Cultural capital and constrained agency in debt-migration for construction work in India

ABSTRACT Debt-migration is endemic in construction work in India, where exploitative social relations are an integral part of businesses’ capital accumulation model that can entrap migrants in forced labour and perpetual indebtedness. Drawing on concepts of cultural capital, ‘constrained agency’, and qualitative interviews with migrants and their wives, this study examines the embodied and material practices that migrants in Telangana use to challenge power hierarchies and embark on different life trajectories. The paper draws out the gendered differences in cultural capital and agential strategies that migrants use to place themselves and their families on an upwardly mobile path. Although still in debt, many migrants had succeeded in buying land, investing in farming, marrying their daughters, releasing themselves from moneylenders, and educating their children privately. A few had transitioned to working and living more permanently in the city of Hyderabad. This research sheds new light on how constrained agency and the use of cultural capital intersects with gender, ethnicity, and caste. It also shows that while the situations of both those who continue to circulate and those who settle in the city remain precarious, small incremental changes can add up to a significant shift in social and economic circumstances.


Introduction
The Indian construction industry employs at least 30 million workers, with more than 80% of the workforce employed informally (Junankar and Shonchoy 2013) including a significant proportion of migrant workers (Guérin 2013). The specific accumulation model of this industry, which depends on the extraction of cheap labour and the bondage of this labour through household debt (Guérin 2013;Pattenden 2016), as well as a recruitment structure relying on chains of contractors and village-level sub-contractors, have been closely linked to workers' economic and political exclusion (Mosse, Gupta, and Shah 2005). The bondage of construction labour is a necessity to ensure that this labour cannot leave during the peak work season (Guérin et al. 2007) which is typically post-Diwali from about October-November until the start of the summer in April in Telangana. The sub-contractors or mestri, are key figures, recruiting labourers, finding work sites and housing, and supervising the workforce (Picherit 2012). Recruitment of migrants typically involves large advance cash payments on the condition that the debt to the broker will be repaid through work. Such debt-migration has been likened to slavery since migrants remain indebted for long periods, forgoing their freedom to repay this debt. Migrant workers are preferred in India as they are unable to unionise and protest against exploitation (Guérin et al. 2007), and the majority are recruited from poorer communities in poorer parts of the country where opportunities for remunerative work are scarce.
The state of Telangana has an established system of debt-migration through chains of brokers, known as the Palamuru labour 1 system and dating back to the 1930s. Nearly all labourers who enter this arrangement belong to the category of 'Other Backward Castes' which includes agricultural labouring castes as well as artisanal castes, and the 'Scheduled Castes' or those who were excluded from the traditional Varna system. 2 Labour recruitment begins with a large government or private sector project for which several 'contractors' apply. When a contract is awarded, the contractor entrusts a broker or gumpumestri (also known simply as mestri) with the task of recruiting a workforce. These mestri may be ex-migrants from the same community and caste as the workers they recruit or outsiders belonging to a higher caste.
In this paper I analyse why male migrants and their wives take the decision to enter debt-migration arrangements, and how these decisions are embedded in their structural caste and gender position in rural agrarian hierarchies. I examine how lower caste men and women engaging in the system of debt-migration use advance payments to mobilise resources for marriages, repaying debts and education for their families in the village. I also scrutinize the strategies that they use individually and as a couple to convert their cultural capital into forms of symbolic capital to reposition their social status in the village. Additionally, I also explore the gendered negotiations and tactics that migrants engage in to access different kinds of work in the city which can offer the greater chances for economic and social mobility. The differing accounts of male and female migrants from similar caste backgrounds show how gender norms related to work and family shape their agential strategies and cultural capital accumulation pathways. The significance of the research is to add to our understanding of how poor, lower caste, and indebted migrants exercise agency despite their subordinate positions and why they choose certain pathways to emancipation based on investment in their children's education, hypergamous marriages, the purchase of land and establishing connections with urban patrons.
Conceptually, I draw on Bourdieusian ideas of capital 3 , social fields, and habitus (Bourdieu 1986(Bourdieu , 1987(Bourdieu , 1991 to understand the symbolic power struggles and positioning of lower caste migrants in relation to others in both the rural fields that they come from and the urban fields that they migrate to. The different kinds of capital that Bourdieu theorized, namely cultural, economic, social, and symbolic capital, are interlinked and strategically utilised by individuals in their 'position-taking' (Bourdieu and Wacquant 1992). Individual agents position themselves in relation to others in the field, which is conceptualised as a multidimensional, hierarchical social space. This positioning is based on the volume and composition of the different kinds of capital that they possess, and the precise social trajectory of individuals depends on the extent to which their capital is valorised by others in their field.
Bourdieu's complex concept of 'habitus' refers to the norms, values, attitudes, and behaviours of a particular social group or class. Thus, a migrant's social positioning within their caste constitutes their caste habitus which is situated within hierarchical social and economic dependency relationships with other castes in their social field. For women, there is an additional layer to navigate as their habitus is also shaped by gender norms in relation to their role within the family, and their culturally assigned care responsibilities (Kaur 2006).
