceremonies (Srivastava 2020a, 169). As non-resident members of households back in Bihar and UP, these men have not settled in Tamil Nadu nor do they intend to do so. Their linked lives include parents, spouses, and siblings back home, to whom they send money regularly. Most men move between factories in the Tiruppur-Palladam region, working for a few months to a few years in each of them.
This reconfigured local labour control regimeinvolving long-distance migrants alongside, but segregated from, local labourershad two aims: to ensure an uninterrupted supply of labour and to suppress wages in order to maximise surplus extraction. Migrant labourers, housed within factory compounds and at the beck and call of employers, constitute a neo-colonial workforce controlled and exploited to a far greater degree than the local labouring population (Shah and Lerche 2020). Regional identity has gradually come to complement caste as the modus of labour recruitment and exploitation. Crucially, this regime also draws on the physical segregation of local and migrant workforces, which prevents the latter from developing local social networks and organising through local unions or other forms of collective action. By the start of the pandemic in 2020, there were at least 100 interstate migrant workers employed in Mannapalayam alone, and many thousands more across the region. It is their accounts that we present below.
Being in the middle of a year-long project (2019-2020) we had interviewed interstate migrant workers, employers, local Dalits and regional migrants when the pandemic struck. We were able to continue collecting material, including accounts of the lockdown, from a range of people, as we had phone contacts for many of our informants. Most interviews with Tamil informants were conducted by the authors and later on by our local research assistant, S. Yuvaraj, who switched to telephone interviews from May 2020. In addition, in the first two weeks of July 2020, Nidhi Subramanyam conducted 15 phone interviews in Hindi with migrant workers who had by then mostly returned to their home villages in Bihar and UP. The migrant narratives collected through these semi-structured interviews were analysed thematically and are presented below.

The immediate aftermath of lockdown
When lockdown began on 25 March 2020, powerloom and sizing units in the region came to a complete halt and migrant workers were instantly without work and income. The local labour control regime immobilised most of them within the company accommodation and many reported being prevented by their employers from returning home or even leaving the factory compound. Aaditya was stuck in Palladam for one and a half months and commented on his own precarity: 'The employers do whatever they want. We are at the owner's mercy. He helps us only if he wants to. He asked us not to go to Palladam (nearest town) saying that we might fall ill'. In the early days of lockdown, factory owners were keen to hang on to their workers hoping to restart production swiftly. Migrant workers themselves, however, were anxious to get home. Anand from Bihar was stuck in a company with one co-worker and was eager to leave, saying 'This virus is so dangerous that if anything happened, I wasn't sure who would take care of us. So, I was desperate to return. My family was also anxious'.
While migrant workers reported being supplied with food by their employers or given money to buy provisions during the lockdown, employers retained tight control over their migrant workers by withholding workers' pay and deducting food expenses from their outstanding wages. Employers subtracted anything from Rs 800 to Rs 2500 per month from workers' unpaid wages. Jeeva from Bihar, who had worked in the region for 8 years, told us that his employer still owed him wages from before lockdown. During lockdown, his employer gave each worker Rs 200 per week and provided them with basic rations, such as rice, wheat, onions and potatoes. Asked if this was adequate, Jeeva replied 'It was not much but what could we do? We couldn't fight, right?' They used the money to buy cooking oil and other items locally. Govind (17), also from Bihar, had been employed for about a year in a company near Palladam, alongside his older brother Raj (19), who had been in the region for about 3 years. When lockdown was announced, they were not allowed to go anywhere apart from the local shop, and stayed in their rooms with about 20 others. While Govind too received basic food supplies for a month, his employer deducted Rs 2500 from the Rs 15,000 he still owed him in arrears. Danvir (26), who had been working in the Palladam area for about 9 years, said he 'sat around in the room for 2 months' along with 15-16 others. His employer gave them Rs 500 per week for food and other essentials. Danvir referred to this money as an 'advance' that would be deducted from future wage payments. As on 12 July 2020, Danvir had not received Rs 6000 of owed wages, minus food deductions, which his employer had promised to deposit into his bank account. Migrant workers appeared totally powerless in the face of employers' almost complete control over their survival, movements and finances.
Employers neither cared nor acted as patrons. The cost of any assistance provided by employers was entirely covered by the workers themselves who saw their outstanding wages docked, often by considerable amounts. When the lockdown started, most workers were still due their March wages, with many even owed wages from as far back as January. The pandemic disruption revealed the usual practice of withholding wages, a practice which is neither exceptional nor unintentional in this industry. It is routinely used by employers as a way of holding on to workers or enticing them back to the company after visits home. Following demonetisation in 2016 and the introduction of the GST Tax in 2019, the textile industry was already flagging in 2019 (Guérin et al. 2017;Harriss-White 2020). Delays in wage payments were more common than ever, with employers typically arguing that they themselves had not been paid and hence were unable to settle workers' wages. Danvir explained this routine practice of withholding wages: In our company, when you want to return home, he will withhold 10 days' pay. Only if you return [to the company], he will pay up. If you say you have an emergency, then he will say that he will deposit Rs 2000-3000 for our expenses on reaching home. But he never deposits it.
During lockdown, employers intensified this strategy of withholding wagesinitially in an attempt to retain their workers and later to entice them back to Tamil Nadu. Migrant workers were thus without cash and wholly reliant on employers' handouts even for basic needs like food. Combined with pandemic-related anxieties, they quickly grew desperate to return home. Initially, however, employers tried to hold on to them and several informants reported being prevented from leaving. Ranjit, from Muzaffarpur district in Bihar, had worked in the same company for about 6 months when lockdown began. Together with his co-workers, they asked their employer for money to return home but he refused and managed to delay their return for about 2 months. Ranjit reflected 'if the owner had allowed us to return [at the start of lockdown], I could have returned with Rs 2500 by train. But he didn't allow us to leave and made us stay forcibly. Since we are poor and helpless, we had to stay'. Below we recount how Ranjit eventually managed to get home.
While most migrants simply hung around in their shared company rooms, some were made to work during lockdown, either with or without pay. This happened to Ranjit and his co-workers, who had to work secretly and unpaid at night. Their previous wages were docked for food and, although by the time he finally left Ranjit was still due unpaid wages, he didn't receive any cash from his employer when he asked for financial support to get home. Other employers, he added, did pay their workers for work performed during lockdown. Anand, for example, worked for his employer and explained: He gave us money for food, and in exchange we would do some small tasks around the farm so that he wouldn't deduct any money from our next payments. I would clean harvested coconuts for him. … We felt that if we were eating, we should work for at least 2-4 hours per day. We shouldn't eat for free. If we ate for free, then there would be a problem.
It is worth noting again that the employer's assistance was not for free. Whilst wages were not deducted in this instance, the cost of food was compensated for by the alternative work migrants performed.
The local state was also of little help. Although Anand told us that they received an 'offer' of rations from the local panchayat if they showed their Aadhaar ID card, this was not mentioned by anyone else and we have no evidence of any migrant workers in Mannapalayam receiving food rations under the PDS. Most reported receiving no assistance from the Tamil Nadu state. Whilst their families back home may have received local state support, that was of no use to their sons in Tamil Nadu. It is clear that a portable PDS system, enabling citizens to access rations wherever they reside, would have been of enormous help to migrant workers (Srivastava 2020b). As well as keeping them fed during the crisis, it would also have reduced migrants' dependency on factory owners who as employers-cum-patrons had almost total control over them. Being already marginal to village life, deprived of local state support, and without money, migrant workers were particularly vulnerable and entirely dependent on their employers for food and shelter.

Journeys home
Within a few weeks, most migrant workers became increasingly desperate to return home, with workers mentioning fear and stress related to money as well as the uncertainty about the wellbeing of their families back home, and how to reach them. In this section, we describe migrants' complex attempts to return home to illustrate two broader points. First, we document how, as soon as it became clear that employers would not be able to restart their units any time soon, their labour control regime changed from holding on to workers to letting them go. This by and large took the form of abandonment as employers provided little or no support for workers' return journeys. Second, we show how stranded migrant workers had to mobilise their 'linked lives' with kin back home to raise funds for their repatriation.