The importance of caste as a religious and ritual institution shaping the division of labour and inter-dependencies in village life has eroded, due to a number of changes including migration and interaction with the outside world. However, caste remains salient in labour recruitment systems of the kind being discussed here. Many of the large construction companies in Hyderabad, the capital city of Telangana, are controlled by the upper castes, who often have roots in the capitalist farms of coastal Andhra (Ratnam 2008). These enterprises recruit lower castes for manual work through caste-based networks which reproduce power and caste identities in urban fields. This pattern of caste-based subjugation has been observed in modern employment sectors across the country (Lerche and Shah 2018). By understanding the specificities of caste interactions in both their rural and urban fields, we can begin to understand the strategies that lower castes use to transcend structural boundaries of their caste. Examining the temporal and spatial dimensions of acquiring, mobilizing, and exchanging capital thus becomes important.
The strength of a Bourdieusian framework over other conceptualisations of capital is its ability to offer insights into differentiation within social groups, their individual trajectories, and why some people are able to benefit from migration and others are not (Samaluk 2016). While Bourdieu has been credited with offering useful concepts for understanding structuration in society and how social exclusion and disadvantage are reproduced, his work has been critiqued for saying little about how forms of capital are mobilised for political struggle and resistance (Erel 2010). Gorringe and Rafanelle (2007) argue that in the Bourdieusian framework one's social status is internalised in a non-reflexive way and any agency to change this status quo occurs only when there is a shock creating a discrepancy between this internalized worldview and the outside world.
In order to address this shortcoming, I deploy the concept of constrained agency (Coe and Jordhus-Lier 2011) to complement Bourdieu's theories, using this concept to chart the struggles of migrant men and women from poorer villages and families in the state of Telangana migrating to the booming construction sector in its capital, Hyderabad. Constrained agency considers agential actions within structural constraints and explicitly recognises the gendered nature of agency. This approach also highlights the importance of taking an intersectional approach to examining specific and grounded processes of resistance through social and cultural relationships (p. 218). Here I examine male and female migrants' motivations, which are based on their understanding of what actions are 'possible, legitimate, and interpretable' derived from their structural positions (Fligstein and McAdam 2012). Their struggles relate to the acquisition of and conversion between different forms of capitalsocial, economic, and symbolicto elevate their status in rural society and establish a foothold in urban areas. I find both the Bourdiuesian framework, and the concept of constrained agency suited to analysing the agency and aspirations of migrants in Telangana and the strategies they use to rework and resist the constraints of their caste and gender habitus.
Interviews with rural migrants show that circular migration to urban locations is perceived to offer multiple possibilities for changing their social and economic position albeit with the constant risk of setbacks due to illness and injury. Given the enormous caste-based structural constraints in both the feudal societies that they come from and urban hierarchies they enter, migrants do not engage in open confrontation and struggle; instead, they undertake a variety of symbolic and material practices in both origin and destination to gradually change their social and economic standing. To analyse these processes, I draw on research by Gidwani and Sivaramakrishnan (2003) and Sunam (2014), who showed how strategic consumption is used to change social identities and rank. Their research demonstrated that rather than reworking or challenging hierarchies in urban areas, migrants are focused on changing social hierarchies in their villages of origin because settling in urban areas is not possible. The dynamics observed in Telangana are similar, in that migrants accept exploitative work in urban areas and do little to challenge such a scenario, instead focusing their spending on conspicuous indicators of wealth and success back home, such as housing and land as well as hypergamous marriages and private education for their children. However, they do pursue, and in some cases, succeed in settling in the city, and I unpack the processes and forms of capital conversion below.
One route through which opportunities for urban work could be accessed by able-bodied poor men and women who lacked social networks in urban areas was through debt-bondage for manual labour on construction sites. Bondage by definition would imply a loss of agency and several authors have emphasised the inability of debt-migrants to extricate themselves from continuing debt because their meagre earnings are used mainly to pay off the borrowing and on basic consumption (Srivastava 2004, Meher 2019. It is also argued that debt is used to pressure labourers to intensify their work rate and extract surplus labour time (Pattenden 2016). Many studies mention the use of debtmigrants' advances and earnings for marriages, health and housing but these uses are not regarded as socially or economically life-changing. For example, Meher (2019) dismisses such uses and contends that migrant construction workers from Odisha were trapped in a 'perennial cycle of debt' with no change in their social or economic situation (p. 776). Similarly, Pattenden (2016) while conceding that workers return more emboldened and assertive because of their reduced dependence on farmers back home, argued that they do not experience gains beyond mere discursive and symbolic shifts.
Only a handful of studies have considered the potential of debt-migration to create the foundations for a better social and economic future for the migrants involved. In her study of debt and migration in Tamil Nadu Guérin (2013) called for consumption to be considered not just as a reason for obtaining advances (from the mestri), but also a mode of resistance and struggle (p. 411). She also showed that consumerist behaviour, by encouraging workers to take larger and larger advances from mestri actually reinforces their own exploitation (p. 413). Her research thus highlighted the paradox of increasing aspirations for equality and integration that were helping to reproduce capitalist exploitation and extraction of surplus value. Marius-Gnanou's (2008) research in Tamil Nadu also demonstrates the ambivalence of debt-migration; although migrants remained in debt, they used advances for expenditure on marriages, housing and land which they regarded as an important step to improving their standard of living. I take these studies as my point of departure to provide insights into the ways in which migrants convert their monetary advances to other forms of capital to embark on a trajectory of upward social mobility, even while they continue to be in debt.

Methods
The research was conducted in Hyderabad, the capital of Telangana state, one village, BN 4 , in Narayankhed Mandal in the Medak district of Telangana, and another, LG, in Makhtal Mandal in Mahbubnagar district, also in Telangana. Both districts rank among the poorest and drought-prone districts in the state have and have a long history of labour migration thus providing a suitable setting for examining how agency is exercised in highly constrained circumstances. Previous research by this author showed that 78% of the households in (BN) and 70% of the households in (LG) included at least one migrant (Deshingkar andStart 2003, Deshingkar 2007).
Field visits and interviews were conducted by the author and research assistants at various points between 2002 and 2009, including 35 semi-structured indepth interviews with adult men and women in rural areas and informal discussions with male and female workers in Hyderabad that were conducted opportunistically during the author's 12-year residence in the city between 1997 and 2009. Informal conversations with men and women in the city took place at a large construction project work site but it was not possible to interview women there due to social restrictions on their interactions with outsiders. Therefore, some of the conversations with women were conducted in places where they could express themselves without being monitored by men, employers or contractors. Six of the women interviewed were domestic workers in the author's neighbourhood and three were working alongside their husbands on construction sites. These conversations extended over several weeks allowing the author to probe decision-making processes, gender relations, lived experiences and aspirations. The author was aware of her positionality as a middleclass and upper caste person who would typically employ lower caste workers and therefore her approach was reflexive and conscious of the impact of the caste/class divide on the interviews. Building trust and rapport helped to mitigate some of these possible impacts. During the interviews, participants recalled the choices they made to give meaning to their trajectory, and this analysis must be read with that in mind.
Additionally, focus group discussions and observation at construction sites were also employed to gain a broader perspective on the migration process. A small number of interviews with mestri, as well as urban employers, NGOs, and activists, were also conducted. These sessions provided additional sources of information and a way of cross-checking the information gathered from migrants' personal biographies. Sites for the interview were selected purposively, starting with discussions with local NGOs. Respondents were selected based on willingness to participate in discussions. Informed consent was obtained verbally from the participants and recorded during the interview as it was felt that asking them to sign a consent form would create feelings of insecurity.

Communities of origin: caste relations and work opportunities
The positions of the lower castes in the social hierarchy of the two villages were first considered, as well as structural constraints within which any struggle would occur. Both villages have a predominantly Hindu and multicaste hierarchy and migrants belonged to landless or marginal farming families from more disadvantaged castes. These included the poorer 'Other Backward Castes' (OBCs), a heterogeneous group of castes including the Mudiraj fisherfolk in Medak and Golla cattle-herders in Mahbubnagar, and Dalits or the erstwhile 'untouchable castes', who were at the bottom of the caste hierarchy. Attention was also paid to gender norms and how these conditioned women's roles within the family, the surveillance of their behaviour and the limits on their control of cash and other assets.
Although local power relations based on caste were somewhat eroded by the Telangana Peasants Armed Struggles of the 1940s and '50s, the ineffectiveness of land reforms of the 1960s, has meant that land ownership and thus wealth from local activities has remained polarised. In BN, the Brahmin landlord reportedly owned 80% of the land (in excess of 300 acres), with the majority of the population being either landless or marginal farmers. In LG, the land, as well as political institutions such as the Gram Panchayat (village council) were controlled by the upper caste, Reddy. There was a wide wealth gap between the ruling classes/castes and the OBCs. In both districts, the poorest and most marginalised were the Dalit or Scheduled Castes.
Relationships of servitude and bondage to the upper castes characterised the work options available to the poorest groups within the village which included poor OBC castes. Since the mid-1970s, migration had become an important part of the livelihood portfolio for the OBCs and Dalits, who have successfully mobilised their caste identities and social capital vis a vis the recruiters and employers whose caste habitus is often different. Migration was perceived to offer a way of escaping caste oppression, even though it involved entering other relationships of servitude and bondage. I argue that this is because debt-migration opened up possibilities beyond the immediate workplace in a variety of different ways, in both rural and urban areas as I discuss through the cases below.

Debt-migration patterns
Diverse migration streams had emerged in both the study villages, which were differentiated largely by caste. Scheduled Tribes were engaged mainly in rural-rural migration for sugarcane harvesting as they were socially constructed as 'unreliable', 'dirty' and 'alcoholics' who were nevertheless seen as suited to backbreaking work in the fields. Dalits were also mostly excluded from debt migration to urban construction work account of being perceived as poor workers who were not committed to working hard. On the other hand, debt-migrants in the construction industry mainly belonged to the OBC castes and a very small number of upwardly mobile Dalits who carried markers of the OBC habitus that the mestri were seeking, namely being hardworking, reliable, and obedient.