In the course of April 2020, as it became clear that the lockdown would not be lifted soon, employers grew eager to let workers go. However, the cancellation of passenger trains from midnight 23 March meant that travel by trainthe most common means for migrants to cover the nearly 2500 km to Biharwas no longer an option (Harriss 2020, 3). Srivastava has argued that central and state governments sought to immobilise migrant workers during the early days of lockdown (2020, 171). It was not until the Indian Railways launched the Shramik Special Trains from 1 May that migrant workers could start heading back. Getting onto a Shramik train, however, was anything but straightforward or indeed cost-free. Some managed to book a Shramik train, such as Anand, who with friends arranged a free travel pass for them all. Screened at the railway station in Tiruppur and provided with food and water, the Shramik train worked well for them. But for many getting a train ticket was an ordeal that required negotiating bureaucratic processes with uncertain outcomes (Deshingkar 2022). It was not just employers but also the bureaucracy that appeared hostile towards migrant workers. Several, like Sanjeev who sought to return to Bihar by train in late May, had to pay a clerk Rs 2000 for what should have been a free ticket. Others, such as Logesh for example, were altogether unsuccessful despite filling in multiple forms online and offline and being sent from pillar to post to try and book a seat.
Many returning migrants thus failed to get home by train, whether because of the opacity of the booking process, trains being diverted, or bookings being made for the wrong trains. A lack of bureaucratic transparency, which oppresses the poor during the best of times, intensified the overwhelming confusion felt by migrant workers during the lockdown and contributed to their stress. Soon stranded migrants turned to other modes of travel to get back to Bihar. In the course of May, as lockdown restrictions began to be lifted and interstate travel became possible, coaches offered an alternative, albeit expensive, means of travel. Migrants from the same home area pooled money and approached bus companies to drive them home. The price seemed to fluctuate around Rs 10,000 for a seat on a coach, including the cost of travel permits. Unsurprisingly, no migrant workers, let alone those with outstanding wages, had such sums to hand. Lacking local social networks, their employers were the obvious people to turn to, not in the least because in India's informal economy employers are generally expected to act as patrons towards their workers, especially in times of crisis (Breman 1993;Carswell et al. forthcoming). Unfortunately, their appeal fell on deaf ears as hardly any employer was willing to settle outstanding wages or advance cash to fund journeys home. Most employers simply absconded their role of patron at this crucial time. The employers' strategy merely shifted from trying to retain migrant workers to effectively letting them go by abandonment. Crucially, stranded workers had to appeal to their own kin hundreds of miles away, thereby reversing the normal flow of remittances.
This was exactly the situation Logesh found himself in. Having failed to board a Shramik train by the middle of May, he went to the Palladam bus stand together with about 30 other men belonging to neighbouring villages in Bihar's Muzaffarpur district. A private coach operator agreed to drive them to Bihar for Rs 8700 per person. Although Logesh's employer still owed him Rs 5000 in arrears, he refused to pay this, only promising to deposit it into Logesh's account by 10 June. When we interviewed Logesh on 8 July, nearly two months after he left Palladam, no money had been forthcoming. Logesh was resigned to not seeing this money unless he returned to Palladam. It was ultimately his mother who sent him Rs 8700 for the bus journey. On leaving, Logesh pleaded with his employer for some money for food on the nearly week-long 3000 km return journey. The employer gave each worker a meagre Rs 500 to Rs 1000 for travel. Danvir faced similar difficulties. He too got a seat on a coach sometime in June, for which he had to request Rs 10,000 from his family back home, who had to borrow the money themselves. Danvir's employer refused to settle his outstanding wages and gave him nothing for the return journey. While Danvir eventually reached home safe and well, he had been without proper food on the 4-day-long journey. Belonging to the same Muzaffarpur district in Bihar,Manish (20) was similarly owed nearly Rs 10,000 in wages by the time he wanted to leave, but his employer too refused to settle any of it, merely promising to pay when he returned to work. While Manish's outstanding wages could have both covered his trip home and supported his family during the pandemic, his employer's refusal to settle these wages meant Manish had to ask his father to send Rs 10,000 for the coach. Ranjit, whose parents too borrowed money to get him home, finally explained when back home: 'we are helpless and struggling to eat. We are not doing anything. We have taken loans here and there. Only if we earn, can we eat'. By July he had already been without income for 4 months, which made life increasingly unbearable. Some families had to borrow tens of thousands of rupees to repatriate several sons from Tamil Nadu.