The typical 'contract' that debt-migrants had in place with the mestri set them on a highly constrained path by agreeing to bondage for advance payment and to repay this advance by working over the 8-9 months of the working season. It was only the poorest families where both the husband and wife migrated together, often leaving older children behind. The mestri was all-powerful in setting wages and usually fixed them well below the open market rate to ensure profits for himself. In 2002, advances paid to a couple for construction work were up to Rs 20,000 and had risen to Rs 25,000 by 2007 (approximately US $625 at 2007 exchange rates), roughly 50 rupees a day, which was much below the open market wage for construction workers. Although aspiring migrants in Mahbubnagar had more options to choose the 'best' deal in terms of the advance being offered, the wages (between Rs 50 and 100 a day in 2007), and the food (whether meat/ chicken provided once or twice a month), they had little control over the long hours that they would be made to work. Workers were sealed off from the wider system of labour protection, citizenship, collective agency, and unionism, and workplace cooperation between them was non-existent.
The mestri represented migrants to urban employers and contractors, drawing on their embodied cultural capital and the group identity of 'Palamuru labour'. This label signalled a habitus to urban employers of being hardworking and obedient, which reassured them that these workers would accept their exploitative working conditions without protest. This kind of representation was also observed by Picherit in his study of migrant construction workers in Andhra Pradesh (Picherit 2019). Rajesh, a construction company owner in Hyderabad said, 'we prefer Palamuru labour recruited by the mestri as they are habituated to heavy physical labour and he can keep them under control.' The cultural capital that the mestri and migrants mobilised with their signalling behaviour towards potential employers and clients was thus not derived only from the embodied cultural capital that they possessed as individuals; it was also derived from their group identities and the expectations of employers. Without the mestri's representation and valorisation of their identities to urban employers, migrants would struggle to find work in the city.

Working conditions in the city
Despite an impressive array of labour laws in India, 5 the working and living conditions of migrant workers were uniformly dire in the urban locations that were visited. For example, in one large luxury residential project, where each flat was retailing for around seven million rupees, a temporary shanty without running water and toilets had been established for the workers.
In the Hyderabad construction industry, the lowest ranking labouring jobs, such as bricklaying, were performed by men belonging to lower castes. Auxiliary services, such as sifting sand and carrying bricks, were performed by women from a similar caste background. Men could move up the job ladder to slightly more skilled positions, but women never graduated to other levels in the construction industry. Highly skilled and managerial jobs were held by people from other parts of the country or higher castes who were regarded as more able to undertake such work.
As noted above, power and control at urban worksites reproduce social hierarchies and agrarian caste identities in rural areas. Those at the bottom of the construction industry hierarchy had learnt from long experience that rebelling against the working conditions would result in dismissal, and their strategy in such heavily constrained conditions was to perform meek behaviour to maintain access to the work. They also co-opt the narrative of the employers to stay in the job. During a group discussion one migrant commented, 'We do whatever the Malak (owner) asks us to do. As Chhote Log (small people) we do not raise our voice. Otherwise, we will lose our jobs.' Notwithstanding the lack of struggle and rebellion in the workplace, other forms of constrained agency were being pursued in parallel by the migrants in urban areas. They were doing this by sending remittances for the social repositioning of their families and also forging new relationships and accumulating new knowledge of the ways of the city and urban people.

Debt migration, family aspirations, and change
Migrant workers' decisions to enter degrading and poorly paid work at their migratory destination were closely linked to major milestones in the life course of the family and aspirations for social and economic mobility. Daughters' marriages, investment in tube wells for groundwater irrigation, and educating children in English-language schools, which were thought to offer better job prospects, were identified as priorities in conversations in rural areas with aspiring and returned migrants, as well as their families. The average spend for a daughter's wedding was Rs 25000 in 2002, which made a Rs 20,000 advance from a broker very attractive. I track the experiences of two lower caste couples, Ramulu and Seetha and Lingaiah and Laxamma and their efforts to improve their social standing through debt-migration. Their experiences are not identical to others from their village and caste background but there are similarities that can help to draw wider inferences.

Ramulu and Seetha
Ramulu and Seetha belong to the OBC caste category who were historically from an artisan caste and continue to be dependent on the local Deshmukh landlord in BN and other farmers with large landholdings for their livelihood. Before they migrated, Ramulu's parents and two older brothers were sharecroppers on extremely disadvantageous terms where they were effectively labourers because the landowner bought all the inputs and took a third of the crop. 'If the rains were good, we did ok, but if the crop failed, we would have to share the losses', he said, explaining the precarity of his family's situation. Seetha, like other lower caste women in the neighbourhood, supplemented the family income through other seasonal work like leaf plate making. This was extremely low paid work, done on a contract basis where the leaves were dropped off by a contractor who would collect the made-up plates in exchange for a few rupees. Other members in the family did other kinds of casual work. Their combined income was not sufficient for survival and borrowing during the lean season or in the event of a health crisis or other shock was common. The family had long been trapped in debts that could not be repaid. Previous research by the author had shown that 68% of the households had borrowed from moneylenders at high interest rates.
After their marriage, following the example of other young men in the village, Ramulu decided to take the significant step of migrating to Hyderabad to establish himself as a responsible male member of the family and mobilise funds to repay the family debt of Rs 30,000 incurred for his sister's wedding. He contacted a mestri who was known in the area and asked him to include him and his wife in the mestri's worker group. Ramulu and Seetha migrated with a mestri, leaving their 10-year-old son behind with his parents. In accordance with gender norms, the mestri negotiated with only Ramulu at the time of recruitment as it was his abilities and skills that were the focus of the arrangement and Seetha was expected to cooperate with her husband's decisions. Ramulu's acceptance of the grinding conditions of migratory work must be understood within the structural constraints suffered by his caste and the cultural, social, and economic capital available to him. For him and others like him, such a move represented one of the best routes to a better future.
Seetha's role as the accompanying wife illustrates a more complicated picture. Her movements were constrained by patriarchal controls on her decision-making and independence, which enforced obeisance and deference to male members of her family. As is the custom in their community, she was married to Ramulu at the young age of 14 and had her first son by the time she was 16. Seetha's body, apart from being subjected to the racialized and caste constraints like her husband, was also a site of patriarchal control and subjugation. In the tradition of their community, the decision to migrate was made for her by her husband and in-laws, and she did not have much say in the matter. Her acceptance of migration for the advancement of the family can be interpreted as being passive and devoid of the ability to exercise even the kind of constrained agency that her husband did. However, as shown later in this paper, migration to urban areas gave her room for finding a job that had possibilities for better earning, saving and renegotiating her relationship with both her husband and the in-law's. Additionally, her participation in the migration was key for the ability the family in the village to challenge their caste habitus through new kinds of material and symbolic capital. The tensions in married women's experiences of migration, work and their agency has been noted by Manohar (2019) in her study of Tamil skilled female migrants in the US.
By Ramulu's calculations, his family's outstanding loans would have been repaid in three years if he and his wife had worked hard in the city. However, the rains were bad, and they could not obtain rice from the village which meant their expenses at their destination were higher than expected. Seetha revealed that one reason for unexpected expenditure was Ramulu's drinking habit which he had developed as he was away from the social controls of elders in the village. He took further small loans from the mestri and this brought unanticipated burdens and worries for Seetha to maintain her earning in case Ramulu was unable to work regularly.
Ramulu's aspirations, shaped by what he had learnt in the city and identified as important for securing a better future, had shifted to providing his son with an English-language education. 6 On their most recent stint, the mestri had given Ramulu an advance of Rs 15,000, and he intended to use some of it to pay for his son's school uniforms, books, and transport to the town, which was 6 km away. Being able to educate the son was a major step for a family where several generations had been illiterate, but it represented a new expense and had meant that the debt had not been repaid in full. Ramulu had no plans to return to the village and had no clear timeframe for repaying the debt. His journey had been extended with changing aspirations and agential strategies, and he had set about accumulating social and cultural capital through his son's education. He felt that the situation of the family had improved through small incremental changes and their future outlook was promising. He was hoping to eventually settle down in Hyderabad, the land of opportunity: 'Amma, no one sleeps hungry in this place. I work hard and I get paid. I can make a good life for my family. ' Seetha's experience of the city and construction work was much more constrained and arduous than Ramulu's. In the city, her burdens and constraints increased because of a lack of her natal family's support, difficulties at the worksite and Ramulu's spending. She was employed at the construction worksite since her labour was cheaper than that of a male but was constantly subjected to sexual harassment by the contractor and male co-workers. The harassment of women construction workers by male contractors and brokers is endemic in the industry (Madhok 2005). Lower caste women are doubly disadvantaged, and their experience is shaped by the social construction of their identities as 'sexually available' and the ideologies of masculinity and caste that give upper-caste men the belief in their right to abuse such women (Olsen and Ramanamurthy 2000, Parry 2014, p. 1249.
At the same time being in the city gave Seetha the opportunity to remove herself from this predicament by looking for other kinds of work. Domestic work was perceived as a better option than construction work because it was indoors and away from predatory male co-workers and contractors. She found an employer through friendships she had developed with other domestic workers from the same rural district, who she had met at shops and tea stalls in the city. Personal recommendations from someone the employer knows, and trusts are crucial for this kind of employment.