It is clear that north Indian men employed in the Tiruppur-Palladam region were simultaneously controlled and exploited by employers who used strategies to keep workers tied in even as they let them go. These strategies included withholding wages, deducting food expenses, refusing to support journeys home, and promising to settle wages on return to work in an attempt to secure a future labour force. While during normal times this labour control regime routinely curtailed workers' freedom in the labour market, during lockdown it became a particularly exploitative regime in which interstate migrant workers were made to rely wholly on their kin back home while remaining tied to employers in the south of the country. Their families back home ended up being their sole source of care and support, with parents sending (largely borrowed) money to repatriate their migrant sons. Migrants' lives remained closely linked across states and these connections to their sphere of social reproduction proved vital for their wellbeing during lockdown. What is equally striking is the almost complete acquiescence by workers. A lack of financial resources, local networks, political support and collective voice to make claims on their employers or the local state explains their powerlessness.

Surviving back in the village
We now turn to their stay back home and to how these young men saw the future. The picture that emerged was one of joblessness, deepening levels of indebtedness, and growing despair, which together explain why many of them were contemplating a return to Tamil Nadu. Most train or coach journeys home were fairly trouble-free and largely uneventful, even though several workers mentioned lacking adequate food on the way. On reaching their home villages, most underwent a health check, with some being asked to self-isolate for 14 days while others were held in quarantine, usually at the local village school. All the workers we spoke to said their families were Covid-free and only one informant mentioned that there was an infected person in their village.
In most informants' households, severalif not allsons had returned, but most were without work. While our insights into their household life in Bihar and UP is limited to what we could gather from the interviews, it was clear that households who owned land were marginally better placed to sustain themselves by farming and less reliant on borrowing to make ends meet. Rajesh's family in UP, for example, managed with stored grains and some farming work, and seemed to have enough food to get by. While Rajesh did not have to take out new loans during lockdown, he did mention struggling to repay older debts and having to borrow small amounts from various local people to repay a moneylender. Anand was in a similar position as his family cultivated their land and his mother managed the householdnow including Anand and other returnee brotherswith some savings she had. Danvir's household, which includes his mother, married elder brother and two sisters, survived on vegetables from a plot of land leased and managed by his mother. Their household had no money coming in and had to borrow to cover Danvir's travel home and to survive.
Most returnees, however, belonged to landless families that were far less fortunate. They had no produce at all and were reliant on their parents' earnings from casual work in the village, which soon became less regular and for which competition rose during lockdown. Unable to find work, Logesh and his brother who had returned penniless from Rajasthan relied on their father's casual labour as a construction worker, but they also had to borrow money to make ends meet. Logesh explained: There is no work here in the village. We cannot do anything. We can only earn if we go out [of the village]. I had planned to go and work with my brother to make money. But there is lockdown everywhere. Some say it will last till March 2021.
Manish's situation was similar. His father and elder brothers had also returned from Delhi emptyhanded, and his household of seven had to survive on the daily wage of Rs 50-100 that their father managed to earn from casual fishing work. Manish commented: We are wage labourers, only if we work, we can eat … we have not earned a single rupee for the last 2-2.5 months. Even now we are struggling for food. Our problems haven't gone away. We are somehow getting by. We take Rs 500-1000 from here and there and run the house.
During the lockdown, most migrants' households survived on irregular wages brought in by their parents, usually from local casual agricultural labour, wood cutting or construction work. Two reversals are noteworthy. One is an intergenerational shift from sons being the main providers towards parents becoming the sole earners as remittances ceased. The other is a spatial shift from rural households relying on remittances from places as far away as Delhi and Tamil Nadu towards local work becoming key to survival. As such, the connections between spatiallydivided spheres of production (employment in Tamil Nadu) and reproduction (the home in Bihar and UP) were not severed but reversed during the pandemic disruption. While inter-state migrants' lives remained linked to kin at home, the nature and direction of flows of care and support were transformed. Rather than being able to support their households, migrant sons became a sudden cost to their family, who lost a major source of household income, had more mouths to feed, and sunk further into debt (Guérin, Joseph, and Venkatasubramanian 2022). Many of our informants mentioned that having no employment and no money was by far the biggest stress at this time, even if they felt relieved to be with their families.