Women like Seetha have to carefully mobilise the habitus that they carry; Seetha had to appeal to prospective employers' expectation of being able to exploit, dominate, and exclude those with a lower caste identity from sharing their amenities. For example, low-caste domestic workers are typically not allowed to use employers' toilets or drink from the same cups. At the same time, Seetha had to reinvent her identity and acquiesce to demonstrate the personal traits that the urban middle classes did not find offensive or threatening, such as dressing 'decently' in ways that suppress sexuality; speaking 'properly' by using urban and polite language; appearing docile and obedient; and appearing clean and free from disease. These kinds of embodied and objective forms of cultural capital were converted to symbolic capital, which provided migrant women with a status distinct to construction workers and with the possibility for improving their living standards through other benefits they could mobilise. At the same time Seetha also described the insulting way in which her employer spoke to her demanding that she do extra jobs without remuneration. She recognised the precarity of not having a formal contract and the exploitative nature of the work but decided not to challenge them because she wanted to stay in the job. Similar to the strategy of male workers at construction sites, she was carefully considering when and how she would change her circumstances.
After two years of employment as a day worker, Seetha was offered a room and a live-in arrangement by her employer. Thus, Seetha was also able to build new forms of cultural capital (Erel 2010) through her relationship with her employer and over time, she used this to negotiate better working hours, leave allowance, and pay, which she could not have done at the construction site. This scenario also represented possibilities for settling in the city as her knowledge of urban ways of living and bodily comportment improved. Even though being a live-in worker meant a reduction in her personal freedom, it also gave her an income of her own that she saved with her employer. She was under pressure from Ramulu to surrender her income, but she kept a part for herself to buy clothes and gifts for her friends and natal family and in that way raise her own social status back home.
Like other migrants from their village, both Seetha and Ramulu had adopted ways of dressing and accessories that marked them out as people with urban connections and resources. It was easy to identify such migrants on visits to the village from the way they spoke, peppering their language with English and Dakhni words, and wearing flashy gold watches, chains, smart clothes, and hairs clips. In doing so they were behaving in ways that would traditionally be available only to the upper castes (Gidwani andSivaramakrishnan 2003, p. 187, Sunam 2014). Although their actions did not completely alter the power balance in rural society, it nevertheless gave them some prestige, which could, in turn, open pathways for gaining other kinds of capitals in the form of wealth, and important contacts and networks (Sunam 2014). These upwardly mobile families and migrants objected to being called Jeetham and displayed body language suggesting that they were not to be regarded as downtrodden. Similar practices were used by migrants in Gidwani and Sivaramakrishnan's study and by migrants returning from the Gulf to Kerala as a way of transgressing social boundaries and hierarchies (Osella and Osella 2000).
At the time of the interview Ramulu said that the money that he and Seetha had sent back home had allowed his father to repay the family debt and start saving towards buying a plot of land. Seetha had become part of a local association of domestic workers negotiating for better wages and work conditions. She had also learnt about child-care and hygiene practices in middle class homes through her work and was exploring the possibility of moving on to a better paid position of child minder or ayah.

Lingaiah and Laxamma
Another example is Lingaiah and his wife Laxamma from LG, who also belonged to a marginalised caste and jointly owned just two acres of land with Lingaiah's three brothers. Due to the shortage of work in the village, they all began to migrate for sugarcane harvesting in the 1980s. Lingaiah recalled many periods of hardship, especially during the drought years. Lingaiah's goal was to buy land, marry off his daughter, and educate his youngest son: 'Madam, I don't want my children to work like me. I am working hard to educate my son. I want him to get a government job so that he can support me and have a good life.' Laxamma had always wanted to migrate to the city and work as a domestic maid as she had seen other women in her village dressing fashionably and commanding respect from both the men and women in the neighbourhood. There was long history of migration in the village and women's migration as well as their work in other people's homes had become socially accepted.
However, she felt her conservative in-laws would object to such a move and she presented her own migration with Lingaiah to them as a familial duty. She engaged in delicate negotiations with them within her role as the family and persuaded them that her migration would enable her to be a better daughter-in-law and mother as her earnings would bring financial benefits to them and investments in her children's future. She also persuaded Lingaiah to take her along, emphasising her care role and the security that he would get from her homemaking and networking with others in the urban neighbourhood.
Although the city involved harsh work and the double burden of working and managing the home without the extended family, Laxamma felt liberated to be away from the strict controls on her personal freedom in the village. During the evenings and off days, she was able to go out to markets and shops on her own and socialise with others of her own choosing. A couple of years after migrating seasonally to Hyderabad and repaying the bulk of their debt with the help of their extended family, Lingaiah and Laxamma moved permanently to Hyderabad and offered a culturally acceptable explanation to his parents that the city offered them multiple work opportunities and better futures for their children.
At first, they found casual work in construction and stayed in temporary shacks erected near the construction site, but after a couple of months, they found accommodation with a cousin who was already living in the city. By relying on this social connection, they reduced the costs and risks of finding accommodation and also had access to information about the availability of work in the city. They stayed with the cousin for four months during which time they learned key Dakhni phrases and how to perform urban behaviours and mannerisms that were regarded as appropriate for their caste. The couple also gathered knowledge about the accommodation system in Hyderabad and were able to rent a small room in an informal settlement. Networks with others from a similar class and racial position as themselves, but who were more established in the city, were also invaluable for Lingaiah's wife Laxamma to secure a position as a domestic worker. The cousin's wife took Laxamma to meet a prospective employer since personal recommendations or a reference from someone the employer knows, and trusts are crucial for employment.