Assistance from the local state in Bihar and UP remained limited to food rations disbursed through the PDS and cash payments of Rs 500 in women's Jan Dhan accounts. While much welcomed, these payments did little to compensate reduced household earnings nor were they available to all. Some said that MGNREGS work had been restarted in their area, but their households had not benefitted from it. Those who were kept in quarantine on their return were being fed by the government and somebut not allmentioned that their household had received 5 kg of free rice per household member. No one reported any additional support in either kind or cash, while some even mentioned struggling to access rations under the PDS.
Promises that wage arrears would be paid into migrants' accounts once they reached home predictably never materialised. The young men themselves were not particularly surprised about this and seemed resigned to the fact that any wages due would either be lost or only settled if they returned to Tamil Nadu. Jeeva, who was one of the last to return by train on 2 June, mentioned that his employer owed each of his workers between Rs 6000 and 9000 in arrears. While Jeeva himself was due wages from before lockdownalready reduced because of the food billshe also worked for 18 days during the month of May when the lockdown was partly lifted and his company restarted production. His employer gave him part of his May wagesabout Rs 4000but had not sent the pre-lockdown wages when we spoke on 8 July. Even when far away, employers in Tamil Nadu retained a grip over their migrant workforce through the withheld wages, which formed an important incentive for migrants to return. In some sense, employers' local labour control regime transpired as not just 'local' at all its tentacles reached across state borders and were kept alive by workers' rapidly depleting social reproduction base at home.

'We will surely return!'
Given the rising levels of debt, joblessness and lack of government support back home, many young men were already contemplating returning to Tamil Nadu in July 2020, partly hoping to recover some of their wage arrears and partly wanting to start earning anew to replete household finances. They were also unsettled by the feeling of wasting time at home and putting their lives on hold. Mohan, who happened to have returned home on 3 March before lockdown, had been without work for 4 months, and was planning to travel back to Palladam as soon as transportation became available. His employer had not yet called him back, but he said 'there is no work here, so we will surely return. How long will we continue farming? In case we fall ill, we will need money for treatment. So, we will need to return'. Yet, feelings about returning to Tamil Nadu were mixed. Most men were unsure about how they would travel back as they had neither cash nor transport, or about what treatment they would receive back in Tamil Nadu. Several were still unclear about when they would be able to return and had no immediate travel plans, but for most it was not a question of whether they would return, but of when and how they would do so.
As soon as they were able to restart production in June 2020, employers changed their strategies once again in an attempt to lure inter-state workers back to Tamil Nadu. Their labour control regime proved to be eminently flexible and adaptable, and at this point several new elements were added to it: phone calls imploring workers to return, a willingness to send transport to collect workers, renewed assurances to settle withheld wages, and promises to hike daily wage rates on return. Taken together, these incentives raised migrants' hopes while also sowing concern about what to believe and how to respond. Jeeva told us on 8 July that his employer had been calling him for the previous 4-5 days asking him to return. Promising to send a coach to bring about 20 workers back to Tamil Nadu, the employer was asking the workers to WhatsApp him a copy of their Aadhaar card so that travel passes could be prepared. But Jeeva was unsure about how to respond, even asking Nidhi for advice on the phone: We are pareshan (stressed), but we don't know what to do. There is no work here, but if we return, we are not sure that it will be easy either. It was already very difficult to return [home] and now we are wondering whether we will face more problems when we get there. We are not sure how long the lockdown will last. Jeeva seemed fairly sure he would return, but was still working out the details of how and when this would be possible. Although he had not yet discussed future wages with his employer, he had heard that those who were already back in Tamil Nadu were receiving higher wages -Rs 490 per shift compared to Rs 450 before lockdown. The fact that he was still due thousands of rupees in unsettled wages no doubt affected his deliberations too. Jeeva ended with a rhetorical question, 'if we don't return, how will our family survive here?' When we phoned Danvir, he was in the middle of negotiations with his employer, who was begging him and some co-workers to return but less willing to fund the journey. Being dismayed about how they had been treated during lockdown, Danvir said he had told his employer that unless he paid them for the two lost months and sent Rs 10,000 for the return trip, they would not go back. The exchange seemed to have reached a stand-off, with the employer unwilling to concede and workers unsure whether to believe their employers' promises. Trust was clearly at the core of the impasse. While most men seemed set on going back to Tamil Nadu and were negotiating the terms of their return journeys, these negotiations were anything but transformative. They might have touched on transport costs and marginal wage increases, but did not address any of the systemic power imbalances that mark this employment relationship.
Manish was similarly hoping to return imminently, knowing that his employer still owed him close to Rs 10,000. Some of his peers had already returned to Palladam a couple of weeks earlier on a coach sent by their employer, but Manish had heard that their wages had not been hiked: Our owners told us that they have taken many loans to buy machines. So, they are asking us to come and work so that they can repay their loans. Some people are trusting these owners and returning, but if we go we will have to pay for food. Those who returned haven't received any pending wages, but I have heard they got Rs 500 as an advance for food [for the journey].
Clearly, mixed messages about cash for travel, payment of arrears, and promised wage hikes raised uncertainty about the wisdom of returning at this point. And yet, even Manish was clearly leaning towards giving it another try, saying 'we have to manage. If we fear the disease, we will die of hunger'. Such a statement reveals how the deepening dependence of impoverished young Bihari and UP men on employment in Tamil Nadu played into the hands of employers. The latter used migrant workers' vulnerability to tighten their grip on them and changed their labour control strategies to lure them back to work.
By mid-July many of the young men were ready to return to Tamil Nadu, but their families were much less keen on letting them go. They wanted to avoid a re-run of the lockdown saga and worried that their sons' return might deepen their financial precarity rather than strengthen their livelihoods. Our informants were clearly torn between the multiple risks of returning to Tamil Nadu and the potential gains to be made in terms of settled wages and new earning opportunities. Many, like Ranjit, worried about what would happen if they got stuck again in Tamil Nadu under another lockdown. While his friends encouraged him to take the 'risk' (using the English word), his family were more cautious, asking him to get a 'guarantee' from his employer first. Logesh too wondered about the risks ahead: What's the point of returning if we won't get paid or they'll cut our wages? If we return and there is another lockdown, what will we do? We had such great difficulty this time. My family has asked me not to return for now.
However, rapidly mounting debts and ongoing joblessness in Bihar and UP also made them more desperate, while promises of wage settlements and wage hikes acted as enticements. It is clear that migrants' experiences of lockdown in Tamil Nadu and of their return journeys home, together with the nature of their economic position at home, were crucial in shaping their willingness to take risks and re-migrate for work.
Whatever individual men ended up deciding, it was clear that employers' labour control regime had a pervasive spatio-temporal reach. It not only impacted migrant workers hundreds of miles away but also kept workers within its grip over time. Crucially, when lockdown was lifted and industry had a renewed need for labour, employers added new elements to their labour control regime: promises of settled wage arrears, wage hikes and sending coaches to lure inter-state workers back to the same employers who had abandoned them only a few months earlier. Migrants' precarity thus emerged from a labour control regime that took advantage of their class-based vulnerabilities, exploited their linked lives with kin back home, and restricted their freedom in the labour market in order to access their labour post-lockdown (Lewis et al. 2015;Shah and Lerche 2020). The pandemic disruption intensified this regime's grip while simultaneously revealing its adaptability to rapidly changing circumstances.

Conclusion
By May 2021, India was hit by a second and much more devastating wave of Covid-19 infections, spreading fear across the country and taking the official dead toll well into the thousands per day. Mirroring the migrant exodus of the previous year, long-distance migrants once again tried to flee Tamil Nadu before a new lockdown was imposed. The north-south circular migration flows that had only just been restarted were once again disrupted as a déjà vu of the previous year. This paper has taken the lens of 'migration disruptions' to shed light on the Covid-19 disruption as simultaneously an unprecedented shock and a typical instance of how migrant lives are more regularly interrupted.