By the time of the interview, Lingaiah and Laxamma had a good understanding of how the casual labour market in Hyderabad operated and where they should position themselves. Laxamma had aspired to bring her daughter to the city to free her from the traditional practice of early marriage and childbearing as she transitioned to adulthood. Due to her own success in finding better work in the city, she was able to use that to influence her inlaws and Lingaiah to bring her daughter to the city. Like her, her daughter's agency in the city was also extremely constrained, being embedded in relations of patriarchy and servitude which were exploitative but offered possibilities for recasting her own identity and life course.
With three incomes in the family, the couple were able to arrange a lavish wedding for their daughter with a man from a wealthier family than theirs, spending 50,000 rupees, (30,000 from their own savings and 20,000 borrowed from one of Laxamma's employers, at 3% interest). For families like theirs, marriage is important to affirm their position in society by establishing a link to a family of higher status than their own. While hypergamous marriages and escalating dowry have been blamed for immiseration and son-preference in India (Bhalotra et al. 2020), their importance for raising social prestige has also been noted (Bloch et al. 2004).
Lingaiah also purchased a plot of fertile land from the village landlord and was planning to send his youngest son to an English-language school 7 near his village to improve his chances of finding a white-collar job. Lingaiah was aware of the symbolic importance of his investments and felt that he could face the ruling caste back in the village without having to show them undue deference as his forefathers did. He was hoping that relationships that he had established with a variety of people including roadside chai carts or bandi wallahs, shopkeepers, mestris and others from his own rural area would help him find a job that was above being a manual labourer. Being a security guard or 'watchman' was a preferred option among his friends as it represented a step up from hard manual labour.
While in the city he had learnt how to speak and carry himself using terms of respect such as 'Sahib', saluting and adopting a humble countenance to demonstrate his respect. These behaviours were valued by the urban middle class and upper caste families that typically inhabited bungalows and blocks of flats that require watchmen. In this way he could, in the future, mobilise these attributes and convert material and embodied forms of capital to symbolic capital, which could allow him to remain in the city.
Laxamma's longer term plan was to help her daughter to find an informal 'factory' job in the city in a small manufacturing unit or similar, as she knew others in the vicinity who had succeeded in doing this. Factory work wages were high enough for the family to repay their debts faster.
In probing the decisions that underpinned labour circulation or settling in the city, it became clear that migrants were constantly assessing future possibilities against the precarious conditions in their past agrarian lives as well as their lives in the city. Their subjectivities and aspirations were continuously changing in response to their experiences and exposure to urban ways of life. In my interviews, some migrants stated that they would eventually like to settle in the city, or that they wanted to return themselves while their children settled in the city. These aspirations were driven by the perception that there was no future in farming, although owning a plot of land in the village was seen as important because of the status it conferred. They pursued a number of strategies within the constraints of urban living and working to start laying the foundations for settling there more permanently or switching to more stable and less physically demanding work.
Given the already precarious position of debt-migrants in urban construction and the entrenched segmentation in Indian labour markets (Papola 2013), which restricts certain kinds of people to certain kinds of jobs, any agency should be considered against the backdrop of these constraints. Indeed, data on urban labour markets in India shows clear division by caste, with the OBCs and Dalits concentrated in labouring jobs and OBCs outnumbering the Dalits (Junankar and Shonchoy 2013). Furthermore, access to urban housing is also notoriously difficult. Urban spaces in India are characterised by corruption and clientelism based on identity links, power, and personal relationships 7 and establishing networks with patrons from more powerful classes is essential for those who belong to less privileged social groups. Migrants from the two villages in this study invested time and effort into developing connections and traits that would help them to establish such connections of trust beyond their own ethnic networks with coworkers and the mestri. Each stint in the city presented new opportunities for interactions and networking with a variety of people in the workplace, at places of leisure, and at marketplaces, through whom connections with urban patrons could be established.

Conclusion
In the two Telangana villages considered here, I have used a Bourdieusian lens together with the concept of constrained agency to analyse the agential strategies of male and female debt-migrants belonging to lower castes. I have unpacked the different ways in which they use their cultural, economic and social capital to advance their own personal goals as well as familial obligations. Men use their embodied cultural capital of backward casteness to signal to mestris that they are suitable for being hired as a debt-migrant. They were using mestri advances to fund the purchase of fertile land, status-enhancing marriages, and education in English-language schools.
In the city, migrants invested in various kinds of embodied and objective forms of capital, engaging in practices to demonstrate their urbanity through the way they dressed, spoke, and behaved to gain acceptance among urban patrons. Men achieved this by tactically mobilising their caste habitus and performing the bodily and behavioural markers of being servile. On the other hand, the agential strategies used by women are much more complicated. They engage in complex caste and gendered practices of obedience, co-opting of patriarchal narratives to become more independent and recast their identities as modern women within their own means. Their actions and strategies show that they are agential in numerous ways and not passive or powerless as often portrayed.
Once they are in a position to release themselves from the mestri, men sought access to less strenuous jobs. Women sought entry into domestic work, which compared to construction work, offered them scope for further loans and higher wages over a period of time. While domestic work limits their personal freedom in many ways it also offers them prospects for both self-development and the advancement of their families as my interviews show.
Migrants' investment in land, housing and education can increase their prestige and material wealth, which in turn can lead to a rise in social and economic status. However these efforts to mobilise symbolic capital and social repositioning must be viewed against the reality of backlashes against lower castes throughout India as they have become more assertive. While in the short term, migrants' efforts to alter the balance of power between themselves and the upper castes were focused on the sending area, corroborating Picherit (2012) and Gidwani and Sivaramakrishnan (2003), my interviews indicated that their aspirations were more fluid, and some changed towards considering urban areas for establishing social networks and more stable employment and accommodation.
These pathways through debt-migration are risky, and migrants remain vulnerable to the reversal of gains through shocks. Yet, in rural settings where access to welfare schemes is based on clientelism, shrinking landbased livelihoods and humiliating patron-client relations, new pathways to progress created by migration were regarded as the best way forward by the lower castes in my study sites. Although the present research did not follow the trajectories of these families to the end, it yielded insights into the nature of migrant agency in highly constrained circumstances and the reasoning behind their choices and actions.
Debt migration does not completely rearrange migrants' social field or radically change structuring relationships in the short term. But their tactics enable them to position themselves into jobs which have the potential to release their families from humiliating caste-based oppression and dependency relations with rural patrons.
All the migrants interviewed were in varying stages of repaying old debts and acquiring new ones. Most had moved away from dependence on village moneylenders, which they regarded as abhorrent and humiliating, to borrowing from mestris or new patrons who they felt offered more prospects for progressing onwards to a better life. There is little doubt that the debt-bondage patterns observed in my study are the result of highly unequal capital-labour power relations that create and perpetuate exploitative working conditions and that migrants' desire for status-enhancing consumption continues to keep them in debt. At the same time, the research shows that this debt can open up alternative livelihood pathways that can place migrants on an upward trajectory over time, albeit with continuing risks due to the failure of the state to protect construction and domestic workers' rights.
The findings presented here extend the debate on debt-bondage by showing how it can be a route to future freedoms and broaden the social fields of migrants from extremely disadvantaged backgrounds, especially women who are subjected to both caste-based and partriarchal restrictions. It does so by tracing the trajectory of debt-migrants, through recall, and adds a much-needed temporal dimension to the analysis of debt-migration and its outcomes. It is hoped that these research findings will stimulate debate and further empirical studies in culturally diverse contexts to add to our understanding of debt-migration as a choice, albeit under highly constraining racial and gendered inequalities. Given the prevalence of debt-migration in other areas of migratory work globally such as domestic and care work; hospitality and the garment industry where these social divisions are evident, this study offers a starting point to interrogate the relationship between debt bondage, migration and social mobility. Furthermore, the research offers a basis for the evaluation of interventions to prevent debt-migration and criminalise migration intermediaries, both of which have met with little success. There is a need to change perceptions away from seeing debt-migration as a poverty trap to understanding how it might bring the kind of change that migrants and their families value. Notes 1. This group was historically from Palamuru, the old name for Mahbubnagar. The system of recruitment through mestris began earlier in this district. 2. The Hindu caste system ascribes four main categories based on social and occupational stratification: Brahmins, Kshatriyas, Vaishyas and the Shudras. The first three are officially grouped under the Forward Castes (FCs) which include the Deshmukhs and Reddys. The Shudras correspond to official category of Other Backward Castes which include artisan communities. Those below these 'clean' castes were the so-called 'untouchable' castes including the Dalits such as the Mala and Madiga that are now grouped under Scheduled Castes (SCs).

Bourdieu identified four different types of capital:
Economic capital-includes income, wealth, financial inheritances, and monetary assets.
Cultural capital-This can exist in three forms: an embodied state, i.e. in the form of long-lasting dispositions of the mind and the body; in the objectified state as cultural goods; and in the institutionalised state, in the form of educational qualifications.
Social capital-Resources based on connections and belonging to networks. This capital is generated through relationships.
Symbolic capital-Different types of capital are converted to symbolic capital once they are recognised as legitimate. Legitimation is the key mechanism in the conversion to symbolic power. 4. The names of villages are not being mentioned here to protect the identity of the participants. 5. The Minimum Wages Act, 1948; the Contract Labour (Regulation and Abolition) Act 1970; the Bonded Labour System (Abolition) Act, 1976; the Equal Remuneration Act, 1976; and the Building and Other Construction Workers (Regulation of Employment and Conditions of Service) Act, 1996. 6. There has been a mushrooming of low-cost private schools all over India that readily accept lower caste students (Tooley and Dixon 2006). Country-wide studies in India establish the importance of education for social mobility among SCs and OBCs (see Dréze and Sen 2013). 7. See Kumar and Landy (2012) for a more detailed discussion on this issue.