A migration disruption perspective highlights how the migration experiences and linked lives of circular migrant labourers are, rather than being marked by stability and regularity, routinely disrupted by unpredictable events. From this perspective, the Covid-19 disruption to north-south migration flows in India offers a unique window on the everyday dynamics of migration and reveals the 'underlying structural conditions which both produce crisis and mask such conditions' (Bylander 2018, 3).
These underlying structural conditions, which generate migrant precarity, are at least twofold. First, the migration disruption lay bare the workings of a tight local labour control regime at destination, which keeps India's interstate circular migrants locked into highly exploitative work environments in Tamil Nadu and which was tightened rather than relaxed during the Covid-19 crisis. Employers mobilised a range of strategies withholding wages, deducting food expenses, and promising future wage settlements and pay risesto first prevent migrant workers from leaving the workplace, then dispose of them when factories lay idle, and finally lure them back when production restarted. Within this labour control regime, migrants' disposability and unfreedom transpired as two sides of the same coin (Lewis et al. 2015). Our study's contribution is to highlight the flexibility and adaptability of a particular labour control regime in India today (as revealed by its shifting dynamics during the pandemic), its grip on migrant labour across space and over time, and its embeddedness in a wilful neglect of the social reproduction of labour (Pattenden 2020;Mezzadri 2020b).
Second, the pandemic disruption revealed how this labour regime is rooted in a form of 'internal colonialism' (Shah and Lerche 2020, 2) that thrives on the recruitment of highly precarious and disposable interstate migrants, marked out by inferior caste and regional identities and separated from linked lives and the sphere of social reproduction back home. The sudden 'care deficit' that India's migrant workers faced during the lockdown was merely one manifestationalbeit an extreme onein a longer history of employers' systemic rejection of responsibility for the care and social reproduction of their informally employed labour. Migrant works ended up being wholly reliant on the support provided by their households hundreds of miles away. This spatio-temporal separation of productive activity from reproductive care and the simultaneous disavowal by employers of any responsibility for the costs of social reproduction enhanced migrant workers' vulnerability during the crisis. However, migrants' accounts of return migration also revealed how reproductive and productive spheres continued to intersect in their 'linked lives' (Mezzadri 2020a). While such economies of care that stretch across space and time tend to remain invisible during normal times (Shah and Lerche 2020), their significance came to the fore during the crisis when trapped migrant workers became fully dependent on their kin in home states. As such, the lockdown disruption revealed capitalism's systemic reliance onand simultaneous neglect ofthe labours of social reproduction in its relentless pursuit of labour surplus extraction (Fraser 2016;Mezzadri 2020a).
Crucially, employers' labour control regime was facilitated by government policies that routinely seek to restrict migrants' access to social welfare outside their home state and protect the interests of capital in maintaining a flexible and readily available workforce (Breman 2020;Srivastava 2020a). This was undeniably reflected in the Indian state's attempt to immobilise migrant workers during the early days of the pandemic (Srivastava 2020a, 171-175) and in the lack of state support available at destination where workers were trapped during the early days of lockdown (Deshingkar 2022). While not formally stripped of their citizenship, interstate migrants see their citizenship rights severely diminished, and often end up being seen as 'aliens' in destination states where antipathy towards them, fuelled by popular political discourse, is widespread Abbas 2016). In the absence of employer incentives to soften their labour regime and support the social reproduction of migrant workers, the onus is on the state to usher in a regulatory framework that better protects migrant workers against similar disruptions in future. State policy towards migrants needs to be underpinned by a better understanding of the nature of labour regimes and migrants' position within them, including the forms of labour control and unfreedom that many migrants are tied into (Deshingkar 2022). Workers circulating between states need to have their citizenship rights protected and be given access to state-provided social welfare at destination so as to reduce their dependence on employers and strengthen their fall-back position when disruptions strike. Such state support would curb the grip of employers on migrant workers, support their reproductive base at destination, and reduce the vulnerabilities engendered by the spatio-temporal separation of productive and reproductive life. Migration disruptions might then become moments of genuine transformation rather than events that entrench the precarity and vulnerability of migrant workers. Note 1. All names of our study villages and informants are pseudonyms.